The New Complex Backpropagation Algorithm and Its Applications: Adaptive Equalization
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
To reduce intersymbol interference (ISI) due to band-limited channels in digital communication, equalization techniques are necessary. Among adaptive equalization techniques, the neural networks having powerful mapping capabilities have been used as an alternative for effectively dealing with the channel distortion, especially the nonlinear distortion. In this paper, the complex BP algorithm for the complex BPN consisting of the new node activation functions with multi-saturated output regions is proposed and analyzed by the benchmark testing. And then the complex Backpropagation Neural Networks (BPN) are proposed as nonlinear adaptive equalizers that can deal with both QAM and PSK signals of any constellation sizes (e.g., 32-QAM, 64-QAM and MPSK). The proposed complex BP equalizer (CBPE) gives, compared with the linear complex equalizers using the least mean squares (LMS) algorithm, an outstanding improvement with respect to Bit Error Rate (BER) when channel distortions are nonlinear.
Abstract
To reduce intersymbol interference (ISI) due to band-limited channels in digital communication, equalization techniques are necessary. Among adaptive equalization techniques, the neural networks having powerful mapping capabilities have been used as an alternative for effectively dealing with the channel distortion, especially the nonlinear distortion. In this paper, the complex BP algorithm for the complex BPN consisting of the new node activation functions with multi-saturated output regions is proposed and analyzed by the benchmark testing. And then the complex Backpropagation Neural Networks (BPN) are proposed as nonlinear adaptive equalizers that can deal with both QAM and PSK signals of any constellation sizes (e.g., 32-QAM, 64-QAM and MPSK). The proposed complex BP equalizer (CBPE) gives, compared with the linear complex equalizers using the least mean squares (LMS) algorithm, an outstanding improvement with respect to Bit Error Rate (BER) when channel distortions are nonlinear.