通信学报

• 学术论文 • 上一篇    下一篇

基于多门限预留机制的自适应带宽分配算法

陈 赓,夏玮玮,沈连丰   

  1. 东南大学 移动通信国家重点实验室,江苏 南京 210096
  • 出版日期:2014-12-25 发布日期:2014-12-15
  • 基金资助:
    国家科技重大专项基金资助项目(2012ZX03004005-003);国家自然科学基金资助项目(61171081, 61201175);江苏省科技型企业技术创新基金资助项目(BC2012006);东南大学移动通信国家重点实验室自主研究基金资助项目(2013A03)

Adaptive bandwidth allocation algorithm based on the multi-threshold reservation mechanism

  • Online:2014-12-25 Published:2014-12-15

摘要: 针对异构无线网络融合环境提出了一种基于多门限预留机制的自适应带宽分配算法,从而为多业务提供QoS保证。该算法采用多宿主传输机制,通过预设各个网络中不同业务的带宽分配门限,并基于各个网络中不同业务和用户的带宽分配矩阵,根据业务k支持的传输速率等级需求和网络状态的变化,将自适应带宽分配问题转化为一个动态优化问题并采用迭代方法来求解,在得到各个网络中不同业务和用户优化的带宽分配矩阵的同时,在带宽预留门限和网络容量的约束条件下实现网络实时吞吐量的最大化,以提高整个异构网络带宽的利用效率。数值仿真结果显示,所提算法能够支持满足QoS需求的传输速率等级,减小了新用户接入异构网络的阻塞概率,提高了平均用户接入率并将网络吞吐量最大提高40%。

Abstract: An adaptive bandwidth allocation algorithm based on multi-threshold reservation mechanism is proposed to guarantee the quality of service (QoS) of multi-services in heterogeneous wireless networks. With the bandwidth reservation mechanism by setting multi-threshold in each network for every traffic, the adaptive bandwidth allocation scheme according to the transmission rate levels requirements for traffic k and the network status can be formulated as an optimal problem with the constraints of the bandwidth allocation matrix for each traffic and all users based on the multi-homing technology. A procedure of the iterative method is presented to solve the optimal problem of the formulated adaptive bandwidth allocation scheme. The optimized bandwidth allocation matrixes for traffics and for all users are obtained for the real-time network throughput maximization under bandwidth reserving thresholds and network capacity constrains and the bandwidth resource utilization efficiency of heterogeneous wireless network can be increased. Numerical simulation results show that the proposed algorithm can support QoS requiring transmission rate grades, decrease the new call blocking probability, increase the average user access rate and improve the network throughput of the heterogeneous wireless network by 40% to the largest extent.

No Suggested Reading articles found!