Research On Blind Estimation Of FH Signal Parameters

Since the advent of the1960s the first practical frequency-hoppingcommunication system BLADES, frequency hopping communication system hasalready experienced50years of continuous development, it has widely used inmilitary tactical wireless communications field, with the rapid development andinvestment application of the large scale integrated circuit technology、microelectronics technology and DFH, broadband receivers, frequency hoppingcommunications system is also leaps and bounds. It is an urgent that we need to findout a feasible and timely detection algorithm for the detection of the frequencyhopping communication signal of the enemy and access to relevant information,then pave the way for either eventual intercept or jamming enemy communications,It is a very important significance for both military operations and our homelandsecurity.Around the theme of Blind Estimation of the parameters of frequency hoppingsignals, first of all, this paper studied the theory of frequency hoppingcommunication system, and has constructed the wide interval of hopping sequencebased on the m sequence as the hopping code, do the simulation of the frequencyhopping signal with noise. And then this paper studied theoretical analysis andsimulation about some time-frequency analysis method which are often used in avariety of frequency-hopping signal parameter estimation, spectrum was selected tobe the time-frequency analysis tool to prerecession of the signal parameterestimation. If the signal for analysis is fast frequency hopping signal, a methodcalled joint FFT-peak detection algorithm was proposed to estimate the frequencyhopping cycle and frequency-hopping time, and the carrier frequency is estimatedby using the FFT method and statistical peak detection method based on thetime-frequency analysis, the assessment of estimation performance of the newalgorithm with the traditional reference algorithm is shown in this paper; If thesignal for analysis is slow frequency hopping signal, a detection algorithm based onthe value of peak segment center of gravity and valley detection algorithm for acertain amount of improvement is proposed to estimate the frequency-hopping cycleand frequency-hopping time. At last, this paper designed a module for the hop ratepre-estimate and it would do a good job to classify the frequency hopping signal bythe frequency-hopping rate, and then the estimation system will select anappropriate algorithm to estimate the parameters. The result of the parameter estimation shows that by the same time of reducingthe estimation error, the new estimation algorithm proposed in this paper can also beextended to the applications case of low SNR, analysis showed that the estimation offast frequency hopping signal parameter will do a good estimation performancewhen the SNR is greater than-7dB;and the estimation of fast frequency hoppingsignal parameter will do a good estimation performance when the SNR is greaterthan-8dB; the hop rate estimate module can help to adaptively select the estimationalgorithm, so the parameter estimation could be an organic whole.