Multi-Carrier CDMA Using Convolutional Coding and Interference Cancellation over Fading Channels

Abstract

The work in this thesis has been motivated by the challenges of third generation mobile communication systems and fixed wireless applications using high data rates. In particular, we have investigated receiver structures that employ the relevant coding techniques and channel equalisation strategies in multi-carrier CDMA systems and multipath faded channels. In fixed wireless access (FWA) and mobile cellular systems it has been shown that the correlation statistics on the link level between the subscriber and basestation can have dramatic effects on the inter-cell interference. The use of highly directional antennas and stationary users has shown that the effects of shadowing depend heavily on the correlation of the fading signals on the uplink. The anticipated inter-cell frequency re-use factor for realistic values of decay index (2 to 4) will be unlikely to drop below 0.85. For mobile cellular applications it has been found that this value can be as low as 0.2, therefore having a significant effect on the overall cellular capacity. The performance of low-rate orthogonal convolutional codes and novel low-rate hyper-orthogonal codes has been studied in a single-user and multi-user AWGN environment and multipath Rayleigh faded channels. Considerable performance improvements have been found for both coding strategies compared to conventional rate 1/2 convolutional coding strategies. The simulation of a rate 1/32, constraint length 5 low-rate orthogonal convolutional coding scheme in a 4-path Rayleigh faded channel with a Doppler of 300Hz has shown to provide extremely good coding gains. Compared to uncoded faded channels, a coding gain in the order of 6dB was found for bit-error rates of 10-5 using low-rate orthogonal convolutional coding structures with BPSK modulation. Using low-rate hyper-orthogonal coding with a constraint length of 8 and rate 1/64 has shown to increase the coding gain to over 10dB at bit-error rates of 10-6. Of course, this is achieved at the expense of increased hardware complexity. The primary advantage of MC-CDMA is its ability to operate in a high user rate system that has only a limited system bandwidth available for spreading. Because MC-CDMA strategically tries to reduce the effects of multi-user interference through sub-carriers and guard intervals, MC-CDMA does not become less efficient in terms of bits/s/Hz when the spreading factor is reduced, as in DS-CDMA systems. Therefore, the goal of MC-CDMA is to address the area of weakness of the DS-CDMA system by supporting higher data-rates. Since a MC-CDMA system is able to adjust the number of sub-carriers per user signal, MC-CDMA is highly flexible with respect to variable rate traffic. This thesis has analysed and simulated the effects of imperfect channel estimation, convolutional coding, time and frequency diversity requirements, equalisation, time-guard bands and interference cancellation using MC-CDMA on the uplink and downlink.