RCNet: Incorporating Structural Information Into Deep RNN for Online MIMO-OFDM Symbol Detection With Limited Training

In this paper, we investigate online learning-based MIMO-OFDM symbol detection strategies focusing on a special recurrent neural network (RNN) – reservoir computing (RC). We first introduce the Time-Frequency RC to take advantage of the structural information inherent in OFDM signals. Using the time domain RC and the time-frequency RC as building blocks, we provide two extensions of the shallow RC to RCNet: 1) Stacking multiple time domain RCs; 2) Stacking multiple time-frequency RCs into a deep structure. The combination of RNN dynamics, the time-frequency structure of MIMO-OFDM signals, and the deep network enables RCNet to handle the interference and nonlinear distortion of MIMO-OFDM signals to outperform existing methods. Unlike most existing NN-based detection strategies, RCNet is also shown to provide a good generalization performance even with a limited online training set (i.e, similar amount of reference signals/training as standard model-based approaches). Numerical experiments demonstrate that the introduced RCNet can offer a faster learning convergence and as much as 20% gain in bit error rate over a shallow RC structure by compensating for the nonlinear distortion of the MIMO-OFDM signal, such as due to power amplifier compression in the transmitter or due to finite quantization resolution in the receiver.

[1]  Lajos Hanzo,et al.  Fifty Years of MIMO Detection: The Road to Large-Scale MIMOs , 2015, IEEE Communications Surveys & Tutorials.

[2]  Ami Wiesel,et al.  Learning to Detect , 2018, IEEE Transactions on Signal Processing.

[3]  Sumei Sun,et al.  A Survey on Power-Amplifier-Centric Techniques for Spectrum- and Energy-Efficient Wireless Communications , 2015, IEEE Communications Surveys & Tutorials.

[4]  Hao Chen,et al.  Artificial Intelligence-Enabled Cellular Networks: A Critical Path to Beyond-5G and 6G , 2019, IEEE Wireless Communications.

[5]  Herbert Jaeger,et al.  The''echo state''approach to analysing and training recurrent neural networks , 2001 .

[6]  Jochen J. Steil,et al.  Analyzing the weight dynamics of recurrent learning algorithms , 2005, Neurocomputing.

[7]  Herbert Jaeger,et al.  Discovering multiscale dynamical features with hierarchical Echo State Networks , 2008 .

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

[9]  Krishna Sayana,et al.  Downlink MIMO in LTE-advanced: SU-MIMO vs. MU-MIMO , 2012, IEEE Communications Magazine.

[10]  Yuichi Nakamura,et al.  Approximation of dynamical systems by continuous time recurrent neural networks , 1993, Neural Networks.

[11]  Helmut Hauser,et al.  Echo state networks with filter neurons and a delay&sum readout , 2010, Neural Networks.

[12]  Yang Yi,et al.  Artificial Intelligence Enabled Internet of Things: Network Architecture and Spectrum Access , 2020, IEEE Computational Intelligence Magazine.

[13]  Yang Yi,et al.  Realizing Green Symbol Detection via Reservoir Computing: An Energy-Efficiency Perspective , 2018, 2018 IEEE International Conference on Communications (ICC).

[14]  Yahong Rosa Zheng,et al.  Brain-Inspired Wireless Communications: Where Reservoir Computing Meets MIMO-OFDM , 2018, IEEE Transactions on Neural Networks and Learning Systems.

[15]  Razvan Pascanu,et al.  On the difficulty of training recurrent neural networks , 2012, ICML.

[16]  Yoshua Bengio,et al.  Hierarchical Recurrent Neural Networks for Long-Term Dependencies , 1995, NIPS.

[17]  Jakob Hoydis,et al.  Adaptive Neural Signal Detection for Massive MIMO , 2019, IEEE Transactions on Wireless Communications.

[18]  Martin Döttling,et al.  Radio technologies and concepts for IMT-Advanced , 2009 .

[19]  Lingjia Liu,et al.  Deep Reservoir Computing Meets 5G MIMO-OFDM Systems in Symbol Detection , 2020, AAAI.

[20]  Tara N. Sainath,et al.  Deep Neural Networks for Acoustic Modeling in Speech Recognition , 2012 .

[21]  Arian Maleki,et al.  Optimality of large MIMO detection via approximate message passing , 2015, 2015 IEEE International Symposium on Information Theory (ISIT).

[22]  Claudio Gallicchio,et al.  Deep reservoir computing: A critical experimental analysis , 2017, Neurocomputing.

[23]  Erik Agrell,et al.  Faster Recursions in Sphere Decoding , 2009, IEEE Transactions on Information Theory.

[24]  Herbert Jaeger,et al.  Reservoir computing approaches to recurrent neural network training , 2009, Comput. Sci. Rev..

[25]  Yoav Freund,et al.  Boosting a weak learning algorithm by majority , 1995, COLT '90.

[26]  Claudio Gallicchio,et al.  Echo State Property of Deep Reservoir Computing Networks , 2017, Cognitive Computation.

[27]  Geoffrey E. Hinton,et al.  Speech recognition with deep recurrent neural networks , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.