Content uploaded by Gabor Paller
Author content
All content in this area was uploaded by Gabor Paller on Jul 20, 2015
Content may be subject to copyright.
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
Sensors & Transducers
c
2015 by IFSA Publishing, S. L.
http://www.sensorsportal.com
Power consumption considerations of GSM-connected
sensors in the AgroDat.hu sensor network
Gábor Paller, Péter Szármes and Gábor Él˝
o
Széchenyi István University, ITOK research group, Gy˝
or, Egyetem tér 1, 9026, Hungary
Tel.: +36 (96) 503-400, fax: +36 (96) 329-263
E-mail: paller.gabor@sze.hu, peter.szarmes@sze.hu, elo@sze.hu
Received: Accepted: Published:
Abstract: The number of large sensor systems are rapidly growing nowadays in many fields. Well-designed Big
Data solutions are able to manage the enormous data flow and create real business benefits. One dynamically
growing application area is precision farming. It requires robust and energy-efficient sensors, because the devices
are placed outdoors, often in harsh conditions, and there is no power outlet “in the middle of a corn field”.
Power efficiency is in general one of the major themes of the Internet of Things (IoT). According to the
IoT vision, embedded sensors send their data to processing units (either located near to the sensor or on some
intermediate ”gateway” device or in the cloud) using heterogeneous transport networks. Some sensors employ
short-range network like Bluetooth and some ”gateway” device like a tablet. Other sensors directly connect to
wide-area networks like cellular networks.
This paper will analyse different communication patterns accomplished over GSM network from the view-
point of the energy consumption of the sensor device with the assumption that the sensor is stationary. The
measurements were done using two different GSM modems designed for embedded systems to ensure that
the results represent a wider picture and not some implementation property of a particular GSM modem.
Recommendations are given about the strategies applications should follow in order to minimize the energy
consumption of their GSM subsystems.
Keywords: agriculture, soil sensor, power efficiency, cellular communication, communication strategies.
1. Introduction
1Internet of Things is often considered a recent
trend but the vision was presented first in 1991 [15].
Weiser envisioned computers that "disappear into the
background" and are connected with wired and wire-
less links. One key element of Weiser’s ubiquitous
computing was the low-power nature of the comput-
ing elements that are able to function for an extended
period of time without recharging otherwise battery
issues would prevent the devices from "disappear-
ing into the background". Low-power and ultra low-
power energy consumption has been a key IoT re-
1This article is an extended version of our SENSORNETS 2015
paper [11]
search theme ever since [14], [16].
IoT systems employ heterogeneous networks to
connect the sensors to the data processing units. Some
solutions are based on short-range networks (e.g. Zig-
Bee, Bluetooth), the data is collected by some "gate-
way" device (e.g. smartphone, tablet, set-top box)
which then connects to a wide-area network. Isolated
sensors that are rarely visited by humans and are far
from any other elements of the ubiquitous network
cannot adopt this solution, these sensors have to con-
nect to the wide-area network directly. The most com-
mon wide-area network with low connectivity cost and
large coverage is the public cellular network.
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
2. The AgroDat project
Today sensors and sensor networks gain more and
more importance in many application areas. Machines
(including cameras, sensors, satellites, imaging de-
vices, etc.) are already generating more data than we,
humans and business processes (Figure 1). These de-
vices often operate in a harsh environment without ac-
cess to electric networks, where robustness and energy
efficiency are very important characteristics.
One such application field is agriculture. Preci-
sion agriculture is an integrated agricultural manage-
ment system incorporating several technologies. This
technology can reduce the cost of producing crops and
the risk of environmental pollution [4]. The Agro-
Dat R&D project with notable industrial and scientific
partners aims to build an agricultural information sys-
tem in Hungary. The system relies on collecting and
analyzing high-volume data about crops and environ-
mental conditions, like soil moisture and temperature,
air temperature, precipitation, solar radiation, etc.
