ArticlePDF Available

Abstract and Figures

The introduction of the Internet of Things (IoT) and Big Data applications have garnered a massive amount of digital data. Processing and analysing these data demand vast computing resources, proportionately. The major downside of producing and using computing resources in such volumes is the deterioration of the Earth's environment. The production process of the electronic devices involves hazardous and toxic substances which not only harm human and other living being’s health but also contaminate the water and soil. The production and operations of these computers in largescale also results in massive energy consumption and greenhouse gas generation. Moreover, the low use cycle of these devices produces a huge amount of not-easy-to-decompose e-waste. In this outlook, instead of buying new devices, it is advisable to use the existing resources to their fullest, which will minimize the environmental penalties of production and e-waste. This paper advocates for using smartphones and smartphone crowd computing (SCC) to ease off the use of PCs/laptops and centralized high-performance computers (HPCs) such as data centres and supercomputers. The paper aims to establish SCC as the most feasible computing system solution for sustainable computing. Detailed comparisons, in terms of environmental effects (e.g., energy consumption, greenhouse gas generation, etc.), between SCC and supercomputers and other green computing initiatives such as Grid and Cloud Computing, are presented. The key enablers of SCC are identified and discussed. One of them is today's computationally powerful smartphones. A comprehensive statistical survey of the various commercial CPUs, GPUs, SoCs for smartphones is presented confirming the capability of the SCC as an alternative to HPC. The challenges involved in realizing SCC are also considered. One of the major challenges is handling the issue of limited battery in smartphones. The reasons for battery drain are recognized with probable measures. An exhaustive survey is presented on the present and optimistic future of the continuous improvement and research on different aspects of smartphone battery and other alternative power sources which will allow users to use their smartphones for SCC without worrying about the battery running out.
Content may be subject to copyright.
Scalable Computing: Practice and Experience
Volume 20, Number 2, pp. 259–283. http://www.scpe.org
DOI 10.12694/scpe.v20i2.1517
ISSN 1895-1767
c
2019 SCPE
GREEN AND SUSTAINABLE HIGH-PERFORMANCE COMPUTING WITH
SMARTPHONE CROWD COMPUTING: BENEFITS, ENABLERS, AND CHALLENGES
PIJUSH KANTI DUTTA PRAMANIK
, SAURABH PAL
,AND PRASENJIT CHOUDHURY
Abstract. The introduction of the Internet of Things (IoT) and Big Data applications have garnered a massive amount of
digital data. Processing and analysing these data demand vast computing resources, proportionately. The major downside of
producing and using computing resources in such volumes is the deterioration of the Earth’s environment. The production process
of the electronic devices involves hazardous and toxic substances which not only harm human and other living beings health but
also contaminate the water and soil. The production and operations of these computers in largescale also results in massive energy
consumption and greenhouse gas generation. Moreover, the low use cycle of these devices produces a huge amount of not-easy-to-
decompose e-waste. In this outlook, instead of buying new devices, it is advisable to use the existing resources to their fullest, which
will minimize the environmental penalties of production and e-waste. This paper advocates for using smartphones and smartphone
crowd computing (SCC) to ease off the use of PCs/laptops and centralized high-performance computers (HPCs) such as data centres
and supercomputers. The paper aims to establish SCC as the most feasible computing system solution for sustainable computing.
Detailed comparisons, in terms of environmental effects (e.g., energy consumption, greenhouse gas generation, etc.), between SCC
and supercomputers and other green computing initiatives such as Grid and Cloud Computing, are presented. The key enablers
of SCC are identified and discussed. One of them is today’s computationally powerful smartphones. A comprehensive statistical
survey of the various commercial CPUs, GPUs, SoCs for smartphones is presented confirming the capability of the SCC as an
alternative to HPC. The challenges involved in realizing SCC are also considered. One of the major challenges is handling the issue
of limited battery in smartphones. The reasons for battery drain are recognized with probable measures. An exhaustive survey
is presented on the present and optimistic future of the continuous improvement and research on different aspects of smartphone
battery and other alternative power sources which will allow users to use their smartphones for SCC without worrying about the
battery running out.
Key words: Green computing, Sustainable computing, Mobile computing, Grid computing, Smartphone computing, HPC,
Cloud computing, Data centre, Energy efficiency, Smartphone battery, Battery research, Future battery, Non-conventional energy,
Wireless charging, Alternative power source
AMS subject classifications. 68M14, 68W10
1. Introduction. In line with the massive increase in data from innumerable sources, the need for high-
performance computing (HPC) has increased enormously. But this has impacted the global environment badly
through various forms such as excessive energy consumption, carbon and greenhouse gas emission, heat gener-
ation, use of toxic materials in production, harmful industrial discharge, non-degradable waste generation, etc.
Device production and operations consume huge energy. For example, nearly 30,000 megajoules of energy is
used in the manufacturing of an average computer. The energy consumption demands more energy production,
which increases the carbon footprint [64]. More requirement of computers leads to more productions and more
uses of computers which means more environmental hazards and pollution. The various environmental hazards
caused by the production and operations of computing devices are listed in Fig 1.
To mitigate the environmental hazards due to computing devices, we need to concentrate on green and
sustainable computing. Several computational measures, as shown in Fig 1, are adopted that helps in min-
imising the energy consumption of the computers. But this is not sufficient for realising sustainable computing
absolutely. One of the strategies to attain sustainable computing to utilise the existing resources optimally
and fully which will minimise the requirement of new devices which ultimately reduce the problems due to the
production process as well as the e-waste. Several approaches have been opted in this regard as mentioned in
Table 1. Among them, Grid and Cloud computing are the most prominent initiatives which have minimised
the requirement of owning personal computer systems considerably. They have also replaced the need for cen-
tralised HPC systems such as supercomputers and mainframes to some extent. Though Grid computing intends
to fully utilise the existing resources (desktops), Cloud computing is not that successful in this aspect. Cloud
computing needs additional resources to provide cloud services. The centralised resources such as data centres,
at the cloud service providers end, consumes massive energy, leading to greenhouse gas emission substantially.
National Institute of Technology, Durgapur, India (pijushjld@yahoo.co.in)
Bengal Institute of Technology, Kolkata, India
National Institute of Technology, Durgapur, India
259
260 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
Fig. 1.1.Environmental hazards of the computing devices.
In fact, data centres capture the one-third share of the total energy consumption of ICT. A well-known report
[45] of the year 2010 stated that the electricity consumption by data centres had risen by 56% in five years;
whereas, during the same period, the overall increase in U.S. electricity usage was only 36%. This statistic
reflects the seriousness of the energy requirements of the data centres, which has become more severe in recent
years. If the right measures are not taken in due course, data centres are sure to become a grave threat to
the environment. Many companies and organizations are doing their best to tackle the energy appetite of the
data centres by adopting different measures. For example, Apple has the largest solar installation for powering
its data centre. Facebooks data centre in Iowa uses wind energy to meet the energy requirements of its data
centre. Google has placed its data centre in Hamina, Finland under the sea for cooling, thus curbing the energy
requirements significantly. Google is using AI for optimizing the energy requirements and consumptions in its
data centres.
The problem with Grid computing is that the desktops are losing popularity; in fact, the same for laptops.
Instead, smartphones are gaining huge acceptance as the new computer with the computing power they offer
thanks to the power-packed hardware. The technological progress of smartphones such as powerful SoCs with
multicore CPUs and GPUs has made them favourable as the primary computing device to many people. In-
dustries are also showing interest in this direction. Initiatives such as Microsofts Continuum1and Samsungs
DeX2are striving to bring the desktop experience on the smartphone. Microsoft has endeavoured to run its full
version of Windows 10 on the ARM chipsets, the most popular chipset for smartphones.
And the great thing about smartphones is that they have become indispensable to our lifestyle. It is not
feasible to restrain ourselves from using that. So, why dont we use these devices of its optimal potential, i.e.,
for computing purposes as well? If these abundantly available public-owned powerful smartphones are used as
1https://www.microsoft.com/en-in/windows/continuum
2https://www.samsung.com/global/galaxy/apps/samsung-dex/
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 261
Fig. 1.2.Measures adopted to minimise energy consumption of the computing devices.
computing resources by forming, like desktop grids, a grid of smartphones, the generated computing power can
well be compared to other HPC systems. We call this computing paradigm as smartphone crowd computing
(SCC). SCC refers to the approach of utilizing the public-owned smartphones (crowd of smartphones) to yield an
HPC facility anywhere and ad hoc basis. A number of powerful smartphone devices, collectively, can offer huge
computing capability. A satisfactory HPC may be achieved by making a grid of smartphones. The cumulative
computing power achieved by such grids of smartphones can tail off the dependency on the data centres and
low-end supercomputers as well [63].
In this paper, we advocate for optimized utilization of smartphones as primary computing devices with the
aim of reducing the two-thirds share of ICT energy consumption on the count of the use of computers and data
centres. The use of smartphones will lessen the environmental hazards because the manufacturing process and
262 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
device operation will be minimized significantly. Due to smaller in size, the less raw material will be used in
production means less exposure to harmful elements and less pollution due to e-waste. If organizations can
avoid buying computing resources, they can save a significant IT investment and operational cost. And less
device in production and operation means a direct benefit to the environment.
Contribution of the paper: This paper offers a fourfold contribution as mentioned below:
The environmental hazards of the growing computing devices and centralised HPC systems are impec-
cably identified.
SCC has been attempted to be established as a better alternative to the existing HPC systems in terms
of sustainable computing.
A detailed survey is presented on the modern smartphones computing capability, highlighting the
capabilities and specifications of the prominent commercial CPUs, GPUs, and SoCs for mobile devices.
An exclusive and state-of-the-art survey has been presented on the recent developments (research and
commercial) on different aspects of smartphone batteries.
Organisation of the paper: Section 2 discusses the environmental hazardous factors especially for cen-
tralised HPCs, in addition to the usual hazards of computers as mentioned in Fig. 1. Section 3 asserts the
focal essence of this paper. It establishes the role of SCC in sustainable computing. In Section 4, the major
factors that are supposed to make the idea of SCC feasible are identified. To endorse that, a comprehensive
statistical review of todays smartphones processing capability is presented in this section. The major challenges
faced in the successful deployment of SCC are recognized in Section 5. Among others, the foremost challenge is
the battery concern. The major reasons for battery drain are identified in Section 6. In this section, a detailed
survey is projected that covers the continuous research and improvement on different aspects of smartphone
battery which should help users to use their smartphone without worrying about the battery drain. The recent
developments, in different segments, towards enhanced energy sources for smartphones are also discussed. And
finally, Section 7 concludes the paper.
2. Factors Behind Environmental Hazards in Centralized HPC Systems like Cloud, Main-
frame, and Supercomputers. Centralized computing facilities such as Cloud services and data centres,
mainframes, and supercomputers have become a real threat to the environment. For instance, the data centres
are responsible for 17% of the global carbon footprint [61]. These large computing systems are notorious for
energy consumption. It is projected that data centres in the US only would be consuming 73 kWh in 2020.
Nearly 30 billion watts of electric power is needed to run the data centres. Among the amount of electricity
consumed by data centres, the servers waste 90% of their total power consumption in keeping the systems
running for the entire day. Besides the operational energy consumption, there are several other factors that are
associated with environmental hazards of the big centralised computing systems, as mentioned below [46]:
Coolant: Computing and allied infrastructures in the data centre produce enormous heat, and this requires
heavy cooling. Most of the air conditioning or cooling system uses coolant for quick and deep cooling. The
coolants used earlier for air conditioning systems are freon/halocarbon or chlorofluorocarbon. These coolants
are mildly toxic in nature and have caused ozone depletion significantly. The new coolants hydrofluorocar-
bon currently used in all air conditioning system is not risky to the ozone layer, but they significantly cause
atmospheric warming. This coolant traps the heat within the atmosphere and causes global warming.
Batteries: Data centres require huge batteries for power backup. These batteries make sure the power
flow is constant, and provide an electric power supply for a few minutes to hours based on the power cut or
power fluctuation. These batteries have a limited life span of 3 to 4 years. The batteries used more often are
the lead-acid battery. These batteries may cause an adverse effect on human health and the environment. The
fumes emitted from the battery are dangerous to human health. Whereas the lead used in batteries is highly
toxic, and if they are not properly disposed of, they may contaminate the soil and water. Presently, in most
power backup systems, dry batteries are used. These batteries use heavy metals like mercury, nickel, cadmium,
and silver. The metals, mercury, nickel, and cadmium are toxic and have a fatal effect on human health. These
batteries cannot be dumped into the open and need to be properly recycled. It is necessary that the batteries are
properly disposed of. There are multiple companies which recycle UPS batteries. 80% of leading manufacturing
units in the U.S. undergo battery recycling.
Cleaning materials: Dusting and cleaning are important in data centres in order to operate efficiently.
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 263
Table 1.1
Sustainable computing approaches, their advantages and problems in terms of sustainability
Working principles Environmental advantages Problems
Grid
com-
pu-
ting
The unused computing
resources of the con-
nected computers are
used opportunistically
providing a massive
virtual computing fa-
cility [62]
Utilising of existing idle resources fa-
cilitates less energy consumption and
minimise environmental hazards due to
the manufacturing and operation of the
computing systems required otherwise.
Needs fast and reliable LAN and WAN connections.
In the volunteered grid, it is difficult to motivate the
resource owner to lend their resources.
In the case of the non-volunteered grid (e.g., commercial
grids), the services might be costly.
Not only setting up and managing the grid resources
but also accessing them often require expertise.
Cloud
com-
pu-
ting
Facilitates a shared
pool of configurable
computing resources
along with quality
services which can
be rapidly provided
on-demand basis [65]
Eliminates the need for private comput-
ers, servers, and data centres which has
significantly contributed to energy sav-
ing.
The data centres, comprising the large servers and com-
puters, and associated cooling systems consume massive
power.
Accessing the cloud service depends on the internet con-
nection. The instability/unreliability and unavailability
of connection hold back accessing the cloud.
Involves security and privacy issues.
It is true that the cloud computing eradicates the big
upfront investment on computing resources but access-
ing the right service is not that cheap either.
Involves data transfer cost.
Since cloud services are typically generic they lack in
flexibility.
The user/client has minimal control of their own ap-
plications as the cloud service infrastructure is entirely
owned, managed and monitored by the service provider.
Switching or migrating to a different cloud service
provider is often complex or infeasible.
Ser-
ver-
less
com-
pu-
ting
Consumers borrow
a third-party server
when needed; thus,
doing away with
purchasing and main-
taining own server
exclusively [36].
