Selecting source image sensor nodes based on 2-hop information to improve image transmissions to mobile robot sinks in search \& rescue operations

Abstract

We consider Robot-assisted Search $\&$ Rescue operations enhanced with some fixed image sensor nodes capable of capturing and sending visual information to a robot sink. In order to increase the performance of image transfer from image sensor nodes to the robot sinks we propose a 2-hop neighborhood information-based cover set selection to determine the most relevant image sensor nodes to activate. Then, in order to be consistent with our proposed approach, a multi-path extension of Greedy Perimeter Stateless Routing (called T-GPSR) wherein routing decisions are also based on 2-hop neighborhood information is proposed. Simulation results show that our proposal reduces packet losses, enabling fast packet delivery and higher visual quality of received images at the robot sink.