An Advanced Total-Zeros Decoding Method Based on New Memory Architecture in H.264/AVC CAVLC

Abstract

A new total-zeros decoding method based on efficient memory architecture for variable-length code tables (VLCTs) is proposed in this paper. In general, a large amount of memory accesses are required for context-based adaptive variable-length coding (CAVLC) decoding in the H.264/AVC standard. It is an important issue for low-power implementation of mobile video applications because the memory access function results in considerable power consumption. Some distinctive features in the VLCTs of the total-zeros have been found through a careful analysis of the codewords. Efficient memory architecture and an advanced decoding are developed in order to remove redundant memory access. The simulation results show that the proposed method achieves an approximately 80%-90% memory access saving, compared with conventional decoding.