Wireless Internet. 9th International Conference, WICON 2016, Haikou, China, December 19-20, 2016, Proceedings

Research Article

An Improved Dynamic Clustering Algorithm Based on Uplink Capacity Analysis in Ultra-Dense Network System

Download
199 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-72998-5_23,
        author={Jie Zeng and Qi Zhang and Xin Su and Liping Rong},
        title={An Improved Dynamic Clustering Algorithm Based on Uplink Capacity Analysis in Ultra-Dense Network System},
        proceedings={Wireless Internet. 9th International Conference, WICON 2016, Haikou, China, December 19-20, 2016, Proceedings},
        proceedings_a={WICON},
        year={2018},
        month={1},
        keywords={Ultra-dense network Uplink capacity Dynamic clustering algorithm},
        doi={10.1007/978-3-319-72998-5_23}
    }
    
  • Jie Zeng
    Qi Zhang
    Xin Su
    Liping Rong
    Year: 2018
    An Improved Dynamic Clustering Algorithm Based on Uplink Capacity Analysis in Ultra-Dense Network System
    WICON
    Springer
    DOI: 10.1007/978-3-319-72998-5_23
Jie Zeng1,*, Qi Zhang1, Xin Su1, Liping Rong1
  • 1: Tsinghua University
*Contact email: zengjie@tsinghua.edu.cn

Abstract

The Ultra-Dense Network (UDN) system is considered as a promising technology in the future wireless communication. Different from the existing heterogeneous network, UDN has a smaller cell radius and a new network structure. The core concept of UDN is to deploy the Low Power Base Stations (LPBSs). With denser cells, the interference scenario is even severer in UDN than Long Term Evolution (LTE) heterogeneous network. Clustering cooperation should reduce interference among the cells. In this paper, we firstly derive the total uplink capacity of the whole system. Then we present a novel dynamic clustering algorithm. The objective of this algorithm for densely deployed small cell network is to make a better tradeoff between the system performance and complexity, while overcome the inter-Mobile Station (MS) interference. Simulation results show that our approach yields significant capacity gains when compared with some proposed clustering algorithms.