Abstract

Digitization of speech and transmission of the generated bit stream is a problem of paramount importance in communications because it addresses the issues of security, jam resistance and compatibility with an all-digital network. The fundamental difficulty is the transmission of the generated bit stream over the available bandwidth which may be narrow because of physical limitations, cost or operational restrictions. For that reason, very low bit rate systems are extremely important, particularly in tactical applications. The process of digital coding of speech can be separated into feature extraction, pattern identification and matching and transmission coding. The simplest systems, such as standard Linear Predictive Coding (LPC). may only include one of these steps. The inclusion of all three, however, can lead to substantial reductions in bit rate requirements, at the expense of increased processing. This paper discusses a 400-2400 b/s system in which feature extraction is performed by standard LPC and pattern matching by vector quantization and trellis coding. Parts of the system have been implemented in hardware and other parts in simulation. Diagnostic Rhyme Test (DRT) test results are presented to indicate the performance of the system at various bit rates.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.