CMA Blind Equalization Algorithm Based on Linear Fitting
Qin Fan, Tielin Zhu, Hao Liu, Zhang Yuying · 2022 International Symposium on Networks, Computers and Communications (ISNCC) · 2022
Aiming at solving the problem of poor stability of measurement and control communication due to the influence of multipath channels when unmanned aerial vehicle (UAV) performs tasks in complex terrain environments such as cities and mountainous regions, a variable step size constant modulus algorithm (CMA) equalization method proposed based on linear fitting in order to improve the anti-interference of UAV links. According to the characteristics of fast convergence of large steps and little error of small steps, a linear relationship between steps and errors is established using the relative value of the residual error as a reference variable for step size varying. The step size is gradually decreased with the convergence of the equalization filter coefficients until maintaining the minimum value to accelerate the convergence speed and reduce the residual error. The simulation results show that the proposed blind equalization algorithm which takes into account the response speed and error performance can effectively suppress multipath interference.