Туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах
IoT based e-Health solutions is an upcoming trend which will revolutionize the healthcare in the near future. IoT has evolved from micro-electro-mechanical systems (MEMS), wireless technologies and Internet which together offer connectivity of systems, microelectronic devices, and medical services a...
Saved in:
| Date: | 2019 |
|---|---|
| Main Authors: | , , |
| Format: | Article |
| Language: | English |
| Published: |
The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute"
2019
|
| Subjects: | |
| Online Access: | https://journal.iasa.kpi.ua/article/view/168445 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Journal Title: | System research and information technologies |
| Download file: |  |
Institution
System research and information technologies| _version_ | 1866302387256295424 |
|---|---|
| author | Pysmennyi, Ihor Kyslyi, Roman Petrenko, Anatoly |
| author_facet | Pysmennyi, Ihor Kyslyi, Roman Petrenko, Anatoly |
| author_sort | Pysmennyi, Ihor |
| baseUrl_str | http://journal.iasa.kpi.ua/oai |
| collection | OJS |
| datestamp_date | 2019-08-07T15:26:27Z |
| description | IoT based e-Health solutions is an upcoming trend which will revolutionize the healthcare in the near future. IoT has evolved from micro-electro-mechanical systems (MEMS), wireless technologies and Internet which together offer connectivity of systems, microelectronic devices, and medical services and allow data processing at the edge. That at the same time allows to save computational resources and avoid unnecessary point of failure, such as centralized synchronization point. Monitoring of patients’ vital signs parameters (measured at home) is achieved by using modern Internet of Things technology which provides networkable connections between portable diagnostic sensors, their cell phones, cloud data storage with patients’ Personal Health Records and professional health providers. This paper explores possibilities of using fog computing approach to shift data processing and computations from cloud to the edge and to build a multi-scope infrastructure for mHealth and citizen-observation system, based on SOA approach. |
| doi_str_mv | 10.20535/SRIT.2308-8893.2019.1.09 |
| first_indexed | 2025-07-17T10:25:08Z |
| format | Article |
| fulltext |
© Ihor Pysmennyi, Roman Kyslyi, Anatoly Petrenko, 2019
118 ISSN 1681–6048 System Research & Information Technologies, 2019, № 1
TIДC
ПРОБЛЕМНО І ФУНКЦІОНАЛЬНО
ОРІЄНТОВАНІ КОМП’ЮТЕРНІ СИСТЕМИ
ТА МЕРЕЖІ
UDC 004.72
DOI: 10.20535/SRIT.2308-8893.2019.1.09
EDGE COMPUTING IN MULTI-SCOPE SERVICE-ORIENTED
MOBILE HEALTHCARE SYSTEMS
IHOR PYSMENNYI, ROMAN KYSLYI, ANATOLY PETRENKO
Abstract. IoT based e-Health solutions is an upcoming trend which will revolution-
ize the healthcare in the near future. IoT has evolved from micro-electro-mechanical
systems (MEMS), wireless technologies and Internet which together offer connec-
tivity of systems, microelectronic devices, and medical services and allow data
processing at the edge. That at the same time allows to save computational resources
and avoid unnecessary point of failure, such as centralized synchronization point.
Monitoring of patients’ vital signs parameters (measured at home) is achieved by us-
ing modern Internet of Things technology which provides networkable connections
between portable diagnostic sensors, their cell phones, cloud data storage with pa-
tients’ Personal Health Records and professional health providers. This paper ex-
plores possibilities of using fog computing approach to shift data processing and
computations from cloud to the edge and to build a multi-scope infrastructure for
mHealth and citizen-observation system, based on SOA approach.
Keywords: wireless sensor networks; personal health systems; cloud services; IoT;
fog computing; service-oriented architecture (SOA); mHealth; communication; in-
frastructure.
INTRODUCTION
Previous developments of medical mobile applications usually were monolithic
ones, designed for a fixed HW infrastructure. The whole application must be de-
veloped and deployed in one piece and the entire tier must be retested and rede-
ployed when something is changed. Hundreds of applications for smartphones
with various operating systems have been created by different providers. We are
mostly the first who start to investigate advantages of service-oriented architec-
ture in mobile medicine. This paper is a generalization of the author's publications
on the possible usage of the service-oriented computing paradigm (SOC) for
building a medical services platform which allows unifying the development of
applications for patients, doctors and the central server by orchestrating and com-
posing web services from a common cloud repository. Due to this approach the
created applications can be adapted to the particular patient, his disease and the
plan of his treatment at home.
