In this paper, we propose an algorithm that outperforms an algorithm used for maximizing barrier coverage in battery harvesting sensor network proposed by DeWitt, Patt, and Shi. The previous work aims to maximize the robustness of the information transmission between two sensor nodes, while guarantee that the amount of energy harvested by each node is enough for its sufficient usage. The algorithm iteratively uses the maximum multiroute flow algorithm proposed by Kishimoto. When
f is a max-flow value, it requires
O(log
f) iterations of Kishimoto’s algorithm to solve the problem. On the other hand, we propose an algorithm based on our previous method, called parametric multiroute flow. The running time of our algorithm asymptotically equals to one iteration of maximum multiroute flow. Our experimental results in random networks show that our algorithm can improve the computation time of the previous work by six times on average, when the number of nodes is 100 and the number of links is 1585.
View full abstract