Practical rendezvous through modified circumcenter algorithms

We present modified nonlinear circumcenter algorithms to achieve "practical" rendezvous when agents take noisy measurements of their neighbors' positions. Assuming a uniform probability distribution of the noise in a disk about the true position, we analyze the algorithms in 1D. In particular, we provide a characterization of the practical stability ball via deterministic and stochastic analysis tools. The higher dimensional cases are discussed in simulation and we propose modified "parallel" circumcenter algorithms that can be used with guaranteed performance.