Asynchronous online ADMM for consensus problems

In this paper, we consider the consensus problem where a set of nodes cooperate to minimize a global cost. In particular, we consider an online setting and propose an online algorithm based on the alternating direction method of multipliers. Besides, we take into account the asynchronous operation of the nodes. In this context, we prove that the algorithm attains sublinear regret on the objective. Finally, we assess numerically the performance of the algorithm in a distributed sparse regression problem