Many specialists in the world believe that patient empowering can transform
medical care. Wireless internet connectivity, cloud computing, mobile devices
Edge computing in multi-scope service-oriented mobile healthcare systems
Системні дослідження та інформаційні технології, 2019, № 1 119
(smartphones and tablets), mobile applications and sensors modernized clinical
trials, internet connectivity, advanced diagnostics, targeted therapies, and other
science enable the individualization of medicine and force overdue radical change
in how medicine is delivered, as well as regulated and reimbursed. Let’s imagine
that every medical sensor (or another data resource) of that ecosystem has its own
URI allowing doctors and patients interact with it via the web browser, and at the
same time every sensor can have the software interface – a set of web services
allowing intelligent software agents to interact with it (analyze the data etc.) on
behalf of doctors and patients. Certainly, the integration of that with the classical
medical record is vital. Mobile devices are being used to capture data at the point
of care and to keep the lines of communication open no matter where the doctor
is, and they’re being used at home to record and send vital health data back to the
health professional and, in turn, to send important healthcare management infor-
mation back to the patient. The personal data is used to track the ups and downs
of patient’s conditions as they go about their lives. This approach implies to the
fog and mist computing paradigms. Fog and mist layers are abstract conceptual
levels at the network edge [1]. While fog connects these edge devices, putting a
lot of storage, configuration, computing and analysis tasks away from cloud to the
edge, mist resides directly within network fabric with ability for end-user devices
to share their available computational and communication capabilities for per-
forming a variety of applications and networking task [2, 3].
Advantages of computations at the edge are listed below:
• Preserving user data privacy by not sending sensitive data to the cloud
and processing it in the local environment.
• Maintenance costs by significantly decreasing required bandwidth and
computing power on the server.
• Reducing latency.
• Heterogeneous by-design architecture.
• Self-awareness.
• Fault-tolerance by-design.
Possible Computing levels of m-Health system in hand are shown on fig. 1
and includes the following levels:
Fig. 1. Computing levels of IoT based system 3 [4]
Ihor Pysmennyi, Roman Kyslyi, Anatoly Petrenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 1 120
• Sensor network and mist computing layer which consists of air quality
sensors and BSN. Filtering and basic data validation is also performed on this
layer.
• Fog-computing layer which is used for gathering and combining data
from different sources and performing some analysis on it (described in sec-
tion IV)
• Fog-gateway layer which provides communication between cloud server
and other fog-networks.
• Cloud analytics layer for storing and processing of aggregated data, per-
forming complex analysis on anonymized data for municipalities.
• PHR integration layer which allows to give user recommendations and
predictions based on external eHealth information.
• Expert analytics layer which assists expert in making decisions and pro-
vides mechanism for feedback.
These levels imply to the security zone approach are proposed in [5]. There
are 4 main challenges in development of such architecture:
• Preserving user privacy [6] as collecting and transferring gathered data
can reveal both user’s identity and sensitive health data with geolocation.
• Efficiency: both in energy consumption and computation power 1.
• Fault tolerance.
• Authentication of system nodes and security against piggybacking at-
tacks, detection of untruthful data.
The rest of the paper is organized as follows - in section 2 we do a deep dive
on the Proposed Architectural Framework (and mainly focus on fault tolerance
and reducing amount of computational resources), followed by communication
issue in section 3 and data storage architecture in section 4.
Novice moving average approach for data preservation during offline phases
is also being introduced in section 3. Our final section Y is the summary and con-
clusions.
The main purpose of such system is to help user monitor and predict devel-
opment of asthma and other respiratory diseases.
PROPOSED ARCHITECTURAL FRAMEWORK
The explosion of affordable sensors and wearable devices lead to growth of Per-
sonal mHealth, self-management of health conditions, and the collection of data,
will radically change how health-care is delivered and information is collected.
