Design Of The CDMA Baseband Signal Generator/Analyzer Man-machine Interface And The Signal Detection Algorithms

TDD-CDMA （Time Division Duplex Code Division Multiple Access） is one of the three access 3G standards, incorporating the advantages of TDD and CDMA, and employing some advanced techniques such as Smart Antenna and Joint Detection to offer a more flexible air interface. By developing a TDD-CDMA simulation software, one can carry out an efficient, powerful and interactive communication system investigation. The TDD-CDMA Signal Generator should be able to perform system parameter settings through its user friendly interface and to generate the corresponding TDD-CDMA signals. The Analyzer can demodulate the received signal, analyze some important parameters and evaluate the performance of the 3^rd or 4^th generation TDD—CDMA systems. The software signal generator and analyzer are the main focus of this thesis.The design and realization of the CDMA baseband signal generator and analyzer man-machine interface are given respectively in Chapter 2 and Chapter 3. The model of TDD-CDMA system transmitter and receiver are presented first. Then, the design scheme for the generator and analyzer are described based on the corresponding block diagram. Finally, the implementation of the software interface is given in detail with the flow chart related to each function module.Joint detection is a key technique in TDD-CDMA system, and is used mainly for eliminating MAI （Multiple Access Interference）, ISI （Inter-Symbol Interference） and Near-Far effects. It can also simplify power control, reduce power control precision, and remedy the negative influence caused by the nonideal crosscorrelation of the orthogonal spreading codes. Consequently, joint detection techniques can improve the system performance, increase the system capacity and extend the cell coverage. In the Chapter 4, we discuss first the mathematical model of the Joint Detection algorithms. Then three major algorithms of Joint Detection are discussed, and their performance and complexity comparison are presented as well.Due to system matrix’ dimension, the matrix inversion makes the implementation of JD （Joint Detection） receiver very complicated. In order to reduce the computational complexity, the approximate Cholesky factorization and Block-Fourier algorithm under ZF-BLE （Zero-Forcing Block Linear Equalizer） and MMSE-BLE （Minimum Mean Square Error Block Linear Equalizer） algorithms are presented in Chapter 5. An improvement is made to Block-Fourier algorithm, resulting a modified Block-Fourier algorithm. Then their performances under different conditions are compared according to the simulation results. It is shown that the modified Block-Fourier algorithm is superior to Cholesky factorization at high SNR region.