In recent years, there has been significant research and development effort for Third-generation or 3G wireless and mobile communication systems. Such systems will be able to provide end-user multimedia services data rates of up to 384 kbps for wide area coverage and up to 2 Mbps for local area coverage. They will also significantly expand the range of options available to users and allow communication, information, multimedia and entertainment services to be delivered via wireless terminals. Of the various air interfaces, Code Division Multiple Access (CDMA) has been regarded as an important part of the third generation systems because of its high frequency utilization and suitability for handling multimedia communications....[ Read more ]

In recent years, there has been significant research and development effort for Third-generation or 3G wireless and mobile communication systems. Such systems will be able to provide end-user multimedia services data rates of up to 384 kbps for wide area coverage and up to 2 Mbps for local area coverage. They will also significantly expand the range of options available to users and allow communication, information, multimedia and entertainment services to be delivered via wireless terminals. Of the various air interfaces, Code Division Multiple Access (CDMA) has been regarded as an important part of the third generation systems because of its high frequency utilization and suitability for handling multimedia communications.

Due to its inherent multiple access interference, the performance of CDMA can be severely degraded as the number of subscribers increases. To fully utilize the potential of wideband CDMA systems, multiuser detection can be used. In recent years, multiuser detection has been subject to extensive research and various schemes have been proposed. In this thesis, we will investigate and analyze the performance of multiuser multistage interference cancellation in wideband CDMA system where the bandwidth expansion factor is very small due to the high rate transmission required in the bandwidth limited wireless channel.

It is observed that the effectiveness of multiuser interference cancellation depends on the correct estimation of the amount of interference. It is therefore clear that we must have good data estimates. To do so, we combine the multistage interference cancellation with Viterbi decoding, which we shall refer to as joint successive interference cancellation (JSIC) where Viterbi decoding is included in each step of the multistage interference cancellation scheme. Obviously, a key disadvantage of this scheme is the receiver complexity and delay. To reduce the amount of complexity and delays, a memory truncated Viterbi decoder must be employed. It turns out that the performance can be significantly degraded by simply employing the memory truncated Viterbi decoder and encoder. To solve this problem, we propose an improved memory truncated decoding and encoding process, which is shown to significantly reduce the decoding delay and complexity while the performance remains about the same.

In addition to the above multiuser interference cancellation method, this thesis also presents a new interference cancellation method which employs a combined coding, successive interference cancellation and random interleaving. In the proposed method, which we shall refer to it as Joint Successive Interference Cancellation with Interleaving (JSICI), different random interleavers are employed for each user so that each user's data bits are re-ordered differently before transmission. At the receiver, a memory truncated Viterbi decoder (MTVD) with memory truncation size S and a channel symbol generator are employed. The MTVD will trace back along the trellis for a depth S to obtain an estimated information bit. The channel symbol generator will make use of this estimated information bit to regenerate the estimated channel symbols. This process is implemented for all the users in parallel. In particular, the receiver will continuously use the latest decoded symbols for all users in order to regenerate the interference. Due to the use of the random interleaver and Viterbi decoding within the SIC scheme, we will demonstrate our ability to obtain much more accurate interference estimates and hence achieve significant capacity and system performance improvement.