Soil sensors (see Figure 2) can measure water po-
tential, electric conductivity, volumetric water content,
soil temperature etc. Electric conductivity correlates
with salt content, influencing plant growth. Water po-
tential refers to the water available for plants. This data
can be used for planning irrigation, forecasting plant
diseases, and analyzing soil aspiration. Light sensors
(see Figure 2) can measure the intensity of photosyn-
thetically active radiation, or the spectrum of incoming
and reflected light in certain bands, which can then be
used to calculate the Normalized Difference Vegeta-
tion Index and Photochemical Reflectance Index [6].
These indexes correlate closely with vegetation and
photosynthetic activity respectively, and they are good
indicators of biomass and plant stress. Sensors can
measure relative humidity, air temperature or vapor
pressure. These values are linked with plant evapo-
ration. Leaf wetness sensors are designed to detect
wetness (presence and duration) and ice formation on
leaf surfaces. The data is useful for forecasting plant
diseases and planning spraying actions.
Combining different sensors into a sensor group
creates synergies, and during the design of such a
sensor unit, low energy consumption and ability to
withstand harsh environmental conditions are impor-
tant objectives. Much of the data needed for the agri-
cultural information system can be collected by these
sensor units, which can make measurements even on
a minute-rate. Data from the sensors across the fields
will be sent via GSM networks onto central servers.
Sensors are very different in terms of their data
communication requirements. Some sensors may send
large amount of data, even in real-time (like streaming
video). The current batch of agricultural sensors being
developed by our project have the following proper-
ties.
•These sensors are stationary. Once installed, they
move very rarely.
•Their environment changes only slowly. For ex-
ample sudden changes in ground temperature or
ground moisture are rare. This means that sensor
values can be sampled with quite long sampling
periods (multiple hours or even daily).
•The quantity of the data to be transmitted is rel-
atively small. One measured quantity is a scalar
value and the sensor equipment measures about
10-20 such quantities.
•These sensors are installed on locations that are
rarely accessed and are far from the usual net-
work infrastructure endpoints. For example one
of our sensors are meant to be installed on large
corn fields. Long, unassisted operation is an im-
portant requirement.
These requirements have led to the following high-
level design decisions.
•The sensors will be connected using ordinary
GSM network directly, without the help of some
"gateway" node. Each sensor will be a GSM end-
point.
•Low-bandwidth data bearers like SMS or GPRS
satisfy the transfer requirements.
•Low energy consumption/long operating time
without on-site service is crucial.
•Remote manageability of the sensor is a must.
The remaining sections of the paper will discuss
the proposed communication architectures, the sensors
supported by the sensor station, the communication al-
ternatives we have evaluated and the evaluation results.
3. The sensor station
The parameters measured by the sensor station
were established according to a risk analysis of the
corn production [5]. The sensor station comes in two
variants. One variant measures only underground pa-
rameters. Except for a plastic dome protecting the
GSM antenna, this station has almost no parts above
the ground (see Figure 3).
In addition to the underground variant, the sensor
station can be equipped with a pole that contains in-
struments above the ground. This extra pole is shown
in Figure 4.
The underground sensor set can measure the fol-
lowing parameters:
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
Fig. 1: Sources of the data growth (source: TDWI)
Fig. 2: Decagon soil and light sensors (source:
Decagon)
•Soil temperature (depth: 5-20-40-60-80 cm)
•Soil moisture (depth: 5-20-40-60-80 cm)
•Concentration of salts in groundwater
•CO2concentration in the ground
The sensor station measures the following param-
eters above the ground.
•Air temperature (height: 20 cm, 2 m)
•Humidity (height: 20 cm, 2 m)
•Rainfall (height: 1 m)
•Wind speed and direction (height: 2 m)
•Solar radiation intensity (height: 2 m)
•Leaf wetness (height: 1 m)
Fig. 3: Underground part of the sensor housing
The construction of the sensor station is modular.
The sensor control (based on a microcontroller and
I2C bus) and the battery units are always present. The
sensor control unit is connected to the communica-
tion unit by means of an asynchronous serial interface.
The modular construction allows the usage of different
communication units.
