A Convex Hull Based Optimization to Reduce the Data Delivery Latency of the Mobile Elements in Wireless Sensor Networks

Recent research shows that significant energy of sensor nodes can be saved in wireless sensor networks with a mobile element called Data Mule (DM) for data collection via short-range communications. However, a major performance bottleneck is the increased data delivery latency, so it is critical to optimize the DM motion, including the path and the speed. In this paper, path selection is formulated as a combined optimization problem where speed control is associated within the problem as constraints. A three-step geometric method is proposed to solve the problem: firstly the convex hull structure is used to define a visiting order of nodes, then the detailed path intervals are adjusted to accord with the optimal speed control plans and finally, the number of turns in the path is reduced through global optimization. The experiments show that the latency is decreased to 80-85% compared with the previous methods, and about 1.3 times of the lower bound minimal value. Besides, the proposed method outperforms the previous methods by an average 10% decrease of the data delivery latency in the overall cases of communication ranges.