Heterogeneous Networks of Remote Monitoring with High Availability and Resilience Application to Wireless Sensor Networks

Heterogeneous Networks of Remote Monitoring
with High Availability and Resilience
Application to Wireless Sensor Networks
Ayoub Marzak 1*
, Mohamed Hamraoui 2*
, Hicham Belhadaoui3*
* RITM Laboratory, CED Engineering Sciences
Ecole Supérieure de Technologie
Hassan II University of Casablanca, Morocco
1
ay.marzak@gmail.com, 2
hamraoui@hotmail.com, 3
belhadaoui_hicham@yahoo.fr
Abstract- Lately, the Wireless Sensors Networks (WSN) have moved to the concept of the hybrids networks in order to
get universal platforms in various types of monitoring and information collecting applications. The work presented in this
paper aims in designing a hybrid remote monitoring architecture, largely secured by a high availability and resilience
WSN. The modeling approach intends to describe the main operation of polling and dispatching between the
communications channels with the purpose of ensuring the information availability and reducing the resilience time. To
achieve our goal, we have realized an experimental platform of measuring, processing and routing data through hybrid
communications technologies. We have illustrated, via curves, the routing of the data measured by a WSN (ZigBee
Technology) to a final user through several communication technologies (HTTPS, SMS, ...).
Keywords- Wireless Sensors Networks, hybrid architecture, communication technology, remote monitoring, resilience,
availability, security.
I. INTRODUCTION
Recently, the technological developments in various fields related to computing and microelectronics have given
rise to a new type of sensor network called WSN [1].
These networks, with no fixed infrastructure, can be deployed widely in critical areas, hostile and difficult to
access by humans. They can detect, calculate and communicate with other devices in order to collect local
information. This information will be used to make decisions regarding the various parameters characterizing the
monitored environment. The structure of a sensor node model and its dynamic aspect were approached in [2]. More
generally, in [3] the authors are interested in the WSN model in its entirety.
Increasingly, the WSN promise a new range of services and a better understanding of the world around us. For
instance, improving the quality of the monitoring service, which can retrieve and continuously transmit the
measurements to a secure data processing server accessible by authorized users. Several works have been devoted to
the problematic of hybrid networks; among others, [4], [5], [13], [14], [15], [16], in which different networks were
merged and federated to get heterogeneous platforms, using various protocols. In this context, a style of architecture
dedicated to distributed hypermedia systems (REST: Representational State Transfer), was created by Roy Fielding
[17]. REST is a hybrid model founded on multiple models and network concepts, combined with additional
constraints. In the same context, [18] presents a monitoring architecture model with WSN based on a Web service
that allows users to access data remotely via an Internet connection. In [19] the authors present a methodology for
designing a hybrid remote monitoring architecture.
In this work, we have modeled and designed a robust prototype with hybrid architecture based on a WSN and
various communication technologies.
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 15, No. 8, August 2017
304 https://sites.google.com/site/ijcsis/
ISSN 1947-5500

