Path Planning for Data Collection Multiagent System with Priority and Moving Nodes in a Sensing Field with Obstacles

Abstract

Nowadays there is a more and more common need for one-off data collection in a specified area. For example, in case of searching for the survivors of disasters or wars, or for skiers who get into trouble. The simplest way of collecting such data is using wireless sensor network (WSN). However, data transmission within the network uses great amount of energy that may significantly reduce the lifetime and operation time of the battery-operated sensors. For this reason, the data transfer between the given sensor nodes is carried out by robots. In order to minimize the latency of data collection, it is possible to use cooperating robots. It frequently occurs that one of the nodes requires more urgent visit than the other ones, for example, the serious injuries require urgent medical attendance. The sensor nodes can also be moved to collect more usable data in a larger field. To solve this, path planning for data collection from priority and moving nodes are also presented here. The algorithm developed here to perform the one-off data collection operates so that the agents start from preliminary determined or random points and then visit all the sensor nodes, download the data from these nodes and after this arrive at a designated point where they upload the collected information. The goal here is the minimization of cost that include not only the walk-through time but also the consequences of the late visit of the higher priority nodes. The optimization is solved using the ant colony optimization.