An Auction-based Cluster Head Selection Approach for Real Wireless Sensor Networks

In this paper, a hierarchical routing approach based on network clustering and using mobile sinks is proposed in WSN. The first, second, and third levels of hierarchy are composed of sensors, cluster heads (CHs) and mobile sinks (gateways), respectively. The most important challenges in the second level of hierarchy are: 1) election of the most suitable node as CH, and 2) reduction of communication overhead of CH election algorithm. Mobile gateway uses different data transfer technologies (e.g. SMS, WiFi, and 3G) and each communication technology has different characteristics in terms of cost, energy consumption pattern, etc. However, the characteristics of available mobile gateway(s) are ignored in designing CH election algorithm in previous studies. Designing CH election algorithm without considering the characteristics of gateways may lead to problems such as increasing data transfer costs and network fragmentation. Thus, unlike previous studies, a new fitness function is designed with respect to local fitness value of a sensor and fitness value of its available mobile sink(s). In addition, an auctionbased method is adopted to control communication overhead of CH election algorithm. The performance of the proposed approach in name DACMS is evaluated in OPNET 14.5 simulation platform. The simulation results show that DACMS outperforms MECA.