II. MODEL OF THE PROPOSED HYBRID REMOTE MONITORING ARCHITECTURE
The objective of this work is to propose a model of implementing a hybrid architecture that aims to improve the
robustness of certain existing architectures. In fact, the proposed architecture model is meant to ensure a local or
remote connection of multiple users to a sensor network through different communication technologies. The
developed architecture should respond to an acceptable level of requirements in terms of availability, reliability and
safety according the case study in question.
A. Proposed platform
We are interested in the Zigbee technology [6], [7], also known as IEEE 802.15.4 [8], [9], which allows to obtain
wireless links with a low energy consumption.
Let's review the main elements of our Platform:
 Zigbee [6], [7]: it is a wireless network protocol, just as Wi-Fi or Bluetooth, which is suitable for the control
and command devices networks, and other applications requiring low debit but high reliability.
 MQTT [10]: (Message Queuing Telemetry Transport) it is a publish-subscribe messaging service based on the
simple and extremely light TCP / IP protocol. It works on the client / server principle. The server, named
broker, collects the information transmitted by the publishers (Communicating objects).
 HTTPS [11]: (Hypertext Transfer Protocol Secure) it is an Internet communication protocol that protects the
integrity and confidentiality of data while transferring information between the client and the server.
 Sockets [12]: it is a model for inter-process communication (IPC) in order to allow various processes to
communicate on the same machine through a TCP / IP network. These sockets will allow managing incoming
and outgoing flows to ensure communication between the client and the server.
Our architecture (figure 1) is composed mainly of a WSN, a processing server and users (clients). The processing
server is composed of two parts. The first part, reserved for the configuration (web services), allows to receive
physical and protocol configuration of each client connected to it to generate a polling program. The second part
deals with polling, monitoring the environment in real time and access to database in order to visualize the collected
data archive. This last part is made between the different protocols (MQTT [10], Sockets [11], HTTP / HTTPS [12]),
and the channels of communication (Wi-Fi [20], Ethernet [21], Internet [22], GSM / GPRS [23]).
In order to carry out our modeling, we opted for the choice of UML modeling language [24]. In this context, we
have established the following diagrams: Use Case Diagram, Class Diagram, Sequence Diagram and Deployment
Diagram. Moreover, we have established a flowchart to better expose the proposed solution.
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

B. Use case of the platform
The use case diagram (figure 2) shows the possible uses related to our system. We have implemented a "client"
interface that allows, on the one hand, to receive live issued data from the WSN, taking into consideration our
strategy of tolerance to technical failures and, on the other hand, to consult the database (processing server). In
addition, the "client" interface allows users to authenticate and select the desired configuration (protocols,
communication channels). Therefore, the processing server generates a polling program (flowchart polling figure 3)
specific to the client's request and then the data are displayed in the specific interfaces to the protocols (figure 9).
This polling program allows the routing of data (packets) from the WSN to the clients.
Figure 1. Hybrid Architecture of the proposed remote monitoring
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

C. Structural and dynamic aspects of the platform
We approached several issues related to the design of this architecture. First of all, we can list the adaptation and
identification of data from the sensor network before storing them in a database. Fault tolerance due to random error
is processed. In case of emergency (critical values, abnormal overflow, fault ...), the information routing is done on a
priority approach (passing by the emergency channel) to the concerned users. Given the confidentiality of the data,
we proposed, for our architecture, a secure authentication solution.
The "Polling" flowchart (figure 3), is established in order to better project our method for the tolerance to the
technical failures.
Figure 2. Use case diagram of the hybrid architecture of remote monitoring
Figure 3. Flowchart treating the polling
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

WSN allows the collection of environmental, physical parameters (temperature, humidity ...) from the sensor
nodes to the base station (sensor gateway). These data (packets) are routed to connected clients taking into account
the algorithm. At the collect of a new packet, the system automatically triggers its issuance. This packet is usually
treated with a treatment interval of urgent packets (priority) that we set according to the application. If, for example,
the detected temperature, for agricultural greenhouses, exceeds the maximum set temperature, the system
automatically performs the sending to the concerned users via the emergency channel (SMS messaging). In the
opposite case, it initializes and resumes the normal cycle of transmission via the last channel used. This one is
systematically exploited as long as the data arrive successfully to destination. In the case of a failure, the channels
testing procedure is triggered to select a new functional channel. If after five test attempts, no channel is detected, the
system calls upon the emergency channel, which is responsible for both the sending of the data and the report on the
breakdowns occurring. All packets are transmitted with acknowledgment of receipt.
Figure 4 shows the class diagram of our architecture:
 SensorGateway: it can acquire data from different sensor nodes.
 OnNewValue: it represents the event of receiving new data.
Figure 4. Class diagram of the hybrid architecture of a remote monitoring
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

