A nonlinear optimization approach to coverage problem in mobile sensor networks

A distributed Voronoi-based sensor deployment approach is proposed to optimize sensor network coverage. It is assumed that each sensor can construct its Voronoi polygon using the information it receives from the neighboring sensors. To increase the local coverage of the sensors, it is desired to find a point in each polygon, such that if the corresponding sensor is placed there, then its covered area within the polygon is maximized. A nonlinear optimization algorithm is proposed based on the gradient projection to find the optimal location for each sensor. The algorithm can be implemented in a distributed fashion with minimum communication among the sensors. Examples are provided to demonstrate the effectiveness of the proposed approach in terms of convergence rate, coverage performance, and energy consumption.