Let us consider these changes on the example of multi-scope infrastructure for
mHealth and citizen-observation system (fig. 2). Air pollution in big cities and
passive way of life combined with the abundant number of stress-factors caused
the growing demand [7] on different health monitoring telecare systems and air-
quality and user-breath monitoring systems in particular.
By gathering, combining and analyzing data collected from weather stations,
Internet of Things (IoT) sensor networks, Body Sensor Networks (BSN), Personal
Health Record (PHR) systems we can find patterns of disease development on
early stages on signalize doctor on escalations.
Edge computing in multi-scope service-oriented mobile healthcare systems
Системні дослідження та інформаційні технології, 2019, № 1 121
The other scope of this system is citizen observation air-quality monitoring
system. In addition to assistance to respiratory-sensitive people we found follow-
ing usages of constant air-quality measurements for municipalities [8]:
• Commuting analyses based on CO2 levels at different time. This can re-
sult in improving city transportation quality and designing better living environ-
ments.
• Participation in life-quality indices for different districts, forecasting their
development.
• Anomaly detection which can be the subject for further investigations.
As shown on Fig. 2 communication between edge-computing level and
cloud is done through fog-computing gateway. The IoT space of terminal end-
points in the discussed infrastructure includes the current smart phones, tablets,
and laptops. While each one is quite an advanced technological piece, including
sensors and cameras, we can ignore their internal complexity and regard them as
simple points, providing connectivity to the person who owns them. Based on
how the devices are connected to the patient, the devices can be classified into
implantable, wearable, unconnected, or connected on need basis. By the end of
2023 95% of population worldwide will have the access to broadband mobile
internet and smartphones [9] (Fig. 3). Respectively, user’s mobile phone seems to
be the perfect platform acting as gateway. This provides the following benefits:
• Available almost everywhere.
• Can aggregate data from different sources, such as internal motion
sensors, GPS systems, provides API for different BSNs (Healthkit for iOS devices
and Google Fit for Android correspondingly), can connect over Bluetooth or
Bluetooth smart to external sensors. These provides location and situation
awareness.
Sensor
network
and
mist
computing
layer
Fog dateway
computing layer
Cloud
computing and
storage layer
Expert layer
Municipatilies
and other
interested
parties
data
Fig. 2. Infrastructure for mHealth and citizen-observation system
Ihor Pysmennyi, Roman Kyslyi, Anatoly Petrenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 1 122
• Has enough computing power for basic data analyses including comput-
ing moving average and even running deep learning models.
• Can send data to cloud storage and processing and share analysis results
with end-user.
• Provides instrumentation to secure user data available locally. Also, we
need to admit, that this responsibility lies on user as he or she is required to lock
personal device.
The key issue is the huge amount of data which is gathered by sensors. As
smart sensor network (SSN) protocols like Bluetooth Low Energy, ANT and
ZigBee have limited bandwidth compared to the volume of collected data it might
be unfeasible to process it on fog-gateway level and store in the cloud. Also, there
are latency limitations in telecare systems, where the delay in making decision
can make it irrelevant and even endanger user. Following solutions are used to
solve these tasks:
• Basic signal processing and feature selection are performed on smart sen-
sor itself, significantly reducing amount of transferred data.
• Multi-party computation (MPC) approach is used for utilizing maximum
resources and reducing latency (fig. 4).
In addition to mentioned above, by applying edge computing we can prevent
sensitive data leakage on cloud level as large part of it is never transmitted to
global network, but it is necessary to keep sensitive data anonymized during local
transactions. For example, while sending air quality stats between wearable de-
vice and citizen-observation module we don’t need to reveal user’s identity. An-
other example is PHR system querying where both patient and hospital would like
to preserve excessive data from each other and third parties. MPC makes possible
sharing data for computation with only analysis result available, without reveal-
ing. For example, when exchanging air quality data only coordinates and features
are shared without connection to the user identity, and parties receive only result-
ing prognosis without revealing each other’s inputs.
The cloud level is designed as pluggable architecture with microservices ap-
proach and consists of:
Fig. 3. Growth of smartphone usage compared to the global population [9, 10]
Edge computing in multi-scope service-oriented mobile healthcare systems
Системні дослідження та інформаційні технології, 2019, № 1 123
• Cloud storage layer for storing data in different security layers with en-
cryption solutions used for sensitive information.
