Research And System Simulation On Digital Down-conversion Of Software Radio

The object of the paper is software radio digital down-conversion. Abouthigh efficiency algorithm of each module in down-conversion, discussion,abstraction and summarization are given in this paper. And any arithmetic is puttogether so that system foundation is carried out in software. The system hascarried out the following operations. They are mixing, filtering, decimation,compensation and shaping of signal. The result shows that the system has achieveddown-conversion of signal and has no distortion phenomena and that the highefficiency arithmetic that is relatively independent each is unified in together. Theauthor carries on the beneficial exploration about reducing the frequencyconversion system operand in view of the ideal software radio bottleneck. Thepaper proposes the method of times-delay-compensated for the shaping filter andsolves the question which highly difficult designed because the order of shapingfilter is too high during down-conversion. This article also involves technicalutilization. They are the direct digital frequency synthesis (DDS), the interpolationsecond order polynomial (ISOP), and means of eliminating system stray and so on.This article main content abstract is as follows:Chapter 1 introduces the background, the significance and the research presentsituation of the article.The traditional communication system based on the hardware has been unableto solve the problem that many wireless communications system coexist currently.The thought of software radio arises. Software radio is basic hardware platform thatis composed by modular, the standardized hardware unit through the busconnection and realizes each kind of wireless correspondence function throughsoftware reloads. This system feasible plan has been produced. Digitaldown-conversion(DDC)is the main part of the software radio system. It alleviatesthe pressure of the DSP processing speed through the digital signal sampling ratetransformation. In the software radio system, digital down-conversion is the largestdigital processing operand part, and it requests the highest operating speed so it ismost difficult part to realize. So whether down-conversion can completely berealized through the software plays the decisive role on the ideal software radiorealization.Chapter 2 introduces theory foundation of digital down-conversion. It includesmilirate signal processing theory, filter design basic theory and highly effectivefilter basic principle.The decimation filtering is the important step in reducing signal samplingfrequency. The function of decimation makes the signal sample number reduced. Inthis paper, it need make the integral multiple to decimate. Therefore it does notinvolve the interpolation theory. The function of filtering reshapes the signal andmakes the frequency spectrum decimated not to be alias. The application of highlyeffective filter may reduce the system operand largely. The impulse response ofhalf-band filter is zero at even position except zero position, so it can save one halfoperands if it is used to realize 2-times decimation. Cascaded Integrator-Comb(CIC) filter suits high times decimation because it has no multiplication operationand the side lobe level weaken insufficiently that can be solved through the cascadeway.Chapter 3 gives the system demand. The project of down-conversion system isdecided. The structure is designed according to theory foundation.According to the function of down-conversion, the system should includethree parts. That is mixing, filtering and decimation. The mixing uses theorthogonal numeral mixing frequency technology;its goal lies in causing the signalfrequency spectrum moving and makes signal frequency reducing so thatdown-conversion to the zero center frequency. The NCO function is that it canproduce the same frequency orthogonal phase sine (cosine) signal. Through thesimulation comparison, the article confirmed that the method of multistagedecimation can greatly reduce the order of filter, its design difficulty and thesystem operand. Therefore, the design uses multistage decimation structure anduses CIC,HBR fully to make the function of down-conversion realized. About thelast level of reshaping filters, highly effective algorithm can not be found in manymaterial, commonly it can be completed by using ordinary FIR, but the FIRrequests high order so that it's difficulty to realize. The chapter 4 proposedtimes-delay-compensated method to solve this problem.The 4th chapter completed the optimized design and simulation analysis aboutfunction module of down-conversion system is made. It's the base of chapter 5.Now, various modules design situation is as follows:1 signal source moduleThe design takes the GSM system as a background;therefore it selects theGSM intermediate frequency signal as signal source.2 NCO moduleNCO module is the important part of digital mixing frequency. Thedown-conversion demands that the sine sample has the following characteristic:high frequency resolution;wide frequency band;quick frequency conversion rate.In order to satisfy the demand, the DDS technology is used. DDS is a newfrequency synthesis technology. It not only has the above characteristic but also hasthe prominent merit of phase continuously and the function expansion convenience.These characteristics conform to the software radio requirement. But during signalprocessing process because the ROM capacity has limit, there is phase quantizationerror. This design eliminates the orderly spurious component by input randomsequence. The simulation result proved that this method is effective. Meanwhile,the article also analyzes the method which judges the spurious free dynamic range(SFDR) of the NCO. Finally the Henning window is chosen to carry on thespectrum analysis3 module of decimation filteringThis module includes CIC, HBR and FIR reshaping filter design. Because thepass band of CIC is weakening oversized, this design uses the interpolation secondorder polynomial (ISOP) to compensate the filter. The article has analyzed thecompensation condition and the performance of ISOP through the simulation, theresult is that the optimization compensates effect has achieved when the value of cis smaller than -2, the c is equal -5.864, the k is equal 4. In order to make signalconform to the GSM system requirement after the frequency conversion, the FIRreshaping filter is been completed. But the high order is needed and realization isdifficultly and the operand is large.Chapter 5 In order to solve the problem educed above, times delay compen-sated design method is proposed.Times-delay-compensated design method lies inthe compression of impulse frequency spectrum. The method make the transitionband of the filter with the same order steeper under the condition that has no effecton the ripple of pass band and stop band. Because other frequency bands similarlycan be compressed, in the first Nyquist sector the pass band quantity will increase,therefore it needs compensation filter to revert the pass band in order to return tooriginal state. The filter only needs small order to be realized because the transitionband is comparatively slow. The article expounds the principle inferentialreasoning, the applicable scope and the design procedure. Compared with theordinary FIR reshaping filter, the method makes filter order reduces k times whileit gets the same reshaping effect. The k is compression multiple. We can see thatthe result of falling order is obvious. The method that is fit for the situation isexactly that traditional method feels difficult. The difficulty is that the useful signalband width is narrow and the signal sampling frequency is high. So the methodoffers a new way for signal shaping. It also provides a new idea for theintermediate frequency signal processing.Chapter 6 builds down-conversion system based on modules that are designedin the 4th chapter and confirms conversion frequency result. The system resultindicated that digital down-conversion is realized and the signal has no distortionphenomenon.Chapter 7 carries on the summary of this article and analyzes insufficiency inthe system. In the end the next research direction is given.