Conference PaperPDF Available

Artificial Intelligence at the Edge in the Blockchain of Things

Authors:

Abstract and Figures

Traditional cloud-centric architectures for Internet-of-Things applications are being replaced by distributed approaches. The Edge and Fog computing paradigms crystallize the concept of moving computation towards the edge of the network, closer to where the data originates. This has important benefits in terms of energy efficiency, network load optimization and latency control. The combination of these paradigms with embedded artificial intelligence in edge devices, or Edge AI, enables further improvements. In turn, the development of blockchain technology and distributed architectures for peer-to-peer communication and trade allows for higher levels of security. This can have a significant impact on data-sensitive and mission-critical applications in the IoT. In this paper, we discuss the potential of an Edge AI capable system architecture for the Blockchain of Things. We show how this architecture can be utilized in health monitoring applications. Furthermore, by analyzing raw data directly at the edge layer, we inherently avoid the possibility of breaches of sensitive information, as raw data is never stored nor transferred outside of the local network.
Content may be subject to copyright.
Articial Intelligence at the Edge in the
Blockchain of Things
Tuan Nguyen Gia1[0000000298519868],
Anum Nawaz1,2[0000000211480084] ,
Jorge Peña Querata1[0000000330913217],
Hannu Tenhunen1[0000000319596513], and
Tomi Westerlund1[0000000217932694]
1Turku Intelligent Embedded and Robotic Systems (TIERS) Group
Department of Future Technologies
University of Turku, Finland
{jopequ, tunggi, hatenhu, tovewe}@utu.fi
https://tiers.fi
2Shanghai Key Laboratory of Intelligent Information Processing
School of Computer Science
Fudan University, China
nanum18@fudan.edu.cn
Abstract. Traditional cloud-centric architectures for Internet-of-Things
applications are being replaced by distributed approaches. The Edge and
Fog computing paradigms crystallize the concept of moving computation
towards the edge of the network, closer to where the data originates. This
has important benets in terms of energy eciency, network load opti-
mization and latency control. The combination of these paradigms with
embedded articial intelligence in edge devices, or Edge AI, enables fur-
ther improvements. In turn, the development of blockchain technology
and distributed architectures for peer-to-peer communication and trade
allows for higher levels of security. This can have a signicant impact on
data-sensitive and mission-critical applications in the IoT. In this paper,
we discuss the potential of an Edge AI capable system architecture for
the Blockchain of Things. We show how this architecture can be utilized
in health monitoring applications. Furthermore, by analyzing raw data
directly at the edge layer, we inherently avoid the possibility of breaches
of sensitive information, as raw data is never stored nor transferred out-
side of the local network.
Keywords: Blockchain; Edge Computing; AI; Edge AI; E-Health; U-
Health; IoT; Internet of Things; ECG Monitoring; ECG Feature Extrac-
tion; Ubiquitous Health; Ethereum;
1 Introduction
With an increasing ubiquity of connected devices penetrating smart homes,
smart cities, smart factories or smart farms, the Internet of Things (IoT) is
2 T. N. Gia et al.
generating vast amounts of data [1,2]. However, many challenges related to IoT
data ownership, security, privacy, and information sharing still remain [3–6].
The increasing integration of third-party services into IoT applications further
increases the risk of security vulnerabilities and cyber attacks [7]. Even with the-
state-of-the-art encryption methods, the IoT presents a non-negligible threat to
users’ privacy and personal data security [8]. While the IoT was born with the
boom in cloud computing, in recent years distributed computing approaches
are extending its potential [9–16]. The edge and fog computing paradigms aim
at migrating computational load towards the edge of the network. Data is pro-
cessed at the local network level or radio access point station and only important
information is transmitted over the network. For example, raw ECG data can
be processed at a smart gateway for extracting important ECG features such as
heart rate, P and T waves. Depending on the applications, raw data or processed
data is stored at distributed edge storage. Edge approaches allow for reduced
latency and more ecient use of both network and computational resources, but
they also raise additional security considerations and requirements [17].
Blockchain technology has seen increasing penetration in multiple technolog-
ical areas in the last decade [18], since its rst introduction as part of the Bitcoin
stack [19]. A blockchain platform can be seen a public and distributed digital
data ledger that allows nodes to record proof of integrity and is unalterable a
posteriori. Blockchain enables a decentralized manner of sharing data, and an
immutable record of transactions, among other benets. Compared to a central-
ized infrastructure, such as most cloud-based IoT Systems, blockchain technology
has the advantage of allowing end-users or devices to exchange information, data
and their assets directly without any intermediate third parties involved in the
process while securing data integrity [20]. With these advantages, blockchain
can be a suitable candidate to deal with some existing security challenges in
many applications [21]. For instance, blockchain can be leveraged as a trading
platform between data producers (i.e., the end-devices in an IoT system or the
edge gateway where sensor node data is being analyzed and processed), and data
consumers or (i.e., third-party applications or end-user applications) [22].
The integration of blockchain technology into the IoT has drawn growing
attention of the research community in recent years. Signicant eorts have
been devoted to propose secured approaches which utilize blockchain technology
to secure M2M transactions in the IoT [23]. An important part of the works
to date is focused on either secure access policies, such as the direct connection
between end-users and smart home appliances [24], or secure machine-to-machine
communication [25]. Although these approaches can indeed provide high levels
of security to IoT platforms [26], their integration within edge-assisted remote
and real-time monitoring applications is not deeply investigated in those works.
In this paper, we present an architecture for the Blockchain of Things that
integrates articial intelligence at the edge (Edge AI) algorithms for ecient
and secure information management and privacy protection in healthcare ap-
plications. The presented system architecture extends our previous work [27],
and it is illustrated in Fig. 1. We also discuss further the potential of this ap-
Articial Intelligence at the Edge in the Blockchain of Things 3
Mo
SENSORLAYER
EDGELAYER
CLOUDLAYER
Data Features
BLOCKCHAIN
EDGEAI
STORAGE THIRDPARTY
SERVICES Encrypted data storage
Third-party services
(data buyers)
Data acquisition
Wireless sensor networks
Bio-signals analysis
Feature extraction
Acess brokerage
Historical records
Metadata
Raw Data
Encrypted Features Metadata
END-USER
APPLICATION
Fig. 1. Proposed System Architecture
plication in various elds. The proposed architecture secures IoT data integrity
with a distributed platform based on the Ethereum blockchain and utilizes Edge
AI for computational ooading at the fog and edge layer. Integrating fog and
edge computing creates new opportunities for enhanced peer-to-peer security
and authorized access [28, 29].
