Wireless Mesh Network Bandwidth Allocation And Adjustment Mechanism

Wireless mesh network (WMN) is a kind of multi-hop and self-organizing network. With the advantages of large capacity, flexible networking and low cost, WMN has become a hot research issue in next-generation broadband wireless access technology. In order to use the limited WMN bandwidth resource in a fair and efficient way, bandwidth allocation and congestion adjustment are studied in this thesis.Firstly, a Queue-Weight-based fair bandwidth allocation algorithm (QWFBA) is proposed. QWFBA works in a flow-based round-robin fashion. The traffic flows requesting or granting bandwidth are buffered in different virtual queues. The weight factor of one queue is updated periodically, and it is equal to a proportion of the bandwidth requested by this queue to the bandwidth requested by all traffic flows. According to the queue weight and the frame grant horizon determined by network control latency, QWFBA allocates more bandwidth to the links with heavier load. Thereby, the load among different traffic flows is balanced, and the fairness is enhanced as well.Secondly, a Hop-by-Hop State-Transition-based mechanism for rate adjustment (HSTRAM) is presented. HSTRAM works in a hop-by-hop mode. The process of rate adjustment is seen as an event triggered finite state mechanism. According to their cache utilization ratios, the mesh nodes transfer among four states and inform their neighbors to change rates to avoid congestion and decrease resource waste. In addition, a hop-by-hop reliability mechanism is incorporated with an end-to-end negative acknowledgment scheme in HSTRAM.Finally, performance evaluation of QWFBA and HSTRAM under different topologies using NS2 is given. The results show that they can achieve high performance in terms of fairness and network throughput.