A New Realistic Approach for the Relay Node Placement Problem in Wireless Sensor Networks by Means of Evolutionary Computation
Jose M. Lanza-Gutierrez, Juan A. Gomez-Pulido and Miguel A. Vega-Rodriguez
At present, the use of the wireless sensor networks is increasing considerably, in more and more fields of application. This encourages many researchers to try to overcome the disadvantages of this type of network. This paper focuses on how to add and place relay nodes to a static wireless sensor network in order to optimize three important factors not considered jointly before in any paper found in the existing literature: energy consumption, average coverage and network reliability. To this end, we tackle a np-hard multi objective optimization problem using two genetic algorithms (NSGA-II and SPEA2) chosen from evolutionary computation. The quality of the results obtained is measured by means of two multi objective indicators (hypervolume and set coverage). All the results obtained are analyzed in depth using a widespread statistical methodology.We conclude that SPEA2 provides better performance on average than NSGA-II, and furthermore that the addition of relay nodes is a good way to optimize traditional wireless sensor networks.
Keywords: Wireless sensor network, router, relay node, coverage, energy consumption, reliability, multiobjective optimization, evolutionary computation, NSGA-II, SPEA2, hypervolume, set coverage.