The rest of this paper is organized as follows. Section 2 overviews previous
works in the use of blockchain technology in the IoT, the Edge AI paradigm
and the use of blockchain for healthcare applications. Section 3 then introduces
the system architecture, and outlines the benets of integrating Edge AI with
blockchain for dierent applications in the IoT. Section 4 presents the exper-
imental data and results. Finally, Section 5 concludes the work and lays out
future work directions.
2 Related Work
In the healthcare IoT domain, it is often recommended that patients should
have the ability to access their own health data. Nonetheless, the data should
4 T. N. Gia et al.
be consistent, protected and unaltered over time by any third parties or patients
themselves [30, 31]. Therefore, it is necessary to have a high level of security
methods which ensure that data transmitted over a network is secure and avail-
able to authorized parties, in addition to having an integrity check that ensures
immutability of the data. Many eorts have been devoted to propose blockchain-
based methods to improve security, transaction speed, and avoid fraud control
in healthcare.
In [32], the authors introduced and discussed dierent access policies to pro-
tect the privacy of private patient’s data. In addition, the authors implemented
deep learning algorithms to extract useful information from private raw data. Al-
though the proposed method and its algorithms focus on healthcare applications,
they can be applied in dierent scenarios including cases in larger perspectives.
In [33], the authors presented a blockchain-based approach for sharing pa-
tient data within a network. In addition, the authors introduced a consensus
algorithm for enabling data interoperability. Dierent measurements of security
on blockchain were carried out and the authors claimed that the blockchain-
based method is a promising solution for avoiding or overcoming problems in
sharing private health data.
In [34], the authors introduced a blockchain-based method for proering
a proof of predened endpoints in clinical trials. They claimed that applying
blockchain methods can provide a high level of reliability while keeping costs
low.
In [35], the authors introduced a framework which has a modied traditional
blockchain method for suiting to IoT applications. The proposed method is suit-
able for resource-constrained devices while it maintains a high level of privacy
and security. The framework ensures that transactions over a blockchain network
are more anonymous and secure.
In [36], Simic et al. showed that it is feasible to apply blockchain into health-
care IoT systems to protect data transmitted over a network. The authors have
examined several possibilities of utilizing smart contracts for healthcare IoT
systems. They claimed that a combination of blockchain and IoT can benet
dierent distributed applications.
In [37], Pham et al. presented a remote health monitoring system utiliz-
ing blockchain. In this system, bio-signal sensor nodes collect and lter patient
data. The useful information extracted from the collected data is written into
blockchain. In case of abnormalities, the extracted information is written im-
mediately to blockchain and a push notication is triggered to inform medical
doctors.
3 Protecting Data Privacy with the Ethereum Blockchain
As compared to the original blockchain platform developed for the bitcoin by
Satoshi Nakamoto [19], the Ethereum platform provides the Ethereum Virtual
Machine (EVM) which is fully autonomous in terms of its system execution by
using smart contracts. Smart contracts are scripts with predened terms and
Articial Intelligence at the Edge in the Blockchain of Things 5
conditions for system transactions. This peer-to-peer (P2P) distributed ledger
relies on its miner nodes. Miner nodes act as validator nodes for every new
transaction block, which are created within certain time intervals. In general, a
single transaction block is a combination of a header block and a data block. The
data block stores the hash of the processed and analyzed data, while the header
contains the hash of previous and current blocks, metadata, timestamp and a
short characterization of the data. If another user or a third party service wants
to access or exchange data, the header data characterization description can be
utilized to see the details of the data block before the transaction is carried out.
The data itself is stored encrypted in a cloud storage solution or in the device
itself if the capacity is enough.
The proposed system architecture is illustrated in Figure 1 and it consists
of three layers. First, the data generation layer, which consists of sensors and
actuators without any computational layer. These sensors and actuators depend
on mining nodes which will collect data from these devices. Sensor and actuators
merely communicate with one miner node which can be used as a gateway to
transfer their data to other gateways or cloud servers. Bluetooth low energy
(BLE) or Wi-Fi is often used this layer. BLE uses less energy whilst Wi-Fi
can transmit a larger data packet size and oer higher bandwidth. Second, the
network layer, where P2P networking is used in this private ethereum network for
communication and data transfer. The distributed ledger topologies are dened
on this layer. Dierent topologies like side chains, shard chains, o-chains can be
used to handle scarce computing-devices issues and scalability. Smart contracts
or scripts run to handle all the processes in a network. Finally, the third layer
is the application layer. Smart applications of the IoT consists of a wide range
of use-cases like smart homes, smart industries, digital medical and many more.
To access these systems, end-users, third parties or control centers need to join
the network rst and then request data via the ethereum network.
3.1 Application Areas
In this section, we give an overview of potential applications for the proposed
architecture. We outline the benets and trade-os of integrating our proposed
platform in dierent IoT domains. We cover the areas of smart homes, smart
cities, industrial applications, connected vehicles with vehicle-to-vehicle (V2V)
and vehicle-to-everything (V2X) communication, and ubiquitous health.
A common problem of IoT devices in all application areas is their security
vulnerabilities in terms of (1) third-party access and control, (2) unauthorized
use of data, and (3) leakage of raw data. While previous works have studied
the problem of protecting access to these devices by integrating blockchain tech-
nology extensively, we will focus on the benets of the proposed for ensuring
that data from sensor nodes is only accessed by authorized third parties, and
that raw data is never made available to these parties through processing at
the local network level and before inclusion in the blockchain. Furthermore, the
blockchain provides an immutable record of all data requests from third parties.
6 T. N. Gia et al.
Smart Home IoT Providers
In the smart home domain, an increasing number of industrial players are in-
troducing a variety of commodities, from voice assistants to smart fridges. How-
ever, these are not exempt from security vulnerabilities [38], and many of them
suppose a serious threat to privacy in users’ homes. Previous works have been
focusing on using blockchain for securing access and control of smart home appli-
ances. This can signicantly reduce the risk of having a spy inside our homes [39].
However, while communication between third-party services is secured, the use
of data gathered by these devices is still being controlled by third-parties.
A smart gateway, which can replace a traditional home Wi-Fi router, serves
as a bridge between sensor nodes and cloud storage or third-party applications,
and at the same time the smart gateway can be utilized to deploy deep learn-
ing analysis and other AI processing which cannot run directly on resource-
constrained devices. Moreover, because the processed data is only stored locally
or encrypted in cloud storage, all data access requests are stored in the blockchain
and therefore access to data is not managed by an external party but by smart
gateways directly.
