Distributed k-Winner-Take-All Network Under Weight-Unbalanced Topology

• Published in: IEEE transactions on circuits and systems. II, Express briefs Vol. 71; no. 2; pp. 717 - 721
• Main Authors: Liang, Siqi, Jin, Long
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.02.2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Convergence; distributed model; Integrated circuit modeling; k-winner-take-all (kWTA); Mathematical models; multi-robot system; Multiple robots; Network topology; Resource management; task allocation; Task analysis; Topology; weight-unbalanced topology
• ISSN: 1549-7747; 1558-3791
• DOI: 10.1109/TCSII.2023.3306732

This brief investigates a distributed <inline-formula> <tex-math notation="LaTeX">k </tex-math></inline-formula>-winner-take-all <inline-formula> <tex-math notation="LaTeX">(k </tex-math></inline-formula>WTA) network that effectively eliminates the lagging error problem. On this basis, a special case that the communication topology of robots is weight-unbalanced in the distributed <inline-formula> <tex-math notation="LaTeX">k </tex-math></inline-formula>WTA problem is considered for the first time. Then a distributed weight-unbalanced <inline-formula> <tex-math notation="LaTeX">k </tex-math></inline-formula>WTA (DWU-<inline-formula> <tex-math notation="LaTeX">k </tex-math></inline-formula>WTA) model is proposed with the aid of a consensus estimator. Theoretical analyses substantiate the usability and exponential convergence of the proposed DWU-<inline-formula> <tex-math notation="LaTeX">k </tex-math></inline-formula>WTA model in encountering directed weight-unbalanced graphs. A simulation based on a multi-robot system further verifies the feasibility of the proposed model in dealing with dynamic task allocation problems.