Reduced Complexity SNR Estimation via Kolmogorov-Smirnov Test

Two complexity reducing schemes are proposed in this letter for the recently presented Kolmogorov-Smirnov (K-S) test based signal-to-noise ratio (SNR) estimator. The K-S test based SNR estimator can work properly over an extended SNR range for various multilevel constellations with limited signal samples, but involves considerably more add operations as a result for the huge amount of reference signals needed for matching operations. The proposed two complexity reducing schemes explore the order characteristic of the SNR matching pool to accelerate the searching procedure. For the situation under consideration, the computational complexities (numbers of add operation) of the two proposed schemes are about 1/5 and 1/20 of the original one respectively. Simulation results have verified these schemes´ effectiveness.