Quality of Service- and Fairness-Aware Resource Allocation Techniques for IEEE802.11ac WLAN

Autor: Summera Nosheen, Jamil Yusuf Khan
Jazyk: angličtina
Rok vydání: 2021
Předmět:
Zdroj: IEEE Access, Vol 9, Pp 25579-25593 (2021)
Druh dokumentu: article
ISSN: 2169-3536
DOI: 10.1109/ACCESS.2021.3051983
Popis: Increased demand for high-quality multimedia applications over IEEE802.11 networks has further motivated service providers to improve the QoS (quality of service) for all types of network services. IEEE802.11ac is one of the advanced WLAN standards that provides wideband transmission channels and higher data rates, which are vital to support very high-throughput services. However, due to the use of the CSMA/CA-based random access packet transmission technique, the QoS and service fairness criteria cannot always be maintained, particularly for variable bit rate services. This paper proposes new packet scheduling algorithms to support consistently high QoS and intra-service fairness for high data rate video streams. We name the proposed algorithms FRA - TXOP (Flow Rate Adaptive Transmission Opportunity) and EAS (Enhanced Access Scheduler). These algorithms offer appropriate QoS and fairness to video services in a CSMA/CA network. The FRA - TXOP algorithm allocates the transmission resources according to the channel congestion levels experienced by video flows to improve and maintain consistent QoS values. At the same time, the EAS algorithm offers service fairness among the video streams by appropriately scheduling them on the downlink channel. The EAS algorithm is implemented on top of the FRA - TXOP algorithm to achieve high QoS and service fairness among video streams. An NS-3 simulation model is developed to evaluate the performance of the proposed algorithms. The performance analysis shows that the proposed algorithms enhance the video flow QoS in terms of throughput and packet delay as well as improve the fairness among the video flows compared to other TXOP algorithms.
Databáze: Directory of Open Access Journals