基于全网时帧改变的TDMA Ad Hoc时隙分配协议

TDMA Ad Hoc time slot assignment protocol based on change of global time frame

针对现有时帧可变的时隙分配协议所存在的平等性差问题,提出一种基于全网时帧改变的时隙分配协议(GASAP)。首先通过研究全网时帧改变的收敛时间,为GASAP设计时帧结构;然后采用"先集中分配再补充分配"方式为网络节点分配控制时隙,使全网时帧结构可随入网节点的数量而变化;另外为了提高协议的信道利用率,又增加了时隙复用过程;最后针对协议中根节点失效问题设计了重组网过程。在GASAP与其他时隙分配协议的对比仿真中,当网络负载未超过协议的传输带宽时,GASAP的端到端时延远小于统一时隙分配协议(USAP),当节点数量不满足2的幂次方时,GASAP的吞吐量大于基于二叉树的帧长可变的动态时隙分配协议(BDSAP),说明了GASAP具有良好的网络传输性能;另外对GASAP的时隙复用进行仿真,结果表明协议在多跳拓扑中的信道利用率比在单跳拓扑中更高;最后对GASAP在根节点失效后的性能进行仿真,失去根节点的网络只需0. 5 s后即可重新完成组网,说明GASAP也具备高效的抗毁能力。

Aiming at the problem of poor fairness in the existing time-slot assignment protocol based on the change of time frame, a Global-frame-change-based Slot Assignment Protocol （GASAP） was designed. By studying the convergence time of frame changes in global network, the time frame structure was designed for GASAP. Then the approach of ＂concentrate distribution and then replenish distribution＂ was adopted to assign slots for network nodes. In addition, in order to improve the channel utilization of the protocol, the slot reuse process was designed. Finally, to solve the problem of root node failure in the protocol, the re-network process was designed. In the comparative simulation of GASAP with other time-slot assignment protocols, when the network load is less than the protocol＇s transmission bandwidth, the end-to-end delay of GASAP is much smaller than that of Unifying Slot Assignment Protocol （ USAP）, and when the number of network nodes does not satisfy the power of 2, the throughput of GASAP is greater than that of Binary-tree-based Dynamic Slot Assignment Protocol （ BDSAP）. Therefore, GASAP has excellent network transmission performance. In addition, the slot reuse performance is simulated, which proves that GASAP has higher channel utilization rate in multi-hop network topology. Finally, the re-network performance of GASAP is simulated. After losing the root node, the network can be re-established in only 0.5 second, which proves that GASAP also has an efficient reassembly capability.