Please use this identifier to cite or link to this item: http://repository.iiitd.edu.in/xmlui/handle/123456789/1918
Title: Proportional fair spectrum sharing in a WiFi network of co-existing WiFi 7 and legacy nodes
Authors: Kumar, Puneet
Maity, Mukulika (Advisor)
Keywords: IEEE802.11be
WiFi7
Multi-Link Operation
MLO Scheduler
Legacy Devices
Channel Occupancy
Issue Date: May-2025
Publisher: IIIT-Delhi
Abstract: Multi-Link Operation (MLO) in WiFi 7 promises to improve throughput and reduce latency for MLDs (multi-link devices). However, the performance of upcoming WiFi networks where the MLDs (Multi-Link Devices) and legacy SLDs (Single Link Devices) will coexist, is largely unexplored. In this work, we explore such coexistence networks; our initial studies indicate that MLDs achieve disproportionately high throughput as compared to SLDs under traditional DCF-based access. This work is thus primarily directed towards achieving proportional fair throughput allocation to all devices by maximizing a weighted log utility objective function. We leverage the adaptability offered by MLO scheduler in WiFi 7 to meet our objective. Our approach involves designing an MLO scheduler that distributes MLD traffic across multiple links to ensure fairness. We propose two ways in which the MLO scheduler can determine traffic splitting parameters: in one, the access point computes and advertises the parameters to stations, and in the other, each station computes its own parameters using a feedback- based distributed algorithm. We introduce channel occupancy metric as feedback for the distributed algorithm. We implemented both approaches in ns-3 and contributed features to ns-3 3.44 version to measure channel occupancy. Through simulations in ns-3, we show that both MLO scheduling approaches can achieve an acceptable level of fairness.
URI: http://repository.iiitd.edu.in/xmlui/handle/123456789/1918
Appears in Collections:Year-2025

Files in This Item:
File Description SizeFormat 
MT23067_Puneet Kumar.pdf1.72 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.