Distributed satisfaction‐aware relay assignment: a novel matching‐game approach. Issue 8 (31st May 2016)

• Record Type: Journal Article
• Authors: Liu, Dianxiong; Xu, Yitao; Xu, Yiwei; Ding, Cheng; Xu, Kun; Xu, Yuhua
• Is Part Of: Transactions on emerging telecommunications technologies. Volume 27:Issue 8(2016:Aug.)
• Page Start: 1087
• Page End: 1096
• Publication Date: 2016-05-31
• Subjects: Telecommunication -- Periodicals; 384.05
• DOI: 10.1002/ett.3047
• Languages: English
• ISSNs: 2161-5748

Abstract

Abstract: Based on matching game theory, this paper puts forward a novel‐distributed relay‐assignment approach to optimise the throughput satisfaction of source nodes, which is scarcely investigated in former works. The assignment problem is modelled as a matching market, where source nodes apply to relay nodes according to their throughput requirements, in the meantime, relay nodes obtain pay offs by assisting source nodes to transmit data. The proposed game appears in the category of matching games with both peer effects and dynamic quotas. The choice of each source node is related to not only the quality of channel gain but also the results of other source nodes choices. Dynamic quotas of matching game are considered because of the limited resource of relay nodes and dynamic requirements of source nodes, which adds difficulty of relay assignment. To address this issue, a novel self‐organising matching approach is proposed. Then, the proposed algorithm is proved to converge to a two‐sided stable matching between source nodes and relay nodes. Shown in the simulation results, the proposed distributed matching algorithm outperforms the existing schemes in terms of both global satisfaction and fairness within a reasonable convergence time. Copyright © 2016 John Wiley & Sons, Ltd. Abstract : Based on matching game theory, this paper puts forward a distributed relay‐assignment approach to optimise the throughput satisfaction of source nodes, which is scarcely investigated in former works. The proposed game appears in the category of matching games with both peer effects and dynamic quotas. We propose a distributed matching algorithm that outperforms the existing schemes in terms of both global satisfaction and fairness within a reasonable convergence time.