Bottleneck nodes avoidance data gathering tree for wireless sensor networks

Abstract

Latest development of electronic and communication technologies enabled Wireless Sensor Networks (WSNs) deployed for data collection from an area of interest. The sensed data is then aggregated and transmitted to the base station (sink), where the collected data is further processed for end-user queries. With the increase of WSNs applications, there has been a growing demand for gathering information from environments for an expected network lifetime. Since the sensor nodes have limited power and it is impractical to replace the energy source of the nodes, an energy efficient data gathering technique need to be design to shrinks the use of radio waves propagation to ensure a desired lifetime. According to the co-operative modes of the sensor nodes, the most popular data gathering techniques in WSNs are cluster-based, chain-based and tree-based. Due to the minimum graph structure and for easy implementation, we consider tree topology based algorithms in our research work. However, most of the existing tree based schemes introduce lower energy bottleneck nodes as intermediate nodes making them overburden during tree construction, shortening the lifetime of the network. To solve the overburden nodes problem, other categories of algorithms perform a load balancing operation to decrease the load of bottleneck nodes. In this thesis, we propose two load balancing tree based data gathering schemes. A mathematical model for load calculation and nodes categorized are also proposed for both the schemes. The first algorithm “Energy aware bottleneck node avoidance data gathering tree (EBA-DG)” is proposed to construct a load balance tree to gather data from environment. Considering the transmission cost with node degree an energy-aware load balancing scheme named “Energy aware bottleneck nodes avoidance shortest-path data gathering tree (EBASP-DG)” is also proposed. In EBASP-DG, a new data structure called data gathering sequence (DGS) is also proposed for the benefits of constructing tree. Theoretical analysis and simulation results show that both EBA-DG and EBASP-DG can achieve better performance than the existing other schemes.