Saves the energy requirement for run-
ning privately-owned servers.
It also scales down quickly by shutting
resources down when they are not in
use. This saves a lot of energy con-
sumption.
Serverless codes not necessarily be run
in any specific server. Hence, the
serverless applications can be deployed
in the edge of the network, i.e., close
to the end users [18]. This will not
only reduce the latency but also saves
a significant amount of energy by elim-
inating the need for unnecessary data
transmission [65].
Not suitable for real-time applications which require low
latencies.
Also, not suitable for applications which need long exe-
cution times.
Everything operates in stateless fashion; hence, han-
dling state using stateless functions is a real issue.
After some time of being idle, the function will require
to go through a cold start which not only takes up to a
few seconds but also consumes energy.
Using
ter-
minal
ser-
vers
The use of terminal
server along with the
thin clients gives users
a feel of computation
being taking place in
the very same terminal,
while the actual com-
putation takesplace in
the terminal server.
The processing and storage require-
ments for client machines are minimal
because a terminal server hosts all the
application logic which also runs on the
server.
The thin client uses up to the 1/8th
amount of energy in comparison to
workstations and thus considerably re-
duces energy consumption.
Running applications on a remote server always involve
performance issues.
There are chances for terminal servers getting bottle-
necked with overloaded requests. Hence, the terminal
server needs to be powerful enough to be able to handle
all connections.
If the terminal server is not backed up, there is a high
risk of downtime due to a single point of failure.
If the communication network is not reliable, the system
will be affected harshly.
Vir-
tua-
liza-
tion
A single physical server
is logically partitioned
into multiple virtual
servers or server in-
stances, sharing the
same hardware re-
sources and allows
processing multiple
applications or jobs
on different virtual
servers.
Effective resource utilisation leads to
fewer production and less e-wastage.
A single physical server acting as multi-
ple server instances consumes consider-
ably less energy in comparison to sepa-
rate dedicated servers.
It requires quantitatively less hardware
to run similar application than dedi-
cated systems which leads to fewer de-
vice production and less e-wastage.
The absence of the usage of the local
hardware or software cuts the overall
energy consumption.
The required hardware specification (e.g., memory, pro-
cessor, etc.)is much higher for the same task executed
in a native computer.
Involves complex troubleshooting, in case of failure.
Degraded performance than a physical server.
Suffers from availability issue which discourages using
virtual servers for mission-critical applications.
Has major security issues.
264 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
Dusting and cleaning require special material that minimizes static discharge and are safe for electronics.
Varieties of cleaning solution are available, and most of them are toxic as they contain bleach, ammonia, or
chlorine. These toxic cleaning solutions have an adverse effect on human health.
Diesel fuel: Another measure of overcoming power failure is the diesel fuel-based power generator. In a
location where there is a recurrent power failure, the data centres heavily depend upon the power generators for
a constant supply of power. Diesel fuel-based power backup is the primary solution for powering data centres
since battery technology is incapable of producing power for long times. Diesel fuel produces an enormous
amount of CO2and other chemicals which have an adverse effect on human and does global warming.
Electronic waste: Electronic equipment have a finite lifespan. Even for functional equipment, the per-
formance degrades with time. Conventionally, the IT products in business firms and data centres such as a
server rack, UPS, desktop computers, servers, monitors, hard drives, etc., are replaced after every 3 to 5 years.
Globally, the scrapped IT equipment for the last 20 years are enormously huge, and it is increasing day-by-day.
The e-wastes are harmful to the environment if dumped on the open landfills. The monitor, CPU, UPS, etc.,
are made up of materials which have negative effects on the environment. It is the responsibility of the business
organizations to either reuse or sell/donate the reusable equipment or recycle them properly.
Fire suppression: It is important to have data intact and safe with the service providers. Due to the
fact, the data centres are dealing with high voltage, and hence, there is a chance of short-circuiting and fire. All
data centres carry extensive and fast fire suppression system which may be environmentally negative. Various
chemicals employed for the fire system may be harmful to the environment contributing to effects like ozone
layer depletion and global warming. These chemicals are toxic and may find its way to underground water or
to rivers, thus, contaminating the water resources. However, as these systems are not used or tested regularly,
the chemical harm is minimized.
Packaging: Business organizations like IT firms and data centres import huge IT equipment every year.
Customers and data centres ship equipment and supplies to and from data centres frequently. The equipment
comes with packaging material like boxes, cardboard, plastic bags, foams, thermocols adds huge waste every
year. A large data centre processes literally tons of boxes each year. Packaging materials like cardboard boxes
can be recycled and are biodegradable, but other materials like foams, thermocols, plastic bag, and plastic
support accessories are nonbiodegradable and need proper recycling. Dumping these on the open area may
harm the environment.
Office zones: IT business firm and data centre office area contribute a lot to emissions and wastes. This
includes wastes due to the inappropriate use of electricity and water resources. Running air conditioners, heating
equipment, and lights throughout the office for the entire day, even if it is not required, causes considerable
electricity wastage. Similarly, unmonitored use of water resources in washrooms and kitchen lead to wastage of
water. Besides, daily office chores produce lots of paper, plastics, and packaging wastes. Floor cleaning, glass
pane, computers, and carpets also produce chemical wastes. Even the fluorescent bulbs used by offices also
carry toxic matters like mercury, lead, etc., and may cause environmental and health hazards.
Overall, it can be said that the centralised HPC systems have a negative impact with respect to nature and
natural resources. But as said there is always room for improvement, which may range from minimizing the
toxic chemicals used to practising the policies of reuse-reduce-recycle across wherever possible. As technology
is developing, new products are coming which causes less emission, less environment hazard and lowers the
greenhouse gas emissions.
3. Smartphone Crowd Computing as a Solution Towards Sustainable HPC. Whatsoever the
negative impacts of computers have on the environment, we cannot head them off from our livings. We need
them in every step of our daily life. Actually, we need more and more powerful computers day by day for
various purposes. In view of that, we need to consider seriously to minimize the environmental externalities of
producing and using computers.
On the other hand, smartphones have become a part of our daily life, and they are getting more resource-
packed and hence, computationally more powerful (see Sect. 4.2). So, it is a good idea to use them as computing
devices because smartphones have less adverse effects on the environment compared to desktops and laptops.
Actually, to many people, smartphones have already become the primary computing device [63]. Because of
portability and easy to use, people are fulfilling most of the daily computing needs using smartphones.
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 265
Since the smartphones are getting close to the sophisticated computers in terms of resources and computing
ability, much like desktop Grid Computing, if a number of such smartphones are connected through a network
and make them act as a virtual single unit of computing resources, it can match the power of high-end powerful
computing systems such as supercomputers and data centres. Like volunteer computing, instead of buying
the required number of smartphones, public-owned smartphones are exploited on a need basis. Hence, this
computing paradigm is termed as smartphone crowd computing (SCC) [63, 67].
3.1. Smartphone Crowd Computing. SCC is a distributive computing framework where a large job is
divided and distributed to the peoples smartphones (hence, crowd computing) to be executed. The philosophy of
SCC is to combine computation power of numerous distributed smartphones to escalate the overall computation
power. SCC paradigm shares the CPU cycle of multiple smartphones in a voluntary or involuntary basis to
resolve a high computational task. The concept is to process a computationally big task by segmenting the
task into smaller microtasks, and each micro-task is processed on the smartphones individually. The processing
is typically carried out opportunistically, i.e., whenever the host smartphone is idle, the crowd computing
application installed on the smartphone will assign the task to the smartphone processor (CPU or GPU). Even
though the computation power of smartphone devices is less, the cumulative processing power of smartphones
is quite high enough to execute complex computational jobs. Fig 3.1 lists the highlights of SCC.
In SCC, a designated coordinator, may be fixed (a desktop computer) or mobile (a smartphone), hosts
the crowd computing application. The coordinator is responsible for dividing and distributing the job, finding
suitable resources (smartphones), allocating and scheduling the jobs to the smartphones, monitoring for fault-
tolerance and collecting the results from each resource. The coordinator keeps microtasks in the job pool, ready
to be dispatched. For distributing the tasks, available crowd devices (smartphones) are searched, and a set of
suitable devices are being selected as crowdworkers (computing resources) for processing the jobs. After the
execution of these tasks, each node submits the completed task with results to a centralized master device,
where all the results are gathered, checked for error to get the final result. It is the job of the master device
(coordinator) to check the error and validity of each task execution.
Since the devices are mobile, the reliability of an available device for a long time is very thin. The coordinator
makes a task group among smartphones which either voluntarily or involuntarily joins the group. SCC provides
quite a flexibility in terms of job allocation and execution as a small task group could be established, comprising
a few available smartphones. Even without an internet connection, the smartphones in the task group can
communicate using Hotspots or Bluetooth. This makes SCC as the ideal alternative of centralised HPCs where
these HPCs are not available or not accessible. The ubiquity and flexibility of SCC is an ideal candidate for
providing the computing facility to applications such as pervasive computing [68] and cognitive IoT [66], where
the data need to be processed near the data source as well as sink.
3.2. How Can Smartphone Crowd Computing Minimize Environmental Hazards?. Object-
oriented programming brought the revolution in software development by introducing the concepts of reusability
and polymorphism, which allow software modules to be used multiple times for multiple purposes. This saves a
significant amount of man-hours and cost. The same is true for crowd computing. Already existing devices are
used for computational and other purposes. This reduces the requirement of buying IT infrastructure separately
because the public will buy phones for their own purposes, anyway. Moreover, the utilization of already
existing smartphones to achieve SCC instead of traditional HPC will reduce the requirement for dedicated large
computers.
If smartphones and SCC are employed in a wide-ranging computing requirement, due to the substantial
reuse, the need for new devices will be much lower. That means unnecessary production is avoided which
will, in turn, minimize the environmental externalities of producing computing devices. The reduction in
the production, operation, and degradation of unnecessary devices will reduce the amount of global e-waste
considerably. In addition to that, being much smaller in size, smartphones produce less e-waste.
Using smartphones for computations will curb energy requirements significantly. The smartphone processor
architectures offer far more energy efficiency than those of desktops and laptops. The energy consumption of a
standard smartphones ranges from a few to 10 kWh per year. Therefore, total energy consumption is 10 TWh
per year considering one billion smartphones are in operation worldwide. This is only 1% of the total energy
consumed by ICT which is typically on the order of 1,000 TWh per year [32].
266 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
Fig. 3.1.Highlights of Smartphone Crowd Computing
Hence, it is apparent that smartphones and SCC will significantly help in protecting the environment along
with saving resources and minimized organizational IT expenditure.
3.3. Environment-friendliness of Smartphone Crowd Computing in Comparison to Other
HPC Systems. In line with our argument that SCC is a viable alternative to the costly HPC systems, in
this section, we shall check out how SCC is more environment-friendlier than other HPC systems viz. desktop
Grid Computing, data centre, and supercomputers.
Due to their computing capacity and power, the energy requirements of supercomputers are gigantic and
might well be equivalent to that of a small city. The correct response relates to electricity; specifically, 17.8
megawatts (MW) of power is required to run Tianhe-2, one of the Top500 ranking supercomputer, boasting
33.9-petaflop through 3.12-million processors. An exaflop (1,000 petaflops) computer needs approximately 500
MW, which is equivalent of the total output of an average-size coal plant, and enough electricity to cater the
needs of all the households in a city like San Francisco [35].
In recent years, the computing services offered through Cloud Computing have got tremendous popularity.
People can rent computing resources on usage and requirement basis. The cloud service providers maintain
big data centres to cater to the computing resource need of the clients. A data centre, abstractly, can be
described as an abundant number of computers stacked together. To make the cloud service available 24x7,
these computers are kept ON throughout the day which makes them very hot. As a result, a huge amount
of power is consumed not only to run these computers but also to keep them cool. About 30 billion watts of
electricity is needed to run the data centres (comparable to the electricity generated from 30 nuclear power
plants) which is nearly 17% of the total carbon footprint caused by technology [61]. Data centres consumed
416.2 TWh of electricity in the year 2015 which is roughly 3% of the global electricity supply [9]. A single data
centre can consume more power than an average town. To provide uninterrupted power supply in case of power
failure, the data centres run generators that emit diesel exhaust. Todays data centres cause approximately 0.3%
of overall carbon emissions [43], which is equivalent to the carbon footprint generated by the airline industry
[9]. Of the total global greenhouse gas emissions, the power-hungry data centres account for nearly 2%. This is
putting an immense impact on the environment leading to global warming. The bad news is, every four years,
this energy requirement is getting doubled and the total energy requirement of the data centres, globally, will
increase threefold in the next decade. By 2025, data centres are expected to use 20% of the worlds energy
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 267
[5]. The efficiency of the data centre is measured in terms of Power Usage Effectiveness (PUE). PUE compares
the non-computing energy to the amount of energy to power actual machines. Data centres operate at 70% of
overhead energy. It means another 0.7 units are used behind the infrastructure of the data centres. So, the
total PUE goes up to 1.7 [61]. Typically, the PUE of the common data centres is about 2.0 [43].
As mentioned in Sect. 1, the desktop Grid is an affordable option for sustainable computing. Here, the
existing devices are used as computing resources instead of buying dedicated HPC. This helps the environment
as this will minimize the requirement for producing extra dedicated HPC systems. Grid Computing can lower
the environmental externalities of supercomputers and data centres considerably. SCC promises to minimize it
further.
Due to the much smaller size, smartphones have much less negative environmental impact than desktops
and laptops and obviously than supercomputers and data centres. So, a grid of smartphones (SCC) will be
more environment-friendly. As people will buy smartphones anyway, SCC can use the existing devices without
requiring additional device production. The environmental advantages of SCC can be summarized as follows:
No need for explicit production of computing devices as people will use smartphones anyway.
Production of smartphones is much environment-friendly compared to large computers.
Due to the small size, the e-waste will be lesser and can be handled more efficiently.
No dedicated cooling systems are required.
No power backups, such as large batteries and generators are required.
Smartphone processors are typically energy-efficient. As a result, they consume much less energy than
other computing systems to perform the same operation.
Table 3.1 statistically demonstrates the advantages of smartphones, in terms of environment-friendliness,
compared to desktops and laptops (Grid Computing), data centres, and supercomputers.
The excessive number of smartphones may be advantageous to SCC, but it is really worrisome considering
the amount of e-waste generated. Therefore, it is extremely crucial to opt for the proper disposal of discarded
devices and try to recycle as much as possible. Fig. 3.2 lists the responsibilities of the smartphone user
suggesting what to be done when they discard their smartphones.
Fig. 3.2.Smartphone users’ responsibility towards e-waste management [51].
