| 摘要: |
| 针对无源异类传感器接收的信号常在持续时间上无交集的情况，提出了一种基于出现-消失时间序列的信号关联算法。首先将时间序列抽象成一个M/G/1排队模型，随后推导了闲期分布，然后依据参考信号的闲期和休假期分布的不同，确定了关联时间窗口，统计N个关联时间窗口内待关联信号出现的频次作为信号关联的依据。仿真实验表明，关联信号与参考信号同时出现的概率（>81%）远高于非关联信号与参考信号同时出现的概率（<39%），验证了所提方法的有效性。 |
| 关键词: 无源异类传感器 信号关联 排队论 闲期分布 出现-消失时间序列 |
| DOI： |
|
| 基金项目: |
|
| Signal association based on time series of appearance and disappearance |
| PAN Hairui,LU Yourong,CAI Qian,XU Gang |
| () |
| Abstract: |
| For the problem that signals from passive-heterogeneous sensors often have no intersection in time,an algorithm based on time series of appearance and disappearance is proposed. Firstly,the time series is abstracted into queuing system M/G/1 and the distribution of idle period is derived.Then the windows associated on time of reference signal are determined by the different distribution between idle period and vacant period. The number of each signal appears within N windows is regarded as the basis of association.Simulation results show that the probability(>81%) of associated signals appearance at the same window is higher than the probability(<39%) of disassociated signals,which proves the effectiveness of the proposed algorithm. |
| Key words: passive-heterogeneous sensor signal association queuing theory distribution of idle period time series of appearance and disappearance |