Cybersecurity in Open Smart Cities
The concept of Smart City mostly relies on IoT. A city is considered to be
smart when it uses large amounts of IoT sensor data to eciently improve the
management of its assets and resources [40]. Another key aspect of smart cities
is openness. By making public all or part of the IoT data that is gathered, city
administrators can engage citizens, local business and large enterprises equally to
develop new products and services based on the data. This benets both the city
management team and the involved parties, with a positive eect on the city’s
economy. In this case, however, it is essential to have a proper methodology for
both sharing data with third parties and ensuring that public datasets are not
misused.
With the implementation of the proposed architecture, administrators can
have full control and monitor the access of third parties of this data. Moreover,
transaction fees and data prices in the ethereum blockchain within the proposed
solution can be used to naturally control the amount of data that each external
user is accessing. In summary, our proposed solution not only provides a secure
and safe way of distributing IoT data gathered around the city to external users
or developers, but it also provides a base for edge computing and local network
analysis and processing. By managing to which level the data is processed in edge
gateways, which information is processed in the gateways, and which information
or raw data is available to external applications.
Modular Smart Factories in Industry 4.0
The fourth industrial revolution, or Industry 4.0, has promised to develop more
agile, modular and smart manufacturing environments where traditional pro-
duction lines are replaced by automated and intelligent lines in which individual
products can be customized on the y [41]. The process towards Industry 4.0
Articial Intelligence at the Edge in the Blockchain of Things 7
requires the integration of the IoT in industrial environments and the installa-
tion of IoT sensor suites and actuators. This will allow managers to gather vast
amounts of data and be able to adjust the manufacturing process dynamically
to improve its eciency.
Although autonomous machines and robots are heavily used in smart fac-
tories, they cannot replace humans completely. In some parts of a production
chain, tight cooperation between machines and humans is unavoidable. There-
fore, it is required that smart factories must guarantee a high safety level for
humans working with autonomous machines. A method for enhancing situational
awareness via intercommunication between everything can be applied in smart
factories to address the target. In detail, a machine such as co-robot communi-
cates and obtains useful information from other machines or even humans. For
instance, a machine in a room can get a position and gesture of engineering
who is walking in an adjacent room and is likely to come close to it. Based on
both the received information and the data collected by the machine itself, it
is able to forecast potential safety-critical situations and react in real-time to
avoid accidents. In such a system, latency and security are essential because a
piece of incorrect information provided by the third party or delayed information
can cause a serious consequence. Therefore, smart factories need an advanced
secured architecture which can guarantee a trusted intercommunication between
machines and human with low latency.
Internet of Vehicles, V2V, and V2x
Nowadays, the number of connected vehicles is increasing signicantly due to
their benets such as improving energy eciency, reducing travelling time, or
avoiding car accidents. The concept of connected vehicles often refers to a num-
ber of communication protocols used to connect the driver with other objects.
For instance, communication in connected vehicles can be categorized into a ve-
hicle to infrastructure (V2I), vehicle to vehicle (V2V), vehicle to Cloud (V2C),
vehicle to pedestrian (V2P) and, in general, vehicle to everything (V2X) com-
munication. In these scenarios, security is essential because incorrect or modied
data introduced in the system by untrusted third parties can cause serious con-
sequences such as a car accident or even death. Conventional security methods
which need a central control system may not be completely suitable for some of
the connected vehicles because those methods can cause an increase in communi-
cation latency. In such vehicle systems, real-time data and reaction are required.
The proposed solution is a potential candidate for such real-time connected ve-
hicle systems as it can provide high levels of security while the latency does
not increase. With the proposed architecture, data related to other vehicles on a
street can be exchanged directly with a connected vehicle through edge gateways
in the near infrastructure. Moreover, the Edge AI opens multiple possibilities for
computational ooading [42, 43]. The benets of our proposed architecture are
in the control of the use of private vehicle data by third parties. In the V2X
scenario, these can be other vehicles (V2V) or infrastructure around the road
(V2I).
8 T. N. Gia et al.
Fig. 2. Sensor Node
Ubiquitous Health
Privacy and private health data must be carefully protected because leaked in-
formation can cause serious consequences. For instance, the leakage information
such as health status can be used for hijacking purposes or spreading false rumors
which causes money and mental damages. It is required that remote and real-
time health monitoring systems must ensure a high level of security. Nonetheless,
there are still many challenges of security issues in these systems. Blockchain can
play an important role in improving a security level in these systems [27]. By
combining blockchain with articial intelligence at the edge of the network, a
system can provide end-to-end protection to users’ privacy. First, sensitive raw
data is processed at the local network level, and therefore the risk of raw data
being leaked is eliminated. With the blockchain utilized to manage an access to
processed data and features, end-users can have full control over their data while
allowing third-party applications to have access only the information that has
been processed already.
4 Experiment and Results
In order to test the feasibility of the proposed architecture, and the possibilities
for deployment and real-time execution, we have targeted a use case of ECG
feature extraction and arrhythmia detection with convolutional neural networks
(CNN). We have used a complete remote health monitoring IoT-based system
utilizing blockchain and edge/fog computing. However, in this paper, we just
focuses on edge gateways which have been used for deploying the advanced
algorithms such as ECG feature extraction and arrhythmia detection with CNN.
Other parts of the system have been discussed in detail in our previous papers
[12, 44, 45].
4.1 Sensor Node
In this paper, ECG is collected by our multi-channel ECG sensor node which
will be described in detail in another work. The sensor node is able to collect
Articial Intelligence at the Edge in the Blockchain of Things 9
Table 1. Loading time of the dierent Arrythmia classication requirements
Execution Time
Loading Numerical Libraries 960 ms
Loading Tensorow and Keras 1478 ms
Loading Trained Model 6683 ms
ECG Feature Extraction 150 ms
Arrythmia Classication 849 ms
Table 2. Blockchain Transaction Average Execution Times
Average Execution Time
Ethereum Transaction Request 17 ms
New data block creation 10 s
16-channel ECG signals with high resolutions (i.e., each channel can collect from
125 samples/s to 1000 samples/s). Then, depending on the requirements of each
application, the data can be pre-processed and kept intact before being sent to a
smart Edge gateway via BLE or Wi-Fi. In this paper, raw ECG data is collected
from the sensor node with a sampling rate of 250 samples/s per channel and sent
to a smart Edge gateway via Wi-Fi. The collected data is not processed at the
sensor node because it is dicult or even not feasible to run heavy computation
methods (e.g., ECG feature extraction based on wavelet transform) at the sensor
node [46–48]. Instead the data will be processed at the Edge gateway which is
capable of running heavy computations while fullling latency requirements [49].
