Lifetime Improvement Based on Event Occurrence Patterns for Wireless Sensor Networks Using Multi-Objective Optimization

• Published in: Wireless personal communications Vol. 125; no. 4; pp. 3333 - 3349
• Main Authors: Mohtashami, Hossein, Movaghar, Ali, Teshnehlab, Mohammad
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.08.2022; Springer Nature B.V
• Subjects: Communications Engineering; Computer Communication Networks; Engineering; Evolutionary algorithms; Multiple objective analysis; Network reliability; Networks; Nodes; Optimization; Optimization algorithms; Placement; Radio communications; Reliability aspects; Sensors; Signal,Image and Speech Processing; Wireless networks; Wireless sensor networks; Wireless sensor network; WSN; Multi-objective evolutionary optimization; Geographical dependent event pattern; Controlled node placement; MOEA; Uniform event pattern
• ISSN: 0929-6212; 1572-834X
• DOI: 10.1007/s11277-022-09712-z

The wide range of wireless sensor network applications has made it an interesting subject for many studies. One area of research is the controlled node placement in which the location of nodes is not random but predetermined. Controlled node placement can be very effective when either the price of the sensor nodes is high or the sensor coverage is of a specific type and it is necessary to provide special characteristics such as coverage, lifetime, reliability, delay, efficiency or other performance aspects of a wireless sensor network by using the minimum number of nodes. Since node placement algorithms are NP-Hard problems, and characteristics of a network are often in conflict with each other, the use of multi-objective evolutionary optimization algorithms in controlled node placement can be helpful. Previous research on node placement has assumed a uniform pattern of events, but this study shows if the pattern of events in the environment under investigation is geographically dependent, the results may lose their effectiveness drastically. In this study, a controlled node placement algorithm is proposed that aims to increase network lifetime and improve sensor coverage and radio communication, assuming that the event pattern is not uniform and has a geographical dependency. The proposed placement algorithm can be used for the initial placement or, for repairing a segmented network over time. In this study, multi-objective evolutionary optimization algorithms based on decomposition (MOEA/D) have been used, and the performance results have been compared with other node placement methods through simulation under different conditions.