A Hybrid Genetic Algorithm for the Minimum Exposure Path Problem of Wireless Sensor Networks Based on a Numerical Functional Extreme Model

Abstract

Searching for the minimum exposure path (MEP) is one of the critical issues in wireless sensor networks (WSNs). The exposure corresponding to the MEP is one indicator to measure the coverage quality of a WSN, reflecting how well a mobile target is monitored through the sensing field. The classical methods, such as the grid-based method and the Voronoi-diagram-based method, are not accurate enough to obtain the MEP, are too complex, and are not applicable to networks with heterogeneous sensor nodes, a large number of sensor nodes, or all-sensor field intensity function. To overcome these challenges, a numerical functional extreme (NFE) model is proposed for the MEP problem in this paper. The NFE model is a high-dimensional and nonlinear optimization problem. To efficiently solve this problem, based on the characteristics of the sensor node coverage, a new crossover operator is designed, a new local search scheme is proposed, and an upside-down operator to escape from local optima is developed. Integrating all of these, a hybrid genetic algorithm (HGA) is proposed for the NFE model, and its global convergence with probability one is proven. An extensive collection of experiments were conducted, and the results indicate that the proposed NFE model and the developed HGA can improve the solution accuracy and can be applicable to not only the case with heterogeneous sensors but the case with a large number of sensor nodes and all-sensor field intensity function as well.