The data rate of 250 samples/s can fulll the requirements of common ECG data
quality standards [50]. In general, BLE is preferred over Wi-Fi because BLE
consumes much less energy than Wi-Fi for a similar transmissions. However,
BLE cannot be chosen for this case because BLE cannot support this large data
rate (i.e., about 3 Mbps for up to 12 channels in each sensor node) [51].
4.2 Gateway
The edge gateways used in our system are Raspberry Pi 3B+ single-board com-
puter (1.4 GHz quad-core processor, 1 GB SRAM, BLE, Wi-Fi). The operating
system running at the gateway is Ubuntu. The gateway is able to store dierent
data and information such as parameters used for algorithms and temporary
health data. The parameters are often kept intact and they are only modied by
a system administrator. The gateway can reserve 20 GB for storing temporary
health data. Raw data is not stored but only the extracted features. If the stor-
age is near its full capacity, then part of the data is encrypted and transferred
to cloud-based storage solutions. All the services (e.g., ECG feature extraction)
run on the gateway. In our experiments, the Pi runs ECG feature extraction
adapted from [52], while a deep learning based arrhythmia classication model
10 T. N. Gia et al.
(a) Raw data over 10 seconds
(b) Extracted cycle template
8
6
7
(c) Heart rate over 10 seconds
Fig. 3. Results of the data analysis at the edge gateway.
adapted from [53] is deployed in the as well. The Pi was used in order to prove
the viability and eectiveness of the proposed architecture. If more computa-
tional resources are required, then this can be replaced by any other hardware
capable of running Ubuntu.
4.3 Performance
To initialize the system, a private ethereum network is created, generating au-
thority and transaction accounts. The rst step is to congure a new genesis
le to build the rst block of the custom ethereum network. Smart contracts
were written in solidity and tested by using Remix IDE. We have analyzed the
execution times of the feature extraction, arrhythmia detection and blockchain
requests in order to assess the possibilities of real-time operation. The execu-
tion times of the dierent processes are shown in Tables 1 and 2. The feature
extraction and arrhythmia classication processes deployed in this use case are
single-threaded and therefore executed within a single core. As the Raspberry
Pi 3B+ has 4 cores, it is possible to concurrently execute the analysis of two
sensor nodes in parallel together with other background processes. The analysis
of ECG data is made in batches of 10 seconds, where an average ECG cycle
template is extracted and the heart rate and other features are calculated. An
Articial Intelligence at the Edge in the Blockchain of Things 11
example of the raw and processed data is shown in Fig. 3. Then, the template
is utilized for arrhythmia classication.
The loading times required for loading numerical libraries, the deep learning
libraries Tensorow and Keras, and the trained model are shown in Table 1.
Taking these into account, we deploy the model in the edge gateway in a way
that the required libraries and the deep learning model are only loaded every
time the gateway is rebooted. Transaction requests time in the ethereum network
was 17 ms as average while using public Wi-Fi. Miner nodes take 10 s as average
to create a new data block.
In summary, since the analysis is carried out every 10 s, a single Raspberry
Pi 3B+ board is able to handle multiple sensor nodes connected via Wi-Fi or
Bluetooth. We can safely assume that around 8 sensor nodes can be handled
in real-time without reaching the maximum level of performance and therefore
allowing for uncertainties in the measurements.
After data processing, the extracted features are encrypted with AES-256
[54,55] and stored in a third party storage solution. A custom distributed storage
solution can be employed instead if tighter control of the data storage is required.
Then, metadata including device ID and type of data are stored in the blockchain
through the execution of a series of smart contracts.
5 Conclusion and Future Work
We have utilized a blockchain-based architecture for managing data security and
integrity in IoT applications, and improved it by integrating Edge AI techniques
to enhance the applications’ security and protect users’ privacy further. This is
of particular interest for mission-critical and data-sensitive applications such as
health monitoring applications in the IoT. We have implemented our proposed
approach using ECG sensor nodes and a Raspberry Pi Model 3B+ as an edge
gateway. The gateway ran a full ethereum node and processed ECG data in real-
time with feature extraction and arrhythmia detection algorithms deployed. We
show that real-time computation with arrhythmia classication is possible with
multiple nodes, and the analysis part utilizes more computation resources than
a typical ethereum deployment.
In future work, we will further integrate how the AI algorithms are executed
together with the smart contracts in an ethereum network. In addition, we will
extend the current system to a larger number of applications in the domain of
ubiquitous health monitoring and others.
References
1. A. Al-Fuqaha et al. Internet of things: A survey on enabling technologies, proto-
cols, and applications. IEEE Communications Surveys Tutorials, 17(4):2347–2376,
Fourthquarter 2015.
2. T. N. Gia et al. Edge ai in smart farming iot: Cnns at the edge and fog computing
with lora. In IEEE AFRICON-2019, 2019.
12 T. N. Gia et al.
3. S.R. Moosavi et al. Session resumption-based end-to-end security for healthcare
internet-of-things. In 2015 IEEE CIT, pages 581–588. IEEE, 2015.
4. J. Gubbi et al. Internet of things (iot): A vision, architectural elements, and future
directions. Future Generation Computer Systems, 29(7):1645 – 1660, 2013.
5. S. R. Moosavi et al. Sea: a secure and ecient authentication and authorization
architecture for iot-based healthcare using smart gateways. Procedia Computer
Science, 52:452–459, 2015.
6. S. R. Moosavi et al. End-to-end security scheme for mobility enabled healthcare
internet of things. Future Generation Computer Systems, 64:108–124, 2016.
7. E. Fernandes et al. Security analysis of emerging smart home applications. In 2016
IEEE Symposium on Security and Privacy (SP), pages 636–654, May 2016.
8. N. Apthorpe, D. Reisman, and N. Feamster. A smart home is no castle: Privacy
vulnerabilities of encrypted iot trac. arXiv preprint arXiv:1705.06805, 2017.
9. M. Ali et al. Intelligent autonomous elderly patient home monitoring system. In
ICC 2019-2019 IEEE International Conference on Communications (ICC), pages
1–6. IEEE, 2019.
