Research On Parameter Estimation Methods Of Hybrid DS/FH Spread Spectrum Signal

The hybrid Direct Sequence/Frequency Hopped Spread Spectrum (DS/FH SS)combines the advantages of the DSSS and the FHSS. With many characteristics such aslow probability of intercept, good anti-jam performance and anti-distance effects, it hasfound wide applications in military communications. This dissertation mainly dealswith the parameter estimation of DS/FH signal. The main contents are as follows:The parameter estimation of DS/FH signal based on cyclic spectrum is analyzed.According to the DS/FH signal characteristics and an existing frequency-jointtechnology, a decomposition-joint technology is developed to transform amulti-component DS/FH signal into a single-component DS signal. On the ground ofthis, a novel method which combines cyclic spectrum with the decomposition-jointtechnology is proposed to estimate the carrier frequency and the chip rate of DS/FHsignal. The proposed method can effectively improve the parameter estimationperformance and it needs no prior information.The S-transform (ST) based parameter estimation of DS/FH signal is studied. Inview of the problems in extracting the S-transform time-frequency ridge line, i.e., thedistortions in high frequency component caused by noise, an effective modifiedalgorithm is proposed. On this basis, two methods are used for the parameter estimationof DS/FH signal:1) original S-transform is improved by adding a parameter to improveits time resolution, and combines with the modified algorithm of the ridge line toestimate the hopping time exactly.2) A hyperbolic window is used to take the place ofthe Gaussian window of original S-transform, which can adjust time-frequencyconcentration. Then the hyperbolic S-transform combines with the modified algorithmof the ridge line to improve the performance of parameter estimation.