Guest Editorial Massive Access for 5G and Beyond—Part II

Xiaoming Chen, Derrick Wing Kwan Ng, Wei Yu, Erik G. Larsson, Naofal Al-Dhahir, Robert Schober
2021 IEEE Journal on Selected Areas in Communications  
T HIS second part of the two-part Special Issue (SI) on massive access for 5G and beyond starts with several papers on massive access techniques, then switches to coverage enhancement approaches, and finishes with a paper on the application of massive access in industrial Internet-of-Things (IoT). The paper "A new path division multiple access for the massive MIMO-OTFS networks," by Li et al., focuses on a new path division multiple access (PDMA) scheme for both uplink and downlink massive
more » ... ple-input multiple-output (MIMO) networks in a high-mobility scenario, where orthogonal time-frequency space (OTFS) modulation is adopted. For the uplink, a path scheduling algorithm is proposed to properly assign angle-domain resources at the user end to achieve inter-user interference-free transmission. For the downlink, a low-complexity beamforming scheme over the angle-delay-Doppler domain is proposed to achieve inter-user interference-free communication. The paper "Nested hybrid cylindrical array design and DoA estimation for massive IoT networks," by Lin et al., presents a novel hybrid uniform circular cylindrical array (UCyA) for massive IoT networks to reduce cost and power consumption while maintaining high network access capability. A nested hybrid beamforming structure based on sparse array techniques is designed, and a corresponding channel estimation method based on second-order channel statistics is proposed. Moreover, a new tensor-based 2-D direction-of-arrival estimation algorithm tailored for the proposed hybrid is provided Xiaoming Chen is with the College to improve estimation accuracy with affordable computational complexity. The paper "Advanced NOMA receivers from a unified variational inference perspective," by Meng et al., studies the multiuser detection (MUD) design at the receiver side for nonorthogonal multiple access (NOMA), so as to improve resource efficiency and support massive access. A unified variational inference (VI) perspective on various universal NOMA MUD algorithms, such as belief propagation (BP), expectation propagation (EP), vector EP (VEP), approximate message passing (AMP), and vector AMP (VAMP), is introduced. Such a unified perspective would not only help the design and adaptation of NOMA receivers but also open the door for the systematic design of joint active user detection and multiuser decoding for sporadic grant-free transmission. The paper "Energy-efficient non-orthogonal multicast and unicast transmission of cell-free massive MIMO systems with SWIPT," by Tan et al., investigates the energy-efficient resource allocation for layered-division multiplexing (LDM) based nonorthogonal multicast and unicast transmission in cell-free massive MIMO systems, where each user equipment (UE) performs wireless information and power transfer simultaneously. To maximize the energy efficiency in the massive access setting, a first-order algorithm is developed to find both an initial point and the nearly optimal solution. Moreover, an accelerated algorithm is designed to improve the speed of convergence. The paper "Generalized user grouping in NOMA based on overlapping coalition formation game," by Chen et al., proposes a novel generalized user grouping (GuG) concept for NOMA from an overlapping perspective, which allows each user to participate in multiple groups but subject to an individual maximum power constraint. A joint power control and GuG algorithm is designed to maximize the system sum rate for the scenario of massive connectivity in 5G systems. The paper "Optimized shallow neural networks for sum-rate maximization in energy harvesting downlink multiuser NOMA systems," by Lee et al., considers a power allocation problem in energy harvesting downlink multiuser NOMA systems in which a transmitter sends desired messages to its respective receivers by using harvested energy. A reinforcement learning approach based on a shallow neural network structure is adopted to find the optimal power allocation.
doi:10.1109/jsac.2020.3018841 fatcat:fjzpbpm5brbmnnpehhsjc4lkoa