10. T. N. Gia et al. Edge ai in smart farming iot: Cnns at the edge and fog computing
with lora. 2019.
11. A. V. Dastjerdi, and R. Buyya. Fog computing: Helping the internet of things
realize its potential. Computer, 49(8):112–116, 2016.
12. T. N. Gia et al. Energy ecient fog-assisted iot system for monitoring diabetic pa-
tients with cardiovascular disease. Future Generation Computer Systems, 93:198–
211, 2019.
13. M. Ali et al. Autonomous patient/home health monitoring powered by energy
harvesting. In GLOBECOM 2017-2017 IEEE Global Communications Conference,
pages 1–7. IEEE, 2017.
14. V. K. Sarker et al. A survey on lora for iot: Integrating edge computing. In 2019
Fourth International Conference on Fog and Mobile Edge Computing (FMEC),
pages 295–300. IEEE, 2019.
15. J. P. Queralta et al. Edge-ai in lora-based health monitoring: Fall detection system
with fog computing and lstm recurrent neural networks. In 2019 42nd International
Conference on Telecommunications and Signal Processing (TSP), pages 601–604.
IEEE, 2019.
16. A. Metwaly et al. Edge computing with embedded ai: Thermal image analysis for
occupancy estimation in intelligent buildings. In INTelligent Embedded Systems
Architectures and Applications, INTESA@ESWEEK 2019. ACM, 2019.
17. R. Roman, J. Lopez, and M. Mambo. Mobile edge computing, fog et al.: A sur-
vey and analysis of security threats and challenges. Future Generation Computer
Systems, 78:680–698, 2018.
18. M. Conoscenti, A. Vetrò, and J. C. De Martin. Blockchain for the internet of things:
A systematic literature review. In 2016 IEEE/ACS 13th International Conference
of Computer Systems and Applications (AICCSA), pages 1–6, Nov 2016.
19. S. Nakamoto. Bitcoin: A peer-to-peer electronic cash system. White Paper, 2008.
20. H. Shafagh et al. Towards blockchain-based auditable storage and sharing of iot
data. In Proceedings of the 2017 on Cloud Computing Security Workshop, CCSW
’17, pages 45–50, New York, NY, USA, 2017. ACM.
21. S. Huh, S. Cho, and S. Kim. Managing iot devices using blockchain plat-
form. In Advanced Communication Technology (ICACT), 2017 19th International
Conference on, pages 464–467. IEEE, 2017.
22. O. Novo. Blockchain meets iot: An architecture for scalable access management in
iot. IEEE Internet of Things Journal, 5(2):1184–1195, April 2018.
Articial Intelligence at the Edge in the Blockchain of Things 13
23. B. Tang et al. A hierarchical distributed fog computing architecture for big data
analysis in smart cities. In Proceedings of the ASE BigData & SocialInformatics
2015, page 28. ACM, 2015.
24. A. Dorri et al. Blockchain for iot security and privacy: The case study of a
smart home. In Pervasive Computing and Communications Workshops (PerCom
Workshops), 2017 IEEE International Conference on, pages 618–623. IEEE, 2017.
25. K. Christidis and M. Devetsikiotis. Blockchains and smart contracts for the internet
of things. IEEE Access, 4, 2016.
26. N. Kshetri. Can blockchain strengthen the internet of things? IT Professional,
19(4):68–72, 2017.
27. A. Nawaz et al. Edge AI and Blockchain for Privacy-Critical and Data-Sensitive
Applications. In The 12th International Conference on Mobile Computing and
Ubiquitous Networking (ICMU), 2019.
28. B. Ndibanje, H-J. Lee, and S-G. Lee. Security analysis and improvements of au-
thentication and access control in the internet of things. Sensors, 14(8):14786–
14805, 2014.
29. A. Bahga, and V. Madisetti. Internet of Things: A hands-on approach. Vpt, 2014.
30. M. Li et al. Securing personal health records in cloud computing: Patient-centric
and ne-grained data access control in multi-owner settings. In SecureComm, pages
89–106. Springer, 2010.
31. K. D. Mandl et al. Public standards and patients’ control: how to keepelectronic
medical records accessible but private. Bmj, 322(7281):283–287, 2001.
32. P. Mamoshina et al. Converging blockchain and next-generation articial intelli-
gence technologies to decentralize and accelerate biomedical research and health-
care. Oncotarget, 9(5):5665, 2018.
33. K. Peterson et al. A blockchain-based approach to health information exchange
networks. In Proc. NIST Workshop Blockchain Healthcare, volume 1, pages 1–10,
2016.
34. G. Irving and J. Holden. How blockchain-timestamped protocols could improve
the trustworthiness of medical science. F1000Research, 5, 2016.
35. A. D. Dwivedi et al. A decentralized privacy-preserving healthcare blockchain for
iot. Sensors, 19(2):326, 2019.
36. M. Simić et al. A case study iot and blockchain powered healthcare. In International
Conference on Engineering and Technology (ICET-2017). June, 2017.
37. H. L. Pham, T. H. Tran, and Y. Nakashima. A secure remote healthcare system
for hospital using blockchain smart contract. In 2018 IEEE Globecom Workshops
(GC Wkshps), pages 1–6. IEEE, 2018.
38. N. Apthorpe et al. Spying on the smart home: Privacy attacks and defenses on
encrypted iot trac. arXiv preprint arXiv:1708.05044, 2017.
39. G. Hernandez et al. Smart nest thermostat: A smart spy in your home. Black Hat
USA, pages 1–8, 2014.
40. V. Albino, U. Berardi, and R. M. Dangelico. Smart cities: Denitions, dimensions,
performance, and initiatives. Journal of urban technology, 22(1):3–21, 2015.
41. H. Lasi et al. Industry 4.0. Business & information systems engineering, 6(4):239–
242, 2014.
42. L. Qingqing et al. Edge Computing for Mobile Robots: Multi-Robot Feature-Based
Lidar Odometry with FPGAs. In The 12th International Conference on Mobile
Computing and Ubiquitous Networking (ICMU), 2019.
43. L. Qingqing et al. Visual Odometry Ooading in Internet of Vehicles with Com-
pression at the Edge of the Network. In The 12th International Conference on
Mobile Computing and Ubiquitous Networking (ICMU), 2019.
14 T. N. Gia et al.
44. T. N. Gia et al. Fog computing approach for mobility support in internet-of-things
systems. IEEE Access, 6:36064–36082, 2018.
