Deep Learning for Wireless Physical Layer: Opportunities and Challenges

Machine learning(ML) has been widely applied to the upper layers of wireless communication systems for various purposes, such as deployment of cognitive radio and communication network. However, its application to the physical layer is hampered by sophisticated channel environments and limited learning ability of conventional ML algorithms. Deep learning(DL) has been recently applied for many fields, such as computer vision and natural language processing, given its expressive capacity and convenient optimization capability. The potential application of DL to the physical layer has also been increasingly recognized because of the new features for future communications, such as complex scenarios with unknown channel models, high speed and accurate processing requirements; these features challenge conventional communication theories. This paper presents a comprehensive overview of the emerging studies on DL-based physical layer processing, including leveraging DL to redesign a module of the conventional communication system(for modulation recognition, channel decoding, and detection) and replace the communication system with a radically new architecture based on an autoencoder. These DL-based methods show promising performance improvements but have certain limitations, such as lack of solid analytical tools and use of architectures that are specifically designed for communication and implementation research, thereby motivating future research in this field.

Improving Expectation Propagation with Lattice Reduction for Massive MIMO Detection

Efficient massive MIMO detection for practical deployment, which is with spatially correlated channel and high-order modulation, is a challenging topic for the fifth generation mobile communication(5 G). In this paper, we propose a lattice reduction aided expectation propagation(LRA-EP) algorithm for massive MIMO detection. LRA-EP applies expectation propagation in lattice reduced MIMO system to approach the distribution of lattice reduced constellation point by iterative refinement on its parameters(mean and covariance). The parameter refinement is based on the lattice reduced, well-conditioned MIMO channel. Numerical result shows that LRA-EP outperforms classic EP based MIMO detection(EPD) with 5～7 dB in terms of required signalto-noise ratio(SNR) for 1% packet error rate in spatially correlated channel for 256-QAM. We also show that LRA-EP has lower computation complexity than EPD.

Design and Implementation of a TDD-Based 128-Antenna Massive MIMO Prototype System

Owing to the dramatic mobile IP growth,the emerging Internet of Things,and cloud-based applications,wireless networking is witnessing a paradigm shift.By fully exploiting spatial degrees of freedom,massive multiple-input-multiple-output(MIMO) systems promise significant gains in data rates and link reliability.Although the research community has recognized the theoretical benefits of these systems,building the hardware of such complex systems is a challenge in practice.This paper presents a time division duplex(TDD)-based 128-antenna massive MIMO prototype system from theory to reality.First,an analytical signal model is provided to facilitate the setup of a feasible massive MIMO prototype system.Second,a link-level simulation consistent with practical TDDbased massive MIMO systems is conducted to guide and validate the massive MIMO system design.We design and implement the TDDbased 128-antenna massive MIMO prototype system with the guidelines obtained from the link-level simulation.Uplink real-time video transmission and downlink data transmission under the configuration of multiple single-antenna users are achieved.Comparisons withstate-of-the-art prototypes demonstrate the advantages of the proposed system in terms of antenna number,bandwidth,latency,and throughput.The proposed system is also equipped with scalability,which makes the system applicable to a wide range of massive scenarios.

Fixed-point Implementation of Approximate Message Passing （AMP） algorithm in massive MIMO systems

In massive multiple input and multiple output （MIMO） systems the challenge is the detection of the individual signals from the composite signal with a large system limit. The optimal detector becomes prohibitively complex. The approximate message passing （AMP） algorithm, designed for compressed sensing, has attracted researchers to counter this problem due to its reduced complexity with a large system limit. For this reason the AMP algorithm has been used for detection in massive MIMO systems. In this paper, we focus on implementing this algorithm in a fixed-point format. To obtain an implementation friendly architecture, we propose approximations for the mean and variance estimation functions within the algorithm. These estimation functions are obtained using the log-sum approximation, then taking the exponent of the result. The log-sum approximation is obtained by the Jacobian logarithm with a correction function. We also provide a modification of the correction function for the approximations that best suits our case. We then transform the algorithm with the approximated functions to fixed-point and provide a BER performance for the algorithm with the variables set to 16-bit word lengths using the hybrid ＂ScaledDouble＂ data types.