Variational autoencoder based receiver for orthogonal time frequency space modulation

Abstract The emerging orthogonal time frequency space (OTFS) modulation is demonstrated to offer reliable communication performance advantages over orthogonal frequency division multiplexing (OFDM) in doubly-dispersive fading channel. However, the existing embedded pilot-aided method to estimate channel impulse response (CIR) requires enormous spectral overhead to avoid the contamination of pilot symbols. In this paper, we present a variational autoencoder (VAE) based receiver for OTFS modulation that achieve a joint estimation and detection without pilot in delay-Doppler (DD) domain. The variational approach is considered to simplify the problem, and evidence lower bound (ELBO) is derived as loss function. In encoder step, an approximate posterior probability is introduced and utilized to minimize the Kullback-Leibler (KL) distance. Then we estimate CIR in decoder step and maximize the ELBO at last. From our simulation results, the proposed VAE based receiver for OTFS modulation enjoys a promising performance with other methods.

[1]  Emanuele Viterbo,et al.  Effective Diversity of OTFS Modulation , 2020, IEEE Wireless Communications Letters.

[2]  A. Chockalingam,et al.  On the Diversity of Uncoded OTFS Modulation in Doubly-Dispersive Channels , 2018, IEEE Transactions on Wireless Communications.

[3]  Diederik P. Kingma,et al.  An Introduction to Variational Autoencoders , 2019, Found. Trends Mach. Learn..

[4]  Sheetal Kalyani,et al.  Design of Communication Systems Using Deep Learning: A Variational Inference Perspective , 2019, IEEE Transactions on Cognitive Communications and Networking.

[5]  Yi Hong,et al.  Interference Cancellation and Iterative Detection for Orthogonal Time Frequency Space Modulation , 2018, IEEE Transactions on Wireless Communications.

[6]  Derrick Wing Kwan Ng,et al.  A Simple Variational Bayes Detector for Orthogonal Time Frequency Space (OTFS) Modulation , 2019, IEEE Transactions on Vehicular Technology.

[7]  Yi Hong,et al.  Embedded Pilot-Aided Channel Estimation for OTFS in Delay–Doppler Channels , 2018, IEEE Transactions on Vehicular Technology.

[8]  Bo Ai,et al.  Towards Realistic High-Speed Train Channels at 5G Millimeter-Wave Band—Part I: Paradigm, Significance Analysis, and Scenario Reconstruction , 2018, IEEE Transactions on Vehicular Technology.

[9]  Andrea Goldsmith,et al.  Neural Network Detection of Data Sequences in Communication Systems , 2018, IEEE Transactions on Signal Processing.

[10]  Lei Zhang,et al.  Low-Complexity Symbol Detection and Interference Cancellation for OTFS System , 2020, IEEE Transactions on Communications.

[11]  G. D. Surabhi,et al.  Low-Complexity Linear Equalization for OTFS Modulation , 2020, IEEE Communications Letters.

[12]  Akbar M. Sayeed,et al.  Orthogonal time-frequency signaling over doubly dispersive channels , 2004, IEEE Transactions on Information Theory.

[13]  Pingzhi Fan,et al.  Channel Estimation for Orthogonal Time Frequency Space (OTFS) Massive MIMO , 2019, ICC 2019 - 2019 IEEE International Conference on Communications (ICC).

[14]  David Burshtein,et al.  Unsupervised Linear and Nonlinear Channel Equalization and Decoding Using Variational Autoencoders , 2019, IEEE Transactions on Cognitive Communications and Networking.

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

[16]  Linglong Dai,et al.  Time-Frequency Training OFDM with High Spectral Efficiency and Reliable Performance in High Speed Environments , 2012, IEEE Journal on Selected Areas in Communications.

[17]  Shashank Tiwari,et al.  Low complexity LMMSE Receiver for OTFS , 2019, IEEE Communications Letters.