Short-Packet Communications in Multi-Hop Networks with WET: Performance Analysis and Deep Learning-Aided Optimization

in IEEE Transactions on Wireless Communications (in press)

FedFog: Network-Aware Optimization of Federated Learning over Wireless Fog-Cloud System

in IEEE Transactions on Wireless Communications (in press)

Federated learning (FL) is capable of performing large distributed machine learning tasks across multiple edge users by periodically aggregating trained local parameters. To address key challenges of ... [more ▼]

Federated learning (FL) is capable of performing large distributed machine learning tasks across multiple edge users by periodically aggregating trained local parameters. To address key challenges of enabling FL over a wireless fogcloud system (e.g., non-i.i.d. data, users’ heterogeneity), we first propose an efficient FL algorithm based on Federated Averaging (called FedFog) to perform the local aggregation of gradient parameters at fog servers and global training update at the cloud. Next, we employ FedFog in wireless fog-cloud systems by investigating a novel network-aware FL optimization problem that strikes the balance between the global loss and completion time. An iterative algorithm is then developed to obtain a precise measurement of the system performance, which helps design an efficient stopping criteria to output an appropriate number of global rounds. To mitigate the straggler effect, we propose a flexible user aggregation strategy that trains fast users first to obtain a certain level of accuracy before allowing slow users to join the global training updates. Extensive numerical results using several real-world FL tasks are provided to verify the theoretical convergence of FedFog. We also show that the proposed co-design of FL and communication is essential to substantially improve resource utilization while achieving comparable accuracy of the learning model. [less ▲]

UAV Relay-Assisted Emergency Communications in IoT Networks: Resource Allocation and Trajectory Optimization

in IEEE Transactions on Wireless Communications (2022), 21(3), 1621-1637

Unmanned aerial vehicle (UAV) communication hasemerged as a prominent technology for emergency communi-cations (e.g., natural disaster) in the Internet of Things (IoT)networks to enhance the ability of ... [more ▼]

Unmanned aerial vehicle (UAV) communication hasemerged as a prominent technology for emergency communi-cations (e.g., natural disaster) in the Internet of Things (IoT)networks to enhance the ability of disaster prediction, damageassessment, and rescue operations promptly. A UAV can bedeployed 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’slimited storage capacity highly hinder practical applicationsof UAV-assisted IoT networks. In this paper, full-duplex (FD)radio is adopted at the UAV to overcome these challenges. Inaddition, half-duplex (HD) scheme for UAV-based relaying isalso considered to provide a comparative study between twomodes (viz., FD and HD). Herein, a device is considered tobe successfully served iff its data is collected by the UAV andconveyed to GW timely during flight time. In this context,we aim to maximize the number of served IoT devices byjointly optimizing bandwidth, power allocation, and the UAVtrajectory while satisfying each device’s requirement and theUAV’s limited storage capacity. The formulated optimizationproblem is troublesome to solve due to its non-convexity andcombinatorial nature. Towards appealing applications, we firstrelax binary variables into continuous ones and transform theoriginal problem into a more computationally tractable form.By leveraging inner approximation framework, we derive newlyapproximated functions for non-convex parts and then develop asimple yet efficient iterative algorithm for its solutions. Next,we attempt to maximize the total throughput subject to thenumber of served IoT devices. Finally, numerical results showthat the proposed algorithms significantly outperform benchmarkapproaches in terms of the number of served IoT devices andsystem throughput. [less ▲]

Learning-Assisted User Clustering in Cell-Free Massive MIMO-NOMA Networks

in IEEE Transactions on Vehicular Technology (2021), 70(12), 12872-12887

The superior spectral efficiency (SE) and user fairness feature of non-orthogonal multiple access (NOMA) systems are achieved by exploiting user clustering (UC) more efficiently. However, a random UC ... [more ▼]

The superior spectral efficiency (SE) and user fairness feature of non-orthogonal multiple access (NOMA) systems are achieved by exploiting user clustering (UC) more efficiently. However, a random UC certainly results in a suboptimal solution while an exhaustive search method comes at the cost of high complexity, especially for systems of medium-to-large size. To address this problem, we develop two efficient unsupervised machine learning based UC algorithms, namely k-means++ and improved k-means++, to effectively cluster users into disjoint clusters in cell-free massive multiple-input multiple-output (CFmMIMO) system. Adopting full-pilot zero-forcing at access points (APs) to comprehensively assess the system performance, we formulate the sum SE optimization problem taking into account power constraints at APs, necessary conditions for implementing successive interference cancellation, and required SE constraints at user equipments. The formulated optimization problem is highly non-convex, and thus, it is difficult to obtain the global optimal solution. Therefore, we develop a simple yet efficient iterative algorithm for its solution. In addition, the performance of collocated massive MIMO-NOMA (COmMIMO-NOMA) system is also characterized. Numerical results are provided to show the superior performance of the proposed UC algorithms compared to baseline schemes. The effectiveness of applying NOMA in CFmMIMO and COmMIMO systems is also validated. [less ▲]

Efficient Federated Learning Algorithm for Resource Allocation in Wireless IoT Networks

in IEEE Internet of Things Journal (2021), 8(5), 3394-3409

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 ... [more ▼]

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. [less ▲]

Energy Efficiency Maximization in RIS-Aided Cell-Free Network with Limited Backhaul

in IEEE Communications Letters (2021), 25(6), 1974-1978

Integrating the reconfigurable intelligent surface in a cell-free (RIS-CF) network is an effective solution to improve the capacity and coverage of future wireless systems with low cost and power ... [more ▼]

