The challenge is to design a way to use feedback information to maximize the rate at which
successes occur, while keeping the delay small.
The most successful algorithms [9], [10], [11] are of the "binary splitting" type. A group of
sources is allowed to transmit. If a "collision" occurs the group is split in two, each subgroup being
allowed to transmit in turn. The process is repeated until all collisions have been resolved.
The excitement has come from the observation that this type of algorithm can achieve success
rates in the vicinity of .5 while keeping the delay bounded no matter the number of sources M.
Larger rates of success are possible but then the delay is not bounded as M increases.
This threshold effect is reminiscent of the concept of channel capacity, and much effort has
been spent using a variety of techniques to characterize it. The threshold is known to be between
48 and .59 (see [12] and [13] and references therein).
Good accounts of the theories of modulation, detection and coding, together with appropriate
credits to their authors, can be found in classical texts like
1. A. J. Viterbi and J. K. Omura, Principles of Digital Communication and Coding, McGraw
Hill, New York, New York, 1979.
2. R. G. Gallager, Information Theory and Reliable Communication, John Wiley & Sons,
Inc., New York, New York, 1968.
Description of their applications to computer networks appears in
3. A. S. Tanenbaum, Computer Networks, Prentice-Hall, Inc., Englewood Cliffs, New Jersey,
Recent papers on equalization, each containing more references to the relevant literature are
4. A. Gersho and T. L. Lim, "Adaptive Cancellation of Intersymbol Interference for Data
Transmission", Bell System Technical Journal., 60, No. 11 (1981).
5. D. Godard, "Channel Equalization Using a Kalman Filter for Fast Data Transmission",
IBM J. Res. Develop., May 1974.
6. J. Proakis, "Advances in Equalization for Intersymbol Interference", Advances in
Communication Systems, 4, Academic Press, 1975.
Previous Page Next Page