Reduced Complexity Sphere Decoding
via a Reordered Lattice Representation
Luay Azzam, Student Member, IEEE, and Ender Ayanoglu, Fellow, IEEE
Center for Pervasive Communications and Computing
Department of Electrical Engineering and Computer Science
University of California, Irvine
Irvine, CA 92697-2625
In this letter, we propose a reordering of the channel representation for Sphere Decoding (SD) where
the real and imaginary parts of each jointly detected symbol are decoded independently. Making use of
the proposed structure along with a scalar quantization technique, we reduce the decoding complexity
substantially. We show that this approach achieves 85% reduction in the overall complexity compared
to the conventional SD for a 2×2 system, and 92% reduction for the 4×4 and 6×6 cases at low SNR
values, and almost 50% at high SNR, thus relaxing the requirements for hardware implementation.
Maximum-likelihood detection, multiple-input multiple-output channels, sphere decoding.
Consider a MIMO system with N transmit and M receive antennas. The received signal is