Performance improvements of video WSN surveillance in case of traffic congestions

The work presented in this paper considers traffic surveillance realized using video camera sensors. These sensors are wireless networked and scattered along the road. They monitor the area of interest and send the most important data to a sink. In case of intense traffic situations the sensors´ field of view (FoV) are obstructed by vehicles. The algorithm proposed here maintains coverage by a dynamic sensor management, based on the concept of redundancy. When the algorithm detects traffic congestion, the sensors´ obstruction gets worse. Therefore the algorithm adjusts the redundancy value in order to increase the number of active sensors. The performance of the proposed algorithm in this situation is analyzed in this paper. The dependability of the method is proved using Monte Carlo simulation.