4. The Enablers of Smartphone Crowd Computing. In this section, we shall look into the key factors
that have helped SCC to be a feasible HPC solution.
4.1. The Mass Adoption of Smartphones. According to the recent research market statistics, globally
smartphone shipments had reached 1.55 billion [75]. It is expected that by 2020, the shipment would reach
6 billion [76, 34]. Fig 4.1 shows the estimation of the number of smartphones that are going to be shipped
globally from 2018 to 2022. Alone in India, in 2017, the number of mobile users had crossed 300 million [38].
268 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
Table 3.1
Environmental externalities comparison of smartphones with data centres, supercomputers, and Grid Computing (desktops
and laptops)
Grid Computing
Data centre Super-
computer
Desktop Laptop Smartphone
Energy
con-
sump-
tion
200 TWh/year [43]. 17.8 MW for
Tianhe-2, the
33.9-petaflop
supercom-
puter with
3.12-million
processors
[35].
100-150 Watts/hour,
600 kWh/year [25].
60
Watts/hour,
300-150
kWh/year
[25].
1.5-3 Wh. An average
phone needs 2kWh/year.
CO2
emission
in the
man-
ufac-
turing
process
171,630 kg CO2[39].
Around 0.3% of overall
carbon emissions [43].
0.175 mil-
lion kg/year
per super-
computer
(equivalent
capacity of
1000 PCs).
380 kg/desktop [31]. 227 to 270
kg/laptop
[1].
16 kg/year [10]. An
average mobile emits 35
kgs of carbon while man-
ufacturing [2].
Other
environ-
mental
hazards
[40, 55,
28]
Along with the common
hazardous materials such
as Fe, Cu, Al, Ag, Au,
Pt, Pd, Pb, Hg, As,
Cd, Se and hexavalent Cr
and BFRs, other harm-
ful elements such as ethy-
lene/propylene glycol for
cooling systems, diesel
fuel for backup genera-
tors, lead-acid batteries
for UPSs, and compressed
gases for fire suppression
makes data centre real
peril to the environment.
Same as data
centres.
The metals contained
in PCs commonly in-
clude Al, Ag, As, Au,
Ba, Be, BFR, Cd, Co,
Cr, Cu, Fe, Ga, Hg,
Mn, Pb, Pd, Pt, PVC,
Sb, Se, and Zn. Most
of them are really haz-
ardous and contami-
nate soil, water, and
air, if not properly dis-
posed of.
Almost all
the haz-
ardous
elements of
desktops are
also found in
a laptop, but
the quantity
is less as
laptops are
typically
smaller than
desktops.
The hazardous metals
such as Al, Ag, Au, Cu,
Fe, Pb, Hg,Cd, etc. are
needed in smartphone
manufacturing also, but
in much less quantity
than desktops and lap-
tops.
E-waste
gener-
ated
32360 metric tons of e-
waste in 2018.
9.3 million
tons/year
[19].
41698.8 metric tons
[79].
3230 metric
tons [79].
In India, out of 650
million mobile users,
40% have changed their
phones in 2017, generat-
ing huge e-waste [52]. In
the USA, yearly, nearly
150 million mobile
phones are discarded.
Weight
fraction
of ma-
terials
(%) [49]
N.A.3N.A.?? 47.2 Fe, 0.9 Cu, 2.8
plastic, 9.4 PCB.
19.5 Fe, 2.4
Al, 1.0 Cu,
25.8 plastic,
13.7 PCB,
14.4 battery.
0.8 Fe, 0.3 Cu, 37.6 plas-
tic, 30.3 PCB, 20.4 bat-
tery.
E-waste
decom-
position
Decomposing is challeng-
ing as a huge volume
of e-waste generated due
to a large scale of com-
ponents. Require large
dumping ground; risk of
toxic metals and chemi-
cals; and contamination
risk. But, since the
data centres are gener-
ally owned by big com-
panies/institutes, they are
expected to follow the sys-
tematic decomposing and
recycling process.
Same as data
centres.
As the number of users
is very large scale, the
proper and systematic
decomposition of e-
waste is really difficult.
Most of the computers
are public-owned or
owned by small or-
ganizations. Most of
them do not follow
the proper decompo-
sition and recycling
processes.
The same
problem, but
moderate
due to less
equipment
as compared
to desktop.
The continuous growth
in smartphone users
with very brief use-cycles
is a great challenge in
terms of decomposing
and recycling as the
lack of awareness and
eagerness among the
public. But if the civic
authorities take active
roles and are able to
convince people the
necessity of proper
disposal of discarded
devices, the problem can
be tackled.
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 269
Fig. 4.1.Worldwide smartphone shipments forecast from 2018 to 2022 [76].
The number of active smartphone connections, worldwide, will reach 6 billion by 2020 [34]. The huge adoption
of smart mobile devices across the world has put a big platform for crowd computing [63, 67].
Factors driving the smartphone market: There is a number of factors influencing the growth of
smartphone use in the global market as mentioned below [34]:
Rapidly falling price of smartphones has accelerated the customers to move from basic and standard
feature phones to smartphones.
Developing smartphone technology has a reason for the increase in the sale of low-end smartphones.
The increasing availability of the highspeed 3G/4G spectrum with increased mobile broadband connec-
tions all around the world.
The availability of highspeed data-centric services and low tariffs have increased the adoption of smart-
phones in both developed and developing societies. Further, the availability of cheap data tariffs tailored
as per the customers needs has also reasoned for smartphone adoption in developing countries.
The concept of tailoring the data tariffs for cost-conscious prepaid consumers can be linked with the
selling rate of smartphones.
Efficient retail channels and supply chain have helped manufacturers to reach customers from every
corner across the globe.
The government policies in support of the growth of the smartphone and subsidiary industries have a
significant role in price slicing and the growth of mobile networks.
4.2. Powerful Smartphones. Over the last couple of years, the smartphone industry has seen an un-
precedented focus on its hardware. Be it CPUs or GPUs or even DSPs, the processing capability of smartphones,
to meet various purposes, has been increased exceptionally. The CPU and memory architectures are designed
and tuned to boost heterogeneous computing. The GPUs are also engineered to enhance general purpose GPU
(GPGPU) computing performance. Advancement on each module, though separately, makes the smartphones,
as a whole unit, a great possibility to become a powerful computing platform. Let us have a run up on the
latest developments on the smartphone hardware.
4.2.1. Powerful CPUs. The first multicore smartphone, announced at the end of 2010, was the LG
Optimus 2X, loaded with the Tegra 2 processor from NVIDIA, with a maximum frequency up to 1.2 GHz
[80, 57]. Since then all smartphone manufacturers have been in a war to load their products with an increasing
number of cores. To buttress this exigence, SoC makers are coming out with more and more powerful smartphone
processors quite regularly.
ARM Processors: ARM is the most popular processor architecture used in smartphones and supported
by most of the major smartphone operating systems. More than 95% of smartphones employ ARM processors
either as main application processors (Cortex processors) or other auxiliary processors such as ARM Embedded
Processors, ARM Embedded Real-time Processors, ARM Secure Processors, etc. [6]. These processors are
particularly preferred in smartphones due to the verity that they yield impressive performance in spite of
consuming significantly less power. High scalability, power efficiency and superior computing performance
270 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
(nearly 30 times better than the smartphones of that period) of ARMv7-A4based smartphones have resulted in
an increase in smartphone sales colossally since its introduction in 2009 [6]. ARMs latest architecture, ARMv8-
A5, having state-of-the-art 64-bit instruction set with superior 64-bit programming model, is the pertinent
extension of its predecessor. It supports the 64-bit general purpose and floating-point registers, each of 32
numbers, promising enhanced performance [6]. ARMs Cortex-A536and Cortex-A577are the first ARMv8-A
based big.LITTLE8pair. As the big processor, Cortex-A57 produces an exceptional degree of performance by
harvesting techniques like out-of-order execution, wide multi-issue capacities with large memories, whereas the
Cortex-A53 focuses on power-efficient operations by means of a simpler pipeline in a smaller configuration while
still delivering handsome performance. In the context of energy efficiency, it is worth mentioning of the Cortex-
A359which is designed for the next-generation smartwatches and probably has set a new standard in energy
efficiency for mobile processors [36]. In 2015, the ARM came with Cortex-A7210, based on 64-bit ARMv8-A, to
thrust the performance up for the next generation of phones. The Cortex-A7311, the smallest ARMv8-A based
processor, has been 30% faster and 20% energy efficient than A72. A7512 came with significantly improved
integer and floating-point performance, and better memory workload management. The latest Cortex-A7613
is designed to deliver laptop-class performance with mobile efficiency. In comparison to the A75, it offers
35% better performance with 40% more power efficiency. The development of ARM processors, in terms of
competence in floating point performance, has made them a serious contender in consideration for a range of
scientific applications. In fact, due to their energy efficiency, the ARM processors are going to be used in the
futures supercomputers also.
4.2.2. Powerful GPUs. In this section, we shall check out some of the latest and most powerful GPUS
for smartphones. Imagination PowerVR: The PowerVR14 Series7XT15 family drives GPU performance of
mobile devices to new heights which can deliver up to 1.6 TFLOPS by employing multiple clusters (2 to 16). All
the GPUs in this family are offered in two modes FP32 and FP16. The significant 8 clustered GT7800 has 256
FP32 ALU cores or 512 FP16 ALU cores. At the clock rate of 800 MHz, the FP32 cores deliver 820 GFLOPS
while the FP16 cores deliver 410 GFLOPS. The highest rung in the series, the GT790016 has 16 clusters either
of 512 FP32 cores that deliver 800 GFLOPS at a clock rate of 800 MHz (for 16nm FinFET+) or of 1024 FP16
cores, capable of delivering an imposing 1.6 TFLOPS at the same clock rate [77]. The company has provided
an optional FP64 ALU for every pipeline that can be added to designs to deal with high-end supercomputing
like applications.
Qualcomm Adreno: Snapdragon 810’s Adreno 43017 GPU has a whopping 288 numbers of shader cores.
It has a clock rate of 600 630 MHz, and it delivers 388.8 408 GFLOPS. The Adreno 53018 packed in Snapdragon
820 and 820A works in the range of 510 - 624 MHz offering 407.4 - 498.5 GFLOPS. It can achieve 588 GFLOPS
working at 736 MHz.
NVIDIA Tegra: Tegra K119, has 192 Kepler cores running at 850 MHz and delivering 326.4 GFLOPS.
The Tegra X120 features impressive 256 Maxwell cores. Though it has fewer shader cores than Adreno 430, it
outsmarts the latter in ground performance. The clock speed of Tegra X1’s GPU touches 1 GHz, and it delivers
512 GFLOPS.
4https://developer.arm.com/docs/ddi0406/latest
5https://developer.arm.com/docs/den0024/latest/armv8-a- architecture-and- processors/armv8-a
6https://developer.arm.com/products/processors/cortex-a/cortex- a53
7https://developer.arm.com/products/processors/cortex-a/cortex- a57
8https://developer.arm.com/technologies/big-little
9https://developer.arm.com/products/processors/cortex-a/cortex- a35
10https://developer.arm.com/products/processors/cortex- a/cortex-a72
11https://developer.arm.com/products/processors/cortex- a/cortex-a73
12https://developer.arm.com/products/processors/cortex- a/cortex-a75
13https://developer.arm.com/products/processors/cortex- a/cortex-a76
14https://www.imgtec.com/graphics- processors/
15https://www.imgtec.com/graphics- processors/powervr-series7xt/
16https://www.extremetech.com/gaming/199933- powervr-goes- 4k-with-gt7900-for-game-\\consoles
17https://www.notebookcheck.net/Qualcomm- Adreno-430.146784.0.html
18https://www.notebookcheck.net/Qualcomm- Adreno-530.156189.0.html
19https://www.nvidia.com/object/tegra- k1-processor.html
20https://www.nvidia.com/object/tegra- x1-processor.html
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 271
ARM Mali: The Mali GPUs from ARM are not really well-known for their performance. Though they
work at the higher clock rate, the GFLOPS output from them, comparatively, on the lower side. The Mali-T880
MP4 used in Helio X25 runs at 850 MHz but offers merely 115.6 GFLOPS while the same GPU used in Kirin
950/955 runs at 900 MHz and offers 122.4 GFLOPS [3]. The Mali-T880 MP12 used by Samsung Exynos 8890
clocks at 650 MHz offering 265.2 GFLOPS. The latest of the series, the Mali-G7121 is 50% faster than the
Mali-T880, attributed to 32 shader cores, twice the amount used on the Mali-T880 while at the same time it is
20% more power efficient [4, 37].
4.2.3. Powerful SoCs. In this section, we shall have a glimpse of some of the latest and most powerful
smartphone processors and SoCs.
Qualcomm Snapdragon: Snapdragon22 is a family of mobile SoC processor, architected and marketed
by Qualcomm Technologies. It leapt a significant step in mobile computing by introducing the first 1 GHz
mobile processor ever which featured in Toshiba TG01 in the year 2009 [42]. Since then there is no stopping
from this SoC leader. Qualcomm Snapdragon 800 processor, launched in the year 2013, packed with four Krait
400 CPU cores and providing up to 2.45 GHz clock speed, outperformed all other processors in the mobile
segment [72]. Next in the line, the octa-core Snapdragon 810 follows the big.LITTLE concept from the ARM
and arranged in two clusters of different clock frequencies, providing an overall performance of 2.7 GHz [50].
The performance-oriented cluster of four 64-bit ARM Cortex-A57 runs at 1.958 GHz, and an energy-efficient
cluster of four ARM Cortex-A53 runs at 1.555 GHz [72]. It also has one of the highly powerful GPUs, the
Adreno 430. The Snapdragon 820, a 64-bit CPU that uses the ARMv8-A instruction set has four custom Kryo
cores arranged as two dual-core setups - one for highly demanding tasks and the other for relatively low-profile
tasks [54]. The performance-oriented cluster clocks at up to 2.15 GHz and the energy-saving cluster runs
at 1.6 GHz. The Snapdragon 820 also packs Adreno 530 GPU, the new Hexagon 680 DSP and the 14-bit
Qualcomm Spectra ISP [70]. The integration of these powerful chips is expected to provide high-performance
mobile computing. To reinforce effective heterogeneous computing, the Adreno GPUs are devised to perform
seamlessly with the Snapdragon CPUs, DSPs, and ISPs to accomplish processing-intensive GPGPU compute
tasks. The new Snapdragon 82123 , loaded with the Qualcomm Kryo quad-core CPU, delivers 10% better
performance than the 820, topping speeds up to 2.4 GHz. The Snapdragon 83524 is 35% smaller and uses
25% less power than its predecessors. The Snapdragon 84525 having an Octa-core CPU (Qualcomm Kryo 385)
provides a clock speed of 2.8 GHz. It also incorporates the Qualcomm Adreno 630 GPU. The latest Snapdragon
85526, built on a 7nm27 technology from Taiwan Semiconductor (TSMC), have an eight-core (one of 2.84 GHz,
three of 2.42 GHz, and four of 1.80 GHz) Qualcomm Kryo 485 CPU and an Adreno 640 GPU.
