2nd International ICST Conference on Communications and Networking in China

Research Article

Characterizing and Evaluating Interference in Multi-Interface Wireless Mesh Networks

  • @INPROCEEDINGS{10.1109/CHINACOM.2007.4469542,
        author={Yunxia Feng and Minglu Li and Min-You Wu and Ling Ding},
        title={Characterizing and Evaluating Interference in Multi-Interface Wireless Mesh Networks},
        proceedings={2nd International ICST Conference on Communications and Networking in China},
        publisher={IEEE},
        proceedings_a={CHINACOM},
        year={2008},
        month={3},
        keywords={Bidirectional control  Communication switching  Computational modeling  Interference  Protocols  Switches  Telecommunication traffic  Traffic control  Wireless mesh networks  Wireless networks},
        doi={10.1109/CHINACOM.2007.4469542}
    }
    
  • Yunxia Feng
    Minglu Li
    Min-You Wu
    Ling Ding
    Year: 2008
    Characterizing and Evaluating Interference in Multi-Interface Wireless Mesh Networks
    CHINACOM
    IEEE
    DOI: 10.1109/CHINACOM.2007.4469542
Yunxia Feng1,*, Minglu Li1,*, Min-You Wu1,*, Ling Ding1,*
  • 1: Dept. of Computer Science & Engineering Shanghai Jiaotong University Shanghai, China
*Contact email: yunxia_feng@sjtu.edu.cn, mlli@sjtu.edu.cn, mwu@sjtu.edu.cn, dingling@sjtu.edu.cn

Abstract

Interference is a key factor that impacts the performance of wireless networks, especially when each node is equipped with multiple interfaces. Based on the classification criterion presented in this paper, we classify current interface assignment strategies into three types: static, dynamic and hybrid. We mainly address the problems of characterizing and evaluating interference effects of static strategy and two hybrid strategies in multi-interface wireless mesh networks. We prove that, if all links are bi-directional, both hybrid strategies described have identical interference characteristics. We derive formula for computing interference node set and channel interference possibility respectively. We also present an extensive evaluation via NS2 simulations. Compared to static strategy, the hybrid strategy yields more throughputs, higher traffic fairness and faster adaptation to traffic changes.