Improving the Performance of Multitask Diffusion APA via Controlled Inter-Cluster Cooperation

Abstract

In this paper, we consider the problem of estimating multiple parameter vectors over a sensor network in a multitasking framework and under temporally-correlated input conditions. For this, an efficient clustered multitask diffusion affine projection algorithm (APA) is proposed that enjoys both intra-cluster and inter-cluster cooperation via diffusion. It is, however, shown that while collaboration in principle is a useful step to enhance the performance of a network, uncontrolled mode of inter-cluster collaboration can at times be detrimental to its convergence performance, especially near steady-state. To overcome this, a controlled form of inter-cluster collaboration is proposed by means of a control variable which helps in maintaining the collaboration in right direction. The proposed controlled multitask strategy attains improved performance in terms of both transient and steady-state mean square deviation (MSD) vis-avis existing algorithms, as also confirmed by simulation studies. We carry out a detailed performance analysis of the proposed algorithm, obtain stability bounds for its convergence in both mean and mean-square senses, and derive expressions for the network level MSD. Simulation results reveal that the proposed scheme performs consistently well even in the absence of cluster information.

Dynamic Incentive Mechanism of Multitask Cooperation in Logistics Supply Chain

“…(1) e implicit reputation effect in dynamic cooperation can effectively improve system incentive efficiency [20] and realize an effective incentive model that combines long-term incentives and short-term incentives. (2) Factors, such as the relevance of the cooperation phase in the dynamic cooperation, the relevance of multiple logistics tasks, and the difference in the contribution of different logistics cooperation to the performance of the LSSC, have a significant impact on the incentive effect.…”

Convex Combination of Diffusion Strategies Over Networks

“…where sgn{x} is the sign function. Compared with the powernormalized scheme, (19) requires only evaluating the sign function.…”

Robust Resilient Diffusion Over Multi-Task Networks Against Byzantine Attacks: Design, Analysis and Applications

“…The single-task one means that all nodes achieve a common objective, whereas in multi-task scenarios the nodes are divided into groups and the nodes within the identical group reach the same objective [11]. Applications of multi-task strategies, where the objectives of different nodes or grouped nodes will be distinct, have been related to multi-target localization and tracking [12], [13], cooperative spectrum sensing [14], [15], node-specific speech enhancement [16], [17], and web page categorization and ranking [18].…”

“…As discussed in [36], it has been proved that the networked MSD performance for all nodes relies on the inter-cluster cooperation over multi-task networks. Utilizing a control variable to maintain inter-cluster cooperation among neighboring clusters, an affine projection-based multi-task diffusion strategy was further proposed to obtain improved convergence performance [17].…”