Joint Model and Data Driven Receiver Design for Data-Dependent Superimposed Training Scheme with Imperfect Hardware

Data-dependent superimposed training (DDST) scheme has shown the potential to achieve high bandwidth efficiency, while encounters symbol misidentification caused by hardware imperfection. To tackle these challenges, a joint model and data driven receiver scheme is proposed in this paper. Specifically, based on the conventional linear receiver model, the least squares (LS) estimation and zero forcing (ZF) equalization are first employed to extract the initial features for channel estimation and data detection. Then, shallow neural networks, named CE-Net and SD-Net, are developed to refine the channel estimation and data detection, where the imperfect hardware is modeled as a nonlinear function and data is utilized to train these neural networks to approximate it. Simulation results show that compared with the conventional minimum mean square error (MMSE) equalization scheme, the proposed one effectively suppresses the symbol misidentification and achieves similar or better bit error rate (BER) performance without the second-order statistics about the channel and noise.

[1]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[2]  Gaoming Huang,et al.  Semiblind Channel Estimation and Symbol Detection for Block Transmission Using Superimposed Training , 2012, 2012 IEEE 12th International Conference on Computer and Information Technology.

[3]  Chaojin Qing,et al.  ELM-Based Frame Synchronization in Nonlinear Distortion Scenario Using Superimposed Training , 2021, IEEE Access.

[4]  Yang-Seok Choi,et al.  Adaptive Nonlinear Digital Self-Interference Cancellation for Mobile Inband Full-Duplex Radio: Algorithms and RF Measurements , 2014, 2015 IEEE Global Communications Conference (GLOBECOM).

[5]  Emil Björnson,et al.  Impact of Residual Transmit RF Impairments on Training-Based MIMO Systems , 2014, IEEE Transactions on Communications.

[6]  Pingzhi Fan,et al.  Data detection and coding for data-dependent superimposed training , 2014, IET Signal Process..

[7]  Vishnu Raj,et al.  Backpropagating Through the Air: Deep Learning at Physical Layer Without Channel Models , 2018, IEEE Communications Letters.

[8]  Christian Fager,et al.  Analysis of nonlinear distortion in phased array transmitters , 2017, 2017 Integrated Nonlinear Microwave and Millimetre-wave Circuits Workshop (INMMiC).

[9]  Paulo Montezuma,et al.  Analytical Performance Evaluation of Precoding Techniques for Nonlinear Massive MIMO Systems With Channel Estimation Errors , 2018, IEEE Transactions on Communications.

[10]  Ananthram Swami,et al.  Channel estimation and symbol detection for block transmission using data-dependent superimposed training , 2005, IEEE Signal Processing Letters.

[11]  Florence Alberge,et al.  Deep Learning Constellation Design for the AWGN Channel With Additive Radar Interference , 2019, IEEE Transactions on Communications.

[12]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[13]  Michail Matthaiou,et al.  Machine Learning-Based Channel Prediction in Massive MIMO With Channel Aging , 2020, IEEE Transactions on Wireless Communications.

[14]  Pankaj Das,et al.  Effects of Channel Estimation Error and Nonlinear HPA on the Performance of OFDM in Rayleigh Channels with Application to 802.11n WLAN , 2008, 2008 IEEE Wireless Communications and Networking Conference.

[15]  Jian Xiong,et al.  Unsupervised Learning-Based Fast Beamforming Design for Downlink MIMO , 2019, IEEE Access.

[16]  Jie Yang,et al.  Data-Driven Deep Learning for Automatic Modulation Recognition in Cognitive Radios , 2019, IEEE Transactions on Vehicular Technology.

[17]  Geoffrey Ye Li,et al.  Power of Deep Learning for Channel Estimation and Signal Detection in OFDM Systems , 2017, IEEE Wireless Communications Letters.

[19]  Mounir Ghogho,et al.  Data Identifiability for Data-Dependent Superimposed Training , 2007, 2007 IEEE International Conference on Communications.

[20]  Jinjun Tang,et al.  Crash injury severity analysis using a two-layer Stacking framework. , 2019, Accident; analysis and prevention.

