Flexible Decoder Based on MAP-Algorithm for Two Different Families of Codes

Many wireless communication systems such as IS-54, EDGE, WiMAX, and LTE have adopted turbo codes and tail-biting convolutional codes as the forward error correcting codes (FEC) scheme for data and overhead channels. However, some releases propose LDPC codes for error-corrections due to the relative complexity of turbo codes decoder implementations as well as the success of LDPC codes in achieving the same performance as turbo codes and in some cases surpassing it with low decoding complexity. As a matter of fact, these new standard releases will work side by side in actual devices with older ones which are based on turbo and convolutional codes. Indeed, these two families of codes are both very good in performance. Consequently, it is mandatory to relate them so as to enhance technology transfer and hybridization between the two methodologies. Moreover, in wireless systems, power consumption as well as die area must be minimized, because of battery life for wireless mobile devices, maximum allowable heat dissipation and available packaging space due to small-size hand-helds. Thus, efficient design of flexible convolutional, turbo and LDPC codes decoder with respect to power, area, and throughput is critical for future wireless system implementations. In the literature, many efficient turbo decoder architectures based on the MAP algorithm have been extensively investigated by many researchers. However, designing a flexible decoder to support LDPC, convolutional and turbo codes based on a unified MAP algorithm is very challenging. Henceforth,, in this paper we aim to provide an alternative solution proposing a combined decoder supporting the three types of codes with a small additional overhead based on a unified MAP algorithm.