We consider the cooperative control of a team of robots to estimate the position of a moving target using onboard sensing. In particular, we do not assume that the robot positions are known, but also estimate their positions using relative onboard sensing. Our probabilistic localization and control method takes into account the motion and sensing capabilities of the individual robots to minimize the expected future uncertainty of the target position. It reasons about multiple possible sensing topologies and incorporates an efficient topology switching technique to generate locally optimal controls in polynomial time complexity. Simulations show the performance of our approach and prove its flexibility to find suitable sensing topologies depending on the limited sensing capabilities of the robots and the movements of the target. Furthermore, we demonstrate the applicability of our method in various experiments with single and multiple quadrotor robots tracking a ground vehicle in an indoor environment.