Design and simulation of soft decision viterbi decoder

The objective of this thesis is the design and simulation of the soft decision Viterbi decoder. An image input is used to verify the functional correctness of the decoder. Viterbi decoder efficiently decodes the outputs of the convolutional encoder by using Viterbi algorithm which is actually equivalent to the maximum likelihood decoding algorithm. The maximum likelihood decoding algorithm finds the shortest path in the trellis diagram of the convolutional encoder. A simple Viterbi decoding system is created by a convolutional encoder and a Viterbi decoder. Convolutional encoder produces symbols from inputs. Then, noises are added to the encoded symbols. Viterbi decoder decodes these symbols to estimate the original inputs. The branch metric unit (BMU), add-compare-select unit (ACSU) and path memory unit (PMU) are main units of the Viterbi decoder. BMU calculates distance between the soft decision outputs and observed outputs to find branch metrics. ACSU adds branch metrics which come from BMU and corresponding path metrics to update path metrics, compares these path metrics and selects the smallest value of the path metric as a survivor path. PMU has memory to store survivor paths which come from ACSU and is used for traceback where original inputs are determined.