Localization of 3D Wireless Sensor Networks Using Improved Bat Algorithm

Node localization is one of the key technologies of wireless sensor networks, while 3D localization has been becoming a hotspot in the field of localization technology research these years. The paper proposes to transfer the 3D localization problem in WSN into global optimization one and use improved bat algorithm (IBA) for the position estimation of unknown node. For the disadvantages of bat algorithm including premature convergence and slow convergence, an IBA algorithm is proposed. The equation of velocities and positions is modified by inertia weight factor and Levy flight strategy, effectively avoiding trapping in local optimum, accelerating the convergence speed, and realizing the balance of exploration and exploitation capabilities. The experiments on Zigbee platform indicated that IBA boasts higher localization accuracy in 3D space and faster convergence.