Research And Realization For Narrow-Band Interference Suppression Technology Based On DSSS Communication System

The technique of spreading spectrum has got rapid progress because of its inherentanti-jamming character, especially DSSS communication system, which is one oftypical spreading spectrum techinques. The anti-jamming ability of a DSSS system isdecided by its processing gain, which can be increased by augmenting the bandwidth.However, processing gain is always finite because there is some restrictions inbandwidth and technique in a practical system.A narrowband interference (NBI) has agreater power spectral density and thus will bring more strong devastation to the DSSSsystem than a broadband one do.So it is difficult for DSSS system to ensure a highquality of communications by itself when strong interference exist, evencommunications may be intermitted. Therefore researching more efficient techniques tosuppress the narrowband interference has great practical significance. Many kinds of filter techniques are presented in this paper, e.g. linear andnonlinear filters in the time domain, filter in the frequency domain and filter usingwavelet packet transform.These traditional filters have some limitations and defects forsuppressing narrowband interference.Commonly, adaptive filters in the time domain useLMS, RLS algorithms, but it is difficuit for these algorithms to complete iteration intime when sampling speed is too high, and which cannot be used to deal with fastchanging interference because of their slow constringency character. Filters based onDFT have truncation error, which make their spectrum have large side petals, sointerference cannot be suppressed entirely. Suppressing interference using waveletpacket transform is too complex, so it is very difficuit to realize real-time processing. Through comparing and summarizing the traditional filters, a novel adaptiveinterference suppression scheme based on DFT and time domain notch filters is putforward. In this scheme, the signal frequency spectrum is analysed fristly and athreshold is calculated according to the statistical characteristics of the analysis results,and then the value of a specific signal is compared with the threshold and theparameters, such as frequency and width, of the interference are distinguished. Since theparameters of interference have been confirmed, the DSP controller controls the notchfilters to suppress the interference, and notch parameters can be adjusted according tothe instance. The notch filters,which adopts the heterodyne-modulation technology, can notch II重庆大学硕士学位论文 英文摘要interference of any frequency and can adjust notch parameters such as width anddepth.The FFT processor adopts pipelining structure and uses an algorithm of N-pointcomplex DFT to compute 2N-point real DFT,which improves the FFT's workingfrequency much. Notch filters and frequency analysis were realized using FPGA. Therealization of threshold algorithm and control of the system are based on DSPs, so thesystem will react rapidly and interference can be suppressed in 100ms.