Efficient Clustering and Path Planning Strategies for Robotic Data Collection Using Space-Filling Curves

Abstract

In this paper, we consider a scenario where a mobile robot is tasked with periodically collecting data from a fixed wireless-sensor network. Our goal is to minimize the total cost of the system, including the motion cost of the robot and the communication cost of the sensors to the robot in realistic fading environments. We propose a strategy that divides the sensors into a number of clusters and uses a mobile robot to visit each cluster in order to wirelessly collect the corresponding data. We then propose a computationally efficient approach to solve this joint path planning and clustering problem by using space-filling curves. More specifically, by utilizing space-filling curves and their locality property, we show how the coupled clustering, stop position selection, path planning, and motion design problems can be solved as a series of convex optimization problems. We further mathematically characterize an upper bound for the total energy consumption of our proposed approach for the case of uniformly distributed sensors, relating it to key motion and communication parameters, such as motor parameters, channel multipath fading and shadowing variances, path-loss exponent, and target bit error rate. Finally, we verify the effectiveness of our framework in a simulation environment. Our results with realistic channel and motion parameters show considerable energy savings.