Formulation and steady-state analysis of diffusion mobile adaptive networks with noisy links

In this study, the effects of noisy links are investigated on the steady-state performance of mobile adaptive networks with diffusion least mean-squares strategies. The authors derive theoretical relations which explain how the steady-state performance metrics, including the steady-state network mean-square deviation and steady-state velocity mean-square-error is affected by noisy links. The provided analysis relies on the spatial-temporal energy conservation argument. The proposed simulation results reveal that although the noisy links degrade the performance of mobile adaptive networks; however, for suitably chosen combination coefficients the mobile adaptive network with noisy links provides a bounded estimation error. Finally, the proposed simulations verify that the derived theoretical analysis closely matches the actual steady-state performance observed in a network.