[IEEE Trans. on Information Theory, December 2004, pp. 3146-3169]
Bit Stuffing Algorithms and Analysis for Run Length Constrained Channels in Two and Three Dimensions
Zsigmond Nagy and Kenneth Zeger
Abstract
A rigorous derivation is given of the coding rate of a variable-to-variable length bit stuffing coder for a two-dimensional
(1,∞)-constrained channel.
The coder studied is ``nearly'' a fixed-to-fixed length algorithm.
Then an analogous variable-to-variable length bit stuffing algorithm for the three-dimensional
(1,∞)-constrained channel is presented,
and its coding rate is analyzed using the two-dimensional method.
The three-dimensional coding rate is demonstrated to be at least 0.502,
which is proven to be within 4% of the capacity.