NVIDIA Tegra: NVIDIA has been the most prominent SoC makers in the market with their powerful
Tegra series. The company has redefined mobile processing with their incredibly advanced SoC, the Tegra
K1, launched in 2014. It encompasses 192 supercomputer-class GPU cores, based on the Kepler architecture
that has accelerated some of the worlds most powerful supercomputers [15]. These prodigious powerful GPUs
generate unbelievable extent of computing potential within an incredibly tiny package with remarkable power
efficiency. As mentioned earlier, NVIDIA compares their latest product the Tegra X1 as yesteryears first teraflop
supercomputer. As part of the CPU, it boasts eight 64-bit ARM cores, divided into two different clusters of
four each to exploit ARM’s HMP scheme. The first cluster is of ARM Cortex-A57 (for enhanced performance)
and the second one of ARM Cortex-A53 (for energy efficiency) cores. At 16-bit floating point operations, the
Tegra X1 nails a peak 1024 GFLOPS. Undoubtedly, the NVIDIA Tegra X1 is the most advanced SoC for any
mobile device by far and considered well ahead of time.
Samsung Exynos: The Exynos series of SoCs is developed by Samsung, indigenously, for their sophis-
ticated high-end smartphones. Samsung Exynos 742028 , the octa-core SoC, features four Cortex-A57 cores at
21https://developer.arm.com/products/graphics- and-multimedia/mali- gpus/mali-g71-gpu
22https://www.qualcomm.com/products/mobile- processors
23https://www.qualcomm.com/products/snapdragon- 821-mobile- platform
24https://www.qualcomm.com/products/snapdragon- 835-mobile- platform
25https://www.qualcomm.com/products/snapdragon- 845-mobile- platform
26https://www.qualcomm.com/products/snapdragon- 855-mobile- platform
27https://www.tsmc.com/english/dedicatedFoundry/technology/7nm.htm
28https://www.samsung.com/semiconductor/minisite/exynos/products/mobileprocessor/exynos- 7-octa- 7420/
272 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
2.1 GHz and four Cortex-A53 cores at 1.5 GHz along with the ARM Mali-T760 MP8, at 772 MHz, as a 12
core GPU which delivers 210 GFLOPS [3]. The latest Exynos 889029, also an octa-core processor, has four
power-efficient A53 running at 1.586 GHz and four Exynos M1 running at 2.29 GHz, the highest in the segment.
If only 1-2 cores are used at a time, the Exynos M1 can peak up to 2.60 GHz [30], which is more than an Intel
Core i5 chip (Intel i5 4302Y30) that has a clock speed of 2.3 GHz. As GPU, Exynos 8890 boasts the Mali-T880
MP1231 which has a clock speed of 650 MHz and delivers astonishing 265.2 GFLOPS.
Apple A9x: Apple Inc. also did not want to be behind in the race as they came up with their indigenous
SoCs, the Apple A series. They have showcased their sophisticated product Apple A9X32 , a 64-bit ARMv-8A
based SoC, into their iPad Pro, which was released on November 11, 2015. The A9X features a 2.26 GHz dual-
core CPU, called Twister. It boasts PVR 12 cluster Series7 GPU (GT7800 model has 8 cores while GT7900
has 16). The A9X also has the M9 motion coprocessor embedded in it. The Apple A10 is claimed to have
better CPU and GPU performance compared to the Apple A9 by 40% and 50% respectively. The successor
of A10, A11 Bionic33 incorporates six cores; out of which two have 25% better performance than A10 while
the performance of the rest is up to 70% better than A10. Additionally, the three-core GPU offers 30% faster
graphics performance than the A10. In the A12 Bionic34, out of six cores, the two performance cores are 15%,
and four energy-efficient cores are 50% better than their counterparts in A11. The four-core GPU of A12 is up
to 50% faster than A11s three-core GPU.
MediaTek’s Helio: High performance in the low-budget is the hallmark of the SoCs from this Chinese
company. It has produced the first-ever deca-core SoC, the Helio X2035 which runs at 2.3 GHz, also featuring
Mali-T880 MP4 GPU of 780 MHz that delivers 106 GFLOPS. Helio X2536 has a Mali-T880 MP437 GPU of
8500 MHz, delivering 115.6 GFLOPS. The latest X3038, also a ten-core chip, is made up of four Cortex-A72
cores at 2.5 GHz, two Cortex-A72 cores at 2.0 GHz, two Cortex-A53 cores at 1.5 GHz and two Cortex-A35
cores with 1.0 GHz.
Huawei HiSilicons Kirin: HiSilicons Kirin 95039 , the first smartphone SoC equipped with ARMs new
Cortex A72 processor is an octa-core 64-bit chip. Following ARM’s big.LITTLE architecture, it boasts four
Cortex-A72, clocking at 2.3 GHz for high performance and four 1.8 GHz Cortex-A53 for power efficiency. The
SoC also bundles a four-core Mali T880MP6 GPU, which is purposefully architectured to process large blocks
of data in parallel. This prodigious GPU runs at 900 MHz, offering 122.4 GFlops. In Kirin 95540 , the clock
rate of A72 processors is increased to 2.5 GHz, though the rest of the architecture in this SoC remains the same
as its predecessor [21]. The latest in the league, Kirin 98041, claimed as the first commercial SoC built on the
7nm42 (TSMC) and also the first SoC to adopt Cortex-A76, has an eight-core CPU (two A76 of 2.6 GHz, two
A76 of 1.92 GHz, and four A55 of 1.8 GHz). Kirin 980 offers a 75% better performance while being 58% more
energy efficient than its predecessor Kirin 970. Its Mali G76 GPU also performs 76% better than Kirin 970.
Moreover, Huawei claims that the performance of its innovative dual Neural Processing Unit (NPU) which is
capable of processing 4500 pictures per minute, is almost two and three times greater than Snapdragon 845 and
Apple A11 respectively.
4.2.4. Final Verdict. In the foreseeable future, more powerful processors with more cores are anticipated.
With 7nm like technologies, more muscle can be put up in a single core which will enable shoving more computing
capacity without compromising the chip size. The future smartphones loaded with these powerful SoCs will be,
29https://www.samsung.com/semiconductor/minisite/exynos/products/mobileprocessor/exynos- 8-octa- 8890/
30http://cpuboss.com/cpu/Intel- Core-i5- 4302Y
31https://www.notebookcheck.net/ARM- Mali-T880- MP12-Benchmarks.160645.0.html
32https://www.anandtech.com/show/9766/the- apple-ipad- pro-review/2
33https://www.apple.com/newsroom/2017/09/iphone- 8-and- iphone-8-plus-a-new-generation-of- iphone/
34https://www.apple.com/iphone- xs/a12-bionic/
35https://www.mediatek.com/products/smartphones/mt6797- helio-x20
36https://www.mediatek.com/products/smartphones/mt6797t- helio-x25
37https://www.notebookcheck.net/ARM- Mali-T880- MP4.159590.0.html
38https://www.mediatek.com/products/smartphones/mediatek- helio-x30
39http://www.hisilicon.com/en/Products/ProductList/Kirin
40https://www.notebookcheck.net/HiSilicon- Kirin-955- Octa-Core-SoC-Benchmarks-and- Specs.163438.0.html
41https://consumer.huawei.com/en/campaign/kirin980/
42https://www.tsmc.com/english/dedicatedFoundry/technology/7nm.htm
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 273
in the true sense, the little computing giants.
5. Challenges for Smartphone Crowd Computing. This section describes several challenges that are
faced, in general, in using a smartphone as a computing device. Below, some of the rudimentary issues are
mentioned.
Motivating the crowd: The challenge to motivate the crowd is always the primary concern of the vendors.
As more people are motivated and participate in SCC, the greater will be the benefit. What one can do, one
must do. Contributing to a social cause gives a person to achieve a feeling of self-actualization. One can
perceive a self-satisfaction by doing what he/she can do within his/her capability. In this process, others may
be benefitted. In other words, the implication of this realization of potential philanthropic results in the greater
good.
Adopting a balanced and effective incentive policy: To attract people to actively participate in SCC,
suitable and attractive incentive policies need to be formulated. The incentive policies should be according to
the particular SCC application rather than generalised.
Security and privacy concerns: Mobile devices hold numerous personal information like financial data,
email, location updates, social media information, and other user id and passwords. For using the mobile as
a crowd computing device, it is utmost important that mobile data are secured from tracking, phishing or
destroyed. The user must be given a sense of data security and privacy. It is this concern for which people are
going to volunteer their devices for SCC. The SCC applications for the mobile devices should have a secured
architecture which ensures confidence that the application is in no way sharing or have access to memory location
and services of other apps.
Variable wireless network: With the majority of wireless services having a variety of different rate of
transmission, both the average rate and rate of oscillation are vital metrics to measure the performance. It is
quite challenging to determine the apt rate of oscillation a service can tolerate obtaining a high average rate.
Here the service satisfaction for a smartphone user may be quantified with an increase in the concave utility
function of the instantaneous transmission rate.
Designing middleware: As per consumer/market demand, smartphone companies are launching new
products very regularly. These devices are widely heterogeneous in terms of hardware platform (CPU, GPU,
SoC, no. of cores, signalling module, etc.), system and application software, manufacturer specifications, user
interface, etc. This diverseness poses a great challenge in designing a generalised middleware and client ap-
plication. The enhancement of interoperability among both the SCC server and the participant smartphones
by incorporating the appropriate middleware that can handle smartphones irrespective of their hardware and
software specification and support efficiently, to be developed and managed.
Maintaining QoS: Alike other distributed and networked computing systems, maintaining the quality of
service (QoS) is challenging in SCC too. It is more challenging in SCC considering the factors such as device
mobility, variable communication quality, diverseness in resources, etc. The criteria of satisfactory QoS have a
vital contribution to the overall success of the SCC applications.
Heat: For smartphones, extreme hot or cold conditions are potentially damaging. The cell phone battery
can be damaged due to excessive heat dissipated. If the CPU is busy for a long time, the phone will be heated.
This may deter users from participating in SCC.
Limited memory: Smartphones have a limited internal memory which cannot be increased or reduced
by the user. They have no additional expansion slots. The external memory option is not present in all the
available smartphones. Therefore, it is vital to use the limited phone memory space intelligently so that it would
not affect the other apps running in memory while importing jobs through SCC.
Battery power: One of the major obstacles in smartphone computing is the limited power of smartphone
batteries. The current batteries are struggling to keep up with users active and ever-increasing smartphone
demands. A whole lot of multiple features, faster processing, high resolution, etc. consume a huge amount of
battery power. In addition to that, SCC will keep busy the smartphone processor which will drain the battery
quickly. This heavily influences the participants willingness to lend their device. To expand the battery lifespan,
the manufacturers are bringing batteries with huge sizes, causing inconvenience to the users. Fortunately, active
research is going on towards solving the battery menace. In the following section, we shall explore some of the
major initiatives.
274 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
6. Future Battery Technologies will Encourage Smartphone Crowd Computing: An Opti-
mistic Study. The most concerning facet regarding smartphone computing is the limitation and fast draining
of battery power. While portable computers, smartphones, wearable computers, and other mobile devices are
growing ever more advanced, both technologically and architecturally, they are still limited by power. The
battery or energy technology has not moved at par in decades, which pulls back the pervasive & ubiquitous
computing revolution. And it is only going to get worse as next-generation 4G networks come online, giving the
phones access to high-speed, always-on connections and torrents of data. The battery is unarguably the biggest
obstacle to smartphone computing. In a recent survey, nearly 70% of respondents stated that battery life is
the single biggest limitation of their mobile phone and most of them are willing to pay more for the phones
which offer extended power [27]. Present smartphone batteries are struggling to keep up with users active and
ever-increasing smartphone usage demands. In spite of several limitations like slow charging, volatility, and
degradation the current Lithium-ion batteries have served us well so far, since its first commercial release by
Sony and Asahi Kasei 25 years ago. But this technology is approaching its limits after which no enhancement
can be made out. According to researchers, potentially it can reach a maximum of 30% above current levels
[23]. Though other classes of battery technologies, especially the lithium-sulphur (Li-S) and the lithium-air
(Li-air) have potential to surpass the lithium-ion, they are not in the mainstream production and uses due to
some fundamental challenges in these technologies.
6.1. Major Battery Drainers. The main reasons for the majority of battery drainage are:
Screen: The major culprit in consuming power in a smartphone is its vibrant screen. The phones are
coming with higher resolutions resulting in more power consumption. Lower-power display technology like
OLED (Organic LED) is expected to take over LCD displays in the coming years, resulting in better battery
life. And as people are more inclined to larger screens (most of the Indians prefer phablets); it indirectly paves
to better battery performance. A larger screen would consume more energy, certainly, but it will allow fitting a
larger battery also. If a phone of the same architecture has a 20% larger screen size, will run 40% longer period
because of the bigger battery.
Power savvy processor: Akin to PCs, the processing capacity of smartphone processors also has increased
consistently by Moors law. But it is not achieved, as it may seem, at the expense of additional battery power.
Actually, improvement in processor technologies has rendered more processing power per watt. For example,
as compared to mid-1980s processors which would consume 100-watt power, todays high-end processors can
carry out more than 100,000 times MIPS, consuming the same amount of watt [56]. So, modern processors
have become not only more powerful but more power-efficient as well. For example, thanks to a proprietary
technology called Asynchronous Multiprocessing (ASMP), Snapdragon processors can dynamically adjust the
voltage and speed of each CPU core independently that leads to decrease in wasted power and heat [6]. Another
major factor which influences power efficiency, especially in mobile devices, is the SoC. SoCs offer smartphones a
better power management capability by making able to put into sleep that particular power-consuming hardware
which is not required at that time; thus, saving a considerable amount of battery power [56].
Signalling module: Similar to processor technology radio/wireless technologies are also improving signif-
icantly. Todays 4G network allows users to talk much longer than earlier older mobile networks by disbursing
equal watt. Another basis of reduced battery drain is lesser transmitting distance. As the number of mobile
phone users has grown enormously, the number of cellular network base stations has been greater than before.
