多信道有限缓存下混合业务接入控制的排队性能研究

Research on queue performance of multi-service admission control under multi-channel and finite buffer

中文关键词  混合业务  网络拥塞控制  有限缓存  离散时间排队

英文关键词  multi-service  congestion control  finite buffer: discrete-time queue

中文摘要

      基于无线通信网的运行机制和实时性、非实时性混合业务对无线通信网络的不同服务质量(QoS)要求, 设 计了有限缓存下实时性业务优先接入控制策略, 并建立了无线通信网络在有限缓存下混合业务带有优先接入控制 的多服务台离散时间排队模型Geom1  Geom2/Geom1, Geom2/s, s(PP)/n + s; s. 精确划分了系统的状态组成, 分析 了系统性能和状态转移概率, 获得了系统的稳态分布、平均队长、损失率、以及信道的平均利用率. 最后, 对所建模 型进行数值模拟, 得到各业务到达率、服务率、信道个数、缓冲器容量与各业务损失率的关系. 针对网络拥塞控制, 本文提出通过增加信道个数、提高优先级业务的服务率能有效地降低两类业务的损失率.

英文摘要

      Based on the operating mechanism of network, and the different QoS requirements of real-time and nonreal- time service for wireless communication network, an admission control scheme with priority for real-time service under finite buffer was designed, and a multi-server discrete-time queuing model with priority under finite buffer Geom1  Geom2/Geom1, Geom2/s, s(PP)/n+s; s was established for multi-service admission control. The state composition of the system was accurately divided, and system performance and state transition probability were analyzed, then the steady-state distribution, average queue length, loss rate and average channel utilization were obtained. Finally, numerical simulation was carried out to analyze the relationship between arrival rate, service rate, channel number, buffer capacity and loss rate. For network congestion control, this paper proposes that the loss rate of the two types of service can be effectively reduced by increasing the number of channels and improving the service rate of the traffic with priority.