Coding and Decoding of perfect Space-time Block Codes: Analysis and Performance Evaluation
No Thumbnail Available
Date
2012-11
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
AAU
Abstract
Multiple-input multiple-output (MIMO) is one of the most significant advances in digital
communication which enabled to increase the data rate as well as improve the reliability
and robustness of the system to as compared to single antenna systems. This is achieved by
deploying multiple antennas both at the transmitter and receiver sides [1]. However,
various assumptions about channel state information (CSI) and channel fading lead to
different capacity results [2]. Fading can be reduced substantially by using diversity
techniques. Space-time diversity is the most economic technique. Space-time block codes
(STBCs) are the most widely employed space-time codes. Perfect space-time block codes
(PSTBCs) are families of STBCs that satisfy all of the following criteria: full diversity, high
rate, good constellation shaping, uniform average transmitted energy per antenna, nonvanishing
determinant (NVD) for increasing spectral efficiency and diversity-multiplexing
tradeoff (DMT).
In this thesis, we investigate and demonstrate the coding and decoding of PSTBCs. We first
give the design criteria of STBCs followed by the mathematical background of cyclic
division algebra (CDA) as a constructing tool for PSTBCs. The PSTBCs are constructed and
investigated to be the most efficient family of STBCs by the design criteria of STBCs.
Having considered conditional optimization maximum likelihood (COML) as a near optimal
yet less complex decoding scheme of PSTBCs, we demonstrated the decoding of Golden
code, a family of PSTBC, as an example. Simulation results show that the symbol error rate
of the COML decoding of the Golden code is almost equal to ML decoding with decoding
complexity of only
2
which is the order of,
V
4
for the conventional ML. Further, we
considered the performance of the COML decoding under transmit, receive and both
transmit and receive antenna correlations and found consistent complexity reduction.
Although there is some deviation, in exact optimality of the COML decoding of the Golden
code as of the previous works in the literature, which arise most probably due to the
inaccuracy of the QAM quantization we used in this thesis, we conclude that an
antenna PSTBC is near optimally COML decodable with a reduction in complexity of the
order,
from its respective ML.
Description
Keywords
STBC, CDA, PSTBC, COML