On the applications of multimedia processing to communications

R.V. Cox, B.G. Haskell, Y. LeCun, B. Shahraray, L. Rabiner
1998 Proceedings of the IEEE  
Fig. 3 . The POTS network with signaling, service-oriented data bases, and connections for facsimile devices and PC's via voice-band modems. Fig. 4. The telecommunications network of today. nonreal-time data among computers and are distinguished by the property that packet communications are routed by address information contained in the data stream itself. The packet network is especially well suited for sending data of various types, including messages, facsimile, and still images. The
more » ... is not well suited for sending real-time communication signals, however, such as speech, audio, and video. The packet network is primarily accessed by client programs in PC's, and so it is inherently PC oriented and client/server driven. It provides access to distributed data bases and, via some excellent search engines, has excellent search capabilities. There are approximately 30 million users of the packet network in the United States, a number that is growing rapidly and will continue to do so over the next decade. Today, the Internet (the largest of the existing packet networks) connects more than 40 million computers in some 140 countries. The growth rate is astounding, with anticipated revenue from content-related Internet services expected to exceed $10 billion by the end of the decade. A key problem with packet networks is that they were not designed to handle voice traffic. With the emergence of the multimedia PC and its concomitant signal-processing capability, a simple solution was proposed, namely, the compression (and decompression) and coding (and decoding) of voice into (from) IP packets using the PC processor and the sound card inherent in the multimedia PC. Unfortunately, associated with this solution was the problem of long delays due to several factors, including: • the algorithmic processing delay required for lowbit-rate compression (on the order of 25-65 ms for high-quality compression algorithms);
doi:10.1109/5.664272 fatcat:yhn4kb3mmbctbbfus7br26oox4