Distributed Adaptive Node-Specific Signal Estimation in Fully Connected Sensor Networks—Part II: Simultaneous and Asynchronous Node Updating

In this paper, we revisit an earlier introduced distributed adaptive node-specific signal estimation (DANSE) algorithm that operates in fully connected sensor networks. In the original algorithm, the nodes update their parameters in a sequential round-robin fashion, which may yield a slow convergence of the estimators, especially so when the number of nodes in the network is large. When all nodes update simultaneously, the algorithm adapts more swiftly, but convergence can no longer be guaranteed. Simulations show that the algorithm then often gets locked in a suboptimal limit cycle. We first provide an extension to the DANSE algorithm, in which we apply an additional relaxation in the updating process. The new algorithm is then proven to converge to the optimal estimators when nodes update simultaneously or asynchronously, be it that the computational load at each node increases in comparison with the algorithm with sequential updates. Finally, based on simulations it is demonstrated that a simplified version of the new algorithm, without any extra computational load, can also provide convergence to the optimal estimators.