Error Performance of Linear Dispersion Codes

In this paper, the authors present a study of error performance of linear dispersion codes. First, an upper bound of average error probability is derived for any signal-to-noise ratio. Based on this bound, a tight upper bound of average error probability at high signal-to-noise ratio is further derived to show the diversity advantage of linear dispersion codes. The tight bound demonstrates the relationship of error probability to constellation size and space-time symbol rate. The bound at high signal-to-noise ratio is verified by simulation results.