4. Communication architecture of
the AgroDat.hu network
As the first version of the AgroDat.hu sensor net-
work will target corn, typical cornfield locations were
considered when designing the communication archi-
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
Fig. 4: Pole with sensors above ground
tecture. Due to large field sizes and the production
area often located far from existing infrastructure, only
a radio technology with large coverage area was ac-
ceptable. There are a number of alternative radio tech-
nologies with this characteristic (e.g. WiMax or cus-
tom VHF/UHF system) but due to its wide availability,
low cost and well-established regulatory framework,
we decided to use the GSM mobile network.
The first version of the sensor network collects data
that change slowly (e.g. soil temperature, soil mois-
ture) and the data representation requires only short
data packets (with our coding format it means 200-
400 bytes of data). This means that the sensor com-
municates on the mobile network relatively rarely (1-3
times a day) and even then only low amount of data is
sent. Part of the sensor stations is only equipped with
underground sensors and minimally protrude above
ground level therefore solar cell-based power supply
was not possible. Energy efficiency was a key re-
quirement when designing the sensor station. Due to
the low amount of data transfer and the energy effi-
ciency requirement, the prorotype was designed using
the low-bandwidth services of the GSM network. This
may mean GPRS or SMS-based data transfer.
We prepared two versions of the communication
architecture. The first is based on GPRS, the second
uses SMS-based data transfer. In case of GPRS the
sensor may access the web infrastucture directly over
the HTTP protocol. This is advantageous from the
point of view of the server side as numerous solutions
providing extreme scalability are available based on
the HTTP protocol.
The relatively low amount of data and the com-
pact binary encoding permits the data transfer over the
Short Message Service (SMS). Even when encoded
into textual format, our measurement data can fit into
3-4 Short Messages (SM). Assuming 3 measurements
per day, this means maximum 12 SMs per day which
is a reasonable cost. From the architectural point of
view, the SMS infrastructure can be connected to as
mobile endpoint or through the application protocol of
the SMS Center (SMSC). Both solutions require an ad-
ditional software component between the SMS infras-
tructure and the application server. This software com-
ponent adapts the SMS interface of the mobile end-
point or the SMSC application protocol interface to the
application server and compared to the variant using
only HTTP, it means a more complicated architecture.
The other consideration is the power consumption of
the sensor unit. As we will demonstrate later, sending
112 bytes requires an order of magnitude less power
when using SMS compared to GPRS. On the other
hand, the battery power required by data transfer oper-
ations is still negligible compared to other elements of
the power consumption, namely keeping the module
registered on the network. We decided that the advan-
tages of the direct HTTP communication exceed the
disadvantage of the slightly higher overall power con-
sumption and we decided to use a GPRS-HTTP-based
architecture. Figure 5 shows the GPRS-HTTP-based
communication architecture.
According to the plans, the sensor network will
consist of 300-1000 sensors. Efficiently operating so
many endpoints cannot be accomplished without a re-
mote management solution. According to the plans,
the sensor network will be managed by two sepa-
rate sensor management systens. The HP Dynamic
SIM Management (DSM) system will manage the SIM
cards which includes the inventory of SIM cards, the
assignment of the SIM cards to sensors, monitoring
the operation of the mobile endpoints and security ser-
vices like checking whether a specific SIM card is still
in its assigned sensor. HP DSM implements a sub-
set of with a SIM Toolkit application installed in the
SIM cards. This SIM Toolkit application communi-
cates with the HP DSM server components using the
SMS infrastructure.
A custom management system is provided to han-
dle the non-telecommunication-specific properties of
the sensor station. For example such properties are the
configuration of the measurement times or the recep-
tion of error reports from the sensor. This sensor man-
agement system also uses the SMS infrastructure for
sending asynchronous messages to the sensor while
management operations requiring larger transfer sizes
are accomplished over the GPRS-HTTP bearer. En-
ergy efficiency issues related to the messages sent from
the server to the sensor will be discussed later in the
paper.
