[Paper Review] Design of Non-orthogonal and Noncoherent Massive MIMO for Scalable URLLC Beyond 5G

He Chen, Zheng Dong+2 more|arXiv (Cornell University)|Jan 29, 2020

Advanced Wireless Communication Technologies40 references4 citations

TL;DR

This paper proposes a non-orthogonal, noncoherent massive MIMO framework for scalable ultra-reliable low-latency communications (sURLLC) in beyond-5G systems. By leveraging large-scale antenna arrays and a novel uniquely-decomposable constellation group (UDCG) differential modulation scheme, it enables reliable user signal separation without instantaneous channel state information, achieving superior error performance over conventional coherent ZF and Euclidean distance-based designs, especially for cell-edge users.

ABSTRACT

This paper is to design and optimize a non-orthogonal and noncoherent massive multiple-input multiple-output (MIMO) framework towards enabling scalable ultra-reliable low-latency communications (sURLLC) in wireless systems beyond 5G. In this framework, the huge diversity gain associated with the large-scale antenna array in massive MIMO systems is leveraged to ensure ultrahigh reliability. To reduce the overhead and latency induced by the channel estimation process, we advocate the noncoherent communication technique which does not need the knowledge of instantaneous channel state information (CSI) but only depends on the large-scale fading coefficients for information decoding. To boost the scalability of the system considered, we enable the non-orthogonal channel access of multiple users by devising a new differential modulation scheme to assure that each transmitted signal matrix can be uniquely determined in the noise-free case and be reliably estimated in noisy cases when the antenna array size is scaled up. The key idea is to make the transmitted signals from multiple users be superimposed properly over the air such that when the sum-signal is correctly detected, the signals sent by all users can be uniquely determined. To further improve the average error performance when the array antenna number is large, we propose a max-min Kullback-Leibler (KL) divergence-based design by jointly optimizing the transmitted powers of all users and the sub-constellation assignment among them. Simulation results show that the proposed design significantly outperforms the existing max-min Euclidean distance-based counterpart in terms of error performance. Moreover, our proposed approach also has a better error performance than the conventional coherent zero-forcing (ZF) receiver with orthogonal channel training, particularly for cell-edge users.

Motivation & Objective

• To address the scalability, reliability, and low-latency requirements of ultra-reliable low-latency communications (URLLC) in beyond-5G wireless networks.
• To eliminate the need for pilot-based channel estimation in massive MIMO by adopting noncoherent communication relying only on large-scale fading coefficients.
• To enable non-orthogonal multiple access (NOMA) in massive MIMO systems while ensuring unique signal recovery at the receiver through a new differential modulation scheme.
• To optimize system performance by jointly designing user power allocation and sub-constellation assignment using a max-min Kullback-Leibler (KL) divergence criterion.
• To improve error performance over existing methods, particularly for cell-edge users, in high-mobility and high-fading environments.

Proposed method

• Introduces a noncoherent massive MIMO framework that decodes signals based solely on large-scale fading coefficients, eliminating pilot overhead and reducing latency.
• Designs a new differential modulation scheme using a uniquely-decomposable constellation group (UDCG) to ensure that individual user signals can be uniquely recovered from the superimposed sum signal.
• Employs a max-min KL divergence-based optimization to jointly allocate transmit powers and assign sub-constellations across users, maximizing the minimum reliability across all users.
• Derives theoretical conditions under which the transmitted signal matrices are uniquely decodable in the noise-free case and robustly estimable in noisy conditions as the array size increases.
• Uses a matrix decomposition approach where the signal covariance matrix is expressed as $ \mathbf{R} = \mathbf{X}_T^H \mathbf{D} \mathbf{X}_T $, ensuring uniqueness of $ \mathbf{X}_T $ under the UDCG condition.
• Proves that the proposed UDCG ensures that different signal matrices produce distinct likelihood functions, enabling reliable maximum-likelihood detection.

Experimental results

Research questions

• RQ1 Can noncoherent massive MIMO with non-orthogonal access achieve scalable, ultra-reliable, and low-latency communication without channel state information at the receiver?
• RQ2 How can user signals be uniquely recovered from a superimposed sum signal in a noncoherent massive MIMO system with large antenna arrays?
• RQ3 What is the optimal power allocation and constellation assignment strategy that maximizes the minimum user reliability in a noncoherent massive MIMO system?
• RQ4 How does the proposed max-min KL divergence-based design compare to conventional max-min Euclidean distance-based designs in terms of error performance?
• RQ5 To what extent does the proposed framework outperform coherent zero-forcing (ZF) receivers with orthogonal training, especially for cell-edge users?

Key findings

• The proposed noncoherent massive MIMO system with UDCG-based differential modulation achieves unique signal recovery in the noise-free case and reliable estimation in noisy conditions as the number of antennas increases.
• The max-min KL divergence-based design significantly outperforms the conventional max-min Euclidean distance-based design in terms of error performance, especially at high signal-to-noise ratios.
• Simulation results show that the proposed scheme achieves better error performance than the conventional coherent ZF receiver with orthogonal training, particularly for cell-edge users suffering from high pathloss and interference.
• The framework enables scalable sURLLC by supporting massive connectivity with reduced pilot overhead and low-latency operation through noncoherent detection.
• Theoretical analysis proves that the signal matrix $ \mathbf{X}_T $ is uniquely determined by the covariance matrix $ \mathbf{R} $, ensuring reliable detection under the UDCG condition.
• The optimization framework ensures fairness by maximizing the minimum reliability across all users, enhancing robustness in heterogeneous channel conditions.