We study the steady-state probability distribution of diffusion and consensus strategies that employ constant step-sizes to enable continuous adaptation and learning. We show that, in the small step-size regime, the estimation error at each agent approache ...
IEEE2013