Diffusion based projection method for distributed source localization in wireless sensor networks

Source localization based on energy measurements is an important problem in wireless sensor networks (WSNs). It is well known that the associated objective function is not convex and may have multiple local optima and saddle points. Most of the existing algorithms can not achieve global optima. In this paper, we formulate the source localization as a convex feasibility problem (CFP) and propose a diffusion based projection method as a solution. In the proposed method, no fusion center is required; the sensor nodes need to communicate only with their closest neighbors and all the sensors update their estimations simultaneously and finally they are able to achieve consensus on a possible minimizer asymptotically. The proposed method has low complexity and achieves global optimality. Theoretical analysis and simulation results show that the proposed method has good estimation performance whether or not the CFP is consistent or inconsistent.