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
Fig. 5: Conceptual communication architecture
5. Evaluated GSM modems
In order to ensure that we are really evaluating the
communication alternatives, we chose to run our mea-
surements from two different GSM modem vendors.
GL865-QUAD is a variant of Telit’s extremely
popular GE865 product family [1]. The module comes
in DUAL (GSM900, DCS1800 frequency bands) and
QUAD (GSM900, DCS1800, GSM850 and DC1900
frequency bands) variants. The module has 2.5G net-
work support which means that it can access GSM
(voice call and SMS) and GPRS network services. 3G
and higher is not supported by this family of modules.
Telit offers 3G modules too but as many sensor ap-
plications can be implemented with 2/2.5G, the lower
cost and power consumption make these 2/2.5G mod-
ules very popular with connected sensors.
Unique property of the Telit modules is that many
of them, including the GE/GL865 family include an
entire runtime for application logic. The modules can
be used in GSM modem mode when the application
code is executed by some external CPU (e.g. a mi-
crocontroller) but a standalone mode is also available
when the application code is executed by the on-chip
Python interpreter. The module offers features of a so-
phisticated embedded platform: non-volatile memory
in the form of a file system, A/D and D/A convert-
ers and general-purpose I/O pins, all accessible from
Python code. The GE/GL865 can therefore implement
the entire sensor control, not just the cellular commu-
nication aspects.
SIMCOM’s SIM900 module [3] was selected to
cross-check the power consumption measurement re-
sults of certain communication scenarios on a different
GSM modem implementation. The SIM900 is a quad-
band GSM modem. It is a more traditional unit in the
sense that SIM900 needs an external CPU to execute
the application logic. SIM900 also has a set of built-in
peripherals, including real-time clock, A/D converter
and general-purpose I/O pins. These peripherals can
be manipulated by custom modem commands.
Power consumption measurements were accom-
plished by inserting a serial 0.1 Ohm resistor into the
power line of the GSM modules. These GSM modules
also include the RF power circuits so the consumption
of the whole communication hardware, including the
RF power amplifiers was measured. Additional inter-
face circuits, e.g. RS232C drivers were not incorpo-
rated into the measurements but these circuits are not
necessarily present in embedded sensors. The voltage
drop on the serial resistor was measured with a digital
multimeter which measured with about 3 Hz sampling
frequency and sent the samples to the PC where the
samples were recorded. The effect of the filter capaci-
tors on the power lines is such that higher frequencies
are filtered out so this relatively slow sampling rate
was acceptable. The samples were further analysed
using the R/R Studio mathematical suite 2.
2http://www.rstudio.com/
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
6. Communication scenarios
6.1. Network registration
This is seemingly the simplest scenario but it
comes with the most complications. Registering on
the network and staying registered involves network
registration and location update procedures but more
importantly, it requires that the GSM module is active
and listens to network events. As we assume station-
ary operation, procedures relevant to mobility manage-
ment e.g. cell handover do not occur but in order to
stay registered on the network, periodic location up-
date has to be executed. Figure 6 shows the power
consumption of the Telit GL865 executing this sce-
nario. The two spikes of power consumption are re-
lated to the network registration and location update
procedures. Location update occurs on the network
used during the measurements (Telenor Hungary) in
about every 55 minutes which is a quite typical value
and can be expected to be between 30 minutes and 2
hours. It is more important to note, however, that the
idle power consumption of the module is about 7mA.
This means that while the actual network procedures
consume 720 mAs ( milliamper-second) for the net-
work registration and 400 mAs for one location update
(with this network, there are about 26 location updates
per day which means about 10400 mAs or 2.89 mAh
cost for location updates), keeping the module opera-
tional costs about 170 mAh for a day.
Note that these values are relatively unaffected
by the received signal strength. The measurements
were done with RSSI=5, RSSI=4 and RSSI=2 signal
strengths and the results were very similar. The reason
of this similarity is that actual network transmission is
very short in these scenarios, transmission power dif-
ference hence averages out.