• Cloud analytics layer for performing analysis on anonymized data using
machine learning and deep learning models.
• PHR and telecare system integration service which is capable for
integrating with external application programming interfaces (API) and providing
the secure way to exchange sensitive data between physician and his patient.
• Public data service which provides the API for accessing and analyzing
public data, such as air quality measurements, generalized medical statistics and
analyses results.
Sensitive data in Cloud storage layer is stored in an anonymized way: set of
states is bound to particular identifiers. Only state owner knows its identifier, so
analysis system is unaware of extraneous data.
COMMUNICATION
As system has the lay`ered structure we should define protocols for communica-
tion inside each layer as well as cross-layer.
As we are using smartphone as fog-gateway device the best option on fog-
level is to use Bluetooth Low Energy (BLE) as it is much more energy-efficient
than communication over Wi-Fi and provides enough bandwidth considering data
preprocessing on smart sensors. BLE application profiles are based on the generic
attribute profile (GATT) [14], so data clients can subscribe only to certain charac-
teristics (features) they are interested in and these features are emitted only when
changed further reducing amount of transferred data.
For exchange between client device (fog-gateway) and cloud server it is pro-
posed to apply to Remote Procedure Call approach and Google’s gRPC frame-
work in particular resulting in the following benefits [15]:
• Reducing data transfer over internet by using binary serialization (by
default, Protobuf is used as interface description language).
SECURITY MULTIPARTY COMPUTATION
DATA PROVIDER DATA PROVIDER
USERS GET RESULTS ONLY
Fig. 4. Multi-party computation (MPC) approach [11]
Ihor Pysmennyi, Roman Kyslyi, Anatoly Petrenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 1 124
• Offers client, server and bidirectional data streaming which is useful for
constantly uploading air quality sensor data or patient’s state from BSN during
exacerbation of the disease.
• Deadline and timeout control allowing RPC be aborted and server
resources freed when computed results are no longer valid to the client.
• Uses SSL/TLS to authenticate the server, and to encrypt all the data ex-
changed between the client and the server out-of-the-box.
This approach can also be used for cross-service communication in cloud
level.
One of the main issues in designing Wireless Sensor Network (WSN) proto-
col is preserving data integrity. While data is kept confidential, we need to pro-
tect it from corrupting and system from men in the middle attacks [16]. Adversary
can use spoofing and provide false data to the system, as his identity is hidden it
might be difficult to track compromised party and prevent further intrusions.
As proposed system is designed in pluggable fashion we assume, that there
could be possible situations that some of the devices will not be able to connect to
the network for some amount of time, we have to be sure that data that was col-
lected would not be lost. Due to insufficient amount of storage place on devices
themselves, we propose to use technique of moving average to preserve valuable
data. We propose implementing it as follows: part of data stream is stored on the
device (depending on devise capacity, but assuming that we collect measurements
with frequency 5 measurements per second, optimal will be to keep couple of
seconds window on the device), and constantly updating while there is a connec-
tion. As the device loses connection, it calculates cumulative moving average [21]
of stored window, and next measurements it calculates with the respect of already
calculated value:
n
xxCMA n
n
++
=
K1 ;
1
1
1 +
+
= +
+ n
CMAnxCMA nn
n . (1)
Using these techniques will allow not to lose all the data collected while
there would be no connection but obtain mean value of measured data through
that time. Also, we have to mention that when connection will appear again, and
mean value would be transferred to server (or other intermediate device), we will
need to put a flag that indicates that it is a mean value, and also 2 timestamps –
beginning and end of the created frame. This frame has variable size, as offline
time is in inverse ratio with data detailing, so sensor readings detailing is always
maximum available.
Proposed temporal storage algorithm consists of following steps:
• Acquiring list of sensor readings with timestamps.
• As we enclose the storage limits of the edge device, first elements are col-
lided using Cumulative moving average (CMA), so the most recent data is also
the most detailed.
• Above repeats after every data append until network connection is ac-
quired again and data is being sent.
• At the same time, this technique should be used only when connection is
lost, because as it can be seen from the (1) when calculating moving average, we
lose measurement accuracy (while not using this feature we obtain no data from
lost period, and overall data accuracy might be lower as a result).
