摘要
直扩是水声网络中常用的通信技术,对非合作直扩通信信号的检测是对其实施干扰的前提.倒谱检测具有计算速度快的优点,得到了广泛应用.传统的倒谱检测方法基于倒谱集平均和单根谱线能量,易受脉冲噪声干扰使得虚警概率高,低信噪比时谱线淹没于噪声中使得检测概率低.新方法利用分段数据的功率谱集平均求解倒谱,提升了倒谱域低端谱线,并以低端一定采样宽度的倒谱幅度平均值与倒谱整体幅度平均值的比值作为检测量,在与传统方法相同的虚警概率下,提高了检测概率.仿真表明了新方法工作性能的提高.
Direct sequence spread spectrum（DSSS） is a widely-used communication technique in underwater acoustic networks.Detection of non-cooperative signals is a precondition when jamming DSSS communications.The cepstrum method is popular in DSSS signal detection because it can be rapidly computed,yet it has serious limitations.The traditional cepstrum approach is based on the cepstrum average and single spectral line energy,thus it is vulnerable to pulse noise which causes a high probability of false alarms.In addition,the spectral line is buried in noise in low signal-to-noise ratio conditions,and this causes a low probability of detection.A novel approach obtains the cepstrum results from a power spectral density averaged result from segmented data,thus lifting spectral lines in the low cepstrum domain.It also takes the ratio of averaged cepstrum amplitudes in the low cepstrum domain and global averaged cepstrum amplitude as a test quantity,thus we obtained a higher detection probability with the same false alarm probability as the traditional approach.The performance improvement was validated by a computer simulation.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2010年第7期863-867,共5页
Journal of Harbin Engineering University
关键词
水声通信信号
直扩
倒谱
非合作信号检测
underwater acoustic communication signal
DSSS
cepstrum
non-cooperative signal detection
