Communication-aware trajectory tracking

This paper investigates the scenario of a robot making a tradeoff between tracking a time-varying reference trajectory and stopping to communicate at points where the radio signal strength is high. Under the assumption that the signal is subject to multipath fading, we formulate this as a hybrid optimal control problem with penalties on tracking error, communication buffer length and control power. The problem is then solved using relaxed dynamic programming, resulting in control laws for the discrete switching sequence and the continuous control. We finally illustrate the results through simulations under non-ideal conditions, confirming that the system maintains a bounded buffer size and zero-mean tracking error.