Edge computing in multi-scope service-oriented mobile healthcare systems
Системні дослідження та інформаційні технології, 2019, № 1 125
For example, proposed approach was used, while testing distributed IoT
network in our experiment with detecting breath patterns using CNN [22]. One
part of it was system capable for air quality measurements — CO2 ppm, dust
level, temperature, pressure and humidity taken second with double precision for
each value and timestamp, so the resulting size of one document is 44 bytes. Sec-
ond module required data gathered from 2 smart accelerometer sensors 3 times
per second and was used for respiration rate measurement – 3 axis values and
timestamp (28 bytes). In the case of the connection loss this makes 158,4+604,8 =
=713,2 KB per hour which exceeds memory resources for most wearable SOCs
(for example, arduino pro mini used in experiment has only 32Kb RAM avail-
able). While recording data from human activity (one of the activities was run-
ning), people moved also in areas with poor internet connection. This caused de-
lays and spaces in collecting data. Proposed approach with running average saved
consistency of data with few accuracy loses (up to 9% in 5 minutes range) for
respiration rate measurement. This proves that proposed approach with moving
average works fine, especially with more static data like weather conditions pre-
serving the most valuable measurements for both current context and historical
analyses. However, when it comes to highly dynamic data like chest moving, it
has main limitation — with time gap increasing, accuracy will decrease.
Researching utilization of different moving average calculation algorithms
for different data types is a subject for further research.
DATA STORAGE
Blockchain might help addressing data integrity and trust problems mentioned
above [2] [19], but there are a lot of limitations that need to be solved including
limited resources that are not enough to properly support mining with keeping
low-latency on low-power devices, large traffic overhead, and scalability issues.
It is obvious that blockchain itself can’t be kept on the edge devices them-
selves, due to lack of computational resources and storage place. So, it is a good
solution to distribute data storing and split it into saving data itself and metain-
formation about data, that could be quite sensitive – like PHR history and loca-
tions coupled with user id, or other personal information that could expose user’s
identity.
One of the possible solutions is to use off-chain database [16] together with
blockchain storing references to data, but not the data itself. These perfectly com-
bines with MPC as we share parts of secret data between computing nodes with
no-party having the full picture. By saving trail of computations in public append-
only bulletin-board as proposed in SPDZ protocol we allow auditing party to
compare tail-of-proofs and check computing result correctness [18].
So, we suggest splitting data in the following way:
• Personal data – user name, sensor device identificator, cached user his-
tory.
• Public anonymized data – data from environmental sensors coupled with
geo location, timestamp, and other aggregated collected data.
With this approach we can benefit from exposing data collected from sensors
as well as anonymized health information to external services and integrating with
them for analysis, research and monetization purposes.
Ihor Pysmennyi, Roman Kyslyi, Anatoly Petrenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 1 126
To match sensitive information with particular patient, secret user key is
proposed as identificator.
CONCLUSIONS AND FUTURE WORK
In this work, we examined the design of multi-scope service-oriented Mobile
Healthcare systems. Proposed multi-layered approach offers pluggable architec-
ture, good scalable and secures user’s sensitive data by isolating their transfer on
different levels. By using user’s smartphone as wearable fog-gateway we
achieved integration with wide range of Bluetooth-enabled smart sensors, situa-
tion and location-awareness combined with computing power sufficient for run-
ning deep-learning models [20]. At the same time we have proposed a novice so-
lution of using moving average for storing data on IoT devices, while there is no
connection to the outer world, that enables not to lose that data, but obtain mean
values from that period of time with more detailing for more recent data.
The main issues of current implementation are:
• Preserving data integrity [13] and intrusion detection [17].
• Vulnerabilities to man in the middle attacks.
• Running computing consuming deep learning models on the edge level.
Planned future research includes:
• Applying blockchain for confirming transactions with preserving privacy
on mist and fog computing levels.
• Designing secure communication protocol on the top of BLE and
Bluetooth 5 using recently introduced Bluetooth mesh networking protocol.
• Optimization of deep learning models to be run on edge.
• Usage of smart sensors as intelligent agents in mist computing.
• Researching different moving average approaches to preserve various
data types while network connection is lost with maximum result usefulness.
• Combining data from body-worn sensors (like heart rate sensor, acceler-
ometer) for recognition of complex human activities and their outcomes regarding
current and prognosed context by building Neural Network Models for it.
