Abstract
Memory polynomial (MP) method has been widely applied to the linearization area in wireless communications, in order to depress the nonlinear and memorial effects caused by the power amplifier. In MP, least square algorithm is usually utilized to calculate the coefficients, which may encounter numerical inaccuracies in high dimensional MP matrix inverse computations when the nonlinear effects are intensified. Aiming at this problem, we propose a memory polynomial with memory elimination (MP-ME) scheme, which is intended to separate the suppression processes of memory and nonlinearity, thus improves the detection performance by reducing the dimension of conventional MP matrix. With the proposed ME structure, we extract the linear component in polynomial’s linearization output by reconstructing and eliminating the memory signal, so the matrix calculation with high dimension in MP is avoided. Computer simulations are carried out considering the nonlinear characteristics of the light emitting diodes in visible light communications, and the results show that MP-ME can improve the bit error rate performance of MP significantly.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
