Applied Mathematics & Information Sciences

A data mule is a mobile device that can traverse a wireless sensor network field to move near stationary sensor nodes that are spatially dispersed for collecting data from them. Using the data mule can significantly reduce energy consumption of sensor nodes compared to common multihop forwarding schemes. However, it also increases the latency of gathering data of all nodes. In this paper, under the assumptions that the data mule can sequentially move from a specific location to another specific location and that sensor nodes can adjust their radios to different power levels, we study the data mule path planning optimization (DMPPO) problem to achieve two goals under one constraint. The two goals are (1) to plan the path for the data mule to move near every sensor node to collect data so that the data mule traversal time (or latency) is minimized, and (2) to adjust the sensor nodes’ transmission ranges so that the total sensor node energy consumption is minimized. The constraint is that the data mule must move near each sensor node at least once for gathering data. The DMPPO problem is a multi-objective optimization problem; it is challenging since a sensor node can shrink its transmission range to reduce energy consumption but the range shrinking requires the data mule to move more for data gathering, which incurs longer latency. We propose a genetic algorithm using heuristics to find Pareto optimal solutions to this problem. We also simulate the proposed algorithm to show its effectiveness.