New Hybrid Concatenated Codes
Abstract
To improve the performance of hybrid concatenated convolutional codes (HCCC); a modified Log-MAP algorithm and an enhanced HCCC are introduced and demonstrated to be efficient and practical by simulation results. The new coding scheme achieves about 1.0 dB additional coding gain, compared to the general turbo coding scheme at a BER = 10-6, with a frame length of 8192-bit. The system complexity and decoding latency of the new scheme is lower than the HCCC proposed by Divsalar and Pollara [6] within acceptable performance degradation. Since the bit-error-rate of the proposed HCCC can be dramatically reduced by slightly increasing signal-to-noise ratio, the new hybrid concatenated coding scheme is very suitable for those communication environments in which high reliability is important.
Keywords
Download Options
Introduction
Turbo codes, proposed in 1993 [1], show outstanding error correction performance and thus are widely used in communication areas, such as third-generation mobile and deep-space communications [2] [3]. The general encoder structure of turbo code is a parallel concatenated convolutional code (PCCC). Two component encoders are used to code the same input bits, but an interleaver is placed between the encoders. The decoder of PCCC is an iterative structure. Two component decoders, two interleavers as well as one de-interleaver are needed. During the iterative decoding procedure, only the extrinsic information of systematic bits is calculated and provided as the prior information of the decoded bits, however, the prior information of the parity bits has not been retrieved. Detail discussions for the kind of turbo codes have been given in [4].
Aiming for excellent BER performance, other concatenated codes, such as serial concatenated convolutional code (SCCC) [5] and hybrid concatenated convolutional code (HCCC) [6], are proposed. In [5] and [6], not only the extrinsic information of systematic bits, but also the extrinsic information of parity bits, is exchanged between the component decoders. Therefore, the performances of SCCC and HCCC have been shown to be superior, in some cases, to a general turbo code.
In this work, modifications of Log-MAP algorithm are presented. Based on the modified Log-MAP algorithm, an enhanced HCCC scheme is also proposed. Using the new scheme, both the extrinsic information of systematic bits and parity bits can be retrieved during iterative decoding. Thus, the decoding performance can be improved dramatically when signal-to-noise ratio (SNR) is slightly raised. With a frame length of 8192-bit, simulation results show that the enhanced scheme achieves an additional coding gain about 1.0dB over the turbo coding scheme, at a bit-error-rate (BER) = 10-6. Moreover, the decoding latency and system complexity is much lower than that of the Divsalar-Pollara HCCC [6] within acceptable performance degradation.
Conclusion
An enhanced hybrid concatenated convolutional code (HCCC) is proposed based on a modified Log-MAP algorithm. Simulation results are given to demonstrate these modifications being efficient and practical to construct the HCCC. Although the BER of the proposed HCCC is slightly worse than a general turbo code when the SNR is very low, it is much better than the latter when the SNR is increased slightly. Comparisons of the proposed HCCC with the Divsalar-Pollara HCCC are given in Table1. From the table, it is easy to see that both the system complexity and the decoding latency of the proposed HCCC are much lower than the latter, with slight BER degradation.