Efficient adaptive strategies over distributed networks

摘要

In this paper, we review some of the computationally efficient adaptive distributed strategies developed using incremental partial update techniques. The schemes mentioned here solve the problem of linear estimation with less number of computations in a cooperative fashion. In a distributed network each node contains local computing equipment which estimates and shares them with other nodes. The resulting algorithms are less complex in computations and in communication because of Incremental partial update algorithms and each node communicate with immediate node only. Computational complexity analysis is evaluated and performance characteristics of each algorithm are given with computer simulations. Simulation results show that with a small degradation in performance, considerable amount of computational complexity is reduced.