On The Research Of High Performance Concatenated Coding

Turbo codes, introduced by C.Berrou etc. in 1993, can be considered as one milestone in the field of channel coding technology within recent ten years, because it can approach the Shannon limit closely with finite code length. The first proposed Turbo codes are parallel concatenation of recursive systematic convolutional codes with iterative decoding structure which adopts soft-in-soft-out decoding algorithms at the receiving end. Nowadays, the concept of Turbo codes has been widely extended. Not only can their component codes be general linear block codes, but also their coding structure employs serial concatenation, hybrid concatenation and even multiple concatenations. At meantime, Turbo Codes have been absorbed into the third generation communication systems and also been written into several communication standards. On the other hand, the Multi-Input-Multi-Output (MIMO) technology has also drawn enormous attention recently, because it can increase the capacity of wireless communication greatly without wasting the limited bandwidth. As one primary technology for MIMO, Space Time Block Coding (STBC) can not only exploit spatial and temporal diversity simultaneously, but also be easily implemented with a linear decoding complexity. However, STBC only shows diversity gain but free of coding gain. Consequently, the error capacity of an STBC design is limited. If one appropriate channel code could be concatenated with STBC, the concatenated space time codes were likely to acquire both diversity and coding gain. Therefore the performance of wireless communication system can be greatly improved.This paper mainly presents two parts. The first part is focused on the research of recent hot channel codes, such as Turbo convolutional codes, Turbo block codes and one special multiple concatenated Turbo block codes which employ single parity check codes as their component codes ( also be called multiple single parity check codes). In this part, the coding and decoding methods are described and detailed simulation resultsare also provided. The second part pays more attention to the concatenation of channel codes and orthogonal space time block codes. We mainly compare the performance of two different concatenation strategies which include Turbo iterative decoding at the receiving end and common decoding structure without any iterative part.We attain main conclusions on the following two aspects: (1) multiple single parity check codes with different coding structures have different but regular performance variation: performance of hybrid coding structure is always lying in between the performance of its parallel and serial counterparts;when code rate increases, the performance crossover between parallel and serial coding structures moves to lower bit error rate region and their performance difference decreases. (2) The concatenated space time codes can provide more coding gains than the one without channel codes. Within the concatenated space time coding system, the one employing Turbo iterative decoding structure at the receiving end outperforms the one only employing common decoding structure.