Performance Analysis Of Mordern Channel Coding And The Simplified Decoding Algorithm

Channel coding is always a hot topic in wireless communications. Promoted by the 3G and 4G wireless communication systems, Turbo code and LDPC code are popularly used in modern systems. These codes have strong forward error correcting capability, while the decoding complexity is also higher than some simple codes. So it's valuable to find a simplified decoding algorithm in order to decrease the delay in decoding process and the power consumption in the system.As the development of modern channel coding and the decoding algorithms, the requirements for performance analysis to these codes become more important. Presently, the analysis technique is mainly the simulation assisted theoretical analysis. However, limited by some factor in a real system, these schemes can rarely be used. So it is significant to find a scheme that is applicable to multiple of real systems.This thesis majored in the performance analysis of modern channel coding and the simplified decoding algorithms. The main contributions are as following:1. Proposed the concept of SR-PDF (Squired Radius Probability Density Function). This concept changes the way to evaluate the average error rate in a system. A simple and flexible way is introduced, which saves a large number of simulations while guarantee the preciseness of the results. Based on this concept, this thesis gave some approximate equations that are useful for theoretical analysis on modern channel coding. Furthermore, after the proposed technique is extended to analyze the decoding threshold of capacity achievable modern channel coding, it is found that SR-PDF provides another way to obtain the Shannon limit for these codes besides the traditional ways.2. Proposed an improved concatenated coding and decoding scheme, which employs SPC as the outer code and modern channel coding as the inner code at the encoder, and employs MRC and iterative decoding algorithms at the decoder. This scheme has better performance and less complexity compared with the traditional concatenated coding and decoding schemes.3. Based on the randomness of the number of iterations, this thesis proposed a parallel decoding scheme that can be used to improve the throughput of decoders for both Turbo codes and LDPC codes. Futhermore, the performance of the parallel decoding is not worse than the traditional scheme.