Tài liệu tham khảo giáo trình tìm hiểu xử lý tín hiệu số bằng tiếng anh ( Understanding digital signal processing ) Chương 7. Lấy mẫu kỹ thuật nâng cao | CHAPTER SEVEN Advanced Sampling Techniques Beyond the low-pass and bandpass sampling schemes discussed in Chapter 2 there are specialized sampling techniques that we re likely to encounter in practice. In this chapter we introduce several of those advanced sampling methods and show why and when they re used. Quadrature Sampling Many digital signal processing applications use the complex data format with its real in-phase and imaginary quadrature phase parts. These applications fall in the general category known as quadrature processing. For example the phase preservation characteristic of complex data representation is exploited in high data rate digital communications systems radar systems and time difference of arrival TDOA processing in radio direction-finding schemes l 2 . The enhanced phase measurement capabilities of quadrature processing are used in coherent pulse measurement systems antenna beamforming applications and single sideband modulators 3-6 . Complex data representation allows us to realize additional processing power through the coherent measurement of the phase of sinusoids comprising an input signal. To obtain the time-domain representation of a continuous signal using complex notation the signal must be digitized by a technique called quadrature sampling Quadrature sampling goes by various other names in the literature such as vector demodulation complex demodulation complex down-conversion quadrature heterodyning I Q sampling and complex sampling 7-9 . Care must be exercised here from a semantic viewpoint. There s a testing scheme referred to as coherent sampling used to characterize the dynamic performance of A D converters that is unrelated to quadrature sampling 10 ll . This clever A D testing technique is discussed in Section . 297 298 Advanced Sampling Techniques Figure 7-1 In-phase 1 f and quadrature-phase q f components of a continuous signal x f . To illustrate this idea let s consider a continuous time-domain signal x f that .