 PollingPool: this is an interface that allows in one hand to manage connected clients and in the other hand the
recording and the outbreak of the diffusions measures.
 PollingPoolImpl: it is an implementation of the PollingPool.
 WebServiceConfig: it is the web service that allows the identification of customers and receipt of their
configuration. This configuration is assigned to PollingPool to generate the adapted polling program.
 SensorValueDao: it saves data in a database implemented in the processing server.
 Polling: it allows publishing the values from the WSN to different clients using the polling.
 SensorValue: it contains the physical measurements from the WSN.
 Channel: an interface that represents a sending channel.
 SocketChannel: a channel for the implementation of Sockets protocol.
 MQTTChannel: a channel for the implementation of the MQTT protocol.
 HTTPSChannel: a channel for the implementation of the HTTPS protocol.
 SMSChannel: a channel for SMS implementation.
After describing the structure of our architecture, we describe the dynamic aspect of our approach using the
sequence diagram (figure 5).
Figure 5. Sequence diagram of the hybrid architecture of a remote monitoring
The events of this diagram are described as follows:
- The sensor nodes send the measured physical measurements to the sensor gateway.
- The sensor gateway collects and routes the data from the WSN to the polling server.
- The polling server allows, the delivering, of data to connected customers and their storing in the database.
- This server is in place for client requests (Reading data in real-time by polling or access to the database).
- A client attempts to connect and authenticates with a username and a password.
- The server configuration allows him to connect and send its configuration via a Web Service (web interface).
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

- The client configuration is received by the polling server. This generates a polling program adapted to the
client's request (this program is unique to each client connected to the architecture).
- The customer receives real-time data from the WSN.
- Access to the database is described in another sequence diagram (figure 6).
Figure 6 represents, in the case of need, the events related to the history access mode:
- First of all, the WSN collects the data by the sensor gateway.
- Thereafter the data are automatically stored in the database.
- The users, who want to access the stored values, must authenticate with a username and password and select the
period of desired values.
- After authentication, the configuration server receives the client request (History according to the period).
- The configuration server starts the playback and diffusion of the historical requested by the client.
D. Components of the platform
The proposed architecture consists mainly of modules illustrated in the deployment diagram in figure 7. The main
component of our system is named PollingPool. It manages all configurations received from the users. These
configurations are secured by an authentication system with login and password, and received by the Controller Web
component. The Sensor Reader component allows the collection of data from the WSN. The polling component
contains a polling program dedicated to each client. Finally, the four redundant and diverse channels represent the
implementation of the communication between the server and the users.
Figure 6. Sequence diagram represents the access to database
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

Number equations consecutively with equation numbers in parentheses flush with the right margin, as in (1). To make your
III. IMPLEMENTATION OF THE PROPOSED HYBRID ARCHITECTURE, EXPERIMENTAL TESTS AND RESULTS
To validate its feasibility, our hybrid architecture of remote monitoring, based on WSN and interconnection
techniques, has been realized in JAVA.
A. Implementation and experimental tests
We have implemented a platform of ZigBee WSN (Crossbow MicaZ) [25] consisted of 40 nodes of sensors based
on the "MPR2400" microcontroller based on "Atmel ATmega128L" of ZigBee-Alliances [6]. The Sensor nodes
("MDA 100" and "MTS 420") are connected to the gateway in Mesh topology.
Firstly, we collect physical measurements such as temperature, humidity, the geo-position of the node in relation
to the gateway and the energy consumed by each node. These physical measurements are stored in a database.
Secondly, we perform various tests by a computer and a smartphone with interfaces designed under JAVA (figures 8,
9, 10). SMS sending is handled by a GSM / GPRS modem [23].
The implemented "client" interfaces allow to receive data from the WSN in real time and consult the database
(figure 10). On the other hand, the "client" interface (figure 8) allows for authenticating and selecting the desired
configuration (Protocols, Communication Channels). Thus, the server generates a polling program specific to the
client's request. Therefore, the data are displayed in interfaces specific to the protocols. In this sense, Figure 9
represents the client interface specific to the MQTT protocol.
Figure 7. Deployment diagram of the hybrid architecture of a remote monitoring
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

