Research On Node Localization Algorithm For Non-uniform Wireless Sensor Networks

Smart home, intelligent medical and other new concepts continue to emerge, wireless sensor networks have also been widely known and used by people. Node localization technology is the basic research issues. However, the sensor nodes have to be deployed in a random way, and it is easy to form a network with uneven node distribution. How to improve the localization accuracy for non-uniform network is an important research direction for wireless sensor networks.Based on the traditional DV-Hop algorithm and analysis of the non-uniform wireless sensor networks, we proposed a new kind of localization algorithm called localization algorithm for non-uniform wireless sensor networks.We propose a beacon-selection-based localization algorithm. In a network with different node densities, the positions of the beacon nodes and the densities of different regions have different effects on the localization error. Based on these two factors, the improved algorithm optimizes existing algorithms locally in two aspects. First, optimize average single-hop distance locally. The beacon nodes select other beacon nodes in the datasheet with small number of hops and high density to calculate their average single-hop distance. Second, optimize the beacon nodes used for localization locally. The unknown nodes select the beacon nodes that are close to them for localization. Simulation results show that the improved algorithm has strong adaptability for non-uniform network, and improves the localization accuracy effectively.In this paper, RSSI technology and DV-hop algorithm are combined, with a correction based on RSSI, a new localization algorithm is proposed. Firstly, we achieve the distance estimation of the unknown nodes that have one-hop distance from the beacon nodes by RSSI technology. Secondly, we further correct the initial estimation position of the unknown nodes in a way based on the constraint. Simulation results show that this improved algorithm has better performance than traditional DV-Hop algorithm in non-uniform networks, and improves the localization accuracy.