A ”multi-user” approach towards a channel decoder for convolutional, turbo and LDPC codes
Steffen Kunze, E. Matus, Gerhard Paul Fettweis, T. Kobori · 2010
In this paper we present the concept of a high-throughput multi-mode channel decoder architecture that consists of a tightly coupled array of independently programmable processing cores. Every core is capable of decoding low-density parity-check (LDPC), convolutional turbo (CTC) and convolutional codes (CC) either independently or jointly with other cores. This approach allows parallel handling of several separate decoding processes in one decoder engine as well as performing a single high-throughput decoding, opening up a new level of flexibility for channel decoding. The multi-mode decoder core as well as the multi-core approach are explained and simulation results presented. A case study of the proposed architecture was implemented in a 65nm-process using an area of 0.44 mm2. At 200 MHz, throughputs of up to 86 Mbps could be reached.