Diffusion LMS Over Multitask Networks
Diffusion LMS Over Multitask Networks
复制标题
DOI：
10.1109/tsp.2015.2412918
复制
发表时间：
2015-06-01
影响因子：
5.4
通讯作者：
Sayed, Ali H.
中科院分区：
文献类型：
--
作者：
Chen, Jie;Richard, Cedric;Sayed, Ali H.
近年来，扩散LMS算法得到了广泛的研究。这种有效的策略允许在节点必须协作估计单个参数向量的情况下解决网络上的分布式优化问题。然而，在实践中存在一些面向多任务的问题，即每个节点的最优参数向量可能并不相同。这就提出了研究扩散LMS算法在多任务环境中有意或无意运行时的性能的问题。本文对违反单任务假设的情况下扩散LMS的随机行为进行了理论分析。我们分析了在多任务环境下影响扩散LMS性能的竞争因素，并使该算法在一些有用的情况下继续提供优于非合作策略的性能。我们还提出了一种无监督聚类策略，该策略允许每个节点通过自适应调整组合权重来选择与其协作的相邻节点，以估计一个公共参数向量。通过仿真验证了理论结果，并验证了所提聚类策略的有效性。
The diffusion LMS algorithm has been extensively studied in recent years. This efficient strategy allows to address distributed optimization problems over networks in the case where nodes have to collaboratively estimate a single parameter vector. Nevertheless, there are several problems in practice that are multitask-oriented in the sense that the optimum parameter vector may not be the same for every node. This brings up the issue of studying the performance of the diffusion LMS algorithm when it is run, either intentionally or unintentionally, in a multitask environment. In this paper, we conduct a theoretical analysis on the stochastic behavior of diffusion LMS in the case where the single-task hypothesis is violated. We analyze the competing factors that influence the performance of diffusion LMS in the multitask environment, and which allow the algorithm to continue to deliver performance superior to non-cooperative strategies in some useful circumstances. We also propose an unsupervised clustering strategy that allows each node to select, via adaptive adjustments of combination weights, the neighboring nodes with which it can collaborate to estimate a common parameter vector. Simulations are presented to illustrate the theoretical results, and to demonstrate the efficiency of the proposed clustering strategy.