Research On The TDOA Estimation Of Frequency Hopping Signal Based On Compressive Sampling

Since the date of the birth of frequency hopping technology, with the advantages of anti-jamming, anti-interception, low probability of intercept, it has been widely used in military communications and has gradually infiltrated the civilian communications. The current state of development of frequency hopping signal is:frequency hopping speed is getting faster and faster, frequency hopping bandwidth is becoming wider and wider and hopping pattern is designed more and more complex. At present, many countries have developed a variety of military frequency hopping radios and applied them into the modern war; this poses a severe challenge for non-cooperative frequency hopping signal detection and localization. So, find an effective positioning method of frequency hopping signals has become an imperative job in the field of Communication countermeasure.Time difference of arrival (TDOA) positioning, with its passive, high positioning accuracy, low cost equipment modification, concealment, and other characteristics, becomes the first choice of many positioning methods. It's well known that high-precision TDOA estimation is the key of TDOA positioning; the study of accurate TDOA estimation is the fundamental problem of passive positioning of frequency hopping signals.There are several difficulties in TDOA positioning of frequency hopping signals. For example, wideband frequency-hopping signal sampling needs extremely high Nyquist rate which leads to problem that monitoring stations have to transmit large amounts of data; and, sorting and identification of frequency hopping signals are also big problems because of the complex frequency hopping communications environment; still, fast and efficient algorithm for TDOA estimation still needs further research.The problems describe above have an effective solution when compressed sampling appears. Compressed sampling can efficiently capture sparse signals using a small amount of non-related measurements and also ensure the reconstruction of original information based on the small number of measurements. Compressed sampling does not only ease the pressure of ADC and digital signal processing algorithms, but also opened up a new direction of signal acquisition and processing. According to the problems of frequency hopping signals positioning, this paper introduced compressed sampling into the acquisition and processing of frequency hopping signals; paper proposed posterior compressive sampling of frequency hopping signal and a new TDOA estimation method which based on AIC information and sparse decomposition.This paper first introduced the mainstream technology of current passive location; analyzed the principle and advantages of TDOA positioning; according to the characteristics of frequency hopping signals, it pointed out TDOA positioning problems of non-cooperative frequency hopping signals. Then paper introduced compressed sampling which is the popular signal processing technology at present; then made a detailed analysis of several major issues of compressive sampling; after that, according to the sparsity of frequency domain of wideband hopping signal, paper applied compressive sampling into the efficient acquisition of frequency hopping signals. Then we introduced the two typical implementation structures of compressive sampling, and did a performance comparison simulation between them. Then, based on the complexity of frequency hopping communications environment, content proposed a method named posterior compressive sampling, both eliminate the interference of fixed frequency signal and ensures the efficient acquisition of frequency hopping signal. The last chapter first analyzed the relationship between TDOA signals when they do sparse representations; then paper proposed a completely new method of frequency-hopping signal TDOA estimation; the new method not only avoided the bottleneck of Nyquist sampling rate, but also completed TDOA estimation of frequency-hopping signals at the expense of Low sampling rate and low complexity; finally, simulation results verified the feasibility of the new method in comparison with another two traditional TDOA estimation methods.