On the self-stabilization of mobile oblivious robots in uniform rings

• We investigate self-stabilizing (SS) algorithms for oblivious robots in rings.
• We propose probabilistic SS algorithms for the gathering and orientation problems.
• Our algorithms assume the SSYNC model and the global-strong multiplicity detection.
• We show these assumptions are weakest to realize probabilistic SS algorithms.

We investigate self-stabilizing algorithms for anonymous and oblivious robots in uniform ring networks, that is, we focus on algorithms that can start from any initial configuration (including those with multiplicity points). First, we show that no probabilistic self-stabilizing gathering algorithm exists in the asynchronous (ASYNC) model or if only global-weak and local-strong multiplicity detection is available. This impossibility result implies that a common assumption about initial configurations (no two robots share a node initially) is a very strong one.On the positive side, we give a probabilistic self-stabilizing algorithm for the gathering and orientation problems in the semi-synchronous (SSYNC) model with global-strong multiplicity detection. With respect to impossibility results, those are the weakest system hypotheses. In addition, as an application of the previous algorithm, we provide a self-stabilizing algorithm for the set formation problem. Our results imply that any static set formation can be realized in a self-stabilizing manner in this model.