Implementation of High Throughput Lossless Canonical Huffman Encoder

This paper presents an improved VLSI architecture for Huffman encoding based on the Canonical Huffman method. Traditional Huffman encoders require scanning input data twice, leading to increased processing time and hardware complexity. Our design reduces encoding time by performing the entire encoding process in a single scan. By using a parallel architecture for frequency sorting and code-size computation, our design achieves higher throughput and efficiency. Experimental results demonstrate that our approach significantly improves encoding speed and reduces hardware requirements, this makes it well-suited of real-time applications, including image and audio compression.