Link Estimation in Robot Networks for Multi-Radio Control and Range Extension

Abstract

Several existing and emerging robotic applications such as ordnance disposal, disaster assessment, and search-and-rescue require high-speed communication links to span large distances. Path loss, combined with other wireless propagation effects, makes this a challenging problem. Directing radio frequency (RF) transmissions toward the intended receiver can mitigate this issue; however, operating smart antenna technology is costly for resource-constrained robots, and its performance degrades in light-of-sight (LOS) conditions, which is common in outdoor scenarios. In order to overcome this challenge, we propose adding a secondary radio, combined with a passive reflector, which is only employed when the primary omnidirectional radio is deemed unreliable. The dynamic switching of the directional radio between sleep and idle states provides communication diversity, while striving to conserve energy for battery-powered systems. An efficient link quality (LQ) estimator based on fuzzy logic provides the intelligence necessary to perform the radio switching action. Our method of passively assessing LQ at the transmitter is shown to be more efficient and agile than other presented LQ estimators, while still remaining effective in dynamic environments.