A Blind Equalization Algorithm Based On Forward Neural Network In Pam System

Ly Zhang,R Lu,Hk Wang,Dg Sha

2003-01-01

Abstract:In this paper, we propose a new Blind Equalization algorithm based on multiplayer forward neural network, we design a new transmission function for the neural network. The computer simulations show that the new algorithm have more fast convergence performance than traditional forward neural network algorithm in PAM system.

What problem does this paper attempt to address?

A Integrated Algorithm of DDLMS and Godard for Blind Equilization

吴禀圣,邹志永,吴禀雅,王匡,朱小富

DOI: https://doi.org/10.3969/j.issn.1007-0249.2003.01.017

2003-01-01

Abstract:In this paper, a new blind equalization algorithm, DDLMS-CMA algorithm, is proposed for M-PAM and M-QAM communication systems. This new algorithm overcomes the shortcoming of large residual error in Sato or Godard algorithm, and preserves the fast convergence performance. Its convergence speed is even faster than BG algorithm, As a stable blind equalization algorithm, it is useful in practice.
Blind Equalization Algorithm of QAM Signal Based on SNR

LI Ang,DUAN Tian-dong,FANG Xiang-qian

DOI: https://doi.org/10.3969/j.issn.1002-8692.2007.z1.007

2007-01-01

Abstract:A modified CMA-DDLMS blind equalization algorithm based on the automatic switch according to the SNR for QAM signals in short wave is discussed.The algorithm has small residual error and fast convergence performance.Its good performance is confirmed by simulation and comparison.
Implementation of blind equalization algorithm for HF QAM based on DSP

Ang Li,Tiandong Duan,Xiangqian Fang,Wei Ren

DOI: https://doi.org/10.1109/WiCOM.2006.163

2007-01-01

Abstract:In this paper, a modified CMA-DDLMS equalization algorithm based on the automatic switch according to SNR is proposed This modified algorithm overcome the shortcoming of large residual error and preserves the fast convergence performance. The simulation results confirm the effectiveness of the proposed algorithm on HF QAM signal. The paper also gives the implementation procedure of the algorithm on TMS320C6711. © 2006 IEEE.
Excess MSE Analysis of the Concurrent Constant Modulus Algorithm and Soft Decision-Directed Scheme for Blind Equalisation

B. Lin,R. He,X. Wang,B. Wang

DOI: https://doi.org/10.1049/iet-spr:20070074

IF: 1.819

2008-01-01

IET Signal Processing

Abstract:The concurrent constant modulus algorithm (CMA) and soft decision-directed (SDD) scheme (CCMA + SDD) blind equalisation achieves a considerable improvement in equalisation performance over the CMA for high-order quadrature amplitude modulation (QAM) channels. Because of its concurrent adaptive filter structure and the complexity in analysing the estimation error signal, the actual steady-state performance has largely been left uninvestigated. By introducing an equivalent blind equalisation algorithm, the steady-state excess mean square error (EMSE) of the CCMA + SDD algorithm in a noise-free environment is studied. Based on the energy preservation approach, two first-order approximation closed-form solutions for the EMSE are derived for real-valued and complex-valued cases, respectively. The experimental simulation results for 16- and 64-QAMs show that the proposed steady-state theoretical EMSE results are reasonably accurate.
A New Dual Mode Blind Equalization Algorithm

XUE Wei,YANG Xiao-niu,ZHANG Zhao-yang

DOI: https://doi.org/10.3969/j.issn.1006-9348.2011.03.040

2011-01-01

Abstract:The constant modulus algorithm(CMA) is a widely used algorithm for blind equalization of QAM signals,but it exhibits slow convergence rate,large steady state mean square error and phase-blind nature.To overcome the defaults of CMA,a new dual mode blind equalization algorithm is proposed based on the modified constant modulus algorithm(MCMA) according to the study on the property of QAM signals,the algorithm switches between modified constant modulus algorithm(MCMA) and decision directed(DD) algorithm according the regions where the signal lies in the constellation plane and utilizes the merits of MCMA and DD algorithm.The simulation results demonstrate that the proposed algorithm can achieve the phase recovery effectively and speed up the convergence rate and decrease the steady state MSE greatly and is suitable for QAM signals.
Deep Learning Equalizer Connected with Viterbi-Viterbi Algorithm for PAM D-Band Radio over Fiber Link

