System design and optimizations for direct -sequence spread -spectrum multiple -access communications
In this thesis, system design and optimizations are considered for direct-sequence spread-spectrum multiple-access (DS/SSMA) communications at the physical layer. In the first work, the receiver optimization for the systems with long-code spreading sequences is performed. From the observation that the multiple-access interference is asymptotically Gaussian after despreading, an optimum spatio-temporal matched filter receiver that minimizes the probability of bit error is derived. The performance of the receiver is also analyzed. In the second work, a new spreading scheme and an accompanying blind adaptive multiuser detector is proposed. The new spreading scheme bridges the gap between the long-code and short-code spreading schemes, and leads to the computationally efficient multiuser detector in a frequency-selective channel which well models the typical wireless channel for DS/SSMA communications. In the third work, system parameter optimizations are performed for long-code systems with the conventional matched filter receiver. First, the optimum chip waveforms, that result in the minimum bit-error-rate for any power and delay profile of users, are found when the system bandwidth, the symbol excess bandwidth, and the chip excess bandwidth are given. Second, the optimum design strategies are derived to determine the processing gain and the chip waveform when the system bandwidth and the symbol excess bandwidth are given. In the final work, closed-form average bit-error-rate formulas are derived for long-code matched filter DS/SSMA systems. The formulas well reflect the filtering effect on the system performance, and make instant and accurate evaluations possible. ^
Major Professor: James S. Lehnert, Purdue University.
Engineering, Electronics and Electrical