Black Widow Optimization for the Node Location Problem in Localization Wireless Sensor Networks.

Local Positioning Systems (LPS) present higher performance and more accurate target localization than traditional GNSS systems in harsh environments. However, the Node Location Problem (NLP) stands as one of the most important problems when designing LPS since the achievement of optimized sensor distributions in space requires addressing this NP-Hard problem. Therefore, it is common the employment of metaheuristics to tackle this problem. In this sense, the fundamentals of the no free lunch theorems state that, in order to obtain improved results for a specific problem, an investigation on the heuristic that best suits for the characteristics of the problem must be considered. Therefore, in this paper, we propose the application of the black widow optimization algorithm for the first time in the literature for the NLP. This metaheuristic allows a more diversified search adapting to the discontinuous landscape fitness of the NLP when considering NLOS links among the positioning signals. The results obtained are compared with those by a canonical genetic algorithm (CGA) introduced in our previous research, outperforming the localization error by 15% and 10% the single-point and multipoint crossover CGAs analyzed.