Citation

BibTex format

@article{Gunduz:2008:10.1109/TIT.2007.911274,
author = {Gunduz, D and Erkip, E},
doi = {10.1109/TIT.2007.911274},
journal = {IEEE Transactions Information Theory},
pages = {116--134},
title = {Joint Source-Channel Codes for MIMO Block Fading Channels},
url = {http://dx.doi.org/10.1109/TIT.2007.911274},
volume = {54},
year = {2008}
}

RIS format (EndNote, RefMan)

TY  - JOUR
AB - We consider transmission of a continuous amplitude source over an L-blockRayleigh fading $M_t \times M_r$ MIMO channel when the channel stateinformation is only available at the receiver. Since the channel is notergodic, Shannon's source-channel separation theorem becomes obsolete and theoptimal performance requires a joint source -channel approach. Our goal is tominimize the expected end-to-end distortion, particularly in the high SNRregime. The figure of merit is the distortion exponent, defined as theexponential decay rate of the expected distortion with increasing SNR. Weprovide an upper bound and lower bounds for the distortion exponent withrespect to the bandwidth ratio among the channel and source bandwidths. For thelower bounds, we analyze three different strategies based on layered sourcecoding concatenated with progressive, superposition or hybrid digital/analogtransmission. In each case, by adjusting the system parameters we optimize thedistortion exponent as a function of the bandwidth ratio. We prove that thedistortion exponent upper bound can be achieved when the channel has only onedegree of freedom, that is L=1, and $\min\{M_t,M_r\}=1$. When we have moredegrees of freedom, our achievable distortion exponents meet the upper boundfor only certain ranges of the bandwidth ratio. We demonstrate that ourresults, which were derived for a complex Gaussian source, can be extended tomore general source distributions as well.
AU - Gunduz,D
AU - Erkip,E
DO - 10.1109/TIT.2007.911274
EP - 134
PY - 2008///
SN - 0018-9448
SP - 116
TI - Joint Source-Channel Codes for MIMO Block Fading Channels
T2 - IEEE Transactions Information Theory
UR - http://dx.doi.org/10.1109/TIT.2007.911274
UR - http://arxiv.org/abs/0711.1161v1
UR - http://hdl.handle.net/10044/1/15311
VL - 54
ER -

Apply for a PhD

The ISN Group is always looking for talented PhD candidates.


To learn more about the application process for postgraduate research, including entry requirements, selection criteria, and financial support, please read the departmental guidance on applying to study for a PhD.