The implementation of equalization algorithms for real transmission channels

摘要

This article concerns the implementation of equalization algorithms of the real transmission channel. It is one of the methods leading to “higher efficiency of the transmission system”, i.e. by increasing the transmission speed in the given frequency bandwidth while maintaining low BER (Bit Error Rate), EVM (Error Vector Magnitude), etc. values. The authors are primarily focused on implementation of equalization algorithms with stochastic gradient adaptation (LMS-Least Mean Squares) for M-QAM (M-ary Quadrature Amplitude Modulation) modulations. Based on simulations conclusions, optimal convergence constant value for each M-QAM state was empirically determined. This set of values was used in the implementation using real hardware (RF VSG NI PXI-5670 vector signal generator and the RF VSA NI PXI-5661 vector signal analyzer, which assumes the role of a software-defined radio, the so-called SDR), where the optimal value of μ (step size or convergence constant) is harder to determine.