Practical Design Of IDMA Systems

The technology of multiple-access is one of the key technologies in wireless communication networks and systems. The problem that the available spectrum is limited becomes more and more serious as the user requirements of high speed wireless services increase. How to reach high spectral efficiency is an important issue for the next generation of wireless communication system. Therefore, interleave-division multiple-access (IDMA), proposed as one of the key technologies in 4G, has become a popular research topic. IDMA could approach not only channel capacity but also low receiver complexity.By now, researches on IDMA are mainly based on theoretic analysis and system simulation evaluation. In order to show its prommised performance in real life systems, we need to optimize the important parameters of IDMA, such as the trade-off between code rate and spreading, the optimization of the special efficiency, the selection and design of optimal codes, and the paralization of interleavers. In this thesis, we do the following two studies to overcome problems for the practical implementation of IDMA.Firstly, we propose the mutual information evolution algorithm (MIEA) based on the concept of exstrinsic information transfer (EXIT) chart. Using MIEA, we can analyze the convergence behavior of IDMA systems by Monte Carlo simulation of parts in IDMA instead of the whole system simulation. The results obtained through the use of MIEA are in good agreement with that obtained using bit error rate (BER) curves. In addition, we can use MIEA to estimate the BER performance of IDMA systems, the result of which is similar to that obtained using system simulations. Besides, using MIEA, we can estimate the asymptotic threshold signal to noise ratio (SNR) and the asymptotic threshold number of users in IDMA systems. The result of MIEA is consistent with that of the system simulation. Moreover, MIEA could guide us to select and design the optimal codes for IDMA systems. Compared to SNR evolution and EXIT chart, the MIEA is more accurate and more general. In one word, the MIEA could be used to guide the practical design of IDMA systems.Secondly, we propose a parallel interleaver design for IDMA systems. For the interleavers proposed in the literature so far, both the interleaving and de-interleaving operations permute sequences serially, which will take many hardware clock periods and lead to much processing latency and low processing throughput. This has been the bottleneck of the system throughput, especially when the iteration numbers are large. In order to solve this problem, we propose a parallel interleaver design method in this thesis. The advantages of our approach are the reduction of the processing latency caused by interleaving and de-interleaving at the receiver, as well as memory requirements, with negligible performance degradations against random interleavers. The use of parallel interleavers results in efficient improvements of system throughput.