Communication-Reducing Algorithm of Distributed Least Mean Square Algorithm with Neighbor-Partial Diffusion

With the development of distributed algorithms, many researchers are committed to the goal of maintaining the long-term stability of the network by reducing the communication cost. However, many algorithms that lessen communication costs often result in a significant decrease in estimation accuracy. In order to reduce the communication cost with less performance degradation, the distributed neighbor-partial diffusion least-mean-square algorithm (NPDLMS) is proposed in this paper. Besides, considering the data redundancy in the network, we offer the distributed data selection NPDLMS algorithm, which further improves the estimation accuracy and reduces the communication cost. In the performance analysis, the stability and the communication cost of the algorithms are given.