Figure 8. A client interface of authentication and configuration choices
Figure 9. A client interface specific to protocol MQTT Figure 10. Database access interface
B. Results and Discussion
The database is powered by sensor data records and their archiving. This allowed us to draw up an evaluation and
validation graph (figure 11) of our polling algorithm. The abscissa axis contains the data Id, while the ordinate axis
includes the sending time of the data by each protocol. In this case, the packets are routed to the client by the sockets
protocol. At ID 138, we notice an increase in the sending time of about 1300 ms which corresponds to the detection
of the failure of the used channel and the switching to the MQTT protocol. The latter takes over the sending cycle
from the non-transmitted data. The same phenomenon occurs at ID 160 between the MQTT protocol and the HTTP /
HTTPS protocol. Another failure occurs at ID 171, the system resumes the test of the communication channels until
the resumption of the transmission by the sockets.
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

Figure 11. The evolution of the polling between channels as a function of time
Figure 12 illustrates the variation of the time transmission to another channel by the polling processing due to the
failure of the used channel. The evolution of this transmission time depends on the transmitted data. We note that,
this time is around 1300 ms as described above.
Figure 12. Data transmission time in case of polling to another channel
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

IV. CONCLUSION
In this paper, we have proposed a hybrid remote monitoring architecture model based on WSN at high availability
and resilience. This architecture allows users to access to data locally or remotely using multiple communication
technologies. We presented the structural and dynamic aspects and the components of our proposed platform for the
main operations: acquisition, storage and distribution of data by polling.
To validate the feasibility of the proposed architecture, we have built a prototype using a ZigBee WSN type and
implemented a platform developed under JAVA.
REFERENCES
[1] J. Yick, B. Mukherjee, D. Ghosal, “Wireless sensor network survey”, Comput. Netw., vol. 52, no 12, p. 2292‑ 2330, août 2008.
[2] A. Sahu, E. B. Fernandez, M. Cardei, M. Vanhilst, « A Pattern for a Sensor Node », in Proceedings of the 17th Conference on Pattern
Languages of Programs, New York, NY, USA, 2010, p. 7:1–7:7.
[3] M. Cardei, E. B. Fernandez, A. Sahu, I. Cardei, « A Pattern for Sensor Network Architectures », in Proceedings of the 2Nd Asian
Conference on Pattern Languages of Programs, New York, NY, USA, 2011, p. 10:1–10:8.
[4] S. Nadir, A. Marzak, K. Lahma, H. Belhadaoui, M. Hamraoui, « Communication protocolaire hybride des données massives distribuées
traitant en parallèle la crédibilité de l’information: Application au WSN », in The 2nd World Congress on Multimedia and Computer Science
2014 (WCMCS), Tunisie, p. 1‑ 6.
[5] A. Marzak, M. Hamraoui, H. Belhadaoui, « Conception et réalisation d’une architecture hybride intégrant des réseaux de capteurs sans fil et
technologies d’information et de communication », Mediterranean Telecommunications Journal (MTJ), Vol. 6, N° 2, June 2016.
[6] « The ZigBee Alliance | Control your World », http://www.zigbee.org/.
[7] F. Shariff, N. A. Rahim, W. P. Hew, « Zigbee-based data acquisition system for online monitoring of grid-connected photovoltaic system »,
