Reserch And Application Of MLC Key Technology

In communication systems, channel coding is a kind of critical method to enhance system reliability and modulation is a necessary step before transferring the digital baseband signal into the transmitted symbols on real channel. However, the channel coding technology is facing a new challenge in the wake of the rapid development of digital communication technology. Currently, there are two main direction of the development of channel coding:one is improving the power efficiency property; and the other one is to design a system which is both power-effectively and frequency effectively combining with modulation in order to meet the needs of next generation mobile communication broadband transferring. At present, there are two kinds of channel coding and modulation schemes, Trellis Coding Modulation (TCM) and Multilevel Coding (MLC).In the MLC schemes, the design of component codes for each level must obey the channel capacity rule. In other words, the rate of each component code must be equal to the capacity of corresponding equivalent channel. Based on the channel capacity rule, the MLC scheme can achieves satisfying performances on both AWGN channel and Rayleigh fading channel, which means the design rule is robust. Also, the channel capacity rule provides us an information theory view of the design of channel coding and modulation system. The MLC scheme supports three kinds of decoding method:Maximum Likelihood Decoding (MLD), Parallel Decoding on Levels (PDL) or Multistage Decoding (MSD). MLD is an optimal decoding approach for MLC, but it is to computationally expensive to apply. Therefore, the suboptimal MSD method is more common in practical. Compared with MSD, the decoding delay of PDL is smaller, and the error propagation can be avoided in PDL, at cost of performance. For high order modulation, there are some new partitioning schemes, which largely simplify the MSD decoding structure of MLC without any loss of performance. At the meantime, the Soft-Input-Soft-Output(SISO) decoder can compensates the information lose in MSD for codes like LDPC.Low-Density Parity-Check (LDPC) code is an important kind of error-correcting codes that is rediscovered recently, whose performance can approach the Shannon limits. LDPC codes outperform the BCH codes used as component codes in existing MLC systems, and the design of LDPC codes is more flexible. Researches show that the performance of LDPC codes can be improved through a degree optimization algorithm based on Extrinsic Information Transferring (EXIT) charts. Thus, it is a tendency to use LDPC codes as component codes in MLC instead of BCH codes.Since SISO decoders are used in the MLC scheme, it is possible to adapt the iterative demodulating and decoding process based on the massage-passing procedures between the demodulator and the MSD decoder, which can erase the effect of error propagation in MSD and compensates the performance loss in PDL. MLC-ID systems of two different patterns are discussed in this paper. Simulation results show that both patterns can bring certain performance gain, and the gain depends on the partitioning schemes.