A new evolutionary based routing protocol for clustered heterogeneous wireless sensor networks

Wireless sensor network (WSN) is a rapidly evolving technological platform with tremendous and novel applications. Recent advances in WSN have led to many new protocols specifically designed for them where energy awareness (i.e. long lived wireless network) is an essential consideration. Most of the attention, however, has been given to the routing protocols since they might differ depending on the application and network architecture. As routing approach with hierarchical structure is realized to successfully provide energy efficient solution, various heuristic clustering algorithms have been proposed. As an attractive WSN routing protocol, LEACH has been widely accepted for its energy efficiency and simplicity. Also, the discipline of meta-heuristics Evolutionary Algorithms (EAs) has been utilized by several researchers to tackle cluster-based routing problem in WSN. These biologically inspired routing mechanisms, e. g., HCR, have proved beneficial in prolonging the WSN lifetime, but unfortunately at the expense of decreasing the stability period of WSN. This is most probably due to the abstract modeling of the EA's clustering fitness function. The aim of this paper is to alleviate the undesirable behavior of the EA when dealing with clustered routing problem in WSN by formulating a new fitness function that incorporates two clustering aspects, viz. cohesion and separation error. Simulation over 20 random heterogeneous WSNs shows that our evolutionary based clustered routing protocol (ERP) always prolongs the network lifetime, preserves more energy as compared to the results obtained using the current heuristics such as LEACH, SEP, and HCR protocols. Additionally, we found that ERP outperforms LEACH and HCR in prolonging the stability period, comparable to SEP performance for heterogeneous networks with 10% extra heterogeneity but requires further heterogeneous-aware modification in the presence of 20% of node heterogeneity. (C) 2011 Elsevier B.V. All rights reserved.