That means transmitting signals must travel a shorter distance from phone to base stations; which results in
less power consumption [56]. Alongside mobile devices and mobile networks, the number of Wi-Fi routers also
has escalated, and it should be noted that Wi-Fi connections consume significantly less energy compared to 3G
mobile networks [78] and of course, even more, less than 4G. By eating up the same amount of power, Wi-Fi
can transfer more than twice the volume of data as 3G [44]. This fact should go in favour of our view of that
the smartphones who participate in volunteer computing should use the local wireless network for necessary
message passing (no one would voluntarily offer their data pack balance for the purpose which least bothers
him, at least immediately).
6.2. Other Causes. The various reasons responsible for the battery drain are [11, 17]:
The GPS: The global positioning system (GPS) in a mobile device helps to identify the geographic location
of the mobile user with the help of orbiting GPS satellites. The mobile device picks the GPS satellite signal to
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 275
determine the mobile position on Earth. Since the GPS satellite signal is a week, the processor needs to perform
a hard task to eliminate noise and decoding the signal; this makes the mobile device to consume energy faster.
The energy consumption increases as mobile user travel and could deplete the battery in one or two hours.
Wi-Fi variations: A major source of battery power consumption comes from the Wi-Fi interface. The
power consumption of the mobile device depends upon how signal strength it receives, weaker the signal more
energy it consumes. The receiving Wi-Fi signal strength varies with the distance from Wi-Fi base station and
also gets affected by obstructions like a wall, door, etc., in between the path of the signal.
The constant search for Wi-Fi networks: Constantly searching for a Wi-Fi station and connecting
to the chosen router through the jungle of signals needs more power; this drains the battery significantly. The
energy consumption increase as the person moves and the mobile device constantly searches for Wi-Fi base
station for networking. Switching off the Wi-Fi interface while not in use would significantly increase the
battery life.
LTEs high data rates: Huge battery is consumed when the phone is used for calling or e-mailing using
the GSM, 3G, or LTE (4G). The smartphone hooks on to a single base station at a time, picking the one
that offers the strongest signal and communicates even when the phone is also not being used. Mobile devices
change the communicating base station for an incoming or outgoing call and for data communication while on
the move. The handing over of communication services from one base station to another takes more energy. The
network handover may also happen even in indoor locations. While moving from one room to another, there
may be a loss of signal strength or network communication due to signal obstructions by walls or other things;
this enforces the devices to search and connect to other base station available for communication causing more
energy consumption. In another case, while the mobile device is not communicating directly, but apparently it
is busy in the background for data communication. Many apps loaded in device get updated in the background
by pulling and pushing data from the web. As this data updation happens in the background it does access
data communication lines without the knowledge of the user. This consumes a lot of network data and thus
enforces mobile devices to continually search and connect to various network connections and hence consumes
a lot of energy. One way to restrict this power drain is to disable LTE, 3G, and Wi-Fi communication when
not used and turning off automatic notifications of apps which are not used much.
The screen brightness is set too high: High screen brightness is a major source of battery consumption.
The brightness can be managed easily by changing the screen brightness level as per the surroundings. For indoor
where surrounding light is very less a low screen brightness is sufficient to view. Whereas at outdoors, where
the surrounding luminosity is high, the screen brightness can be changed as per suitability. The other way of
reducing the amount of battery power consumption is shortening the display time and turning off the display
when not used.
Apps are staying busy in the background: One of the major reasons for profuse battery drain in a
smartphone is the apps that run in the background. Mobile applications like memory booster, file cleaning,
Facebook, email, etc., run in the background continually. This keeps draining the battery. The apps such as
email, news and the apps for social media keep refreshing themselves in the background, discharging the battery
automatically. Even the battery saver applications do not help; most of them actually consume more power
than they save. While an application is installed, the background refresh change setting should be changed
from automatic to manual. This will skip the automatic data update or feed every time the application starts
and will improve the mobile battery life.
Using resource-intensive apps: One of the major sources of battery drain is continuously using resource-
intensive applications for a long time. Gaming applications which are resource intensive in terms of processor
cycle, display, sound, memory, and network keep continuously consuming energy. Frequent use of heavy apps
such as weather apps and the Google Map which constantly locate users locations causes power leakage. The
same happens in the case of overusing the apps which require GPS. Also, changing the application setting for
refresh and updates frequently, does not help in battery conservation. To save battery life, the resource intensive
application must be constraints for longer use.
Software updating: Updating the OS or application may add updates and fixes or bugs. Updating
software causes connecting to the internet and continuously downloading data. This may contribute to battery
usage.
276 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
The phone has a hardware issue: A phone suffering from hardware glitches is prone to quick battery
drain.
Battery-draining activities: Phones that are over-engaged with battery-drainer activities are bound to
drop battery power quickly. For example, the constant use of hotspot, Wi-Fi, and Bluetooth or continuous
streaming of music and videos are the major reasons for faster batter loss. Also, making phone calls while
travelling needs frequent hands-off, which requires energy. Other battery-losing activities are watching videos
and playing graphics-heavy games for a long duration, using the phone camera frequently, etc.
Automatic syncing of the apps and data: The auto backup of the data (e.g., photos, videos, or
documents) and synchronization of apps account for huge battery drain.
Notifications are going off all day: Constant lighting, vibration, or ringing notification sounds may be
a reason why the battery dies quickly. The constant stream of notification sound is annoying; hence users often
prefer to substitute it by vibration resulting in rapid battery drain.
Charging using a slower charger or a faulty cable: Charging a mobile device using a faulty or cheap
cable or slower charger or a replica charger for a long time could harm the battery. This reduces the charge
retention capability of the battery and could lead to excessive power draining. Therefore, to increase the battery
life, it is required to charge the phone with the original charger or the charger which is shipped with the mobile
device.
Killing recent apps or using task managers to free memory or kill apps: The recent Android
version dynamically allocates memory to all apps running in the background. In these versions, the memory of
apps is managed dynamically. For if the system requires memory, it automatically takes the memory allocated
to apps running in the background without taking too much of the resources. Closing an app which is running in
the background may close some of the system processes with the app. This causes the application to be restarted
again from scratch next time when the application is started. This would cause unnecessary wastage of CPU
resources and thus consumes battery. Hence, it is better to keep the application running in the background
than to restart it again.
Using a ton of widgets, third-party customizations and health trackers: Using a huge number of
widgets on an Android home screen may experience a drain of battery in smartphones. For example, widgets
like step counters, calorie burn calculators, etc. are quite notorious as a battery burner.
Live wallpapers or bright wallpapers on AMOLED displays: Live wallpapers or bright wallpapers
on AMOLED display look amazing, but they consume a great deal of battery power. Using sharp contrast and
vivid coloured wallpapers on the AMOLED screen definitely juice more battery power. LCD, in contrast to
AMOLED, consumes less battery power, no matter what wallpaper is displayed.
Unnecessary push notifications: The push notifications from different corners of the digital world are
being hurled every day, which causes the phone to light up multiple times and make the transceiver antenna
busy which consumes battery.
Replace the original old battery with a nonstandard one: Smartphones older by 2 or 3 years many
often have issues with the battery. The battery either dies out or retains a charge for very less time. Replacing
the old battery with a replica or nonstandard one is often cost-effective but may have serious problems. The
nonstandard battery goes faulty very quickly and does not retain charge much longer; this needs frequent
recharging, causing a lot of power consumption. Replacing the original battery with nonstandard ones is not
an effective solution for the long run and may also cause harm to the mobile device also.
6.3. Advancement in Battery Technologies. It is untoward that the lack of advancement in the field
of battery technologies has prevented smartphones being used as per their potential. But the good news is that
we are on the verge of a power revolution. Research groups in universities and organizations are coming up
with innovative ideas either to extend battery life or to minimize the recharging time and may be to accomplish
both. Some are experimenting with alternate power sources like solar, body heat or air. In this section, we shall
explore some of the promising new developments (present and future) which may encourage users to use their
mobile devices more comprehensively without worrying about power limitation.
6.3.1. Extended Durability. Majority of the researches in battery technology are focusing on to max-
imise the duration between two charges, i.e., to minimise the charging frequency. The notable endeavours
towards this direction are mentioned below:
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 277
Intelligent Energy, a British energy technology company, aspires to revolutionize smartphones by re-
placing traditional lithium-ion batteries with hydrogen fuel cells. They are developing cells based on
embedded fuel cell technology which will keep smartphones powered for more than a week after just a
single charge [26].
Rechargeable lithium metal batteries offer nearly 10 times of energy capacity than lithium-ion batteries.
A Li-metal anode batterys capacity could be as high as 3,860 mAh/g, whereas a typical Li-ion battery
with graphite on the anode can store around 380 mAh/g [41]. But these batteries could not be used
in mobile devices so far due to safety hazards [29]. Scientists at Cornell University have developed a
new safe way to use rechargeable lithium metal batteries at room temperature, an achievement which
may lead to significantly extended power backup in mobile devices [69].
The solid-state batteries found by MIT scientists along with Samsung offer more power and last longer.
In these batteries, power density is improved by 20 to 30% which means more power to the device. The
founders claim that the life of these batteries is thousands of cycles before degrading [24].
Fuji Pigment has come out with their exciting aluminium-air Alfa batteries which have 40 times more
capacity than lithium-ion and can be recharged by simply being topped up with water (with or without
salt) [24]. According to the company, these batteries should last up to 14 days.
Samsung also has claimed that they may have figured out a way to nearly double the battery life of
smartphones by adding graphene on top of silicon anodes. When paired with a commercial lithium
cobalt oxide cathode, the silicon carbide-free graphene coating allows the full cell to reach volumetric
energy densities of 972 and 700 Wh at first and 200th cycle, respectively, 1.8 and 1.5 times higher than
those of current commercial lithium-ion batteries [73].
To tackle with the battery consumption by the smartphone screen, the main culprit as battery killer, a
British start-up has come up with smart glass, a unique and innovative technology where the display
will no longer rely on the battery to illuminate the screen, as it will use electrical pulses instead, allowing
a device’s battery life to last for a week [48]. Dr. Peiman Hosseini, founder of Bodle Technologies,
enlightened that the smart glass technology, based on the technology that is used for rewritable DVDs,
uses electrical pulses to create vivid, hi-tech displays that require no power and can be viewed clearly,
even in direct sunlight [14].
6.3.2. Larger Capacity and Longer Battery Life. Larger battery capacity is one way to ensure the
mobile devices run a long time without frequent recharging. Manufacturers are now going for batteries with a
larger capacity which can hold huge power and thus less recharging. Companies are coming up with batteries
with huge capacities. A couple of such examples is given below:
Oukitel has introduced battery with a capacity of 10000mAh and which is increasingly on the standard
battery capacity of 2000-3000mAh. The Oukitel battery of 10000mAH once charged can go 15 days
without recharging.
A group of physicists from the University of Missouri has developed a material with a unique structure
(a honeycomb lattice) that are claimed to enable increasing the life of batteries by more than a hundred
times [74].
Batteries with larger capacity are bigger in size and thus making the mobile device bigger in size. Companies
are struggling to keep a balance between battery size and battery power. New battery technology is needed
which seriously increases the battery power without compromising the battery size and thus the mobile design
[7].
6.3.3. Faster Charging. Battery capacities are increasing, but the time it takes for charging makes the
manufacturer and the users frustrating alike. Everybody hates waiting incessantly to charge ones mobile phone.
Who has the time for this pesky task! Contentedly for us, we’re starting to get smartphone batteries that not
only last longer but also recharge awfully fast. Fortunately, the researchers are working and exploring a new
way to charge batteries in faster ways. In the near future, we are going to have a breakthrough in charging
technology.
Some of the efforts being explored by the researcher around the globe are:
278 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
Quick Charge43, a new charging technology, introduced by Qualcomm for its next generation of Snap-
dragon processors, has facilitated users to pull off a considerable amount of battery life with a short
initial burst of charging. The introductory version (Quick Charge 1.0) could charge up to 40% faster
than conventional charging while Quick Charge 2.0 augmented the charge time gain up to 75% [33]. The
company claims the latest version, i.e., Quick Charge 3.0 is four times faster than conventional charging
and 38% more efficient than Quick Charge 2.0. By their laboratory tests, it enables smartphones to
attain from 0 to 80% charge in 35 minutes [22].
Super VOOC Flash Charge44, used in latest Oppo smartphones, thanks to the all-new 5V low-voltage
pulse-charge algorithm, can charge an average smartphone battery (2500mAh) up to 100% in just 15
minutes [47].
Lyte Systems has come up with a portable battery called Lumopack45 that can be loaded in just 6
minutes with enough power to charge an iPhone 6 fully, thanks to its charge rate at 140W.
A team at Stanford University has developed an ultrafast rechargeable high-performance, a long-lasting
aluminium battery that can be fully charged in one minute and able to withstand more than 7,500 cycles
without any loss of capacity in comparison of 1000 cycles of a typical lithium-ion battery [71].
A team of scientists at MIT has created nanoparticles based ”yolk and shell” battery with a titanium
dioxide ”shell” and aluminium ”yolk” (anode) [16]. They claim that this battery can be charged from
zero to full in just six minutes. Moreover, according to them, these batteries can hold three times
the capacity of the current lithium-ion batteries and also degrade slower, giving the longer life of the
battery [58].
Scientists at Rice University have made a breakthrough in micro-supercapacitors using which in batteries
will allow them to charge 50 times faster than current batteries and discharge even slower than current
supercapacitors [60].
StoreDot46, a start-up, born from the nanotechnology department at Tel Aviv University developed
a biological semiconductor based superfast charger called StoreDot charger that can charge a mobile
phone battery from empty to full in only thirty seconds.
Harvard student Eesha Khare developed an award-winning tiny battery. The battery can store plenty
of charge despite its small dimensions. The battery is a supercapacitor energy storage device made
up of carbon fibre and metal oxide. The nanotechnology used in fabrication has made charging faster
than previously possible. It not only charges faster but holds the charge for much longer time. Eesha
is working further to reduce the supercapacitor battery charging time less than a minute [7].
Kansas based engineer Shawn West claimed to develop a prototype rechargeable AA cell which recharges
fully in 26 seconds, unlike other rechargeable battery which takes 34 hours for full charging. The battery
is a lithium-ion capacitor with a voltage regulator and monitor circuitry attached to it. The circuitry
ensures the correct voltage is being applied for charging and the right voltage is given away as output
supply [7].
