The Internet of Drones (IoD) integrates drone technology with the Internet of Things, enabling efficient data collection and communication applications. Federated learning (FL) in IoD networks facilitates collaborative model training while preserving data privacy but imposes significant computational and communication demands on resource-constrained drones. Federated knowledge distillation learning (FedKD) addresses this challenge by training both a large teacher model and a smaller student model locally but only updating the smaller student model, thereby reducing communication overhead. This article tackles the resource allocation problem in FedKD within IoD networks, focusing on optimizing CPU computing resource, wireless transmission power, and bandwidth allocation to minimize overall drone energy consumption. We formulate this as an optimization problem, considering constraints on latency, computing resource, bandwidth, and power. To effectively address this problem, we design a low-complexity algorithm. Extensive simulations validate our approach, showing it reduces energy consumption by an average of 85% compared to FedKD and 94% compared to FedAvg (a standard FL algorithm).

Advanced Air Mobility (AAM) seeks to establish a next-generation air transportation system by leveraging autonomous unpiloted aerial vehicles (UAVs) to transport passengers and cargo between locations previously underserved or unserved by traditional aviation. Achieving AAM at scale requires overcoming significant challenges in airspace management, classification, and traffic control to safely accommodate the increasing volume of UAV operations. This paper presents a comprehensive design for air corridors to facilitate efficient aerial transport and formulates a multi-UAV coordination problem within these corridors. The objective is to enable each UAV to autonomously make control decisions based on local observations gathered from onboard sensors. This decentralized control approach is modeled as a multi-agent partially observable Markov decision process (POMDP), aiming at minimizing UAV travel time while ensuring adherence to corridor boundaries and collision avoidance. To address the complexities posed by varying state dimensions and types, we propose a novel Hybrid Transformer-based Multi-agent Reinforcement Learning (HTransRL) architecture. HTransRL integrates a customized transformer model into an actor-critic network, effectively processing both sequential and non-sequential observed states of varying sizes while capturing their correlations. This enables safe and efficient UAV navigation. Simulation results show that in test environments similar to or simpler than training scenarios, HTransRL achieves a successful arrival rate exceeding 90% in worst-case test scenarios. In test environments more complex than training scenarios, HTransRL demonstrates superior scalability compared to two baseline methods, achieving higher arrival rates and comparable travel times.

With the evolution of wireless sensing technology, WiFi-based human identification has demonstrated tremendous potential in human-computer interaction and home security. However, most existing research operates in controlled environments, overlooking the complexities of real-world scenarios, such as signal fading, signal interference and uncertainty, and the diversity of application requirements. This article presents a comprehensive analysis of a series of representative research articles on WiFi-based human identification and summarizes five major challenges in achieving high-precision identification in complex application scenarios. Non-line-of-sight (NLOS) user sensing, coexistence user sensing, dynamic group user sensing, cross-domain user sensing, and multi-task sensing. Additionally, this article proposes a series of current solutions, including improving the signal-to-noise ratio (SNR) of NLOS sensing from the perspective of communication signals and sensing models, addressing coexistence sensing issues, designing adaptable tasks for dynamic user groups, leveraging techniques like adversarial learning and transfer learning for cross-domain problems, and employing modular deep learning architectures. By providing a comprehensive overview of WiFi-based human identification, this survey not only offers insights into current research but also charts a roadmap for future investigations. It is anticipated that this survey will stimulate innovative research endeavors and foster the expansion of wireless sensing technology across diverse application domains.

Generative Adversarial Networks (GANs) are deep learning models that learn and generate new samples similar to existing ones. Traditionally, GANs are trained in centralized data centers, raising data privacy concerns due to the need for clients to upload their data. To address this, Federated Learning (FL) integrates with GANs, allowing collaborative training without sharing local data. However, this integration is complex because GANs involve two interdependent models—the generator and the discriminator—while FL typically handles a single model over distributed datasets. In this article, we propose a novel asynchronous FL framework for GANs, called AsyncFedGAN, designed to efficiently and distributively train both models tailored for molecule generation. AsyncFedGAN addresses the challenges of training interactive models, resolves the straggler issue in synchronous FL, reduces model staleness in asynchronous FL, and lowers client energy consumption. Our extensive simulations for molecular discovery show that AsyncFedGAN achieves convergence with proper settings, outperforms baseline methods, and balances model performance with client energy usage.

