Compressive and collaborative mobile sensing for scalar field mapping in robotic networks

In this paper, we propose a compressive and collaborative sensing (CCS) algorithm for distributed robotic networks to build scalar field map. A collaborative control law is utilized to steer the robots to move on the field while avoiding collision with each other and with obstacles. At each time instant, the robots collect, add measurements within their sensing range and exchange data with their neighbors to form compressive sensing (CS) measurements at each robot. After a certain times of moving and sampling, each robot can achieve that number of CS measurements to be able to reconstruct all sensory readings from the positions that the group of robots visited to build a scalar map. We further analyze and formulate the total communication power consumption associated with the number of robots, sensor communication range and provide suggestions for more energy saving.