Training-Embedded, Single-Symbol ML-Decodable, Distributed STBCs for Relay Networks
Abstract
Recently, a special class of complex designs called Training-Embedded Complex Orthogonal Designs (TE-CODs) has been introduced to construct single-symbol Maximum Likelihood (ML) decodable (SSD) distributed space-time block codes (DSTBCs) for two-hop wireless relay networks using the amplify and forward protocol. However, to implement DSTBCs from square TE-CODs, the overhead due to the transmission of training symbols becomes prohibitively large as the number of relays increase. In this paper, we propose TE-Coordinate Interleaved Orthogonal Designs (TE-CIODs) to construct SSD DSTBCs. Exploiting the block diagonal structure of TE-CIODs, we show that, the overhead due to the transmission of training symbols to implement DSTBCs from TE-CIODs is smaller than that for TE-CODs. We also show that DSTBCs from TE-CIODs offer higher rate than those from TE-CODs for identical number of relays while maintaining the SSD and full-diversity properties.
Keywords
Cite
@article{arxiv.0909.1525,
title = {Training-Embedded, Single-Symbol ML-Decodable, Distributed STBCs for Relay Networks},
author = {J. Harshan and B. Sundar Rajan and Are Hjørungnes},
journal= {arXiv preprint arXiv:0909.1525},
year = {2009}
}
Comments
7 pages, 2 figures