45. M. Jiang et al. Iot-based remote facial expression monitoring system with semg
signal. In 2016 IEEE Sensors Applications Symposium (SAS), pages 1–6. IEEE,
2016.
46. T. N. Gia et al. Fog computing in healthcare internet of things: A case study on
ecg feature extraction. In 2015 IEEE CIT, pages 356–363. IEEE, 2015.
47. A. P-Enriquez and V. Ponomaryov. Feature extraction based on wavelet trans-
form using ecg signal. In 2013 International Kharkov Symposium on Physics and
Engineering of Microwaves, Millimeter and Submillimeter Waves, pages 632–634.
IEEE, 2013.
48. T. N. Gia et al. Fog computing in body sensor networks: An energy ecient
approach. In Proc. IEEE Int. Body Sensor Netw. Conf.(BSN), pages 1–7, 2015.
49. T. N. Gia et al. Customizing 6lowpan networks towards internet-of-things based
ubiquitous healthcare systems. In 2014 Norchip, pages 1–6. IEEE, 2014.
50. C. Steinberg et al. A novel wearable device for continuous ambulatory ecg record-
ing: Proof of concept and assessment of signal quality. Biosensors, 9(1):17, 2019.
51. V. K. Sarker et al. Portable multipurpose bio-signal acquisition and wireless
streaming device for wearables. In 2017 IEEE Sensors Applications Symposium
(SAS), pages 1–6. IEEE, 2017.
52. C. Carreiras et al. BioSPPy: Biosignal processing in Python, 2015–. [Online;
accessed Aug. 2019].
53. T. J. Jun et al. Ecg arrhythmia classication using a 2-d convolutional neural
network. arXiv preprint arXiv:1804.06812, 2018.
54. I. B. Dhaou et al. Low-latency hardware architecture for cipher-based message au-
thentication code. In 2017 IEEE International Symposium on Circuits and Systems
(ISCAS), pages 1–4. IEEE, 2017.
55. T. N. Gia et al. Low-cost fog-assisted health-care iot system with energy-ecient
sensor nodes. In 2017 13th International Wireless Communications and Mobile
Computing Conference (IWCMC), pages 1765–1770. IEEE, 2017.
... By participating in a blockchain network, all the participants with permission can access the same information simultaneously. Due to blockchain's merit, blockchain is applied in many applications such as healthcare, cryptocurrency exchange, and trading [12]- [15]. ...
Conference Paper
Full-text available
With the development in information and communications technology (ICT) and drones such as Internet-of-Things (IoT), edge computing, image processing, and autonomous drones, solutions supporting search and rescue (SAR) missions can be developed with more intelligent capabilities. In most of the drone and unmanned aerial vehicle (UAV) based systems supporting SAR missions, several drones deployed in different areas acquire images and videos that are sent to a ground control station (GCS) for processing and detecting a missing person. Although this offers many advantages, such as easy management and deployment, the approach still has many limitations. For example, when a connection between a drone and a GCS has some problems, the quality of service cannot be maintained. Many drone and UAV-based systems do not support flexibility, transparency, security, and traceability. In this paper, we propose a novel Internet-of-Drones (IoD) architecture using blockchain technology. We implement the proposed architecture with different drones, edge servers, and a Hyperledger blockchain network. The proof-of-concept design demonstrates that the proposed architecture can offer high-level services such as prolonging the operating time of a drone, improving the capability of detecting humans accurately, and a high level of transparency, traceability, and security.
... The healthcare system is a vast market [13] and it also faces scalability issues [8,[14][15][16][17]. Besides, with the advancement of Artificial Intelligence [18] and Machine Learning [15], patients' sensitive data become much more valuable to researchers, doctors, pharmaceutical companies [19], and hence to hackers [20]. Therefore, highly effective encryption scheme and good authorization framework are required to preserve the identity and privacy of the data and manage access control of varying degrees [47]. ...
Article
Full-text available
Personalization in Health Care services has gained popularity in recent years with the advancement and rapid deployment of wearable devices, smartphones with powerful sensors, Internet of Things, and Telemedicine solutions. It has now become necessary that the patient keeps the ownership of the personal medical data, securely organizes and shares them with the remaining anonymous in the system. In this paper, we propose a blockchain, InterPlanetary File System (IPFS), and encryption based highly secure network framework responsible for medical data input, transmission, distribution, analysis, and management among patients, doctors, and remote servers resourced with automatic report generation. To make the system highly scalable and reduce the cost of each transaction in Blockchain incurred by cryptocurrency, BigchainDB and IPFS have been utilised which are P2P networks and do not require cryptocurrency for transactions. We have also implemented a divisible token-based payment system where patients transfer ownership of certain amounts of tokens to doctors for getting feedback on reports. Patients can input data, view transactions, generate reports, request for doctor’s feedback, and manage payments. Doctors, registered by the system, may decide to access reports, provide feedback, and check their payment status. Users (Doctor, Patients) will remain completely anonymous to each other and their activity and data will be encrypted and immutable in the proposed system. A smartphone-based application has been developed as a demonstration that can record audio representing phonocardiography (PCG) from patients, send it to the remote server for report generation and to a doctor for receiving feedback while maintaining data integrity. The remote server can also be equipped with AI based automated diagnosis and thus the proposed framework can further be scaled in the near future. We evaluated the performance of our proposed framework through BigchainDB-based Byzantine Fault Tolerance (BFT) testbed that confirms a nearly 20 fold reduction of execution and validation time for 2000 concurrent transactions, and higher throughput compared to the existing literature.
... Moreover, safety-critical applications can benefit from the enhanced security that blockchains and other DLTs provide. These include the automotive sector with vehicle to everything communication routed at the edge [49,50], and the healthcare sector [34,51,52]. ...
Preprint
Mobile edge computing (MEC) and next-generation mobile networks are set to disrupt the way intelligent and autonomous systems are interconnected. This will have an effect on a wide range of domains, from the Internet of Things to autonomous mobile robots. The integration of such a variety of MEC services in a inherently distributed architecture requires a robust system for managing hardware resources, balancing the network load and securing the distributed applications. Blockchain technology has emerged a solution for managing MEC services, with consensus protocols and data integrity checks that enable transparent and efficient distributed decision-making. In addition to transparency, the benefits from a security point of view are evident. Nonetheless, blockchain technology faces significant challenges in terms of scalability. In this chapter, we review existing consensus protocols and scalability techniques in both well-established and next-generation blockchain architectures. From this, we evaluate the most suitable solutions for managing MEC services and discuss the benefits and drawbacks of the available alternatives.