Expert Syst. Appl., vol. 42, no 3, p. 1730‑ 1742, févr. 2015.
[8] « IEEE-SA - The IEEE Standards Association », http://standards.ieee.org/.
[9] C. Suh, Z. H. Mir, Y.-B. Ko, « Design and implementation of enhanced IEEE 802.15.4 for supporting multimedia service in Wireless Sensor
Networks », Comput. Netw., vol. 52, no 13, p. 2568‑ 2581, sept. 2008.
[10] « IEEE Xplore Abstract - MQTT-S — A publish/subscribe protocol for Wireless Sensor Networks ».
[11] R. Fielding, J. Gettys, J. Mogul, H. Frystyk, L. Masinter, P. Leach, T. Berners-Lee, « Hypertext transfer protocol--HTTP/1.1. RFC 2616 »,
June, 1999.
[12] K. L. Calvert, M. J. Donahoo, « TCP/IP sockets in Java: practical guide for programmers », Morgan Kaufmann, 2011.
[13] M. Cardei, A. Marcus, I. Cardei, T. Tavtilov, « Web-based heterogeneous WSN integration using pervasive communication », in
Performance Computing and Communications Conference (IPCCC), 2011 IEEE 30th International, 2011, p. 1‑ 6.
[14] D. Barata, G. Louzada, A. Carreiro, A. Damasceno, « System of Acquisition, Transmission, Storage and Visualization of Pulse Oximeter
and ECG Data Using Android and MQTT », Procedia Technol., vol. 9, p. 1265‑ 1272, 2013.
[15] H. Huang, S. Xiao, X. Meng, Y. Xiong, « A Remote Home Security System Based on Wireless Sensor Network and GSM Technology », in
2010 Second International Conference on Networks Security Wireless Communications and Trusted Computing (NSWCTC), 2010, vol. 1, p.
535‑ 538.
[16] S. Nadir, A. Marzak, K. Lahma, H. Belhadaoui, M. Hamraoui, « Design and complexity analysis of algorithm treating the credibility of the
information: Application to WSN », NNGT Int J Netw. Comput., vol. 2, févr. 2015.
[17] R. T. Fielding, « Architectural Styles and the Design of Network-based Software Architectures », University of California, Irvine, 2000.
[18] A. Marcus, M. Cardei, I. Cardei, E. Fernandez, F. Frati, E. Damiani, « A Pattern for Web-based WSN Monitoring », J. Commun., vol. 6, no
5, août 2011.
[19] A. Marzak, H. Belhadaoui, M. Hamraoui, A. Mektoubi, « Modeling a surveillance architecture giving a high-availability and resilience
based on wireless sensor networks », The 5th International Conference on Multimedia Computing and Systems 2016 (ICMCS), Septembre
29-30, Marrakech, Morocco.
[20] P. Mühlethaler, O. Salvatori, « 802.11 et les réseaux sans fil », Eyrolles, 2002.
[21] R. M. Metcalfe, D. R. Boggs, « Ethernet: distributed packet switching for local computer networks », Commun. ACM, vol. 19, no 7, p.
395‑ 404, 1976.
[22] J. Postel, « Internet protocol », 1981.
[23] T. Halonen, J. Romero, J. Melero, « GSM, GPRS and EDGE performance: evolution towards 3G/UMTS », John Wiley & Sons, 2004.
[24] J. Rumbaugh, I. Jacobson, G. Booch, « Unified Modeling Language Reference Manual », The 2Nd Edition. Pearson Higher Education,
2004.
[25] « Crossbow, W. S. N. Professional Kit », http://www. xbow. com. Products/productdetails. aspx.
Vol. 15, No. 8, Augus 2017
ISSN 1947-5500

Heterogeneous Networks of Remote Monitoring with High Availability and Resilience Application to Wireless Sensor Networks

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Heterogeneous Networks of Remote Monitoring with High Availability and Resilience Application to Wireless Sensor Networks

Similar to Heterogeneous Networks of Remote Monitoring with High Availability and Resilience Application to Wireless Sensor Networks (20)

Recently uploaded

Recently uploaded (20)

Heterogeneous Networks of Remote Monitoring with High Availability and Resilience Application to Wireless Sensor Networks