Đang chuẩn bị nút TẢI XUỐNG, xin hãy chờ
Tải xuống
This paper proposes a new approach to combine the convex of two single-kernel adaptive equalizers with different convergent rates and different efficiencies in order to get the best kernel equalizer. This is the Gaussian multikernel equalizer. | MULTI-KERNEL EQUALIZATION FOR NON-LINEAR CHANNELS MULTI-KERNEL EQUALIZATION FOR NON-LINEAR CHANNELS Minh Nguyen-Viet Posts and Telecommunications Institute of Technology Hanoi City Vietnam Abstract Nonlinear channel equalization using kernel equalizers is a method that has attracted lots of attention today due to its ability to solve nonlinear equalization problems effectively. Kernel equalizers based on Recursive Least Squared K-RLS are successful methods with high convergent rate and overcome the local optimization problem of RBF neural equalizers. In recent years some simple K-LMS algorithms are used in nonlinear equalizers to further enhance the flexibility with the adaptive capability of equalizers and reduce the computational complexity. This paper proposes a new approach to combine the convex of two single-kernel adaptive equalizers with different convergent rates and different efficiencies in order to get the best kernel equalizer. This is the Gaussian multikernel equalizer. Keywords Adaptive equalization kernel equalizer multi-kernel filter nonlinear channel. I. INTRODUCTION Single-kernel adaptive filters are used widely today to identify and track the non-linear systems 1 2 3 . The developments of kernel adaptive filters enable us to solve non-linear estimation problems using linear structures. In this paper we use kernel adaptive filters for equalizations of non-linear wireless channels such as satellite channels. Wireless channels with their time-variant Correspondence Minh Nguyen-Viet email minhnv@ptit.edu.vn Communication received Mar. 3 2016 revised May 6 2016 accepted May 30 2016. Tạp chí KHOA HỌC CÔNG NGHỆ L86 THÔNG TIN VÀ TRUYỀN THÔnG SÔ 1 năm 2016 parameters cause non-linear and linear distortions to the transmitted signal. Channel equalizers are used to minimize these distortions. Commonly channel equalizers can be considered as reverse filters which have characteristics must repeat the structure and the conversion rule of the channel. To .