6.3.4. Wireless Chargers for Hassle-free Charging. In the near future, wireless chargers will replace
traditional inconvenient wired chargers which require mobile devices are to be connected to a power socket or
USB port. Wireless chargers can make us loose of this binding. The idea can be compared to how mobile
devices access the Internet through Wi-Fi. The user can walk around the room with his phone or can work on
it while it charges over the air, provided it should be in the range of the transmitter. Below some examples are
cited of this kind of initiative:
A start-up company uBeam47 has discovered a wireless charging technology that allows a phone to be
charged in the air from a transmitter, may be attached to the wall, using ultrasound for transmitting
electricity. The uBeam technology48 is based on the pioneering attempt of turning power into the sound
43https://www.qualcomm.com/products/snapdragon/quick- charge
44http://www.oppo.com/en/technology/vooc
45lytesystems.com
46https://www.store- dot.com/
47https://ubeam.com/
48https://ubeam.com/technology/
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 279
waves to be transmitted and then converted back to power on reaching the device.
Ukrainian company XE has come up with a wireless charging prototype that can charge mobiles up to 16
feet away from the charging point [13]. One XE charger can charge four mobile devices simultaneously.
Researchers from KAIST in South Korea have developed a new Dipole Coil Resonant System (DCRS)
that can charge up to 40 smartphones from 5 meters away [8].
The Cota project from Ossia Inc. enables multiple mobile devices to be charged wirelessly over a
distance of 30 feet [7]. The charging can be done across walls, doors, and clothes.
Many of todays high-end phones (e.g., Apple iPhone 8 and X series, Samsung Galaxy S7, S8, and S9
series, LG G7 Thinq and V30, Sony Xperia XZ series, Nokia 8 Sirocco and Lumia 735 and above, etc.,
to name a few) incorporate wireless charging facility. Most of them follow the Qi wireless charging
standard which is capable of providing 5-15 watts of power to small personal electronic devices [20].
The technology is still in its early stage. Many issues are to be addressed. For instance, most of the energy
is lost on the way from the charging base to the device. Also, the wirelessly transmitted power is very limited,
unable to charge multiple devices continuously. Nevertheless, wireless chargers are getting popularity among
users. Modern furniture, especially the work desks and coffee tables are incorporating wireless charging pads
inbuilt. This will eventually allow users to get rid of mobile chargers.
6.3.5. Alternative Power Sources for Ease of Charging. Researchers have tried to explore different
non-conventional sources of energy to charge batteries of mobile devices as described below:
Solar charged batteries: Batteries, rechargeable by solar rays are also being experimented by the smart-
phone makers like Alcatel [53] and Tag Heuer [12] where a transparent solar panel over the screen would let
users charge their phone by simply placing it in the sun or even some artificial light. X-Play5 from Vivo, a
Chinese mobile company, features the solar charging facility.
Batteries charged by human body heat: The idea of charging batteries from human body heat has
been fascinated by the researchers for a long time. Recently a research team at Korea Advanced Institute of
Science and Technology (KAIST) has developed a thin and light thermoelectric generator which collects body
heat and converts it into energy, supplying the gadget (mainly wearables) it is attached to a never-ending stream
of power [59].
Sound energy for charging phone: Nokia in collaboration with British engineers is creating a prototype
phone which gets automatically recharged by background noise [7]. The phone will literally never go out of
power. Further Nokia also has launched a first of its kind a wireless charging trouser. The trouser has an
inductive charging plate which charges the Nokia phone on the move.
6.3.6. Other Innovative Way to Maximise the Battery Life. Other innovative approaches are being
tried by researchers in preserving the battery as long as possible by developing technology for mobile devices
which consumes less battery power. One such example is mentioned below [7]:
The innovative e-ink phone: Smartphone manufacturer Yota Devices has developed dual screen display
for YotaPhone 249 mobiles to extend the mobiles battery life. An extra screen is added to the back side of
the mobile device which works using e-ink technology causing very minimal battery usage. The e-ink screen
like Amazons Kindle notebook is used for reading a text message, SMS, notifications, mail, news, etc.; thus,
avoiding the LED screen usage and saving a huge amount of energy. For viewing video, web page, and managing
a desktop application, the colour LED screen is used. However, it is found that separating the view into two
displays would not much of the difference in overall battery life. A saving of two and a half days is gained in
between charging. It might not be considered enough, but a good initiative to start.
6.4. Final Verdict. It is very unlikely that all the above-mentioned endeavours, towards better power
backup for mobile devices, should go to the production. Many of them are not commercially viable. Some are
not practicable. But considering all the research works going on, with different approaches, we are very much
optimistic that the days are not far when the users will no further be haunted by the horror of low battery.
7. Conclusion. Smartphone crowd computing (SCC) has been seen as an ideal alternative to the existing
high-performance computing (HPC) systems in attaining green and sustainable computing. There will be
49http://yotaphone.com/gb- en/
280 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
no additional productions because people will be using smartphones anyhow. Due to the smaller size, the
environmental hazards and amount of e-waste will be substantially lesser. For the successful employment of
SCC, several challenges need to be addressed. The most concerning facet regarding using smartphones is the
limitation and fast draining of battery power. Battery technologies have not developed as per with the processor,
memory, and storage technologies. But hopefully, some recent researches promise to solve the battery woos in
smartphone uses. Various technologies such as long-lasting batteries, quick-charging batteries, wireless-charged
batteries, batteries charged through unconventional means such as light or body temperature, etc. are on the
verge of commercial launch. Some of them are already successfully introduced. Provided the challenges are
resolved and with sufficient public awareness, SCC is expected to be as successful as Grid and Cloud computing
as a cost and energy-efficient sustainable HPC paradigm.
REFERENCES
[1] Factory is where our computers eat up most energy. https://phys.org/news/2011-04-factory-energy.html, April 2011.
Accessed 22 February 2019.
[2] The global footprint of mobiles. https://therestartproject.org/the-global- footprint-of-mobiles/, 2018. Accessed 24
February 2019.
[3] Gpu gflops. https://gflops.surge.sh/, March 2018. Accessed 27 February 2019.
[4] F. Alan. Arm introduces cortex a-73 chip and mali-g71 graphics chip; higher performance, en-
ergy efficient. http : / / www . phonearena . com / news / ARM-introduces- Cortex-A- 73-chip- and-Mali-
G71-graphics- chip-higher- performance-energy- efficient id81620, May 2016. Accessed 8 June 2016.
[5] A. S. Andrae. Total consumer power consumption forecast. In Nordic Digital Business Summit, Helsinki, Finland, 2017.
[6] ARM. Arm and qualcomm: Enabling the next mobile computing revolution with highly integrated armv8-a based socs.
Technical report, ARM/Qualcomm, 2014.
[7] C. Barraclough. How mobile phone battery life is going to seriously improve in 2016. https ://recombu.com/mobile/
article/smartphone-mobile- phone-battery- life-better- longer#, December 2015. Accessed 30 September 2016.
[8] C. B. Bautista. Wireless charging is coming: Researchers juice up 40 smartphones from 5 meters away. http://www.
digitaltrends . com / mobile / farewell-cords- new-tech- can-wirelessly-charge-40-smartphones- 5-meters- away/,
April 2014. Accessed 25 June 2016.
[9] T. Bawden. Global warming: Data centres to consume three times as much energy in next decade,
experts warn. https : / / www . independent . co . uk / environment / global- warming-data- centres-to- consume-
three-times- as-much- energy-in- next-decade-experts-warn-a6830086 .html, January 2016. Accessed 23 February
2019.
[10] Lovefone Blog. How much co2 does it take to make a smartphone? https: //www .lovefone .co. uk/ blogs/ news /
how-much- co2-does- it-take- to-make-a-smartphone, March 2018. Accessed 22 February 2019.
[11] J. Bolluyt. 9 reasons your phone battery dies so quickly. http : / / www . cheatsheet . com / gear-style /
reasons-phone- battery-dies- quickly.html/?a=viewall, February 2017. Accessed 17 July 2017.
[12] A. Boxall. This luxury phone’s screen has an amazing built-in solar charging panel. http://www.digitaltrends.com/
mobile/tag-heuers- merediist-infinite- phone-has- a-solar-charging-screen/#!Elqeb, April 2014. Accessed 12 July
2016.
[13] J.-P. Buntinx. Long-distance wireless charging could boost mobile and bitcoin payments. https://news.bitcoin.com/
long-distance- wireless-charging- boost-mobile- bitcoin-payments/, December 2015. Accessed 30 July 2016.
[14] R. Burn-Callander. Is this the end of constant smartphone recharging? http: // www.telegraph .co . uk/ finance/
businessclub/ technology /12008390 / Is-this- the-end- of-constant- smartphone-recharging. html, November 2015.
Accessed 12 July 2016.
[15] B. Caulfield. The story behind pro ject 192: How a salinas, calif., crop circle became a worldwide puzzle. https://blogs.
nvidia.com/blog/2014/01/05/salinas-crop- circle-and- project-192/, January 2014. Accessed 24 June 2016.
[16] D. L. Chandler. yolks and shells improve rechargeable batteries. http : / / news . mit . edu / 2015 /
yolks-and- shells-improve- rechargeable-batteries- 0805, August 2015. Accessed 21 February 2019.
[17] Deccan Chronicle. 8 reasons why your smartphone battery could die faster. http://www.deccanchronicle.com/technology/
mobiles-and- tabs/260516/8-reasons- why-your- smartphone-battery-could-die-faster.html, May 2016. Accessed 17
July 2017.
[18] Cloudflare. How are serverless computing and platform-as-a-service different? — paas vs. serverless. https://www.
cloudflare.com/learning/serverless/glossary/serverless-vs- paas/, 2019. Accessed 28 March 2019.
[19] C. Conger. Can my computer poison me? https://computer.howstuffworks.com/computer-poison1.htm., August 2009.
Accessed 22 February 2019.
[20] Wireless Power Consortium. Qi - mobile computing. https://www.wirelesspowerconsortium.com/qi/, 2018. Accessed
17 February 2019.
[21] I. Cutress and A. Frumusanu. Huawei launches the p9 and p9 plus. http: / / www.anandtech . com/show / 10231 /
huawei-launches- the-p9- and-p9- plus, April 2016. Accessed 7 June 2016.
[22] C. d. Looper. Qualcomm’s quick charge 3 takes smartphones from 0 to 80 percent battery in 35 minutes. http://www.
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 281
techtimes.com / articles / 84999 / 20150915 /qualcomms-quick- charge-3- takes-smartphones- 0-80- percent-battery.
htm, September 2015. Accessed 11 March 2016.
[23] P. Daniel. 5 future battery technologies to make your phone last longer and charge faster. http://www.phonearena.com/
news/ 5-future- battery-technologies-to-make-your-phone-last- longer-and- charge-faster id67357, March 2015.
Accessed 30 September 2016.
[24] L. Edwards. Future batteries, coming soon: charge in seconds, last months and power over the air. http://www.pocket-lint.
com / news / 130380-future- batteries-coming- soon-charge-in- seconds-last- months-and- power-over-the- air, De-
cember 2015. Accessed 4 March 2016.
[25] Energuide. How much power does a computer use? and how much co2 does that represent? https://www.energuide.be/en/
questions-answers/how-much- power-does- a-computer- use-and-how-much-co2- does-that- represent/54/, 2019. Ac-
cessed 22 February 2019.
[26] Intelligent Energy. Intelligent energy launches a customer funded 5.25m project to deliver embedded week-
long mobile phone power. http: / / www . intelligent-energy. com / news- and-events / company-news/2016 / 02 / 08 /
intelligent-energy- launches-a- customer-funded- 525m-project-to-deliver-embedded- week-long- mobile-phone-
power/, February 2016. Accessed 25 June 2016.
[27] Intelligent Energy. Survey: 70devices. http://www.intelligent-energy.com/news- and-events/company- news/2016/03/
10/survey-70-of- consumers-drained- by-battery- limitations-of- mobile-devices/, March 2016. Accessed 25 June
2016.
[28] ERIdirect. How long does it take electronic waste to decompose? ht tps : / / eridirect . com / blog / 2015 / 11 /
how-long- does-it- take-electronic- waste-to-decompose/, November 2015. Accessed 22 February 2019.
[29] T. Fleischman. Room-temperature lithium metal battery closer to reality. http://news.cornell.edu/stories/2016/02/
room-temperature- lithium-metal- battery-closer- reality, February 2016. Accessed 25 June 2016.
[30] A. Frumusanu. Early exynos 8890 impressions and full specifications. http:// www . anandtech . com / show / 10075/
early-exynos- 8890-impressions, February 2016. Accessed 14 September 2016.
[31] Fujitsu. Life cycle assessment and product carbon footprint. Technical report, 2010.
[32] E. Gelenbe and Y. Caseau. The impact of information technology on energy consumption and carbon emissions. Ubiquity,
2015(June):1–15, 2015.
[33] G. Gordon. Introducing quick charge 3.0: next-generation fast charging technology. https :// www .qualcomm. com /
news/snapdragon/2015/09/14/ introducing-quick- charge-30- next-generation- fast-charging- technology, Septem-
ber 2015. Accessed 10 March 2016.
[34] GSMA. Smartphones to account for two thirds of world’s mobile market by 2020 says new gsma intelligence study. https://
www.gsma.com/newsroom/press-release/smartphones-account- two-thirds-worlds-mobile-market-2020/, September
2014. Accessed 17 August 2018.
[35] M. Halper. Supercomputing’s super energy needs, and what to do about them. https : / / cacm .acm.org / news /
192296-supercomputings- super-energy- needs-and- what-to-do-about-them/fulltext, September 2015. Accessed 22
February 2019.
[36] E. Herrmann. The fastest mobile processors compared. https://www.androidpit.de/die- schnellsten-handy- prozessoren,
February 2016. Accessed 11 March 2016.
[37] J. Hruska. Arm announces new cortex-a73 cpu, mali g71 gpu. http : / / www . extremetech . com / computing /
229328-arm- announces-new- cortex-a73- cpu-mali-g71-gpu, May 2016. Accessed 13 September 2016.
[38] IANS. Number of smartphone users crosses 300 million in india as shipments grew 18 percent.
https : / / www . firstpost . com / tech / news-analysis / number- of-smartphone- users-crosses-300-million-
in-india- as-shipments- grew-18- percent-3696457.html, January 2017. Accessed 17 August 2018.
[39] IEA. Co2 emissions from fuel combustion highlights. Technical report, 2013.
[40] Research Triangle Institute. Hazard assessment of the electronic component manufacturing industry. Technical report,
U.S. Department of Health and Human Services, Ohio, 1985.
[41] D. Johnson. Nanomembrane may bring rechargeable lithium-metal batteries back. http://spectrum.ieee.org/nanoclast/
semiconductors/materials/ nanomembrane- may-bring- rechargeable-lithiummetal-batteries-back, February 2016.