The results are very similar with the SIM900 mod-
ule (Figure 7). Short power consumption spikes re-
lated to the network registration and location update
procedures can be observed but it is more important to
note the idle current consumption of the module which
is close to 20 mA. While the network registration pro-
cedure costs only 834 mAs and 26 location updates
cost 2.71 mAh, keeping the module operational costs
456 mAh for a day.
Both modules offer custom power saving modes.
The idea behind these modes is that only the functional
units executing GSM procedures remain operational,
units communicating with the application CPU (and in
case of the Telit module, units executing the applica-
tion logic) are switched off. In case of the Telit GL865
this mode is activated by the MOD.powerSaving()
Python call that places the module into power-save
mode for the specified duration of time. If an event
(e.g. incoming network event) occurs during this pe-
riod, the power-save mode is exited and the application
logic may start processing the event. As SIM900 does
not have an application execution environment, power-
save mode is offered differently. The AT+CSCLK
("slow clock") command with parameters of 1 or 2
switches off the interface with the application proces-
sor with slightly different wake-up mechanisms. Ei-
ther the communication interface is re-activated when
DTR is low (AT+CSCLK=1) or the module is re-
activated when there is data available on the serial in-
terface serving the application logic (AT+CSCLK=2).
With these power saving modes, the idle con-
sumption of the devices decreases quite dramatically.
For both the GL865 and the SIM900, the idle power
consumption falls below 1 mA. Specifically, for the
GL865 the power consumption needed to keep the
module operational for a day is about 11 mAh while
network procedures cost additional 2.9 mAh, resulting
a total of 14 mAh for a day. For the SIM900, the idle
power consumption for a day is about 23.3 mAh and
network procedures add 2.71 mAh, resulting a total of
about 26 mAh for a day.
These measurements show the importance of
implementation-specific power-save modes and high-
light the fact that the Telit GL865 is about twice more
efficient than the SIM900 when it comes to low-power
operation. It is a much more important observation,
however, that even with power-save modes active, the
continuous operation of the module has by far the
highest cost. For the Telit GL865, only 20% of the
power budget is spent on actual network procedures,
the remaining 80% is the cost of keeping the module
operational. The difference is more dramatic for the
less power-efficient SIM900: only 10% of the daily
power budget is spent on network operations, the re-
maining 90% is needed to simply keep the module ac-
tive.
6.2. Data communication
So far only the cost of being registered on the net-
work was calculated. Data communication comes with
additional costs. Our sensors send relatively small
amount of data (10-20 scalar values) relatively rarely
(1-2 times a day) so network bearers with lower band-
width were analysed. A wide variety of data encod-
ings have been proposed for IoT applications but the
area is far from settled. XML-based formats [13] and
publish-subscribe frameworks are being proposed for
IoT [8].
Our intention was to keep the amount of data trans-
mitted, the power needed for data transmission and
the CPU cycles needed to encode/decode packets low
so we adopted a size-efficient data encoding based on
ASN.1 and Basic Encoding Rules (BER) [2]. These
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
Fig. 6: Telit GL865 power consumption (initial registration and location update)
BER data structures were then sent to the server us-
ing HTTP. Specialized protocols like CoAP have been
proposed for IoT applications to replace the widely de-
ployed Internet protocol suite (HTTP, FTP, TCP ...)
[10] but CoAP is not that attractive on GSM transport
networks where the limitations of 802.15.4 do not ap-
ply. Both GL865 and SIM900 supports TCP by cus-
tom modem commands. Using the modules’ native
TCP support, HTTP was implemented in the applica-
tion logic.
The power consumption was measured with in-
creasing amount of data items (16 bit values) using
the BER encoding mentioned earlier. With regards to
PDP context handling, two different approaches were
implemented. The first approach activates the PDP
context, sends the packet then deactivates the context.
This is closer to our data communication scenarios
when we send data packets only rarely. In order to
evaluate the cost of PDP context activation, the second
approach activates the PDP context once, sends all the
test packets then deactivates the context after all the
packets are sent. Table 1 shows the results for the first
approach while Table 2 shows the results for the sec-
ond approach using the GL865 module. It can be ob-
served that PDP context activation adds a constant but
not too significant power cost to the communication
scenario.
