An Improved Search Algorithm for WSN with Charging Nodes

Abstract

The existing semi-centralized stochastic election approach, Stochastic Election of Appropriate Range Cluster Heads (SEARCH) is analyzed and features are improved which, boasts low time cost, improved stable period, and eliminates the unfavorable scenario, that no cluster head emerges during some specific rounds. However, the implementation of SEARCH in WSNs with solar charging nodes offers very marginal improvement in a stable period. This proves the inefficiency of the algorithm in handling these charging nodes. The improvement offered aims to make the algorithm, more energy aware and hence comparatively well organized in handling the charging nodes in the network. This improvement, equipped with dynamic tracking of instantaneous node energy shows a promising way to extend clustering algorithms to networks with chargeable nodes. In addition, the existing SEARCH algorithm’s cluster head election procedure lacks a futuristic view of the minimum energy requirement for successful operation throughout an entire round. Intermediate cluster head death can be detrimental, to network performance by paralyzing the corresponding clusters, resulting in total data transmission failure of all nodes in that cluster. Implementing the improved SEARCH algorithm drastically reduces the bit loss by 95%, with increasing the transmitted bits successfully by 4%. The overall lifetime of the WSN improves by around 30% and the number of cluster heads dying prematurely in the middle of round operation is reduced by 89%. This paper analyzes and refines a current clustering algorithm, thereby promisingly enhancing its capacity to boost the overall network performance.