Improved technique for quick error rate estimation of multi-dimensional communication schemes

A new approach for the quick estimation of error rates for complex communication schemes is presented. The purpose is to decrease the computations needed for analysing and simulating the performance in terms of error rates. By conditioning the error rate on the instantaneous noise energy, the error rate can be partly analysed by numerical calculations. Simulations, in the form of importance sampling, find the remaining part with lower computational complexity than has been reported previously. The gain is several orders of magnitude for the examples given. It is believed that the gain is also large for other systems. By deeper analysis of particular codes further improvements may be achieved.