1st International ICST Conference on Communications and Networking in China

Research Article

A Weighted K-Subset Load Balancing Algorithm for Distributed Slee in Next Generation Network

  • @INPROCEEDINGS{10.1109/CHINACOM.2006.344642,
        author={Yao Zhao and Fangchun  Yang and Dongwei Bai},
        title={A Weighted K-Subset Load Balancing Algorithm for Distributed Slee in Next Generation Network},
        proceedings={1st International ICST Conference on Communications and Networking in China},
        publisher={IEEE},
        proceedings_a={CHINACOM},
        year={2007},
        month={4},
        keywords={NGN distributed system load balancing ksubset algorithm stale information},
        doi={10.1109/CHINACOM.2006.344642}
    }
    
  • Yao Zhao
    Fangchun Yang
    Dongwei Bai
    Year: 2007
    A Weighted K-Subset Load Balancing Algorithm for Distributed Slee in Next Generation Network
    CHINACOM
    IEEE
    DOI: 10.1109/CHINACOM.2006.344642
Yao Zhao1,2,3, Fangchun Yang1,2,3, Dongwei Bai1,2,3
  • 1: State Key Laboratory of Networking and Switching Technology,
  • 2: Beijing University of Posts and Telecommunications,
  • 3: Beijing, China

Abstract

Load balancing with stale information is becoming an increasingly important problem for distributed systems. And k-subset algorithm is promising to address this problem. In this paper, the load balancing problem for distributed SLEE in NGN is studied. Considering the heterogeneity of the clustering system, a load model is designed, and a weighted k-subset (WK) algorithm is proposed. WK improves k-subset algorithm with a weight defined on the process capabilities and the resource capacities of the nodes. Experiments show that WK balances system load effectively, and outperforms k-subset in every situation by about 10%.