Tangyao Xie,Qiang Sheng,Jianguo Yu

DOI: https://doi.org/10.3390/s23249773

2023-12-12

Abstract:D-band (110-170 GHz) has been regarded as a potential candidate for the future 6G wireless network because of its large available bandwidth. At present, the lack of electrical amplifiers operating in the high frequency band and the strong nonlinear effect, i.e., the D-band, are still important problems. Therefore, effective methods to mitigate the nonlinear issue resulting from the ROF link are indispensable, among of which machine learning is considered the most effective paradigm to model the nonlinear behavior due to its nonlinear active function and structure. In order to reduce the computation amount and burden, a novel deep learning neural network equalizer connected with typical mathematical frequency offset estimation (FOE) and carrier phase recovery (CPR) algorithms is proposed. We implement D-band 45 Gbaud PAM-4 and 20 Gbaud PAM-8 ROF transmission simulations, and the simulation results show that the real value neural network (RVNN) equalizer connected with the Viterbi-Viterbi algorithm exhibits better compensation ability for nonlinear impairment, especially when dealing with serious inter-symbol interference and nonlinear effects. In our experiment, we employ coherent detection to further improve the receiver sensitivity, so a complex baseband signal after down conversion at the receiver is inherently produced. In this scenario, the complex value neural network (CVNN) and RVNN equalizer connected with the Viterbi-Viterbi algorithm have better BER performance with an error rate lower than the HD-FEC threshold of 3.8 × 10-3.
Improved Weighted Multimodulus Blind Equalization Algorithm

XUE Wei,YANG Xiao-niu,ZHANG Zhao-yang

DOI: https://doi.org/10.16182/j.cnki.joss.2011.03.019

2011-01-01

Abstract:An improved algorithm in which the error function was switched between decision directed mode and weighted multimodulus mode was proposed based on the weighted multimodulus blind equalization algorithm.The proposed algorithm switches between decision directed mode and fixed weighted multimodulus mode according to the comparison of the signs of error functions,which can utilize the benefits of the two modes to achieve better constellation-matching performance and avoid the weight adjustment in dynamic weighted multimodulus algorithm.The simulation results demonstrate that compared with dynamic weighted multimodulus algorithm,the proposed algorithm can achieve faster convergence rate and lower steady state mean square error(MSE) with little addition in computation and is suitable for QAM signals.
High-speed PAM4 Transmission with a GeSi Electro-Absorption Modulator and Dual-Path Neural-Network-based Equalization

Xiaoke Ruan,Fan Yang,Lei Zhang,Hao Ming,Yanping Li,Fan Zhang

DOI: https://doi.org/10.1364/ol.401242

IF: 3.6

2020-01-01

Optics Letters

Abstract:Equalization based on artificial neural networks (NN) has proved to be an effective way for nonlinearity mitigation in various kinds of optical communication systems. In this Letter, we propose a novel methodology of dual-path neural network (DP-NN)-based equalization. By combining a linear equalizer with an input-pruned NN equalizer, DP-NN can effectively reduce the computation cost compared to a conventional NN equalizer. We confirm its feasibility through 4-ary pulse amplitude modulation (PAM4) transmission at a gross(net) bitrate of 160 Gb/s (133.3 Gb/s), based on a GeSi electro-absorption modulator operating at C-band. After a 2 km transmission, the bit error rate is below the 20% hard-decision forward-error-correction threshold of 1.5×10-2 with the DP-NN equalization, which outperforms the Volterra equalization and is comparable to conventional NN-based equalization.
Adaptive Partial-Response Neural Network Equalization for Bandwidth-limited PAM Transmission in Intra-datacenter Interconnect

An Yan,Sizhe Xing,Zhongya Li,Guoqiang Li,Aolong Sun,Penghao Luo,Jianyang Shi,Hongguang Zhang,Dong Wang,Xi Xiao,Zhixue He,Nan Chi,Junwen Zhang

DOI: https://doi.org/10.1109/jlt.2023.3348482

IF: 4.7

2024-01-01

Journal of Lightwave Technology

Abstract:High-speed optical transmission systems suffer from severe intersymbol-interference (ISI) caused by limited bandwidth. To address this challenge, previous studies on bandwidth limited signals have explored the use of a post filter after full-response equalization to whiten colored noise and enhance system performance. However, the post filter requires additional training or scanning processes to the appropriate coefficient values, which introduces additional overhead and time consumption. A low-complexity duobinary neural network equalization (DB-NNE) has been investigated recently, which is trained with the target of the duobinary version of the original symbol. Although this approach achieves both channel equalization and noise suppression, it may not achieve the optimal coefficient value for noise suppression. In this research, we propose an adaptive partial-response neural network equalization (APR-NNE) for bandwidth-limited signal processing and nonlinearity mitigation. Our approach seamlessly integrates channel equalization and noise suppression, achieving an optimal balance between these two functions. To verify the effectiveness of the proposed scheme, we experimentally demonstrate the transmission of beyond-150-GBaud PAM-4 signals in a system with -10-dB bandwidth of 57 GHz. The experimental results show that the APR-NNE can compensate the bandwidth-limitation induced ISI effectively, enabling the achievement of the 7% HD-FEC threshold for 155-GBaud PAM-4 signals over a 0.5-km SSMF in C-band.
Blind and Channel-agnostic Equalization Using Adversarial Networks

