In hostile environments, localization often suffers from malicious attacks that may distort transmit power and degrade positioning accuracy significantly for wireless sensor network.A robust semidefinite relaxation secure localiza-tion algorithm RSRSL Mist was proposed to improve the location accuracy against malicious attacks.On the assumption of unknown transmit power, which is undoubtedly approximate to the fact of WSN, a novel secure location probability model was introduced for single-target and multi-target sensor networks, respectively.Taking the computational complexity of RSRSL Hand Bell into account, the nonlinear and non-convex optimization problem was simplified into a semidefinite programming problem.
According to the results from both simulations and field experiments, it is clearly demonstrated that the proposed RSRSL has better performance on location accuracy, in contrast to the conventional localization algorithms.