... Nonetheless, streaming data to the cloud has the potential to significantly reduce the overall system reliability with uncontrolled latency or unstable network connection [18,19]. We extend the recent trend in the IoT towards more decentralized network architectures with the fog and edge computing paradigms [20][21][22]. Edge computing crystallizes the idea of keeping the data processing as close as possible to where the data originates. With this approach, raw data is processed at the local network level instead of the cloud, decreasing the latency and optimizing the network load [23]. ...
Conference Paper
Full-text available
Fleets of autonomous mobile robots are becoming ubiquitous in industrial environments such as logistic warehouses. This ubiquity has led in the Internet of Things field towards more distributed net-work architectures, which have crystallized under the rising edge and fog computing paradigms. In this paper, we propose the combi-nation of an edge computing approach with computational offload-ing for mobile robot navigation. As smaller and relatively simpler robots become more capable, their penetration in different domains rises. These large multi-robot systems are often characterized by constrained computational and sensing resources. An efficient com-putational offloading scheme has the potential to bring multiple operational enhancements. However, with the most cost-effective autonomous navigation method being visual-inertial odometry,streaming high-quality images can induce latency increments with a consequent negative impact on operational performance. In this paper, we analyze the impact that image quality and compression have on the state-of-the-art on visual inertial odometry. Our results indicate that over one order of magnitude in image size and network bandwidth can be reduced without compromising the accuracy of the odometry methods even in challenging environments.This opens the door to further optimization by dynamically assessing the trade-off between image quality, network load, latency and performance of the visual-inertial odometry and localization accuracy.
... Nonetheless, streaming data to the cloud has the potential to significantly reduce the overall system reliability with uncontrolled latency or unstable network connection [12,33]. We extend the recent trend in the IoT towards more decentralized network architectures with the fog and edge computing paradigms [2,3,22]. Edge computing crystallizes the idea of keeping the data processing as close as possible to where the data originates. ...
Conference Paper
Full-text available
Fleets of autonomous mobile robots are becoming ubiquitous in industrial environments such as logistic warehouses. This ubiquity has led in the Internet of Things field towards more distributed network architectures, which have crystallized under the rising edge and fog computing paradigms. In this paper, we propose the combination of an edge computing approach with computational offload-ing for mobile robot navigation. As smaller and relatively simpler robots become more capable, their penetration in different domains rises. These large multi-robot systems are often characterized by constrained computational and sensing resources. An efficient computational offloading scheme has the potential to bring multiple operational enhancements. However, with the most cost-effective autonomous navigation method being visual-inertial odometry, streaming high-quality images can induce latency increments with a consequent negative impact on operational performance. In this paper, we analyze the impact that image quality and compression have on the state-of-the-art on visual inertial odometry. Our results indicate that over one order of magnitude in image size and network bandwidth can be reduced without compromising the accuracy of the odometry methods even in challenging environments .This opens the door to further optimization by dynamically assessing the trade-off between image quality, network load, latency and performance of the visual-inertial odometry and localization accuracy.
Chapter
Mobile edge computing (MEC) and next-generation mobile networks are set to disrupt the way intelligent and autonomous systems are interconnected. This will have an effect on a wide range of domains, from the Internet of Things to autonomous mobile robots. The integration of such a variety of MEC services in an inherently distributed architecture requires a robust system for managing hardware resources, balancing the network load and securing the distributed applications. Blockchain technology has emerged a solution for managing MEC services, with consensus protocols and data integrity checks that enable transparent and efficient distributed decision-making. In addition to transparency, the benefits from a security point of view are evident. Nonetheless, blockchain technology faces significant challenges in terms of scalability. In this chapter, we review existing consensus protocols and scalability techniques in both well-established and next-generation blockchain architectures. From this, we evaluate the most suitable solutions for managing MEC services and discuss the benefits and drawbacks of the available alternatives.
Conference Paper
Full-text available
With the rise of the IoT, there has been a growing demand for people counting and occupancy estimation in Intelligent buildings for adapting their heating, ventilation and cooling systems. This can have a significant impact on energy consumption at a global scale as such systems consume about 40% of electricity and create about 36% of the CO2 emissions in Europe. Previous approaches to occupancy estimation either utilize methods that do not ensure people's privacy when obtaining high accuracy estimations, such as RGB cameras, or utilize thermal or radar sensors with lower accuracy. Thermal vision for people detection has several advantages. It protects people's privacy while being less affected by changes in the environment. In addition, most previous approaches relying on image processing stream data to the cloud to be analyzed. However, with the development of the more distributed network paradigms edge and fog computing, there has been a trend in moving computation towards the edge of the network. This process of embedding intelligence into end-devices enables more efficient energy consumption and network load distribution. In this work, we present an embedded algorithm for room occupancy estimation based on a thermal sensor with accuracy over the state-of-the-art. We study the performance of a variety of deep learning models on different embedded processors. We achieve a prediction accuracy of 98.9% for people counting estimation with a minimal 2 KB RAM utilization. Furthermore, the proposed algorithm has very low latency achieving execution times under 14 ms.
Conference Paper
Full-text available
A recent trend in the IoT is to shift from traditional cloud-centric applications towards more distributed approaches embracing the fog and edge computing paradigms. In autonomous robots and vehicles, much research has been put into the potential of offloading computationally intensive tasks to cloud computing. Visual odometry is a common example, as real-time analysis of one or multiple video feeds requires significant on-board computation. If this operations are offloaded, then the on-board hardware can be simplified, and the battery life extended. In the case of self-driving cars, efficient offloading can significantly decrease the price of the hardware. Nonetheless, offloading to cloud computing compromises the system's latency and poses serious reliability issues. Visual odometry offloading requires streaming of video-feeds in real-time. In a multi-vehicle scenario, enabling efficient data compression without compromising performance can help save bandwidth and increase reliability.
Conference Paper
Full-text available
The edge and fog computing paradigms enable more responsive and smarter systems without relying on cloud servers for data processing and storage. This reduces network load as well as latency. Nonetheless, the addition of new layers in the network architecture increases the number of security vulnerabilities. In privacy-critical systems, the appearance of new vulnerabilities is more significant. To cope with this issue, we propose and implement an Ethereum Blockchain based architecture with edge artificial intelligence to analyze data at the edge of the network and keep track of the parties that access the results of the analysis, which are stored in distributed databases. A use case of edge AI for ECG feature extraction and real-time support of multiple sensor nodes is analyzed in the experiments.
