基于聚类分析的抗时延抖动时隙分配算法

信息传输的实时性是战术数据链的突出特征,要求数据链通信不仅有较好的时延特性,还要具备可靠的时延抖动特性。随机报文流的产生具有随机性、突发性的特点,需按照需求动态地分配时隙,对算法的实时性要求较高。因此,提出基于聚类分析的抗时延抖动时隙分配算法,应用聚类的方法将空闲时隙划分成时隙簇,简化了时隙分配的分析过程。实验仿真表明:该算法分配的时隙能够满足时延抖动的要求,而且算法的复杂度小,可以有效处理周期性随机报文流的时隙分配问题。

De-jitter Dynamic Slot Allocation Algorithm Based on Cluster Analysis

Instantaneity of information transmission is a prominent feature of tactical data link.Data chain communication is requited possessing good delay feature as well as reliable delay jitter performance.The generation of random message flow possesses characteristics of randomness and abruptness,and should dynamically allocate timeslot in accor-dance with demands,which requires a high real-time capability for algorithm.Therefore,a de-jitter dynamic slot allocation algorithm based on cluster analysis was proposed and the clustering method was adopted to divide idle timeslot into slot cluster,which can thus simplify the analysis process of slot allocation.The simulation result shows that timeslot allocated by the algorithm can meet the requirement of delay jitter and the algorithm is of low complexity which can settle timeslot assignment of periodic random message flow efficiently.