Reduced complexity detection of spatially multiplexed data transmissions under frequency flat fading is considered. Noting that the optimal detection is the generation of marginal distributions of a joint probability distribution, we use the approximate marginal generating algorithm termed the treebased reparameterization. The resulting system complexity is having an order which is less than quadratic in the number of transmit antennas. Simulation of a 8 × 8 BPSK system shows a coded bit error rate of 10−3 at an SNR level of only 1dB greater than required by the optimal method.
|Translated title of the contribution||Tree-based reparameterization for symbol detection in spatially multiplexed MIMO systems in frequency flat fading|
|Title of host publication||IEEE 7th Workshop on Signal Processing Advances in Wireless Communications, 2006 (SPAWC '06), Cannes|
|Publisher||Institute of Electrical and Electronics Engineers (IEEE)|
|Pages||1 - 5|
|Number of pages||5|
|ISBN (Print)||0780397118, 1424497118|
|Publication status||Published - Jun 2006|
|Event||7th Workshop on Signal Processing Advances in Wireless Communications - Cannes, France|
Duration: 1 Jun 2006 → …
|Conference||7th Workshop on Signal Processing Advances in Wireless Communications|
|Period||1/06/06 → …|
Bibliographical noteConference Proceedings/Title of Journal: IEEE 7th Workshop on Signal Processing Advances in Wireless Communications, 2006 (SPAWC '06)
Rose publication type: Conference contribution
Sponsorship: The authors would like to thank Toshiba Research Europe Limited, Bristol, U.K. for financially supporting this work. The first author would like to thank Mr. Yugang Jia for the helpful comments.
- approximate inference
- belief propagation,
- tree-based reparameterization.