A Studey On Mitigation Of Multiple Access Interference In Optical Code Division Multiple Access Communication Systems

As a scheme of multiple access, optical code division multiple access(OCDMA) technique has become a hot topic for local area and optical access network for excellent anti-interference, high security, large capacity and fully making use of bandwidth in optical fibers. Since unipolar code sequences are not completely orthogonal, OCDMA systems suffer from multiple access interference (MAI) which results in deterioration in the system performance. To reduce the impact of MAI on OCDMA systems efficiently, the thesis investigate two methods, parallel interference cancellation and decorrelating detection.To reduce the effect of MAI on one-dimensional OCDMA systems, parallel interference cancellation technique is adopted. The bit error rate is developed detailedly. The system performance is analyzed and is compared with that of correlation receiver. The modified PIC receiver and the PIC receiver with optical hard limiter are proposed to eliminate MAI more effectively. The bit error rate is derived and the system performance is analyzed. The minimal code length with different kinds of receivers under BER≤10^-9 are compared.To reduce the effect of MAI on two-dimensional OCDMA systems, decorrelating detection is adopted. Taking in account the effect of MAI, the dark current, and the Poisson characteristics of the optical direct detection process, the bit error rate of two-dimensional OCDMA systems with decorrleating receivers is derived by adopting the maximum posteriori probability rule. The system performance is analyzed detailedly. To reduce MAI more efficiently, the parallel decorrelaing receiver is proposed.It is shown that: in one-dimensional OCDMA systems, the PIC receiver, the modified PIC receiver and the PIC receiver with an optical hard limiter can effectively reduce the impact of MAI, lower the bit error rate and enhance system capacity; the performance of the two-dimensional OCDMA system can also be greatly improved by adopting the decorrelating detection and the parallel decorrelating detection. This is because the parallel interference cancellation and the decorrelaing detection recover data signals of the desired user by making use of information about undesired users which cause interference on the desired user.