10th EAI International Conference on Communications and Networking in China

Research Article

Adaptive Link Selection and Power Allocation Buffer-aided Relay Networks with Multiple Sources

  • @INPROCEEDINGS{10.4108/eai.15-8-2015.2260904,
        author={Rongkuan Liu and Petar Popovski and Gang Wang},
        title={Adaptive Link Selection and Power Allocation Buffer-aided Relay Networks with Multiple Sources},
        proceedings={10th EAI International Conference on Communications and Networking in China},
        publisher={IEEE},
        proceedings_a={CHINACOM},
        year={2015},
        month={9},
        keywords={adaptive link selection power allocation buffer-aided relaying},
        doi={10.4108/eai.15-8-2015.2260904}
    }
    
  • Rongkuan Liu
    Petar Popovski
    Gang Wang
    Year: 2015
    Adaptive Link Selection and Power Allocation Buffer-aided Relay Networks with Multiple Sources
    CHINACOM
    IEEE
    DOI: 10.4108/eai.15-8-2015.2260904
Rongkuan Liu1, Petar Popovski2, Gang Wang1,*
  • 1: Harbin Institute of Technology
  • 2: Aalborg University
*Contact email: gwang51@hit.edu.cn

Abstract

We consider a relay network with M nodes (terminals) that act as sources. The nodes send uplink data to a common destination through a relay that is equipped with a buffer. There are no direct links between each terminal and the destination. The nodes are backlogged and always have data to send. We develop a protocol which can make full use of the involved links adaptively, including the link selection and power allocation, based on the channel state information(CSI). We jointly optimize the link selection policy and power allocation for the average throughput maximization by limiting the total average power budget for all the nodes in the considered network. The simulation results show that the proposed protocol exceeds the buffer-aided multiuser relaying protocol without power allocation, especially in the region of a low power budget. It is also shown that power allocation deals efficiently with the bottleneck problem.