Vincent Lauinger,Manuel Hoffmann,Jonas Ney,Norbert Wehn,Laurent Schmalen

DOI: https://doi.org/10.48550/arXiv.2209.07277

2022-09-15

Abstract:Due to the rapid development of autonomous driving, the Internet of Things and streaming services, modern communication systems have to cope with varying channel conditions and a steadily rising number of users and devices. This, and the still rising bandwidth demands, can only be met by intelligent network automation, which requires highly flexible and blind transceiver algorithms. To tackle those challenges, we propose a novel adaptive equalization scheme, which exploits the prosperous advances in deep learning by training an equalizer with an adversarial network. The learning is only based on the statistics of the transmit signal, so it is blind regarding the actual transmit symbols and agnostic to the channel model. The proposed approach is independent of the equalizer topology and enables the application of powerful neural network based equalizers. In this work, we prove this concept in simulations of different -- both linear and nonlinear -- transmission channels and demonstrate the capability of the proposed blind learning scheme to approach the performance of non-blind equalizers. Furthermore, we provide a theoretical perspective and highlight the challenges of the approach.

Signal Processing,Information Theory,Machine Learning
Low Complexity Joint Neural Network Equalizer in a 248 Gbit/s VSB PS-PAM8 IM/DD Transmission System

Chen Wang,Kaihui Wang,Yuxuan Tan,Junjie Ding,Bohan Sang,Feng Wang,Bowen Zhu,Miao Kong,Wen Zhou,Jianjun Yu

2022-01-01

Abstract:We propose a novel joint neural network equalizer in a 248 Gbit/s VSB PS-PAM8 transmission system at the C-band. The proposed joint neural network equalizer outperforms the conventional neural network equalizer with significant MACC calculation complexity deduction. (C) 2022 The Author(s)
Trigonometric Coordinate Transformation Blind Equalization Algorithm Based on Bi-Direction Long and Short-Term Memory Neural Networks

Na Liu,Zuoxun Wang,Haiwen Wei

DOI: https://doi.org/10.1109/access.2024.3368857

IF: 3.9

2024-03-06

IEEE Access

Abstract:Aiming at the problem that the feedforward neural network blind equalization algorithm has a slow convergence rate and a large steady-state error when equalizing the high-order non-constant modulus signals, a trigonometric coordinate transformation blind equalization algorithm based on Bi-direction long and short-term memory (BLSTM) neural networks (BLSTM-TCT-CMA) is proposed. First, the BLSTM neural network has a strong processing ability for one-dimensional long-sequence signals, which is suitable for high-order signal sequence information. Secondly, a triangular coordinate transformation method was introduced in the BLSTM neural network loss function to transform the statistical modulus of the non-constant modulus signal into a constant modulus value, which speeds up convergence and further reduces the steady-state error. It was observed through the simulation that compared with the constant modulus blind equalization algorithm (CMA), the square contour blind equalization algorithm based on BP neural networks (SCA-BP-CMA) and the tunable activation functions blind equalization algorithm based on complex BP neural network (TAF-CBP-CMA). When the BLSTM-TCT-CMA equalized the 32QAM signal, the steady-state error was -13dB, and the loss function converged at 800 steps. When the 64QAM signal was equalized, the steady-state error was -10.5dB, and the loss function converged at 1200 steps. It is concluded that both indicators were optimal, and the CT-GRUNN-CMA output signal constellation was the clearest.

computer science, information systems,telecommunications,engineering, electrical & electronic
A Stop and Go Blind Equalization Algorithm for QAM Signals

Xue Wei,Yang Xiaoniu,Zhang Zhaoyang

DOI: https://doi.org/10.1109/wicom.2010.5600253

2010-01-01

Abstract:The constant modulus algorithm (CMA) has been widely used in the blind equalization of quadrature amplitude modulation (QAM) signals, but it has slow convergence rate and large steady state mean square error and the phase blind nature. In this paper, a modified constant modulus algorithm which incorporates a stop and go technique is proposed. The proposed algorithm can speed up convergence rate and decrease steady state mean square error and correct phase error and frequency offset simultaneously. The simulation results demonstrate the effectiveness of the proposed algorithm in improving convergence behavior and bit error rate (BER) performance.
A Multilayer Complex Neural Network Training Algorithm and Its Application in Adaptive Equalization

