基于时分多址(TDD)的无线mesh网络是实现无线多跳网络和宽带无线接入的一种关键技术.为用户提供一种可以保障稳定通信的有效的分布式算法是影响基于TDD的无线mesh网络性能的关键因素.本文针对TDD的无线mesh网络提出一种随机自学习分布式调度算法,这种算法是一种基于信息学习的随机选择算法.在网络中,任一节点都要根据其邻居节点控制消息中所携带的数据,学习邻居节点的调度信息,从而判断可用时隙.若上次信息传输成功,则节点仍然沿用上次的传输时隙;反之,节点在可用时隙中随机选择一个时隙发送控制消息.仿真结果表明,与IEEE802.16标准中定义的算法相比,提出的随机自学习分布式调度算法能实现更高的吞吐量.
Time division duplex based wireless mesh networks(TDD-based WMNs) have emerged as a crucial technology for wireless multihop networks and broadband wireless access(BWA).The resource management in TDD-based WMNs is important to the network performance and one of the main challenges is to provide an efficient distributed scheduling scheme to achieve robust communication for all nodes.In this paper,a randomized learning scheduling scheme is proposed for TDD-based WMNs.The scheme is a random selection scheduling scheme based on information learning. In the network,the node decides the available slots according to the slots information learning from its neighbor and selects a control slot from its available slots to transmit the scheduling message.Numerical results show that the proposed scheduling scheme achieves higher throughput than the scheme presented in IEEE 802.16 standard.