Conclusion is that data size/data format optimiza-
tion does matter when trying to lower power consump-
tion. To significantly increase power consumption,
Table 1: Power consumption of data communication,
PDP context activated/deactivated for each packet
Data items Packet size
(bytes)
Power
consumption
(mAs)
16 287 2370
32 511 2595
64 1981 2945
128 4157 3307
256 8509 3951
however, data sizes must be several times larger than
the baseline data size. Optimization of data sizes may
be more relevant for ensuring data transfer in case the
radio path between the base station and the sensor is
not very optimal and hence the bit error rate is higher
which is a common case for our agricultural sensors,
some of them deployed at remote locations with less
than optimal network coverage.
6.3. SMS bearer
Data may also be sent using short messsages, pop-
ularly called SMS. Binary SMS is often filtered by
operators so we employed Base64 encoding and sent
the ASN.1 BER content in textual format. Figure 8
shows the power consumption using the SMS bearer
with a 112 byte long data packet (which is actually 154
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
Fig. 7: SIM900 power consumption (initial registration and location update)
Table 2: Power consumption of data communication,
PDP context activated only once
Data items Packet size
(bytes)
Power
consumption
(mAs)
16 287 1987
32 511 2180
64 1981 2590
128 4157 3270
256 8509 3570
character long after Base64 encoding) and Figure 9
shows the sending of the same packet using GPRS. In-
tuitively, it seems that SMS requires much less power
and it is indeed the case: GPRS requires 2347 mAs
power while SMS needs only 247 mAs power. The
large difference is caused by the fact that SMS uses
signalling radio channels that are already allocated
when the module registered with the network while
GPRS has to allocate (and deallocate) additional radio
channel for the data transfer. SMS is therefore attrac-
tive due to its much lower power consumption require-
ment but quite frequently the pricing of the subscrip-
tion prevents using SMS extensively for data transfer.
6.4. Push bearer
One strong requirement for our remotely placed
sensors is manageability because physically accessing
the sensors’ location is not always feasible. Manage-
ment operations are usually initated by the manage-
ment server operator asynchronously, independently
of the sensor’s scheduled operations. This requires a
push bearer that can be used to instruct the sensor to
contact the management server.
If the sensor is not registered to the network, such
an operation is impossible. The management server
operator may have to wait for the sensor to contact the
server when the sensor sends in its scheduled batch
of data and may send its management commands in
the context of the sensor data sending session. While
this approach is attractive due to its low power con-
sumption, it makes life of management operators much
harder because configuration updates or other manage-
ment commands cannot be executed at any time, only
after the sensor contacts the server for scheduled data
sending. Also, in case of doubt (e.g. when an acci-
dent damaging the sensor is suspected), the sensor’s
health cannot be verified immediately which may pre-
vent timely maintenance operations. Management re-
quirements create a strong incentive to register the sen-
sor to the mobile network continuously.
If the GSM module is registered continuously,
short message service (SMS) may be used to send alert
to the sensor to connect to the management server for
management operations. As we have seen, SMS is
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
Fig. 8: Power consumption of the data transfer with SMS
Fig. 9: Power consumption of the data transfer with GPRS
very power-efficient and management operations are
infrequent enough so that SMS pricing is not so much
of an issue. Another option is to simulate the push
bearer using TCP.
TCP-based push bearer simulation relies on the
sensor to maintain a TCP connection to the manage-
ment server. When the server wants to send a man-
agement packet, it may use the duplex nature of TCP
streams to send the packet to the sensor. Timeout is-
sues, however, make this solution tricky to implement.
