In this brief, we propose a reduced communication diffusion algorithm for distributed estimation over multi-agent. In the proposed algorithm, agents cooperatively optimize a global least-squares cost function, while each agent is allowed to share information only with a subset of its neighbors. We demonstrate that our proposed algorithm, termed reduced-communication diffusion recursive least-squares algorithm provides a trade-off between communication burden and estimation performance. We analyze the mean and mean-square stability of the algorithm and derive a closed-form theoretical expression for its steady-state mean-square deviation. Numerical simulations are provided to support the theoretical findings.
