Research On Qos Scheduling Algorithm In Wimax Mesh Network

With the development of wireless communication technology, broadband wireless access has drawn considerable attention. WiMAX has been widely concerned due to its greater coverage, higher transmission rate and more reliable transmission performance. WiMAX Mesh network, with its multi-hop and self-organization capability, is becoming the mainstream of the WMN network. However, many difficulties still exist in resource scheduling and allocating, as well as QoS guarantees. Although the standard defines a series of QoS guarantees in PMP network, there is no specific definition for business classification and QoS levels in Mesh network. As a result, it is difficult to achieve the end to end QoS requirements for users.IEEE802.16standard for WiMAX Mesh network defines two kinds of scheduling mechanism:the centralized scheduling and the distributed scheduling. In centralized scheduling, all of the bandwidth requests of nodes should be gathered to BS, then scheduled and allocated by BS. In distributed scheduling, nodes negotiate bandwidth allocation by each other, using three-way handshake to establish connection. However, the bandwidth scheduling in both of these two mechanisms burdens the overall network performance. In this thesis, we analyze the two scheduling algorithms respectively for Mesh network, and propose new algorithm to improve the performance.This thesis takes advantage of both the minimum number and the minimum conflict routing algorithm for centralized scheduling, proposes a routing remodeling method to construct the centralized scheduling tree. The purpose is to reduce the conflict, while at the same time increase the possibility of concurrent transmission between nodes, so as to improve the overall network throughput. Then we propose a packet-based centralized QoS scheduling algorithm on the basis of scheduling tree. The PMP packet scheduling is involved into the centralized scheduling by distinguishing the priority of business data flow. The weight of each node is calculated to ensure the high-transmission and low-transmission requirements for fairness in the business flow of the same category. Compared with the HRF and the LTD algorithm, the simulation results show that our proposed centralized scheduling algorithm meets the different needs for business QoS, and performs better on both delay and throughput.A QoS slot allocation algorithm based on resource reservation is proposed for distributed scheduling. This algorithm categorizes the business data into high-priority and low-priority groups. A proper time slot is reserved in each frame of high-priority business. The size of the slot can be adjusted in real time in accordance with the network status. The advantage is that while the QoS of high-priority business data is guaranteed, the business failure rate and the time slot utilization rate can also be assured. The simulation results show that our proposed method has better performance in the time utilization rate and the delay compared with several known time slot reservation algorithms.