This project consists of the development of an embedded voice codec for telecommunication networks, with the goal of compressing voice data so as to significantly reduce the required bandwidth for transmission of the resultant signal without introducing detectable noise or time delay by applying burstiness constraints. The codec consists of Differential Pulse Code Modulation and Huffman algorithms to encode and decode the voice data, reducing statistical redundancy with parameters optimized to reduce token bucket size at a given token rate.
Worcester Polytechnic Institute
Major Qualifying Project
Restricted-WPI community only
Electrical and Computer Engineering