Li Chunguang,Liao Xiaofeng,Wu Zhongfu,Yu Juebang

DOI: https://doi.org/10.1007/s11767-001-0046-z

2001-01-01

Abstract:In this paper, the layer-by-layer optimizing algorithm for training multilayer neural network is extended for the case of a multilayer neural network whose inputs, weights, and activation functions are all complex. The updating of the weights of each layer in the network is based on the recursive least squares method. The performance of the proposed algorithm is demonstrated with application in adaptive complex communication channel equalization.
Blind Fuzzy Adaptation Step Control for a Concurrent Neural Network Equalizer

Kayol S. Mayer,Matheus S. De Oliveira,Candice Müller,Fernando C. C. De Castro,Maria C. F. De Castro

DOI: https://doi.org/10.1155/2019/9082362

2019-01-10

Wireless Communications and Mobile Computing

Abstract:Mobile communications, not infrequently, are disrupted by multipath propagation in the wireless channel. In this context, this paper proposes a new blind concurrent equalization approach that combines a Phase Transmittance Radial Basis Function Neural Network (PTRBFNN) and the classic Constant Modulus Algorithm (CMA) in a concurrent architecture, with a Fuzzy Controller (FC) responsible for adapting the PTRBFNN and CMA step sizes. Differently from the Neural Network (NN) based equalizers present in literature, the proposed Fuzzy Controller Concurrent Neural Network Equalizer (FC-CNNE) is a completely self-taught concurrent architecture that does not need any training. The Fuzzy Controller inputs are based on the estimated mean squared error of the equalization process and on its variation in time. The proposed solution has been evaluated over standard multipath VHF/UHF channels defined by the International Telecommunication Union. Results show that the FC-CNNE is able to achieve lower residual steady-state MSE value and/or faster convergence rate and consequently lower Bit Error Rate (BER) when compared to Constant Modulus Algorithm-Phase Transmittance Radial Basis Function Neural Network (CMA-PTRBFNN) equalizer.

computer science, information systems,telecommunications,engineering, electrical & electronic
A Nonlinear ANN Equalizer with Mini-Batch Gradient Descent in 40gbaud PAM-8 IM/DD System

Pengqi Gou,Jianjun Yu

DOI: https://doi.org/10.1016/j.yofte.2018.09.015

IF: 2.8

2018-01-01

Optical Fiber Technology

Abstract:We propose and validate a novel nonlinear artificial neural network (ANN) equalizer for PAM-8 transmission in IM/DD system. Mini-batch gradient descent is introduced to efficiently train ANN equalizer. Using the proposed ANN equalizer, we successfully transmit a 40Gbaud PAM-8 signal over 4-km SMF with BER under the threshold of 3.8 x 10(-3) and over 10-km SMF with BER under the threshold of 1 x 10(-2). We also elaborately compare the proposed ANN equalizer with other methods including LMS equalizer, Volterra equalizer and look-up table (LUT). Experimental results indicate that ANN achieves the best performance that is slightly superior to Volterra equalizer with computational complexity half reduced. To the best of our knowledge, this is the first time to adopt mini-batch gradient descent to train ANN equalizer in IM/DD system.
Fully-blind Neural Network Based Equalization for Severe Nonlinear Distortions in 112 Gbit/s Passive Optical Networks

Vincent Lauinger,Patrick Matalla,Jonas Ney,Norbert Wehn,Sebastian Randel,Laurent Schmalen

2024-01-18

Abstract:We demonstrate and evaluate a fully-blind digital signal processing (DSP) chain for 100G passive optical networks (PONs), and analyze different equalizer topologies based on neural networks with low hardware complexity.

