Traffic Management in Wireless ATM Networks

这个研究合作项目由NCR-9714702、加州州立大学北岭分校的Jaime B. Kim和NCR-9714700、圣何塞州立大学的W. Melody Moh共同完成，旨在研究无线ATM网络中的流量管理。它包括以下两个阶段：第一阶段：基本流量控制在这个初始阶段，我们将重点关注无线ATM网络的两个基本流量管理问题：提出了一种能够为微蜂窝无线ATM网络中承载的多媒体业务提供高度QoS保证的接入控制方案。通过不断监测当前网络流量负载和调整预留带宽，可以适应不断变化的网络条件。2. 我们提出了两种新的支点控制机制：ISQ（忽略慢速请求）和WS（加权和）方案。这两种方案都可以很容易地应用于现有的多播ABR流控制机制，其中大多数采用LSQ （Listen the最慢请求）原则。我们还为非实时流量提出了两种完整的多播流控制机制，每种机制都包括交换机、分支点、目的地和源的特定控制。第二阶段：基础交通控制的整合与扩展这一阶段将重点关注基础交通控制机制的整合，以实现完整交通管理方案的最终目标。为此，第一期基本交通管制项目将向以下方向扩展：组播连接的端到端准入控制实时流量组播流控制结合准入控制、动态带宽分配和流量控制的综合流量管理对于两个阶段的每个子项目，我们都确定了需要调查的研究问题，并提出了分析和仿真方法。我们相信这项研究对下一代宽带无线网络至关重要。

This research collaboration project between NCR-9714702, Jaime B. Kim, California State University, Northridge, and NCR-9714700, W. Melody Moh, San Jose State University, aims to study traffic management in wireless ATM networks. It includes the following two phases: PHASE 1: BASIC TRAFFIC CONTROL In this initial phase, we will focus on two basic traffic management issues for wireless ATM networks: 1. Admission Control We propose to develop admission control schemes which can provide a high degree of QoS guarantee for the multimedia traffic carried in micro-cellular, wireless ATM networks. The proposed schemes can also adapt to the changing network conditions by constantly monitoring current network traffic load and adjusting the amount of reserved bandwidth. 2. Multicasting Flow Control We propose two new branch-point control mechanisms: the ISQ (Ignore the Slow reQuests) and the WS (Weighted Sum) schemes. Both schemes can be readily applied to the existing multicasting ABR flow control mechanisms, most of which adopt the LSQ (Listen to the Slowest reQuests) principle. We also propose two complete multicasting flow control mechanisms for non-real-time traffic, each includes specific controls at switches, branch-points, destinations, as well as the source. PHASE 2: INTEGRATED AND EXTENSION OF BASIC TRAFFIC CONTROL This phase will focus on integration of basic traffic control mechanisms to achieve the ultimate goal of a complete traffic management scheme. For this purpose, extensions on basic traffic control projects of Phase 1 will be carried out in the following directions: 1. End-to-end admission control for multicasting connections 2. Multicasting flow control for real-time traffic 3. Integrated traffic management by combining admission control, dynamic bandwidth allocation, and flow control For each of the sub-projects in both phases, we have identified research issues for investigation and presented analytical and simulation me thods. We believe that the proposed research is vital to the next generation of broadband wireless networks.