Depending on particular requirements the Distributed multi-scope service-
oriented Mobile Healthcare system can be scaled from the corporate (national)
scale of patients care to the scale of supporting profile patients in a particular
region.
REFERENCES
1. Jones Jake. Edge Computing - The Cloud, The Fog And The Edge / Jake Jones. —
Solidrun, 2017. — Avaliable at: https://www.solid-run.com/edge-computing-
cloud-fog-edge/.
2. Rahmani Amir M. Fog Computing In The Internet Of Things / M. Rahmani Amir et
al. — Cham, Springer, 2017. — 169.
3. Maier Martin. Context- And Self-Awareness In Fog And Mist Computing / Maier
Martin, Mohammad Hossein Same. — 2017. — Avaliable at:
http://www.zeitgeistlab.ca/doc/context_and_self_awareness_in_fog_and_mist_co
mputing.html.
Edge computing in multi-scope service-oriented mobile healthcare systems
Системні дослідження та інформаційні технології, 2019, № 1 127
4. Solidrun. Fog And Mist Computing Levels. — 2017. — Avaliable at: https://www.
solid-run.com/wp-content/uploads/2017/04/edge-computing-600x450.png.
5. Diogenes Yuri. Internet Of Things Security Architecture / Yuri Diogenes. —
2017. — Avaliable at: https://docs.microsoft.com/en-us/azure/iot-suite/iot-
security-architecture.
6. Song Tianyi. A Privacy Preserving Communication Protocol For Iot Applications In
Smart Homes / Tianyi Song // IEEE Internet Of Things Journal. — Vol 4,
N. 6. — 2017. — P. 1844–1852. doi:10.1109/jiot.2017.2707489.
7. Global Structural Health Monitoring Market Demand, Growth & Revenue Opportu-
nity (2016-2023). — 2017. — Avaliable at: https://www.researchnester.com/ re-
ports/structural-health-monitoring-market-global-demand-analysis-opportunity-
outlook-2023/173.
8. CITI-SENSE: Development Of Sensor-Based Citizens’ Observatory Community For
Improving Quality Of Life In Cities. — 2017. — Avaliable at: http://www.citi-
sense.eu
9. Ericsson Mobility Report November 2017 – Ericsson. — 2017. — Avaliable at:
https://www.ericsson.com/en/mobility-report/reports/november-2017.
10. World Population Prospects: The 2017 Revision, DVD Edition. — United Nations,
Department of Economic and Social Affairs, Population Division, 2017. —
Avaliable at: https://esa.un.org/unpd/wpp/Download/Standard/Population/
11. Secure Multiparty Computation. — 2017. — Avaliable at: https://partisia.com/ se-
cure-simple-efficient/
12. Secure Multiparty Computation. — 2017. — Avaliable at: https://partisia.com/ se-
cure-simple-efficient/
13. Prosanta Gope. BSN-Care: A Secure IoT-Based Modern Healthcare System Using
Body Sensor Network / Prosanta Gope, Tzonelih Hwang // IEEE Sensors Jour-
nal. — Vol 16, N 5. — 2016. — P. 1368–1376.
14. Bluetooth Topology Options. — 2017. — Avaliable at: https://www.bluetooth.com/
bluetooth-technology/topology-options.
15. Grpc / Grpc Concepts. — 2017. — Avaliable at: https://grpc.io/docs/guides/ con-
cepts.html.
16. Zyskind G. Enigma: Decentralized Computation Platform with Guaranteed Privacy/
G. Zyskind, O. Nathan, A. Pentland. — arXiv:1506.03471v1 [cs.CR], Jun. 2015.
17. Stojmenovic Ivan. The Fog Computing Paradigm: Scenarios And Security Issues /
Ivan Stojmenovic, Wen Sheng // Proceedings Of The 2014 Federated Conference
On Computer Science And Information Systems, 2014. — P. 1–8. — IEEE,
doi:10.15439/2014f503.
18. Damgrd Ivan. Practical covertly secure MPC for dishonest majorityor: Breaking the
SPDZ limits / Ivan Damgrd et al. // Computer SecurityESORICS 2013. Springer
Berlin Heidelberg, 2013. — P. 1–18.
