Robot deployment with end-to-end communication constraints

Abstract

During deployment, mobile robots must form an ad-hoc wireless communication network guaranteeing reliable communications between agents and possibly with some fixed base stations. First, when only connectivity constraints need to be enforced, we discuss deployment algorithms whose execution does not require a model of the communication channels, but exploits possibly random and time-varying channel gain measurements between the robots to maintain a connected network. We then turn our attention to the more realistic situation where the communication network linking the robots must support certain rates between possibly distant terminals. For this problem, we propose deployment algorithms based on a projected-gradient scheme that provide end-to-end bandwidth guarantees, assuming a channel model with deterministic exponential path loss. In addition to setting the robot positions to optimize the deployment objective, these algorithms adjust the transmission powers at the wireless nodes and route communication packets through the network to support the desired flow rates.