DIRECT sequence code-division multiple-access (DSCDMA) has been selected as the fundamental signaling technique for third generation (3G) wireless communication systems, due to its advantages of soft user capacity limit and inherent frequency diversity. However, it suffers from multiple- access interference (MAI) caused by the nonorthogonality of spreading codes, particularly for heavily loaded systems. Therefore, techniques for mitigating the MAI, namely multiuser detection, have been the subject of an intensive research effort over the past two decades. It is well known that multiuser detection can substantially suppress MAI, thus improving system performance.
Maximum likelihood (ML) multiuser detection was proposed in the early 1980s, and achieves the optimal performance at the cost of prohibitive computational cost when the number of users is large. For practical implementation, suboptimal algorithms, such as the linear minimum mean square error (LMMSE) detector or decorrelator, allow a tradeoff between complexity and performance. It should be noted that, with the development of interference cancellation (IC) techniques, multiuser detection is being applied in practical systems, such as the EV-DO Revision A systems in recent years, the turbo principle, namely the iterative exchange of soft information among different blocks in a communication system to improve the system performance, has been applied to combine multiuser detection with channel decoding. In such turbo multiuser detectors, the outputs of channel decoders are fed back to the multiuser detector, thus enhancing the performance iteratively.
Turbo multiuser detection based on the maximum a posteriori probability (MAP) detection and decoding criterion has been proposed in together with a lower complexity technique based on interference cancellation and LMMSE filtering. Further simplification is obtained by applying parallel interference cancellation (PIC) for multiuser detection, where the decisions of the decoders are directly subtracted from the original signal to cancel the MAI.
No comments:
Post a Comment