Mobile ad-hoc networks are used to extend and adapt the area covered by a static network of antennas. This is the case of an autonomous agent moving in a cluttered environment that needs to be connected to a fixed, limited-range, base station. The wireless network can be obtained by resorting to a platoon of mobile robots carrying each one antenna. Self-configuration of the robots' platoon is achieved by a singularity-robust task-priority inverse kinematics algorithm via the definition of suitable task functions. The effectiveness of the proposed algorithm has been extensively verified via numerical simulations involving an autonomous mobile robot moving in a cluttered environment.