TCP timeouts on the sensor and on the server-side can
be controlled by the implementation but mobile and
backbone networks between the mobile network gate-
ways to the servers often employ Network Address
Translators (NATs) that remove IP address associa-
tions for TCP streams that look idle. The problem
was demonstrated with a test program implemented
on both the GL865 and SIM900 modules and a test
server application deployed on a cloud-based Win-
dows Server. The GSM modules attached to the mo-
bile network (Telenor Hungary), opened a TCP con-
nection to the server and left the connection idle. Af-
ter a timeout expired, a packet was sent from the server
to the GSM module. It was found that the maximum
safe timeout period was 2 hours which is consistent
with the recommendations in [7]. Longer timeout re-
sulted in the server and the GSM module to be silently
disconnected by some NAT on the network without ei-
ther of the communicating parties being aware of the
disconnection. The results were consistent with both
GSM modules, demonstrating that this behaviour is
the property of the network between the GSM mod-
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
ule and the cloud-based server. Without deeper inves-
tigation of the full network topology, it is hard to say
where the NAT was located that terminated the con-
nection.
Reliable implementation of the TCP-based push
bearer must use a heuristic algorithm [12], [9] to es-
timate the timeout between the GSM module and the
server by sending test packets with different timeouts.
The heuristic algorithm must also be prepared for the
fact that this timeout may also change dynamically,
due to changes in the network topology. Once that
timeout is known, a keepalive packet must be sent in
any direction over the TCP stream to prevent any NAT
that may be present between the GSM module and the
server to terminate the connection. This keepalive op-
eration has a power consumption cost.
Both modules are able to wake up from power-save
mode when an incoming data packet arrives on a TCP
connection that has been opened previously. The pro-
totype for this test relies again on the built-in TCP
stack in case of both modules. For the GL865, the
reception of one such packet costs 942 mAs. Using
2 hour timeout (hence 12 such packets per day), the
daily power consumption cost is about 3.1 mAh. The
SIM900 performs better in this test, the cost of one
keepalive packet was 570 mAs which means 1.9 mAh
for a day. This means that the power consumption cost
of maintaining one TCP connection is comparable to
the cost of the location update operations that keep
the module registered on the mobile network. For the
Telit GL865, such keepalive procedure increases the
daily power consumption by 22%. For the SIM900,
the increase is only 7% due to the higher baseline
power consumption of the module and the better TCP
packet reception power cost. It must be noted that the
GL865 also executed the application logic for this test
but the SIM900 acted only as a modem. This means
that additional power consumption cost must be calcu-
lated for the CPU that runs the application logic for the
SIM900.
TCP-based push bearer comes with other problems
on the server-side like keeping a large amount of TCP
connections open at the same time but these issues are
not discussed in this paper.
7. Conclusions
Directly connecting a remotely located, battery-
powered sensor to the GSM network comes with a set
of compromises. In our case, the power consumption
and manageability requirements were in direct conflict
with each other. From the power consumption point of
view, the best solution would be to attach the sensor to
the mobile network only for the duration of sending the
scheduled measurement data package. This would also
decrease the load on the mobile network infrastructure
in case of a large number of sensors. This approach
would make the sensors more complicated to manage,
however. In order to send a management operation, the
management server operator should wait until the sen-
sor connects back to the server for the scheduled data
sending operation. If this never happens (e.g. the sen-
sor is damaged, vandalized or misconfigured) then the
sensor’s deployment location must be visited which
may not be simple for a remotely placed sensor.
The compromise may be the power-saving mode
of the GSM modules. Both GSM modules we eval-
uated have such mode even though these features are
non-standard and are specific to the particular GSM
module. A daily consumption of 15-30 mAh means
80-160 days of operation with a low-cost 2400 mAh
battery pack. As special, high capacity batteries are
now commercially available, this operational time may
be increased dramatically.
Push bearer is required for asynchronous manage-
ment operations. SMS offers an attractive alternative.
TCP-based push bearer is possible to implement with
relatively minor increase of the power consumption
but is problematic to make reliable due to NAT issues
and limitations of the number of the TCP streams on
the server side.
We aim to proceed in this research project with
adding more sensors to the sensor station. Currently
we are investigating imaging type of sensors that gen-
erate much more data and we expect that these new
requirements lead to a better understanding of energy-
efficient sensor communication on cellular networks.