19. Petrenko А. Blockchain as a service for medical records / А. Petrenko, R. Kyisliy,
I. Pysmennyi // System Research & Information Technologies. — 2017. —
N 1. — P. 7–11.
20. Baidu Mobile Deep Learning. — 2017. — Avaliable at: https://github.com/
baidu/mobile-deep-learning.
21. Weisstein Eric W. Moving Average / Eric W. Weisstein // Wolfram MathWorld. —
2017. — Avaliable at: http://mathworld.wolfram.com/MovingAverage.html
22. Petrenko A. Human Respiration Pattern Detection Using Deep Convolutional Neural
Networks / A. Petrenko, R. Kyslyi, I. Pysmennyi // Eastern-European Journal of
Enterprise Technologies. — 2018. — N 4. — P. 6–13.
Received 22.01.2019
From the Editorial Board: the article corresponds completely to submitted manuscript.
|
| id | journaliasakpiua-article-168445 |
| institution | System research and information technologies |
| keywords_txt_mv | keywords |
| language | English |
| last_indexed | 2025-07-17T10:25:08Z |
| publishDate | 2019 |
| publisher | The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" |
| record_format | ojs |
| resource_txt_mv | journaliasakpiua/3f/736a31484ce33dbb546d654af02bc73f.pdf |
| spelling | journaliasakpiua-article-1684452019-08-07T15:26:27Z Edge computing in multi-scope service-oriented mobile healthcare systems Туманные вычисления в многоцелевых сервис-ориентированных мобильных медицинских системах Туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах Pysmennyi, Ihor Kyslyi, Roman Petrenko, Anatoly бездротові сенсорні мережі (WSN) персональні системи охорони здоров'я хмарні сервіси IoT туманні обчислення сервіс-орієнтована архітектура (SOA) мобільна медицина зв'язок інфраструктура беспроводные сенсорные сети (WSN) персональные системы здравохранения облачные сервисы IoT туманные вычисления сервис-ориентированная архитектура (SOA) мобильное здравоохранение связь инфраструктура wireless sensor networks personal health systems cloud services IoT fog computing service-oriented architecture (SOA) mHealth communication infrastructure IoT based e-Health solutions is an upcoming trend which will revolutionize the healthcare in the near future. IoT has evolved from micro-electro-mechanical systems (MEMS), wireless technologies and Internet which together offer connectivity of systems, microelectronic devices, and medical services and allow data processing at the edge. That at the same time allows to save computational resources and avoid unnecessary point of failure, such as centralized synchronization point. Monitoring of patients’ vital signs parameters (measured at home) is achieved by using modern Internet of Things technology which provides networkable connections between portable diagnostic sensors, their cell phones, cloud data storage with patients’ Personal Health Records and professional health providers. This paper explores possibilities of using fog computing approach to shift data processing and computations from cloud to the edge and to build a multi-scope infrastructure for mHealth and citizen-observation system, based on SOA approach. Электронные медицинские системы, основанные на технологиях интернета вещей (IoT), — это тенденция, которая в ближайшем будущем произведет революцию в сфере здравоохранения, возникла с развитием микроэлектромеханических систем (MEMS), беспроводных технологий и интернета, которые обеспечивают возможность интеграции систем, микроэлектронных устройств и медицинских услуг, позволяя обрабатывать данные на границе сети (максимально близко к устройству). Данный подход позволяет экономить вычислительные ресурсы и избегать ненужных точек отказа. Мониторинг параметров жизненно важных функций пациентов (измеряемых в домашних условиях) достигается с помощью современной технологии IoT, которая предполагает интеграцию между портативными диагностическими датчиками, сотовыми телефонами, облачным хранилищем данных с электронными картами пациентов и профессиональными поставщиками медицинских услуг. В работе рассматриваются возможности использования подхода туманных вычислений для переноса обработки данных и вычислений из облака к границе сети для создания многоцелевой инфраструктуры для мобильного здравоохранения и системы наблюдения за гражданами на основе сервис-ориентированного подхода. Електронні медичні системи, побудовані