Accessed 25 June 2016.
[42] N. Jones. Battle of the 1ghz snapdragon super-phones. http : / / www . knowyourmobile . com / products / 8324 /
battle-1ghz- snapdragon-super- phones, March 2010. Accessed 23 June 2016.
[43] N. Jones. How to stop data centres from gobbling up the worlds electricity. https: // www .nature . com/articles /
d41586-018- 06610-y, September 2018. Accessed 24 February 2019.
[44] G. Kalic, I. Bojic, and M. Kusek. Energy consumption in android phones when using wireless communication technologies.
In 35th International Convention MIPRO, Opatija, Croatia, 2012.
[45] J. Koomey. Growth in data center electricity use 2005 to 2010. https://www.missioncriticalmagazine.com/ext/resources/
MC/Home/Files/PDFs/Koomey Data Center.pdf, August 2011. [Accessed 1 September 2018].
[46] J. Kozlowicz. 8 ways data center environmental impact goes beyond emissions. https://www.greenhousedata.com/blog/
data-center- environmental-impact- goes-beyond- emissions, November 2015. Accessed 18 August 2018.
[47] A. Kumar. Oppo wows with fast-charging super vooc, image stabilization tech. http://www.pcmag.com/article2/0,2817,
2499860,00.asp, February 2016. Accessed 26 February 2016.
[48] V. Lanaria. Seven-day battery life on your smartphone? it could be coming soon. www.techtimes.com/articles/110852/
20151126/seven-day- battery-life- on-your- smartphone-it-could-be-coming- soon.htm, November 2015. Accessed 12
July 2016.
[49] S. Liu. Analysis of electronic waste recycling in the united states and potential application in china. Technical report,
282 Pijush Kanti Dutta Pramanik, Saurabh Pal, Prasenjit Choudhury
Columbia University, New York City, 2014.
[50] D. Luis. Tech war: Nvidia tegra x1 takes on snapdragon 810 with raw gpu power. http: //www .phonearena. com/
news/Tech-war- Nvidia-Tegra- X1-takes- on-Snapdragon-810-with-raw- GPU-power id64748, January 2015. Accessed 04
March 2016.
[51] MCMC. Mobile e-waste: Old phone, new life. https://mobileewaste.mcmc.gov.my/en-my/about-mobile- e-waste, 2015.
Accessed 25 February 2019.
[52] R. Mihindukulasuriya. Your mobile phone is a major contributor to toxic e-waste in the country. https://theprint.in/
science/your-mobile-phone-is- a-major- contributor-to- toxic-e- waste-in-the-country/141430/, October 2018.
Accessed 25 February 2019.
[53] S. Miles and I. Morris. Transparent solar panel display charges your phone through the screen. http://www.pocket-lint.
com/news /126245- transparent-solar- panel-display- charges-your-phone-through-the- screen, January 2014. Ac-
cessed 12 July 2016.
[54] J. Mundy. Snapdragon 820 vs exynos 8890: Which galaxy s7 cpu is best? http:/ /www.trustedreviews.com/opinions/
snapdragon-vs- exynos, March 2016. Accessed 11 March 2016.
[55] S. Needhidasan, M. Samuel, and R. Chidambaram. Electronic waste an emerging threat to the environment of urban
india. Journal of Environmental Health Science & Engineering, 12(36), 2014.
[56] Deloitte Network. Smartphone batteries : Better but no breakthrough. Technical report, Deloitte Touche Tohmatsu
Limited, London, 2015.
[57] NVIDIA. The benefits of multiple cpu cores in mobile devices. Technical report, NVIDIA Corporation, 2010.
[58] J. Parsons. A battery that charges in 6 minutes? mit breakthrough could power smartphones of the future. http://www.
mirror.co.uk/news/technology- science/science/battery-charges- 6-minutes- mit-6231611, August 2015. Accessed
25 June 2016.
[59] Phys.org. Thermoelectric generator on glass fabric for wearable electronic devices. http: / / phys . org / news /
2014-04- thermoelectric-glass- fabric-wearable- electronic.html, April 2014. Accessed 25 June 2016.
[60] Phys.org. Scientists see the light on microsupercapacitors: Laser-induced graphene makes simple, powerful energy storage
possible. http://phys.org/news/2015- 12-scientists-microsupercapacitors.html, December 2015. Accessed 25 June
2016.
[61] P. Prakash. Environmental impact of internet searches and data centers. https: / / www . linkedin . com / pulse /
environmental-impact- internet-searches- data-centers- pranav-prakash/, July 2017. Accessed 18 August 2018.
[62] P. K. D. Pramanik and P. Choudhury. Iot data processing: The different archetypes and their security & privacy as-
sessments. In S. K. Shandilya, S. A. Chun, S. Shandilya, and E. Weippl, editors, Internet of Things (IoT) Security:
Fundamentals, Techniques and Applications, pages 37–54. River Publishers, 2018.
[63] P. K. D. Pramanik, P. Choudhury, and A. Saha. Economical supercomputing thru smartphone crowd computing: An
assessment of opportunities, benefits, deterrents, and applications from indias perspective. In 4th International Conference
on Advanced Computing and Communication Systems (ICACCS - 2017), Coimbatore, India, 2017. IEEE.
[64] P. K. D. Pramanik, B. Mukherjee, S. Pal, T. Pal, and S. P. Singh. Green smart building: Requisites, architecture,
challenges, and use cases. In A. Solanki and A. Nayyar, editors, Green Building Management and Smart Automation.
IGI Global, 2019.
[65] P. K. D. Pramanik, S. Pal, A. Brahmachari, and P. Choudhury. Processing iot data: From cloud to fog. its time to
be down-to-earth. In P. Karthikeyan and M. Thangavel, editors, Applications of Security, Mobile, Analytic and Cloud
(SMAC) Technologies for Effective Information Processing and Management, pages 124–148. IGI Global, 2018.
[66] P. K. D. Pramanik, S. Pal, and P. Choudhury. Beyond automation: The cognitive iot. artificial intelligence brings sense
to the internet of things. In A. K. Sangaiah, A. Thangavelu, and V. M. Sundaram, editors, Cognitive Computing for Big
Data Systems Over IoT: Frameworks, Tools and Application, pages 1–37. Springer, 2018.
[67] P. K. D. Pramanik, S. Pal, G. Pareek, S. Dutta, and P. Choudhury. Crowd computing: The computing revolution.
In R. Lenart-Gansiniec, editor, Crowdsourcing and Knowledge Management in Contemporary Business Environments,
pages 166–198. IGI Global, 2018.
[68] P. K. D. Pramanik, B. Upadhyaya, S. Pal, and T. Pal. Internet of things, smart sensors, and pervasive systems: Enabling
the connected and pervasive health care. In N. Dey, A. Ashour, S. J. Fong, and C. Bhatt, editors, Healthcare Data
Analytics and Management, pages 1–58. Academic Press, 2018.
[69] PTI. Superior batteries for laptops, smartphones in the offing. http://www.business-standard.com/article/pti- stories/
superior-batteries- for-laptops- smartphones-in- the-offing-116020800581 1 .html, February 2016. Accessed 25
June 2016.
[70] M. Shedd. Snapdragon 820 and kryo cpu: heterogeneous computing and the role of custom compute. https://www.qualcomm.
com / news / snapdragon / 2015 / 09 / 02 / snapdragon-820- and-kryo- cpu-heterogeneous- computing-and-role-custom,
September 2015. Accessed 11 March 2016.
[71] M. Shwartz. Aluminum battery from stanford offers safe alternative to conventional batteries. http://news.stanford.edu/
2015/04/06/aluminum-ion- battery-033115/, April 2015. Accessed 25 June 2016.
[72] R. Smith and A. Frumusanu. The qualcomm snapdragon 820 performance preview: Meet kryo. http://www.anandtech.
com/show/9837/snapdragon-820- preview, December 2015. Accessed 08 June 2016.
[73] In Hyuk Son, Jong Hwan Park, Soonchul Kwon, Seongyong Park, Mark H R¨
ummeli, Alicja Bachmatiuk, Hyun Jae
Song, Junhwan Ku, Jang Wook Choi, Jae-man Choi, et al. Silicon carbide-free graphene growth on silicon for
lithium-ion battery with high volumetric energy density. Nature communications, 6:7393, 2015.
[74] J. Sossamon. New device could increase battery life of electronic devices by more than a hundred-
fold. https : / / munews . missouri . edu / news-releases / 2018 / 0516- new-device- could-increase-battery-
Green and Sustainable High-Performance Computing with Smartphone Crowd Computing 283
life-of- electronic-devices- by-more- than-a-hundred-fold/, May 2018. Accessed 17 February 2019.
[75] S. Srivastava. Global smartphone shipments reached record 1.55 billion units in cy 2017. https : / / www .
counterpointresearch.com/global-smartphone- shipments-reached- record-1-55-billion-units- cy-2017/, February
2018. Accessed 17 August 2018.
[76] Statista. Global smartphone shipments forecast from 2010 to 2022 (in million units). https: //www .statista . com/
statistics/263441/global-smartphone- shipments-forecast/, May 2018. Accessed 17 August 2018.
[77] A. Voica. Powervr gt7900: redefining performance efficiency in graphics. https : / / imgtec . com / blog /
powervr-gt7900- redefining-performance- efficiency/, February 2015. Accessed 13 September 2016.
[78] Y. Xiao, R. S. Kalyanaraman, and A. Yla-Jaaski. Energy consumption of mobile youtube: Quantitative measurement
and analysis. In Second International Conference on Next Generation Mobile Applications, Services, and Technologies,
Cardiff, UK, 2008.
[79] A. Zeenat. A Study and Development of an Efficient E-waste Management System for Minimizing the Risks of Environ-
mental Pollution. PhD thesis, Savitribai Phule Pune University, Pune, 2016.
[80] C. Ziegler. Lg optimus 2x: first dual-core smartphone launches with android, 4-inch display, 1080p video recording. https:
//www.engadget.com/2010/12/15/lg- optimus-2x- first-dual- core-smartphone- launches-with-android/, December
2010. Accessed 23 June 2016.
Edited by: Anand Nayyar
Received: Mar 5, 2019
Accepted: Mar 25, 2019
... 10,11 In contrast to conventional HPC systems, such as cloud data centers and supercomputers, the MCC serves as an exceptionally energy-efficient and cost-effective alternative. 12 In addition to reducing their IT expenditures, the implementation of MCCs at the enterprise level would enable them to implement ecological and sustainable computing. 13,14 ...
Article
Due to their increasing computational power and energy-efficient hardware, today’s smart mobile devices (SMDs) are replacing desktops and laptops as casual computing devices. Moreover, a cluster of such powerful SMDs can garner substantial high-performance computing (HPC). Such an HPC is achieved by utilizing publicly owned SMDs in mobile crowd computing (MCC). Here, a large computing-intensive task is divided and scheduled for the available SMDs for execution, and the results are recollected. This approach provides an economical and sustainable HPC. However, battery-powered constrained energy is a great hindrance to achieving this goal. Therefore, in the MCC, it is crucial to minimize the overall energy consumption to complete the task. This can be achieved to some extent by optimizing task scheduling to the appropriate SMDs. However, considering only energy efficiency might lead to an enormous load imbalance among SMDs, i.e., the most energy-efficient SMDs would be overloaded most of the time. Considering this, in this paper, we present a modified particle swarm optimization (PSO)-based scheduling algorithm to minimize the overall energy consumption among a set of SMDs designated to execute a set of MCC tasks while maintaining a satisfactory load balance level. Extensive simulations with both synthetic and real data sets are carried out to analyze and validate the proposed method. The work was compared with popular heuristic (minimum completion time (MCT), MinMin, MaxMin, and preconditioned progressive iterative approximation (PPIA)) and metaheuristic (genetic algorithm (GA)) optimization algorithms, which yields significant improvements over others in terms of the considered objectives. In addition, an analysis of variance (ANOVA) test is conducted to provide further evidence regarding the distinctiveness of the proposed algorithm.
... Recent advancements in wireless networking, mobile computing, and sensing technologies have transformed smartphones into IoT-connected platforms to enable crowdsourcing, augmented reality, and geospatial computing for a variety of research and educational applications [31,32,33,34]. Equipped with a number of built-in sensors that measure geo-location, acceleration, orientation, and angular velocity, smartphones can provide the essential hardware for a variety of mobile applications (i.e., smartphone apps) to support the smart management of urban mobility and transportation systems [35,30,36]. ...
Article
The stop-and-go traffic pattern on urban roads often results in excessive energy consumption because of unnecessary vehicle braking, idling, and accelerations. With the widespread and increased use of automobiles, this traffic pattern creates many negative impacts (e.g., delayed travel time, air pollution, and additional carbon emission) on the sustainability of our cities. Taking advantage of the recent emerging Internet of Things (IoT) and edge computing paradigms, we propose a mobile edge computing framework that integrates the capability of real-time vehicle-to-infrastructure communication and intelligent speed optimization algorithms into a mobile app to optimize individual vehicles' driving speed at signalized intersections. The optimization aims to mitigate the stop-and-go traffic pattern and its undesirable consequences in urban transportation systems. The framework consists of (1) a cyberinfrastructure-enabled dynamic messaging system for retrieving and delivering real-time traffic and signal phase and timing information from IoT-connected signal controllers and sensors, (2) a real-time speed optimization algorithm for generating intelligent speed advisory using vehicle's information (e.g., GPS and driving directions from mobile sensing) and corresponding signal and traffic information, and (3) an ad-hoc mobile computing environment that converts drivers' smartphones into edge devices to host the speed optimization algorithms for enabling intelligent advisory on the vehicle's driving speed within signalized corridors. The paper presents the design and implementation of the proposed framework. We demonstrate the feasibility, usefulness, and energy-saving benefits of our proposed framework and its prototyping mobile app on urban transportation systems through traffic simulation, real-vehicle laboratory experiments, an evaluative survey, and field communication tests. The simulation-based energy evaluation results show that the 100% usage of the mobile app can achieve 24% energy savings in the transportation system.
... . MCC provides not only a cost-effective HPC but also a significantly energy-efficient HPC option compared to other traditional HPC systems such as supercomputers and cloud data centres [12]. Besides reducing their IT expenditure, the organisation-level adoption of MCC would allow them to achieve green and sustainable computing [13]. ...