In wireless Internet of Things (IoT) networks, Federated Reinforcement Learning (FRL) has emerged as a decentralized strategy for data-driven decision-making, enabling devices to learn directly from real-time environmental interactions, sidestepping the need for labeled data. This method promises enhanced data privacy and finds practical applications in autonomous driving, smart grids, and industrial automation. However, the integrity of FRL can be compromised by malicious clients injecting false data, underlining the need for a fault-tolerant mechanism to sustain the robustness and accuracy of the learning phase. Moreover, the inherent client heterogeneity within IoT networks propels the demand for judicious client selection, optimizing computational and communication resources. This paper investigates client selection problem within a fault-tolerant FRL framework for wireless IoT networks. Our objective is to explore the tradeoff between maximizing client participation and minimizing energy consumption of IoT devices. We formulate our problem as a mixed-integer linear programming (MILP) model and design an efficient algorithm with low computational complexity to address it. Extensive simulations are conducted to demonstrate the superiority of our proposed algorithm.

Vehicular mobile edge computing (vMEC) and non-orthogonal multiple access (NOMA) have emerged as promising technologies for enabling low-latency and high-throughput applications in vehicular networks. In this paper, we propose a novel multi-agent deep deterministic policy gradient (MADDPG) approach for resource allocation in NOMA-based vMEC systems. Our approach leverages deep reinforcement learning (DRL) to enable vehicles to offload computation-intensive tasks to nearby edge servers, optimizing resource allocation decisions while ensuring low-latency communication. We introduce an attention mechanism within the MADDPG model to dynamically focus on relevant information from the input state and joint actions, enhancing the model’s predictive accuracy. Additionally, we propose an attention-based experience replay method to expedite network convergence. The simulation results highlight the effectiveness of multi-agent reinforcement learning (MARL) algorithms, such as MADDPG with attention, in achieving better convergence and performance in various scenarios. The influence of different model parameters, such as input data volumes, task load levels, and resource configurations, on optimization results is also evident. The decision making processes of agents are dynamic and depend on factors specific to the task and environment.

Recent advances in deep learning have accelerated its use in various applications, such as cellular image analysis and molecular discovery. In molecular discovery, a generative adversarial network (GAN), which comprises a discriminator to distinguish generated molecules from existing molecules and a generator to generate new molecules, is one of the premier technologies due to its ability to learn from a large molecular data set efficiently and generate novel molecules that preserve similar properties. However, different pharmaceutical companies may be unwilling or unable to share their local data sets due to the geo-distributed and sensitive nature of molecular data sets, making it impossible to train GANs in a centralized manner. In this paper, we propose aGraphconvolutional network inGenerativeAdversarialNetworks viaFederated learning (GraphGANFed) framework, which integrates graph convolutional neural Network (GCN), GAN, and federated learning (FL) as a whole system to generate novel molecules without sharing local data sets. In GraphGANFed, the discriminator is implemented as a GCN to better capture features from molecules represented as molecular graphs, and FL is used to train both the discriminator and generator in a distributive manner to preserve data privacy. Extensive simulations are conducted based on the three benchmark data sets to demonstrate the feasibility and effectiveness of GraphGANFed. The molecules generated by GraphGANFed can achieve high novelty$(\approx 100 )$and diversity$(\gt 0.9)$. The simulation results also indicate that 1) a lower complexity discriminator model can better avoid mode collapse for a smaller data set, 2) there is a tradeoff among different evaluation metrics, and 3) having the right dropout ratio of the generator and discriminator can avoid mode collapse.

