Showing papers on "Codebook published in 1983"

Fundamental issues of multiple accessing

Abstract: : Fundamental issues of multiple accessing are identified. These issues include transmitter asynchronism, variability of the set of active users, feedback, and degree of codebook knowledge among the users. Various multiple access schemes are examined under these issues. These issues are subsequently modeled and analysed using an information theoretic framework. We discover that the capacity region of the asynchronous multiple access channel is different from that of the synchronous channel. For communication systems with users having random message generation time, we demonstrate that its channel capacity resembles that of an asynchronous channel even though the users are synchronous, if decoding delay is constrained to be much smaller than the message inter- arrival time. We investigate communication with restricted decoder structure. New information theoretic quantities that incorporate the decoding metric used are discovered and examined. Using these quantities, we provide a rigorous and novel treatment for the theory of jamming. These mathematical techniques provide insight for achieving reliable communication in a multiple access environment where each user may not know the codebook of the other users and a jammer may be present. We then apply the general theory developed to three specific asynchronous channel without feedback, namely the OR channel, the spread spectrum channel and the collision channel. Practical and novel coding schemes are suggested. Maximum throughput, error rate and decoding complexity are analysed.

A generalization of isolated word recognition using vector quantization

Abstract: The use of vector quantization (VQ) for isolated-word speech recognition of a 20-word vocabulary was shown in previous work to achieve more than 99% accuracy for speaker-dependent recognition and 87% accuracy for speaker-independent recognition. Separate VQ codebooks were designed for each word in the recognition vocabulary, and input words were classified by performing VQ and finding the codebook that achieves the smallest average distortion. The method obviates time-normalization and makes no use of time-sequence information. This paper presents results for a generalization that incorporates time-sequence information. The generalization, which was motivated by work of Martinez, Riveria, and Buzo, is more accurate and faster than the previous method. Words in the training and input sequences are normalized linearly to the same length and then divided into sections. Separate VQ codebooks are designed for each section of each vocabulary word. Each vocabulary word is then represented by a "multisection" codebook - a time-dependent sequence of section-codebooks. New words are classified by performing VQ and finding the multi-section codebook that achieves the smallest average distortion. Initial tests on a twenty-word vocabulary resulted in accuracies greater than 97% for speaker-independent recognition.