energy spent by other nodes to transmit through node 62 will be high and hence a low cluster-head election chance. Similarly, although the central node has a very low centrality value but it also has a very low energy which does not make it suitable for being elected. The best node 108 on the other hand has all the three descriptors suitable for being elected as the cluster-head with a maximum chance of 75 for the current scenario.
The further assumptions to compare our strategy. In the case of the random deployment, a systematic sampling over the area which equals the square grid. Taking a smaller granularity does not significantly change the results, so they c hose 0. 5m for the experiments. The network size is varied from 10 to 50 nodes for deferent scenarios. Due to space restrictions, only the result of the 50-node network is shown in figure2. For all experiments, 50m sensing range is considered. The distributions of the exact k-coverage are relatively the same in all scenarios. The exact k-coverage varies from 0 to 8. In all scenarios, 5% of the network is not covered by any nodes. Most of the area is covered by exact 1 to 4 coverage and exact 3 coverage has the highest covered area which varies from 21. 6% to 23% of the network. The random deployment has an average 3. 00 coverage with a standard deviation of 1. 7.