FPGA Design and Implementation of a Convolutional Encoder and a Viterbi Decoder Based on 802.11a for OFDM

In this paper, a modified FPGA scheme for the convolutional encoder and Viterbi decoder based on the IEEE 802.11a standards of WLAN is presented in OFDM baseband processing systems. The proposed design supports a generic, robust and configurable Viterbi decoder with constraint length of 7, code rate of 1/2 and decoding depth of 36 symbols. The Viterbi decoder uses full-parallel structure to improve computational speed for the add-compare-select (ACS) modules, adopts optimal data storage mechanism to avoid overflow and employs three distributed RAM blocks to complete cyclic trace-back. It includes the core parts, for example, the state path measure computation, the preservation and transfer of the survivor path and trace-back decoding, etc. Compared to the general Viterbi decoder, this design can effectively decrease the 10% of chip logic elements, reduce 5% of power consumption, and increase the encoder and decoder working performance in the hardware implementation. Lastly, relevant simulation results using Verilog HDL language are verified based on a Xinlinx Virtex-II FPGA by ISE 7.1i. It is shown that the Viterbi decoder is capable of decoding (2, 1, 7) convolutional codes accurately with a throughput of 80 Mbps.