3rd International ICST Conference on Quality of Service in Heterogeneous Wired/Wireless Networks

Research Article

Multi-constrained soft-QoS provisioning in wireless sensor networks

  • @INPROCEEDINGS{10.1145/1185373.1185392,
        author={Xiaoxia  Huang and Yuguang  Fang},
        title={Multi-constrained soft-QoS provisioning in wireless sensor networks},
        proceedings={3rd International ICST Conference on Quality of Service in Heterogeneous Wired/Wireless Networks},
        publisher={ACM},
        proceedings_a={QSHINE},
        year={2006},
        month={8},
        keywords={},
        doi={10.1145/1185373.1185392}
    }
    
  • Xiaoxia Huang
    Yuguang Fang
    Year: 2006
    Multi-constrained soft-QoS provisioning in wireless sensor networks
    QSHINE
    ACM
    DOI: 10.1145/1185373.1185392
Xiaoxia Huang1,*, Yuguang Fang1,*
  • 1: Department of Electrical & Computer Engineering, University of Florida.
*Contact email: xiaoxiah@ufl.edu, fang@ece.ufl.edu

Abstract

Due to the inexpensive cost and small size of the sensor node, sensor networks are densely deployed for most applications. In the application oriented wireless sensor networks, traffic is usually mixed with time-sensitive packets and reliability-demanding packets. Hence, routing regardless of the packet characteristics is not efficient. Our goal is to provide soft-QoS to different types of packets since accurate path information can be hardly obtained in wireless networks. In this paper, we utilize the multiple paths between the source and sink pairs for QoS provisioning. Unlike E2E QoS schemes, soft-QoS mapped into links on a path is determined based on local link state information. Through the estimation and approximation of path quality, traditional NP-complete QoS problem is split into many small problems. The idea is to formulate the problem as a probabilistic programming, then based on some approximation technique, we convert it into an integer programming, which is much easier to solve. The resulting solution is also one to the original probabilistic programming. Simulation results demonstrate the effectiveness of our approach.