Unmanned aerial vehicle (UAV) communication has emerged as a prominent technology for emergency communications (e.g., natural disaster) in the Internet of Things (IoT) networks to enhance the ability of disaster prediction, damage assessment, and rescue operations promptly. A UAV can be deployed as a flying base station (BS) to collect data from time-constrained IoT devices and then transfer it to a ground gateway (GW). In general, the latency constraint at IoT devices and UAV’s limited storage capacity highly hinder practical applications of UAV-assisted IoT networks. In this paper, full-duplex (FD) radio is adopted at the UAV to overcome these challenges. In addition, half-duplex (HD) scheme for UAV-based relaying is also considered to provide a comparative study between two modes (viz., FD and HD). Herein, a device is considered to be successfully served if its data is collected by the UAV and conveyed to GW timely during flight time. In this context, we aim to maximize the number of served IoT devices by jointly optimizing bandwidth, power allocation, and the UAV trajectory while satisfying each device’s requirement and the UAV’s limited storage capacity. The formulated optimization problem is troublesome to solve due to its non-convexity and combinatorial nature. Towards appealing applications, we first relax binary variables into continuous ones and transform the original problem into a more computationally tractable form. By leveraging inner approximation framework, we derive newly approximated functions for non-convex parts and then develop a simple yet efficient iterative algorithm for its solutions. Next, we attempt to maximize the total throughput subject to the number of served IoT devices. Finally, numerical results show that the proposed algorithms significantly outperform benchmark approaches in terms of the number of served IoT devices and system throughput.

Federated learning (FL) allows multiple edge computing nodes to jointly build a shared learning model without having to transfer their raw data to a centralized server, thus reducing communication overhead. However, FL still faces a number of challenges such as non-iid distributed data and heterogeneity of user equipments (UEs). Enabling a large number of UEs to join the training process in every round raises a potential issue of the heavy global communication burden. To address these issues, we generalize the current state-of-the-art Federated Averaging (FedAvg) by adding a weight-based proximal term to the local loss function. The proposed FL algorithm runs stochastic gradient descent in parallel on a sampled subset of the total UEs with replacement during each global round. We provide a convergence upper bound characterizing the trade-off between convergence rate and global rounds, showing that a small number of active UEs per round still guarantees convergence. Next, we employ the proposed FL algorithm in wireless Internet-of-Things (IoT) networks to minimize either total energy consumption or completion time of FL, where a simple yet efficient path-following algorithm is developed for its solutions. Finally, numerical results on unbalanced datasets are provided to demonstrate the performance improvement and robustness on the convergence rate of the proposed FL algorithm over FedAvg. They also reveal that the proposed algorithm requires much less training time and energy consumption than the FL algorithm with full user participation. These observations advocate the proposed FL algorithm for a paradigm shift in bandwidth-constrained learning wireless IoT networks.

A communication system is considered consisting of a full-duplex multiple-antenna base station (BS) and multiple single-antenna downlink users (DLUs) and single-antenna uplink users (ULUs), where the latter need to harvest energy for transmitting information to the BS. The communication is thus divided into two phases. In the first phase, the BS uses all available antennas for conveying information to DLUs and wireless energy to ULUs via information and energy beamforming, respectively. In the second phase, ULUs send their independent information to the BS using their harvested energy while the BS transmits the information to the DLUs. In both the phases, the communication is operated at the same time and over the same frequency band. The aim is to maximize the sum rate and energy efficiency under ULU achievable information throughput constraints by jointly optimizing beamforming and time allocation. The utility functions of interest are nonconcave and the involved constraints are nonconvex, so these problems are computationally troublesome. To address them, path-following algorithms are proposed to arrive at least at local optima. The proposed algorithms iteratively improve the objectives with convergence guaranteed. Simulation results demonstrate that they achieve rapid convergence and outperform conventional solutions.

In-band full-duplex (FD) operation is practically more suited for short-range communications such as WiFi and small-cell networks, due to its current practical limitations on the self-interference cancellation. In addition, cell-free massive multiple-input multiple-output (CF-mMIMO) is a new and scalable version of MIMO networks, which is designed to bring service antennas closer to end user equipments (UEs). To achieve higher spectral and energy efficiencies (SE-EE) of a wireless network, it is of practical interest to incorporate FD capability into CF-mMIMO systems to utilize their combined benefits. We formulate a novel and comprehensive optimization problem for the maximization of SE and EE in which power control, access point-UE (AP-UE) association and AP selection are jointly optimized under a realistic power consumption model, resulting in a difficult class of mixed-integer nonconvex programming. To tackle the binary nature of the formulated problem, we propose an efficient approach by exploiting a strong coupling between binary and continuous variables, leading to a more tractable problem. In this regard, two low-complexity transmission designs based on zero-forcing (ZF) are proposed. Combining tools from inner approximation framework and Dinkelbach method, we develop simple iterative algorithms with polynomial computational complexity in each iteration and strong theoretical performance guaranteed. Furthermore, towards a robust design for FD CF-mMIMO, a novel heap-based pilot assignment algorithm is proposed to mitigate effects of pilot contamination. Numerical results show that our proposed designs with realistic parameters significantly outperform the well-known approaches (i.e., small-cell and collocated mMIMO) in terms of the SE and EE. Notably, the proposed ZF designs require much less execution time than the simple maximum ratio transmission/combining.

This paper addresses the problem of minimising latency in computation offloading with digital twin (DT) wire-less edge networks for industrial Internet-of-Things (IoT) environment via ultra-reliable and low latency communications (URLLC) links. The considered DT-aided edge networks provide a powerful computing framework to enable computation-intensive services, where the DT is used to model the computing capacity of edge servers and optimise the resource allocation of the entire system. The objective function is comprised of local processing latency, URLLC-based transmission latency and edge processing latency, subject to both communication and computation resources budgets. In this regard, the minimum latency is obtained by jointly optimising the transmit power, user association, offloading portions, the processing rate of users and edge servers. The formulated problem is highly complicated due to complex non-convex constraints and strong coupling variables. To deal with this computationally intractable problem, we propose an iterative algorithm which decomposes the original problem into three sub-problems and resolve this problem in the fashion of alternating optimisation approach combined with an inner convex approximation framework. Simulation results demonstrate the effectiveness of the proposed method in reducing the latency compared with other benchmark schemes.

The throughput of users with poor channel conditions, such as those at a cell edge, is a bottleneck in wireless systems. A major part of the power budget must be allocated to serve these users in guaranteeing their quality-of-service (QoS) requirements, hampering QoS for other users, and thus compromising the system reliability. In non-orthogonal multiple access (NOMA), the message intended for a user with a poor channel condition is decoded by itself and by another user with a better channel condition. The message intended for the latter is then successively decoded by itself after canceling the interference of the former. The overall information throughput is thus improved by this particular successive decoding and interference cancellation. This paper aims to design linear precoders/beamformers for signal superposition at the base stations of NOMA multiple-input multiple-output multi-cellular systems to maximize the overall sum throughput subject to the users’ QoS requirements, which are imposed independently on the users’ channel conditions. This design problem is formulated as the maximization of a highly nonlinear and nonsmooth function subject to nonconvex constraints, which is very computationally challenging. Path-following algorithms for its solution, which invoke only a simple convex problem of moderate dimension at each iteration, are developed. Generating a sequence of improved points, these algorithms converge at least to a local optimum. Extensive numerical simulations are then provided to demonstrate their merit.