на технологіях інтернету речей (IoT), — це тенденція, яка найближчим часом зробить революцію у сфері охорони здоров’я, виникла з розвитком мікроелектромеханічних систем (MEMS), бездротових технологій та інтернету, які уможливлюють інтеграцію систем, мікроелектронних пристроїв та медичних послуг і дозволяють обробляти дані на межі мережі (в максимальному наближенні до пристрою). Такий підхід дає змогу економити обчислювальні ресурси і уникати непотрібних точок відмови. Моніторинг параметрів життєво важливих функцій пацієнтів (вимірюваних у домашніх умовах) досягається за допомогою сучасної технології IoT, яка передбачає інтеграцію між портативними діагностичними датчиками, стільниковими телефонами, хмарним сховищем даних з електронними картами пацієнтів і професійними постачальниками медичних послуг. У роботі розглянуто можливості використання підходу туманних обчислень для перенесення оброблень даних і обчислень з хмари до меж мережі для створення багатоцільової інфраструктури для мобільної охорони здоров’я та системи спостереження за громадянами на основі сервіс-орієнтованого підходу. The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" 2019-03-25 Article Article application/pdf https://journal.iasa.kpi.ua/article/view/168445 10.20535/SRIT.2308-8893.2019.1.09 System research and information technologies; No. 1 (2019); 118-127 Системные исследования и информационные технологии; № 1 (2019); 118-127 Системні дослідження та інформаційні технології; № 1 (2019); 118-127 2308-8893 1681-6048 en https://journal.iasa.kpi.ua/article/view/168445/168262 Copyright (c) 2021 System research and information technologies |
| spellingShingle | бездротові сенсорні мережі (WSN) персональні системи охорони здоров'я хмарні сервіси IoT туманні обчислення сервіс-орієнтована архітектура (SOA) мобільна медицина зв'язок інфраструктура Pysmennyi, Ihor Kyslyi, Roman Petrenko, Anatoly Туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах |
| title | Туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах |
| title_alt | Edge computing in multi-scope service-oriented mobile healthcare systems Туманные вычисления в многоцелевых сервис-ориентированных мобильных медицинских системах |
| title_full | Туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах |
| title_fullStr | Туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах |
| title_full_unstemmed | Туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах |
| title_short | Туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах |
| title_sort | туманні обчислення в багатоцільових сервіс-орієнтованих мобільних медичних системах |
| topic | бездротові сенсорні мережі (WSN) персональні системи охорони здоров'я хмарні сервіси IoT туманні обчислення сервіс-орієнтована архітектура (SOA) мобільна медицина зв'язок інфраструктура |
| topic_facet | бездротові сенсорні мережі (WSN) персональні системи охорони здоров'я хмарні сервіси IoT туманні обчислення сервіс-орієнтована архітектура (SOA) мобільна медицина зв'язок інфраструктура беспроводные сенсорные сети (WSN) персональные системы здравохранения облачные сервисы IoT туманные вычисления сервис-ориентированная архитектура (SOA) мобильное здравоохранение связь инфраструктура wireless sensor networks personal health systems cloud services IoT fog computing service-oriented architecture (SOA) mHealth communication infrastructure |
| url | https://journal.iasa.kpi.ua/article/view/168445 |
| work_keys_str_mv | AT pysmennyiihor edgecomputinginmultiscopeserviceorientedmobilehealthcaresystems AT kyslyiroman edgecomputinginmultiscopeserviceorientedmobilehealthcaresystems AT petrenkoanatoly edgecomputinginmultiscopeserviceorientedmobilehealthcaresystems AT pysmennyiihor tumannyevyčisleniâvmnogocelevyhservisorientirovannyhmobilʹnyhmedicinskihsistemah AT kyslyiroman tumannyevyčisleniâvmnogocelevyhservisorientirovannyhmobilʹnyhmedicinskihsistemah AT petrenkoanatoly tumannyevyčisleniâvmnogocelevyhservisorientirovannyhmobilʹnyhmedicinskihsistemah AT pysmennyiihor tumanníobčislennâvbagatocílʹovihservísoríêntovanihmobílʹnihmedičnihsistemah AT kyslyiroman tumanníobčislennâvbagatocílʹovihservísoríêntovanihmobílʹnihmedičnihsistemah AT petrenkoanatoly tumanníobčislennâvbagatocílʹovihservísoríêntovanihmobílʹnihmedičnihsistemah |