Integrating the reconfigurable intelligent surface in a cell-free (RIS-CF) network is an effective solution to improve the capacity and coverage of future wireless systems with low cost and power consumption. The reflecting coefficients of RISs can be programmed to enhance signals received at users. This letter addresses a joint design of transmit beamformers at access points and reflecting coefficients at RISs to maximize the energy efficiency (EE) of RIS-CF networks, taking into account the limited backhaul capacity constraints. Due to a very computationally challenging nonconvex problem, we develop a simple yet efficient alternating descent algorithm for its solution. Numerical results verify that the EE of RIS-CF networks is greatly improved, showing the benefit of using RISs. [less ▲]

Resource Allocation for AF Relaying Wireless-Powered Networks with Nonlinear Energy Harvester

in IEEE Communications Letters (2021), 25(1), 1089-7798

This letter considers a relay-based wireless-powered communication network to assist wireless communication between a source and multiple users. In particular, the relay adopts a nonlinear energy model to ... [more ▼]

This letter considers a relay-based wireless-powered communication network to assist wireless communication between a source and multiple users. In particular, the relay adopts a nonlinear energy model to harvest energy from a power beacon and subsequently uses it for information transmission over timedivision multiple access. Aiming at the maximization of end-toend (e2e) sum throughput, we formulate a novel optimization problem that jointly optimizes the power and time fraction for energy and information transmission. For a simple yet efficient solution for the nonconvex problem, we first convert it to a more computationally tractable problem and then develop an iterative algorithm, in which closed-form solutions are obtained at each iteration. The effectiveness of our proposed approach is verified and demonstrated through simulation results. Moreover, the results reveal that the source should transmit with its maximum allowable power budget to obtain the optimal e2e sum throughput. [less ▲]

UAV Relay-Assisted Emergency Communications in IoT Networks: Resource Allocation and Trajectory Optimization

E-print/Working paper (2020)

Unmanned aerial vehicle (UAV) communication has emerged as a prominent technology for emergency communications (e.g., natural disaster) in Internet of Things (IoT) networks to enhance the ability of ... [more ▼]

Unmanned aerial vehicle (UAV) communication has emerged as a prominent technology for emergency communications (e.g., natural disaster) in Internet of Things (IoT) networks to enhance the ability of disaster prediction, damage assessment, and rescue operations promptly. In this paper, a UAV is deployed as a flying base station (BS) to collect data from time-constrained IoT devices and then transfer the data to a ground gateway (GW). In general, the latency constraint at IoT users and the limited storage capacity of UAV highly hinder practical applications of UAV-assisted IoT networks. In this paper, full-duplex (FD) technique 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 successfully served iff its data is collected by UAV and conveyed to GW within the flight time. In this context, we aim at maximizing the number of served IoT devices by jointly optimizing bandwidth and power allocation, as well as the UAV trajectory, while satisfying the requested timeout (RT) requirement of each device and the UAV’s limited storage capacity. The formulated optimization problem is troublesome to solve due to its non-convexity and combinatorial nature. Toward appealing applications, we first relax binary variables into continuous values 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 the amount of collected data. [less ▲]

Resource Allocation for Energy Efficiency in OFDMA-Enabled WPCN

in IEEE Wireless Communications Letters (2020), 9(12), 2049-2053

This letter considers a wireless powered communication network (WPCN), where an energy-constrained device directly uses harvested energy from a power transfer source to transmit independent signals to ... [more ▼]

This letter considers a wireless powered communication network (WPCN), where an energy-constrained device directly uses harvested energy from a power transfer source to transmit independent signals to multiple Internet of Things (IoT) users using orthogonal frequency division multiple access (OFDMA). Our goal is to maximize the system energy efficiency (EE) by jointly optimizing the duration of energy harvesting (EH), subcarrier and power allocation. The formulated problem is a mixed integer nonlinear programming (MINLP) problem due to the presence of binary assignment variables, and thus it is very challenging to solve it directly. By leveraging Dinkelbach method, a very efficient iterative algorithm with closed-form solutions in each iteration is developed, where its convergence is guaranteed. Numerical results show that the proposed algorithm obtains a fast convergence and outperforms baseline algorithms. Notably, they also reveal that the power source should transmit its maximum allowable power to obtain the optimal EE performance. [less ▲]

Hybrid User Pairing for Spectral and Energy Efficiencies in Multiuser MISO-NOMA Networks with SWIPT

in IEEE Transactions on Communications (2020), 68(8), 4874-4890

In this paper, we propose a novel hybrid user pairing (HUP) scheme in multiuser multiple-input single-output nonorthogonal multiple access networks with simultaneous wireless information and power ... [more ▼]

In this paper, we propose a novel hybrid user pairing (HUP) scheme in multiuser multiple-input single-output nonorthogonal multiple access networks with simultaneous wireless information and power transfer. In this system, two information users with distinct channel conditions are optimally paired while energy users perform energy harvesting (EH) under non-linearity of the EH circuits. We consider the problem of jointly optimizing user pairing and power allocation to maximize the overall spectral efficiency (SE) and energy efficiency (EE) subject to userspecific quality-of-service and harvested power requirements. A new paradigm for the EE-EH trade-off is then proposed to achieve a good balance of network power consumption. Such design problems are formulated as the maximization of nonconcave functions subject to the class of mixed-integer non-convex constraints, which are very challenging to solve optimally. To address these challenges, we first relax binary pairing variables to be continuous and transform the design problems into equivalent non-convex ones, but with more tractable forms. We then develop low-complexity iterative algorithms to improve the objectives and converge to a local optimum by means of the inner approximation framework. Simulation results show the convergence of proposed algorithms and the SE and EE improvements of the proposed HUP scheme over state-of-the-art designs. In addition, the effects of key parameters such as the number of antennas and dynamic power at the BS, target data rates, and energy threshold, on the system performance are evaluated to show the effectiveness of the proposed schemes in balancing resource utilization. [less ▲]