Preprint
Full-text available
Due to their increasing computational power and energy-efficient hardware, today's smart mobile devices (SMDs) are replacing desktops and laptops as casual computing devices. Moreover, a cluster of such powerful SMDs can garner substantial high-performance computing (HPC). In mobile crowd computing (MCC), public-owned SMDs are utilised to attain such HPC. Here, a large computing-intensive task is divided and scheduled to the available SMDs for execution and the results are recollected. This provides an economical and sustainable HPC. However, battery-powered constrained energy is a great hindrance to achieving this. Therefore, in MCC, it is crucial to minimise the overall energy consumption to complete the task. This can be achieved to some extent by optimising the task scheduling to the appropriate SMDs. However, considering only energy efficiency might lead to a huge load imbalance among SMDs, i.e., the most energy-efficient SMDs would be overloaded most of the time. Considering this, in this paper, we present a PSO-based scheduling algorithm to minimise the overall energy consumption among a set of SMDs designated to execute a set of MCC tasks while maintaining a satisfactory load balance level. The proposed method is analysed and validated by extensive simulations with synthetic as well as collected datasets. The work is compared with popular heuristic (MCT, MinMin, MaxMin, and PPIA) and metaheuristic (GA) optimisation algorithms, displaying significant improvements over others in terms of the considered objectives. Furthermore, a statistical hypothesis test, ANOVA, is carried out to demonstrate the distinctiveness of the proposed algorithm.
... For instance, to make a computer with a 17-in. CRT monitor, on average, 1500 L of water, 240 kg of fuel, and 22 kg of chemicals are needed, which costs approximately a total of 1.8 tons of material [186,187]. Besides, it generates a significant amount of hazardous e-wastes contaminating the earth. MCC utilises already existing resources, i.e. the public's SMDs. ...
Article
Full-text available
Owing to the enormous advancement in miniature hardware, modern smart mobile devices (SMDs) have become computationally powerful. Mobile crowd computing (MCC) is the computing paradigm that uses public-owned SMDs to garner affordable high-performance computing (HPC). Though several empirical works have established the feasibility of mobile-based computing for various applications, there is a lack of comprehensive coverage of MCC. This paper aims to explore the fundamentals and other nitty–gritty of the idea of MCC in a comprehensive manner. Starting with an explicit definition of MCC, the enabling backdrops and the detailed architectural layouts of different models of MCC are presented, along with categorising different types of MCC based on infrastructure and application demands. MCC is compared extensively with other HPC systems (e.g. desktop grid, cloud, clusters and supercomputers) and similar mobile computing systems (e.g. mobile grid, mobile cloud, ad hoc mobile cloud, and mobile crowdsourcing). MCC being a complex system, various design requirements and considerations are extensively analysed. The potential benefits of MCC are meticulously mentioned, with special discussions on the ubiquity and sustainability of MCC. The issues and challenges of MCC are critically presented in light of further research scopes. Several real-world applications of MCC are identified and propositioned. Finally, to carry forward the accomplishment of the MCC vision, the future prospects are briefly elucidated.
Conference Paper
Mobile crowd computing (MCC) has emerged as a promising paradigm to leverage the underutilised computational resources of smart mobile devices (SMDs). However, the energy constraints of these devices pose a significant challenge in efficiently executing computational tasks. This study analyses the energy efficiency of various scheduling algorithms for MCC environments. The study compares the performance of seven scheduling algorithms: PPIA, RAS, PSO, GA, MaxMin, MinMin, and MCT. These algorithms were evaluated in terms of their energy efficiency when executing different sets of tasks with varied instruction lengths. The experiment used a simulated MCC environment with real data from five SMDs. The results demonstrate that the PSO algorithm consistently outperformed the other scheduling schemes regarding energy efficiency across all task sets. The GA algorithm also showed promising results. The MinMin algorithm exhibited the lowest energy efficiency. The findings of this study provide valuable insights for designing and implementing energy-efficient task scheduling strategies in MCC systems. Analysing the different algorithms under varying task loads can help MCC service providers make informed decisions when selecting appropriate scheduling techniques to optimise the energy utilisation of SMDs.
Article
Full-text available
The widespread adoption of utility-based real-time applications has placed the necessity of widescale deployment of edge computing infrastructure. Crowdsourced edge computing is deemed a suitable way out. On the other hand, a collection of today’s powerful smart mobile devices (SMDs) can cumulatively offer high-performance computing. The public-owned SMDs are utilized opportunistically to form a dynamic ad-hoc computing grid known as mobile crowd computing (MCC). This paper aspires to establish a proof-of-concept for the feasibility and use of MCC as a sustainable edge computing solution (MCC-edge). A typical smart HVAC system of an office building has been considered for the experiment case. We aim to process the HVAC data in real-time using the MCC-edge setup within the building for auto adjustment of the AC controller and error notifications. To maintain the ideal comfort level of the occupants, we present an extensive calculation using the dew point and heat index of the room. A high-level layered architecture of the MCC-edge for HVAC is presented along with a general framework of the MCC-edge. We report the module-wise design and implementation procedures with exhaustive details. The performance of MCC-edge is statistically compared with the commercial edge and cloud computing solutions in terms of cost, energy consumption, latency, and environmental impact, showing a significant advantage over the two. Every procedural detail of each module's design, development, and implementation is meticulously presented, which would aid interested readers and researchers in rebuilding such an application.
Article
Full-text available
The impact of digital finance on enterprise economic development and ecological environment has been widely debated and researched. The purpose of this paper is to evaluate and deconstruct the relationship between digital finance and the synergy of enterprise economic development and ecological environment. The paper examines the benefits and drawbacks of digital finance and its influence on the economy, as well as its impact on the ecological environment. Through a comprehensive literature review and empirical analysis, the study investigates the mechanisms through which digital finance affects the synergy of enterprise economic development and ecological environment, and evaluates the effectiveness of current measures and policies to address these impacts. The results of this study provide valuable insights for policymakers, researchers, and practitioners on how to achieve a balance between digital finance and sustainable development. By deconstructing the relationship between digital finance and the synergy of enterprise economic development and ecological environment, this paper aims to contribute to the development of sustainable finance models and to foster a deeper understanding of the role of digital finance in shaping a sustainable future.
Chapter
Full-text available
Non-sustainable buildings have threatened the ecosystem globally. In this chapter, a comprehensive discussion on the green and smart building is presented, considering how the buildings are made green and smart and how they support in developing sustainable cities. Though smart buildings are the positive catalyst towards sustainability, the excessive use of electronic devices puts a check in attaining the overall green goal. This chapter suggests merging green and smart technologies to have green smart building (GSB) with the aim of offering the populations a smart and eco-friendly living. Promises and challenges in attaining this goal are meticulously explored. The GSB concept is discussed in details, suitably supported with the architectural models of overall and the various components of a GSB. The communication architecture is also presented emphasizing on various entities and activities in different levels of communication between various digital components of a GSB. A few cases have been presented showing practical applications of green and smart technologies in buildings.
Chapter
Full-text available
Undoubtedly, in recent years, Internet of Things (IoT) has created the maximum ripples in the IT industry. Large-scale implementation of IoT will generate a massive amount of data. To obtain actionable knowledge, these data must be processed. But the complex nature of these data and the IoT architecture as well have made the processing a complicated task. This chapter looks into the fact, how it becomes challenging to the computing fraternity to process the diverse types of dynamic data generated from diverse heterogeneous IoT devices. This chapter will help readers in having an overview of the different architectures for IoT data processing. Each architecture is discussed decoratively along with the associated challenges. The open and ubiquitous nature of IoT makes it more vulnerable to the security threats. The issues pertaining to security and vulnerabilities that are faced by the architectures are discussed specifically. Along with a general discussion on the security and privacy issues in IoT, the authors have also portrayed a different view on the overhyped concerns for the security and privacy aspects in IoT. The purpose of this chapter is to make readers able to decide on the suitable data processing architecture for their IoT applications considering different factors such as cost, response time, etc. with a special focus on security and privacy.
Chapter
Full-text available
The power of crowd always has brought wonders. The same applies to the computing as well. The accumulated idle CPU cycles of millions of personally owned devices are capable of producing huge computing capacity. We have termed it as crowd computing. Though this very concept has been nurtured in the past through grid computing, in the age of powerful smartphones and tablets, it deserves to have a fresh look. In this chapter, the authors aim to present crowd computing in a modern approach. Readers will be able to gain a fair comprehension of the various aspects of crowd computing and have an insight of the ecosystem of this computing paradigm. The characteristics, benefits, issues, implementational challenges, applications, and examples of crowd computing are portrayed elaborately. To clear the air, crowd computing has been distinguishably compared to other analogous computing systems such as P2P computing, cloud computing, supercomputing, and crowdsourcing. The business values of crowd computing as well as the scope of offering crowd computing as a service have also been explored.
Chapter
Full-text available
Cloud Computing has been used for processing Internet of Things (IoT) data. This works fine for the analytical and batch processing jobs. But most of the IoT applications demand real-time response which cannot be achieved through Cloud Computing mainly because of inherent latency. Fog Computing solves this problem by offering cloud-like services at the edge of the network. The computationally powerful edge devices have enabled realising this idea. Witnessing the exponential rise of IoT applications, Fog Computing deserves an in-depth exploration. This chapter establishes the need for Fog Computing for processing IoT data. Readers will be able to gain a fair comprehension of the various aspects of Fog Computing. The benefits, challenges and applications of Fog Computing with respect to IoT have been mentioned elaboratively. An architecture for IoT data processing is presented. A thorough comparison between Cloud and Fog has been portrayed. Also, a detailed discussion has been depicted on how the IoT, Fog, and Cloud interact among them.
Presentation
Full-text available
This presentation outlines an estimation of the global electricity usage that can be ascribed to Communication Technology (CT) in the coming decade. The scope is two scenarios for use and production of consumer devices, communication networks and data centers. Two different scenarios— best and expected—are set up, which include annual numbers of sold devices, data traffic and electricity intensities/efficiencies. For the first time AR and VR devices will be included in a CT power trend analysis. I will emphasize the potential development of total Global Data Center IP Traffic and data center electric power consumption. The effect of 5G adaption rate on the total CT power consumption will also be addressed. The likely share of CT of the total global electric power consumption will be discussed in the light of CTs potential to reduce consumption.
Conference Paper
Full-text available
Buying and maintaining supercomputers are exceedingly costly affairs. In a country like India, not many organisations can afford to have one, due to which many compute-intensive works and research remain undone. On the other hand, the revolution of smartphones has brought unforeseen computing power to our fingertips. Today’s smartphones can deliver equivalent or even better performance than supercomputers of yesteryears. We argue that the collective computing power of these mighty smartphones, much like of desktop grid computing, can provide us the computing power of potentially equivalent to a modern supercomputer without paying out much. We have endorsed our assessment with the statistics of smartphone’s computing potential and user development of smartphones in India. Some obvious restraining issues, as well the future promises to come over them, have been discussed. Some specific application areas in Indian perspective have been broached also.
Article
Full-text available
In this article the authors evaluate the impact of different sectors of information and communication technologies (ICT) on energy consumption of society, and on CO2 emissions. ICT is understood to cover computer and peripheral equipment including local area networks, telecommunication equipment and networks, and data centers. In 2012, close to 4.7 percent of the world’s electrical energy was consumed by ICT, releasing into the atmosphere roughly 1.7 percent of the total CO2 emissions. These numbers are growing, although the improving energy efficiency of ICT equipment leads to a slower growth of these metrics than the increase of ICT's usage worldwide. On the other hand, ICT can contribute to a reduction in energy consumption and CO2 impact in other sectors, such as transportation, smart buildings and virtual work and learning. Regarding transportation, studies show in the United States the average number of miles traveled in private vehicles has decreased significantly since 2007 despite a sustained growth over many decades; the drop has persisted after the end of the most recent economic crisis, suggesting that these energy savings may be attributed to changes in the modes of work and commerce induced by ICT.
Chapter
The advancement in smart medical devices and pervasive systems is setting the connected health care into a prospering platform for pervasive health care. The Internet of Things (IoT), smart sensors, and the wearables have augmented the health care system for enabling to remote monitoring and supporting the medical condition of the patient in and out of clinics. This chapter aims to serve as a brief primer on the development of smart and pervasive health care systems. The chapter starts with an overview of IoT, smart sensors, and pervasive systems and also their interrelationship. The differences between IoT and pervasive systems are meticulously laid down to clear the confusion between these two computing paradigms. It highlights the challenges faced by today's health care systems and analyses how to overcome these with the help of pervasive health care. It discusses the basics of mobile and pervasive health care including context-aware and connected health care while distinguishing it from telemedicine. It elaborates on the role of IoT in health care while discussing a number of sensors and smart devices used in health care, fitness, and medical care units. The benefits of and challenges involved in pervasive health care are reviewed comprehensively. Several real-life applications and use cases of IoT and smart sensors are presented showing the effectiveness of pervasive health care. The chapter also presents an assessment of the current and future IoT health care market along with the listing of the key players in the IoT health care market.
Article
The energy-efficiency drive at the information factories that serve us Facebook, Google and Bitcoin. The energy-efficiency drive at the information factories that serve us Facebook, Google and Bitcoin.
Chapter
The Internet of Things (IoT) has already been infiltrated to our everyday life in the forms of smartphones, smart TVs, fitness trackers, health monitoring systems, smart watches, vending machines, smart meters, city traffic, building security systems and much more like this. It has steered the automation to a new high. But IoT alone has limited capability. To reap the actual benefit of IoT, it has to be intelligent. In this chapter, we have reasoned why IoT needs Artificial Intelligence (AI). The intelligent IoT that we discuss here, is termed as the Cognitive IoT (CIoT) . Cognitive IoT uses a new computing paradigm called Cognitive Computing, often popularly dubbed as the third era of computing. Cognitive Computing will make IoT more sophisticated, more intelligent, and more interactive. This chapter essentially focuses on the convergence of IoT with Cognitive AI. In addition to introducing a preliminary idea of Cognitive Computing, the chapter discusses several aspects of CIoT. It explains how Cognitive Computing has been used in IoT through different forms such as Cognitive Network, Cognitive Devices, and Cognitive Analytics. The desirable properties of the CIoT are mentioned. A special discussion is presented on how CIoT has taken automation to a new level. Realizing CIoT is not straightforward for the reason that it is a complex system and is quite different from other computing systems we have been familiar with. The implementation challenges along with the societal and ethical apprehensions of CIoT have been identified and discussed in detail. The business world is showing great interest in IoT and subsequently in CIoT. Considering that, we have highlighted the business values of CIoT and some prospective application areas along with a few prominent use cases. But, will excessive adoption of AI and tendering human-like intellect to inorganic devices be a threat to the human race? Do we need to be warned? A rational discussion on that has been presented before winding up.