Thursday, February 17, 2011
Coverage Problems in Wireless Ad hoc sensor networks
In this paper the author introduced the concept of optimization of wireless adhoc sensor networks by discussing one of the fundamental issues named coverage. Coverage is the measure of quality of service in the sensor network. In this paper, author addresses coverage probelem using two key methodologies: Voronoi Diagrams and Delaunay triangulation. In addition, author also describes two types of coverages: deteministic and stochastic. In deterministic coverage is deployed in the predefined shapes. The two methods of deployment which have been discussed are uniformily and weighted deployment. In stochastic coverage, sensors are randomly deployed. One method to analyze that coverage is using best case and worst case scenarios. Worst case coverage is based on Voronoi diagram i.e max edges are calculated using breach weight method following the line segments. the max edges reflect the worst case coverage since they are placed at maximum distance from the sensor. Similarly in best case coverage, a path is found using the metric of closest proximity to the sensor. This method is based on Delaunay triangulation and helps in calculating the min edge and hence the best signal strength. In the end with the help of sensor deployment heuristics, it is shown that using breach and support improvement can be increased by the addition of extra sensors in the randomly placed sensor network.
Subscribe to:
Post Comments (Atom)
The usefulness of computational geometry techniques for coverage in sensor networks have been very well introduced in this paper. But what I feel that could have been addressed are, the computaional power expended for these methods could have been projected and how does these techniques handle obstacles.
ReplyDeleteInfluence of obstacles, environmental conditions, radio range, and noise on coverage is not explained clearly.This algorithm does not work for distributed control systems.Algorithm works only for uniform distribution of sensor nodes..Algorithm may not provide the optimal coverage in real scenario.
ReplyDeleteThis is an interesting paper. But the proposed scheme may not work for dense deployment. Even though simulation results are satisfactory but the Breach-coverage graph is contradictory. Overall it is a well explained paper.
ReplyDeleteAll calculations are based solely on geometric properties. However, in real world, issues like interference, direction of antennas, etc, which come into picture, are not considered. However, the paper is well explained and shows how geometric principles could be exploited to get a close estimate of the coverage in the network.
ReplyDelete