Signal Processing,Machine Learning
Blind Equalization Algorithm with Global Convergence

CUI Wei-ning,YU Li-xin,TANG Yin-hui

2010-01-01

Abstract:In order to solve the problem that traditional adaptive equalization technology needs to transmit training sequences,using blind equalization technology,the equalizer converges by making use of characteristics of received signal.A new blind equalization algorithm(C&D algorithm) is proposed in this paper.The algorithm combines the CRIMNO algorithm and the Decision Directed (DD) algorithm,and makes use of irrelevance among signal sequences,so that it has global convergence.The experimental results show that,compared with the CRIMNO algorithm,under the condition of the same signal-to-noise ratio,the C&D algorithm has faster convergence speed,less computation,and smaller remaining error.
Frequency-Domain Complex-Valued NN Equalizer Based on Deep Reinforcement Learning for Photonics-Assisted W-Band 16QAM-DMT Signals System over 50 M

Wen Zhou,Siqi Wang,Sicong Xu,Qihang Wang,Zhihang Ou,Yuan Ma,Jingwen Lin,Jingtao Ge,Jianjun Yu

DOI: https://doi.org/10.1109/jlt.2024.3502116

IF: 4.7

2024-01-01

Journal of Lightwave Technology

Abstract:Photonics-assisted millimeter wave (MMW) can overcome the bandwidth bottleneck of electronic devices, but it also faces nonlinear damage caused by optoelectronic devices. Nonlinear equalization algorithms based on neural networks (NN) have gradually been introduced in the receiving end of 6G MMW communication, but the high computational complexity limits the real-time performance of the equalizer, and there is a tradeoff between the good equalization performance and low complexity. This paper proposes an Automatic Search Algorithm (ASA) based on deep reinforcement learning (DRL), which learns the functional relationship between equalizer structure parameters and optimization objectives through DRL, in order to quickly search for the optimal structure of various equalizers. The experiment demonstrated the W-band 16QAM discrete multi-tone (DMT) signal transmission in an intensity modulation and direct detection (IM/DD) system over a 50 m wireless link. We implement real-valued time-domain neural network equalizer (TDNNE), and complex-valued frequency-domain neural network equalizer (FDNNE), to solve the nonlinear issues respectively. Subsequently, ASA and greedy search algorithm were applied to TDNNE and FDNNE respectively, and the superiority of ASA was verified based on the bit error rate (BER) and the number of Multiply-Accumulate Operations (MACs) consumed by the equalizer. Finally, the experiment verified that using ASA to optimize the pruning of FDNNE can greatly reduce the complexity with almost no BER loss, while retaining the good equalization performance of NN equalizers and overcoming the disadvantage of high computational complexity. Thanks to our proposed complex-valued FDNNE based on DRL, the effective data rate reaches 11.4Gb/s with 0.5 dB receiver sensitivity improvement compared to VNLE and TDNNE. At the same time, after pruning optimization, the average number of MACs consumed by FDNNE at various input optical powers into photodiode (PD) reduced by 37%. We believe ASA is an important application of using DRL to optimize neural networks in the field of nonlinear equalization. It can efficiently explore the performance limits of equalizers and has great potential for high-precision reception in future 6G wireless communications.

A Blind Equalization Algorithm Based On Forward Neural Network In Pam System

A Integrated Algorithm of DDLMS and Godard for Blind Equilization

Blind Equalization Algorithm of QAM Signal Based on SNR

Implementation of blind equalization algorithm for HF QAM based on DSP

Excess MSE Analysis of the Concurrent Constant Modulus Algorithm and Soft Decision-Directed Scheme for Blind Equalisation

A New Dual Mode Blind Equalization Algorithm

Deep Learning Equalizer Connected with Viterbi-Viterbi Algorithm for PAM D-Band Radio over Fiber Link

Improved Weighted Multimodulus Blind Equalization Algorithm

High-speed PAM4 Transmission with a GeSi Electro-Absorption Modulator and Dual-Path Neural-Network-based Equalization

Adaptive Partial-Response Neural Network Equalization for Bandwidth-limited PAM Transmission in Intra-datacenter Interconnect

Blind and Channel-agnostic Equalization Using Adversarial Networks

Low Complexity Joint Neural Network Equalizer in a 248 Gbit/s VSB PS-PAM8 IM/DD Transmission System

Trigonometric Coordinate Transformation Blind Equalization Algorithm Based on Bi-Direction Long and Short-Term Memory Neural Networks

A Stop and Go Blind Equalization Algorithm for QAM Signals

A Multilayer Complex Neural Network Training Algorithm and Its Application in Adaptive Equalization

Blind Fuzzy Adaptation Step Control for a Concurrent Neural Network Equalizer

A Nonlinear ANN Equalizer with Mini-Batch Gradient Descent in 40gbaud PAM-8 IM/DD System

Fully-blind Neural Network Based Equalization for Severe Nonlinear Distortions in 112 Gbit/s Passive Optical Networks

Blind Equalization Algorithm with Global Convergence

Frequency-Domain Complex-Valued NN Equalizer Based on Deep Reinforcement Learning for Photonics-Assisted W-Band 16QAM-DMT Signals System over 50 M