[21]  Rui Dinis,et al.  Turbo equalization with cancelation of nonlinear distortion for CP-assisted and zero-padded MC-CDM schemes , 2009, IEEE Transactions on Communications.

[22]  Wan-Jen Huang,et al.  A Precoding Scheme for Eliminating Data Identification Problem in Single Carrier System Using Data-Dependent Superimposed Training , 2019, IEEE Access.

[23]  Geoffrey Ye Li,et al.  Deep Learning-Based Downlink Channel Prediction for FDD Massive MIMO System , 2019, IEEE Communications Letters.

[24]  Adel A. M. Saleh,et al.  Frequency-Independent and Frequency-Dependent Nonlinear Models of TWT Amplifiers , 1981, IEEE Trans. Commun..

[25]  Xiaojun Yuan,et al.  CNN-Based Signal Detection for Banded Linear Systems , 2018, IEEE Transactions on Wireless Communications.

[26]  Fumiyuki Adachi,et al.  Deep-Learning-Based Millimeter-Wave Massive MIMO for Hybrid Precoding , 2019, IEEE Transactions on Vehicular Technology.

[27]  Matthieu Crussière,et al.  On Providing the Theoretical EVM Limit for Tone Reservation PAPR Reduction Technique , 2020, 2020 IEEE 21st International Workshop on Signal Processing Advances in Wireless Communications (SPAWC).

[28]  Han Zhang,et al.  Linearly time-varying channel estimation for MIMO/OFDM systems using superimposed training , 2010, IEEE Transactions on Communications.

[29]  Ahmad Nimr,et al.  Enhancing Least Square Channel Estimation Using Deep Learning , 2020, 2020 IEEE 91st Vehicular Technology Conference (VTC2020-Spring).

[30]  Dabal Pedamonti,et al.  Comparison of non-linear activation functions for deep neural networks on MNIST classification task , 2018, ArXiv.

[32]  Chaojin Qing,et al.  Deep Learning for CSI Feedback Based on Superimposed Coding , 2019, IEEE Access.

[33]  Hsueh-Jyh Li,et al.  Investigation on Data Identification Problem for Data-Dependent Superimposed Training , 2012, 2012 IEEE 75th Vehicular Technology Conference (VTC Spring).

[34]  Hsueh-Jyh Li,et al.  Elimination of Data Identification Problem for Data-Dependent Superimposed Training , 2015, IEEE Transactions on Signal Processing.

[35]  Claude Oestges,et al.  The COST 2100 MIMO channel model , 2011, IEEE Wirel. Commun..

[36]  Geoffrey Ye Li,et al.  Practical Considerations on Channel Estimation for Up-Link MC-CDMA Systems , 2008, IEEE Transactions on Wireless Communications.

[37]  Geoffrey Ye Li,et al.  Convolutional Neural Network-Based Multiple-Rate Compressive Sensing for Massive MIMO CSI Feedback: Design, Simulation, and Analysis , 2019, IEEE Transactions on Wireless Communications.

[38]  Guan Gui,et al.  Deep Learning for Super-Resolution Channel Estimation and DOA Estimation Based Massive MIMO System , 2018, IEEE Transactions on Vehicular Technology.

[39]  Geoffrey Ye Li,et al.  ComNet: Combination of Deep Learning and Expert Knowledge in OFDM Receivers , 2018, IEEE Communications Letters.

[40]  Yoshua Bengio,et al.  The Benefits of Over-parameterization at Initialization in Deep ReLU Networks , 2019, ArXiv.

[41]  Youyun Xu,et al.  Hardware Impairments Aware Transceiver for Full-Duplex Massive MIMO Relaying , 2015, IEEE Transactions on Signal Processing.

[42]  Jitendra K. Tugnait,et al.  Self-Interference Suppression in Doubly-Selective Channel Estimation Using Superimposed Training , 2007, 2007 IEEE International Conference on Communications.

[43]  Chaojin Qing,et al.  ELM-Based Frame Synchronization in Burst-Mode Communication Systems With Nonlinear Distortion , 2020, IEEE Wireless Communications Letters.

[44]  Hui Zhang,et al.  An Enhanced Partial-Data Superimposed Training Scheme for OFDM Systems , 2020, IEEE Communications Letters.