Coverage in wireless sensor networks (WSNs) is a well-studied problem. In most of the situation WSNs have two-dimensional bounded continuous domain. In practice, sensors usually dropped randomly from air on previously determined points (called, vertices) of the domain (called Region of Interest or simply ROI). But since the deployment is random in manner the sensors will not place on the target vertices in most of the times. Hence ROI will not cover by the deployed sensors. The question is, how we reduced the area which is not covered by the sensors? Usually, extra sensors are dropped on some randomly chosen vertices to minimize the uncovered area. In our previous work we developed an alternative strategy, we reduced the distance between two adjacent vertices and drop exactly one sensor on each vertex. The amount of reduction depends on the number of extra sensors used. In that paper we compare uncovered area for the two strategies (the old one and the alternative one), for two distributions (uniform and normal), and for several number of excess sensors. Simulation result shows that alternative strategy is better for lower variance of the randomness but old one is better for higher variance. In this paper, we partitioned the ROI in regular hexagons and develop a new strategy, which is mixing of the above two strategies, for deployment of extra sensors. We divide the total number of extra sensors in two parts. One part is used for reducing the distance between the two adjacent vertices and other parts is used for deploying two sensors on some randomly selected vertices. Simulation results suggest the proper balancing between these two parts with respect to the variance of two distributions (uniform and normal).

Keywords : Sensor Coverage; Wireless Sensor; Sensor Networks; Random Deployment.