Methods and Experience of Using Matlab and FPGA For Teaching Practice in Digital Signal Processing Zhang Yuxi, Kang Li, Wang Jun, Sun Jinping, Wang Zulin School of Electronic and Information Engineering Beihang University (BUAA) Beijing, P. R. China E-mail:yxzhang.mr@gmail.com Abstract—Digital Signal Processing is a curriculum closely integrated by theory, implementation and application. With the development of microelectronics technology in recent years, the emergence of variety of chips makes digital signal processing widely used in various fields. Therefore, almost all of the electronic and computer engineering departments are now offering the digital signal processing courses. The main content of this course is abstract signal processing and transforming, involving a large number of mathematical knowledge and basic theory. The Matlab software can make the signal be presented in the form of visualized graphic image. Besides, FPGA, DSP and other high-performance chips can make digital signal processing technology widely used in various fields. Leading Matlab and FPGA into the teaching of digital signal processing is useful to improve the students’ mastery of the knowledge points and enhance interest in learning. In this paper, for the purpose of radar real-time processing system application, methods of using Matlab and FPGA for teaching knowledge points such as bandpass sampling, polyphase filters, FFT processing, etc. are introduced. Keywords-Digital Signal Processing; Matlab; FPGA; FFT; Pulse Compression I. INTRODUCTION The term RADAR is the acronym for Radio Detection And Ranging, namely using electromagnetic waves to detect and identify the range. A radar system is formed by transmitter and receiver. The transmitter emits radio-frequency waves. When they come into contact with an object they are partly reflected back. The weak echo signal returned to the receiver is amplified through RF amplifier and converted to IF signal through signal down-conversion circuit. Then ADC turns the analog signal into digital signal. The distance to the object and other information is obtained by digital signal processing afterwards. With the emergence of high-resolution radar, the linear frequency modulation (LFM) signal is widely used, not only improving the accuracy of radar range, but also increasing the maximum range [1, 2]. τ i τ Transmit Signals Echo Signals IF Sampling IF Signal Distance to target Digital I/Q DDC Digital Pulse Compression Figure 1. Signal Processing of the Radar System In radar system, the part starting from the ADC processing belongs to the scope of digital signal processing. ADC completed the conversion from analog signal to digital signal, and the distance information of the target is obtained after digital down-conversion (DDC), pulse compression. Now, with the DDC and pulse compression, the bandpass sampling, polyphase filter and FFT are explained as follows. II. POLYPHASE FILTER AND FFT THEORY A. Bandpass Sampling and Polyphase Filter According to Nyquist bandpass sampling theorem, in order to avoid signal spectrum aliasing, the sampling rate should be 2 s f B ≥ , and ( ) c 4 2 1 s f f m = + . Where m presents any positive integer, c f presents the center frequency, B presents the signal bandwidth. Now suppose the input signal is () ( ) cos(2 ( )) c rt At ft t π φ = + , and sample the signal with the rate S f to get the sequence as () ( ) cos[2 ( )] (2 1) ( ) cos[2 ( )] 4 (2 1) ( ) cos ( ) cos( ) 2 (2 1) ( )sin ( ) sin( ) 2 (2 1) (2 1) ( ) cos( ) ( ) sin( ) 2 2 c s f rn An n n f m An n n m An n n m An n n m m In n Qn n π φ π φ φ π φ π π π = + + = + + = + - + + = - (1) Where [] [ ]cos[ ( )] In An n ϕ = is the in-phase component of the EDUCATION REFORM FUND NUMBER OF BEIHANG UNIVERSITY: 40024906 2010 International Conference on Education and Management Technology（ICEMT 2010） 414 978-1-4244-8618-2/10/$26.00  2010 IEEE