Radar Data Processing Based On Wavelet And DSP

This thesis is focused on radar signal processing based on wavelet and DSP. At the beginning, we study the characteristics of radar signal processing and the analysis tool(wavelet) and the digital signal processing tool(DSP), Then we propose an full system for radar image compression and transmission. This system consists of four main parts: data acquisition, A/D conversion, DSP processing system and PCI communication, which are integrated into a card. Each part of the system is profoundly studied respectively.Data acquisition module consists of high speed data conversion(A/D) and memory, The system adopts the scheme of four-channel A/D composite converter, which reduces the difficulty of the circuit implementation and increases the stability of the system , and with which we have solved the universal problem that the speed of data acquisition was too slow. The interface method of direct memory map is also applied for purposes of realizing the sharing of the extended acquisition data memory between high speed data acquisition system and DSP system.DSP data processing module based on DSP performs such factions as preprocessing radar data, wavelet transform, encoding algorithm of the SPIHT(set partitioning in hierarchical trees) and adaptive arithmetic coding. Firstly, in this part the author investigates the characteristics of DSP and its superiority for Radar Signal Processing System. The TMS320C6000 series chips of Tl Company is introduced. The highest clock frequency can reach 200MHz and processing capability is as high as 1600MIPs (1600 million times interaction per second) when all eight functional units are running. So it can meet the demands of real-time processing of radar signal basically. Secondly, it introduces how to realize radar data preprocessing by merging radar scan lines into radar image based on its cycle, according to the characteristics of radar data and the demands of the wavelet transformation.Then, by in-depth research and analysis on the theory of wavelet transformation, the author found that the Mallat algorithm of wavelet transformation is better for radar image coding. Finally, a further research is made for the EZW and the SPIHT by analyzing the instances and comparing the concrete procedures of these two encoding algorithmis, which concludes that under the same conditions, SPIHT is not only simpler but also using less memory. Therefore the author adopts SPIHT to encode the wavelet coefficients.PCI communication module is PCI interface in a domination/subordination pattern, which the C6000 series of DSP contains. In this paper, by using PCI bus it resolves the bottleneck problem of communication between the DSP and the hostThe software part illustrates how to get the radar reconstructed images under different compression ratio by loading the programming algorithm to DSP programming memorizer and executing it through CCS - the integrated developmental environment provided by Tl Company. The programming algorithm is evaluated by analyzing the reconstructed imagesBy analyzing and comparing the compression algorithms mentioned in this thesis with that from other reference, the last part shows that using the above method will generate higher RSNR than using EZW method under the condition of low bit ratio.The research on radar video data processing in this thesis is of significant applicational value and prospect. It provides a method to resolve the bottleneck problem of microwave and satellite communications .