Summary: IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 53, NO. 12, DECEMBER 2007 4457
Joint Source­Channel Coding Error Exponent
for Discrete Communication Systems
With Markovian Memory
Yangfan Zhong, Student Member, IEEE, Fady Alajaji, Senior Member, IEEE, and L. Lorne Campbell, Fellow, IEEE
Abstract--We study the error exponent, EJ , for reliably trans-
mitting a discrete stationary ergodic Markov (SEM) source QQQ
over a discrete channel WWW with additive SEM noise via a joint
source­channel (JSC) code. We first establish an upper bound for
EJ in terms of the Rényi entropy rates of the source and noise
processes. We next investigate the analytical computation of EJ by
comparing our bound with Gallager's lower bound (1968) when
the latter one is specialized to the SEM source­channel system.
We also note that both bounds can be represented in Csiszár's
form (1980), as the minimum of the sum of the source and channel
error exponents. Our results provide us with the tools to system-
atically compare EJ with the tandem (separate) coding exponent
ET . We show that as in the case of memoryless source­channel
pairs EJ 2ET and we provide explicit conditions for which
EJ > ET . Numerical results indicate that EJ 2ET for many

Source: Alajaji, Fady - Department of Mathematics and Statistics, Queen's University (Kingston)

Collections: Engineering