Enhanced Hybrid TIM-TDMA to Maximize the Total Throughput With Fairness Constraints

Autor: David Jia, Xavier Leturc, Christophe J. Le Martret, Mohamad Assaad
Jazyk: angličtina
Rok vydání: 2024
Předmět:
Zdroj: IEEE Open Journal of the Communications Society, Vol 5, Pp 6780-6801 (2024)
Druh dokumentu: article
ISSN: 2644-125X
DOI: 10.1109/OJCOMS.2024.3480325
Popis: This paper addresses the problem of topological interference management (TIM) in ad hoc networks operating under a frequency-flat Rayleigh fading channel in single-input single-output (SISO). Assuming that the nodes have access to the statistical channel state information of the direct and interfering links, we introduce an extension of TIM referred to as enhanced TIM (eTIM) that maximizes an approximation of the expected average sum rate (EASR) when all the nodes communicate on the same channel. The drawback of eTIM is that it leads to poor fairness. By fairness, we mean in this paper ensuring a minimum average data rate for each user. To address this issue, we provide a second extension of the TIM, called eTIM-hybrid, by combining eTIM with time-division multiple access (TDMA) to improve both the EASR and the user fairness. The main idea of eTIM-hybrid involves partitioning the links into two groups: the “good links” on which we apply eTIM and “bad links” on which TDMA is used. Subsequently, a scheduling optimization phase optimizes the time allocation for each link within a given frame. We provide the closed-form expression of the optimal solution of the scheduling optimization to maximize eTIM-hybrid’s EASR while ensuring an expected average minimum user rate (EAMUR) for each link in the network. Additionally, we extend eTIM-hybrid to scenarios with two available channels by combining a novel channel allocation strategy with eTIM-hybrid, and called this method joint channel allocation eTIM-hybrid (JCA-eTIM-hy). Numerical results show that both eTIM-hybrid and JCA-eTIM-hy outperform the simulated state-of-the-art approaches in various scenarios in terms of EASR while satisfying an EAMUR constraint.
Databáze: Directory of Open Access Journals