Coverage Hole Detection with Image Analysis and Heuristic for healing

Abstract

Coverage holes in WAN are of significant issue. They reduces impact overall performance of WAN. Early detection of coverage holes in a ROI thus has great significant as it helps to plan a healing strategy. Various detection approaches has been proposed in earlier papers however they are all computationally intensive or require prior knowledge of the sensor location. In this paper we propose a image analysis based hole detection approach. The method has the advantage that it require minimum prior knowedge of the ROI and can be used in non homogenous ROI. The information extracted from the detection is used as heuristic to heal larger holes in ROI quickly. The algorithm has similar performance to existing algorithms as the no of sensors reaches 50 however has a better performance as the sensor count increase beyond 50.