Conference Paper
Full-text available
Offloading computationally intensive tasks such as lidar or visual odometry from mobile robots has multiple benefits. Resource constrained robots can make use of their network capabilities to reduce the data processing load and be able to perform a larger number tasks in a more efficient manner. However, previous works have mostly focused on cloud offloading, which increases latency and reduces reliability, or high-end edge devices. Instead, we explore the utilization of FPGAs at the edge for computational offloading with minimal latency and high parallelism. We present the potential for modelling feature-based odometry in VHDL and utilizing FPGA implementations.
Conference Paper
Full-text available
The agricultural and farming industries have been widely influenced by the disruption of the Internet of Things. The impact of the IoT is more limited in countries with less penetration of mobile internet such as sub-Saharan countries, where agriculture commonly accounts for 10 to 50% of their GPD. The boom of low-power wide-area networks (LPWAN) in the last decade, with technologies such as LoRa or NB-IoT, has mitigated this providing a relatively cheap infrastructure that enables low-power and long-range transmissions. Nonetheless, the benefits that LPWAN technologies enable have the disadvantage of low-bandwidth transmissions. Therefore, the integration of Edge and Fog computing, moving data analytics and compression near end devices, is key in order to extend functionality. By integrating artificial intelligence at the local network layer, or Edge AI, we present a system architecture and implementation that expands the possibilities of smart agriculture and farming applications with Edge and Fog computing and LPWAN technology for large area coverage. We propose and implement a system consisting on a sensor node, an Edge gateway, LoRa repeaters, Fog gateway, cloud servers and end-user terminal application. At the Edge layer, we propose the implementation of a CNN-based image compression method in order to send in a single message information about hundreds or thousands of sensor nodes within the gateway's range. We use advanced compression techniques to reduce the size of data up to 67% with a decompression error below 5%, within a novel scheme for IoT data.
Conference Paper
Full-text available
Remote healthcare monitoring has exponentially grown over the past decade together with the increasing penetration of Internet of Things (IoT) platforms. IoT-based health systems help to improve the quality of healthcare services through real-time data acquisition and processing. However, traditional IoT architectures have some limitations. For instance, they cannot properly function in areas with poor or unstable Internet. Low power wide area network (LPWAN) technologies, including long-range communication protocols such as LoRa, are a potential candidate to overcome the lacking network infrastructure. Nevertheless, LPWANs have limited transmission bandwidth not suitable for high data rate applications such as fall detection systems or electrocardiography monitoring. Therefore, data processing and compression are required at the edge of the network. We propose a system architecture with integrated artificial intelligence that combines Edge and Fog computing, LPWAN technology, IoT and deep learning algorithms to perform health monitoring tasks. In particular, we demonstrate the feasibility and effectiveness of this architecture via a use case of fall detection using recurrent neural networks. We have implemented a fall detection system from the sensor node and Edge gateway to cloud services and end-user applications. The system uses inertial data as input and achieves an average precision of over 90\% and an average recall over 95\% in fall detection.
Conference Paper
Full-text available
Increased automation and intelligence in computer systems have revealed limitations of Cloud-based computing such as unpredicted latency in safety-critical and performance-sensitive applications. The amount of data generated from ubiquitous sensors has reached a degree where it becomes impractical to always store and process in the Cloud. Edge computing brings computation and storage to the Edge of the network near to where the data originates yielding reduced network load and better performance of services. In parallel, new wireless communication technologies have appeared to facilitate the expansion of Internet of Things (IoT). Instead of seeking higher data rates, low-power wide-area network aims at battery-powered sensor nodes and devices which require reliable communication for a prolonged period of time. Recently, Long Range (LoRa) has become a popular choice for IoT-based solutions. In this paper, we explore and analyze different application fields and related works which use LoRa and investigate potential improvement opportunities and considerations. Furthermore, we propose a generic architecture to integrate Edge computation capability in IoT-based applications for enhanced performance.
Article
Full-text available
Diagnosis of arrhythmic disorders is challenging because of their short-lasting, intermittent character. Conventional technologies of noninvasive ambulatory rhythm monitoring are limited by modest sensitivity. We present a novel form of wearable electrocardiogram (ECG) sensors providing an alternative tool for long-term rhythm monitoring with the potential of increased sensitivity to detect intermittent or subclinical arrhythmia. The objective was to assess the signal quality and R-R coverage of a wearable ECG sensor system compared to a standard 3-lead Holter. In this phase-1 trial, healthy individuals underwent 24-h simultaneous rhythm monitoring using the OMsignal system together with a 3-lead Holter recording. The OMsignal system consists of a garment (bra or shirt) with integrated sensors recording a single-lead ECG and an acquisition module for data storage and processing. Head-to-head signal quality was assessed regarding adequate P-QRS-T distinction and was performed by three electrophysiologists blinded to the recording technology. The accuracy of signal coverage was assessed using Bland-Altman analysis. Fifteen individuals underwent simultaneous 24-h recording. Signal quality and accuracy of the OMgaments was equivalent to Holter-monitoring (84% vs 93% electrophysiologists rating, p = 0.06). Signal coverage of R-R intervals showed a very close overlay between the OMsignal system and Holter signals, mean difference in heart rate of 2 5 bpm. The noise level of OMgarments was comparable to Holter recording. OMgarments provide high signal quality for adequate rhythm analysis, representing a promising novel technology for long-term non-invasive ECG monitoring.
Conference Paper
This paper presents the implementation of an intelligent home-based elderly patient monitoring system. Four patient's physiological parameters are being continuously monitored, namely, temperature, glucose, and 3D accelerometer and gyroscope data for fall detection. Contextual sensors are mounted across the home to observe the patient's surrounding environment such as temperature and humidity. All sensors, wearable and contextual, transmit their measured data to smart gateways (fog layer) via nRF communication protocol. At the fog layer, diverse functions are being carried out, from collected measurements transfer to health care providers for further processing and analysis via Internet (cloud layer), sending push notifications and reports to patient's mobile phone, to alerting ambulance or civil defence authorities in case of an emergency. To insure power autonomy and eliminate the need for frequent sensor node battery replacement, an efficient thermal energy harvesting system is developed. Associated with a boost converter, the thermal energy harvesting system is able to sustain 3.3 OCV leveraging a temperature difference of 20 degree C between patient's body and room temperature, while achieving an efficiency of 82.6%.