On the Performance of Diagonal Lattice Space-Time Codes for the Quasi-Static MIMO Channel

There has been tremendous work done on designing space-time codes for the quasi-static multiple-input multiple-output (MIMO) channel. All the coding design to date focuses on either high-performance, high rates, low complexity encoding and decoding, or targeting a combination of these criteria. In this paper, we analyze in detail the performance of diagonal lattice space-time codes under lattice decoding. We present both upper and lower bounds on the average error probability. We derive a new closed form expression of the lower bound using the so-called sphere-packing bound. This bound presents the ultimate performance limit a diagonal lattice space-time code can achieve at any signal-to-noise ratio (SNR). The upper bound is simply derived using the union-bound and demonstrates how the average error probability can be minimized by maximizing the minimum product distance of the code.