References
[1] GL865-QUAD. In Retrieved August 19 2014
from http:// www.telit.com/ products/ product-
service-selector/ product-service-selector/
show/product/ gl865-quad/.
[2] ITU-T X.690, ASN.1 encoding rules: Specifi-
cation of basic encoding rules (ber), canonical
encoding rules (cer) and distinguished encoding
rules (der).
[3] SIM900 specification. In Retrieved Au-
gust 19 2014 from http:// wm.sim.com/ down-
loaden.aspx? id=2972.
[4] R. Earl, P. N. Wheeler, B. S. Blackmore, and R. J.
Godwin. Precision farming – the management of
variability. In Landwards. Winter, 1997, pp. 18-
23, 1997.
[5] G. Él˝
o, K. Koppány, N. Kovács, J. Szabó, and
P. Szármes (editor). Precíciós gazdálkodás kock-
ázatmenedzsment. 2015.
Sensors & Transducers, Vol. 0, Issue 0, Month 2015, pp.
[6] R. S. Garrity, L. A. Vierling, and K. Bickford. A
simple filtered photodiode instrument for contin-
uous measurement of narrowband ndvi and pri
over vegetated canopies. In Agricultural and
Forest Meteorology, Vol. 150, Issue 3, 15 March
2010, pp. 489-496, 2010.
[7] S. Guha, K. Biswas, B. Ford, S. Sivakumar, and
P. Srisuresh. Nat behavioral requirements for
TCP. In RFC 5382, 2008.
[8] A. Hakiri, P. Berthou, A. Gokhale, D. Schmidt,
and T. Gayraud. Supporting end-to-end scala-
bility and real-time event dissemination in the
omg data distribution service over wide area net-
works. In Elsevier Journal of Systems Software
(JSS), 86(10):2574–2593, Oct. 2013., 2013.
[9] H. Haverinen, J. Siren, and P. Eronen. En-
ergy consumption of always-on applications in
WCDMA networks. In Vehicular Technology
Conference, 2007. VTC2007-Spring. IEEE 65th,
22-25 April 2007, Dublin, 2007.
[10] K. Kuladinithi, O. Bergmann, T. M. B. Pötsch,
and C. Görg. Implementation of CoAP and its
application in transport logistics. In Proceedings
of the Workshop on Extending the Internet to Low
power and Lossy Networks, 2011.
[11] G. Paller, P. Szármes, and G. Él˝
o. Efficient
power consumption strategies for stationary sen-
sors connected to gsm network. In SENSOR-
NETS 2015: Proceedings of the 4th International
Conference on Sensor Networks., At Angers,
France, Volume: pp. 63-68., editors: César
Benavente-Peces, Patrick Plainchault, Octavian
Postolache, 2015.
[12] R. Price and P. Tino. Adapting to NAT time-
out values in P2P overlay networks. In 2010
IEEE International Symposium on Parallel &
Distributed Processing, Workshops and Phd Fo-
rum (IPDPSW), Atlanta, GA, 2010.
[13] X. Su, J. Riekki, J. K. Nurminen, J. Nieminen,
and M. Koskimies. Adding semantics to inter-
net of things. In Concurrency Computat.: Pract.
Exper.. doi: 10.1002/cpe.3203, 2014.
[14] H. Sundmaeker, P. Guillemin, P. Friess, and
S. Woelffle. Vision and challenges for realising
the internet of things. In Cluster of European Re-
search Projects on the Internet of Things-CERP
IoT, 2010., 2010.
[15] M. Weiser. The Computer for the 21st cen-
tury. In Scientific American 265(3):66–75 (Jan-
uary 1991), 1991.
[16] M. Zorzi, A. Gluhak, S. Lange, and A. Bassi.
From today’s INTRAnet of things to a future
INTERnet of things: a wireless- and mobility-
related view. In Wireless Communications, IEEE
(Volume:17 , Issue: 6 ), December 2010, 2010.