A Projection Based Fully Distributed Approach for Source Localization in Wireless Sensor Networks
Wei Meng, Wendong Xiao and Lihua Xie
Source localization based on energy measurements is an important problem in wireless sensor networks (WSNs). It is well known that a solution to the problem is generally associated with a non-convex optimization which may have multiple local optima and saddle points. Most of the existing algorithms can not achieve a global optimum. In this paper, we formulate the source localization problem as a convex feasibility problem (CFP) and propose a fully distributed method for it. In the proposed method, sensor nodes only need to communicate with one-hop neighbors and update their estimates simultaneously based on a projection algorithm. Finally they are able to achieve consensus on a possible minimizer asymptotically. The proposed method has low complexity and can achieve global optimality fast. Theoretical analysis and simulation results show that the proposed approach has good estimation performance, regardless of whether the CFP is consistent or not.
Keywords: Wireless sensor network, source localization, convex feasibility problem, nonlinear least-squares method.