Internet of drones (IoD), where drones act as the Internet of things (IoT) devices, makes IoT networks much more flexible and responsive because of high mobility of drones. Machine learning (ML) techniques can be applied in IoD to facilitate multiple applications such as object tracking and traffic surveillance, where ML data samples are collected and analyzed in the edge servers at the ground base station (BS). However, aggregating all data samples incurs huge wireless network traffic and potential data privacy leakage. Federated learning (FL) is then proposed to address these challenges by performing local training in drones and aggregating model parameters at the BS without sharing raw data samples. The FL performance in IoD networks is greatly affected by limited drone batteries which power FL local training, wireless data transmission, and drones’ movements. This paper hence investigates the energy-efficient FL in IoD networks to optimize CPU frequencies of drones’ on-board computing units such that total energy consumption of all the drones in the FL process can be minimized, while satisfying the FL training time requirement. We formulate the problem as a non-linear programming problem and then design an algorithm with polynomial time complexity to derive the optimum solution. Extensive simulations are conducted to demonstrate the performance of our proposed algorithm.

Advanced Air Mobility (AAM) is envisioned to bring a new mode of air transportation with promising benefits to the society, at large. AAM services such as air taxis and air ambulances are expected to bridge the digital divide between urban and rural/tribal regions. For AAM services to be deployed at large-scale, there are several technical challenges that need to be addressed. Reliable communications to support autonomous airspace operations is one such challenge. This paper presents a comprehensive architecture for ground-based communication support for AAM vehicles. The architecture includes: (1) air corridor design, (2) unit air cell that serves as a building block for air corridor, and (3) novel antenna designs for the base station and the AAM vehicle. Detailed analysis is presented for these architectural elements to demonstrate their viability for practical implementation.

To reap the benefit of big data generated by the massive number of Internet of Things (IoT) devices while preserving data privacy, federated learning (FL) has been proposed to enable IoT devices to train machine learning models locally. That is, instead of sharing the local data sets, different clients in terms of IoT devices only need to upload their local models to a centralized FL server. Client selection in FL is critical to maximize the number of qualified clients, who can successfully upload their local models to the FL server before the predefined deadline. Normally, client selection is coupled with wireless resource management owing to the fact that different clients need to share the same spectrum to upload their local models. The existing solutions of joint optimizing client selection and resource management are designed based on frequency-division multiple access (FDMA) or time-division multiple access (TDMA), which do not consider the dynamics of the clients and lead to low bandwidth utilization. In this article, we propose the Nonorthogonal-Multiple-Access (NOMA)-based resource allocation for client selection in FL to dynamically and jointly optimize client selection for each global iteration as well as the transmission power of each selected client in each time slot within a global iteration. We design the deep-reinforcement-learn-based client selection in NOMA-based federated learning (DREAM-FL) algorithm to solve the problem. Extensive simulations are conducted to demonstrate that DREAM-FL can select more qualified clients and has higher model accuracy than FDMA and TDMA-based solutions.

In urban environments, tall buildings or structures can pose limits on the direct channel link between a base station (BS) and an Internet of Thing device (IoTD) for wireless communication. Unmanned aerial vehicles (UAVs) with a mounted reconfigurable intelligent surface (RIS), denoted as UAV-RIS, have been introduced in recent works to enhance the system throughput capacity by acting as a relay node between the BS and the IoTDs in wireless access networks. Uncoordinated UAVs or RIS phase shift elements will make unnecessary adjustments that can significantly impact the signal transmission to IoTDs in the area. The concept of Age of Information (AoI) is proposed in wireless network research to categorize the freshness of the received update message. To minimize the Average Sum of AoI (ASoA) in the network, two model-free deep reinforcement learning (DRL) approaches—Off-Policy deepQ-network (DQN) and On-Policy proximal policy optimization (PPO)—are developed to solve the problem by jointly optimizing the RIS phase shift, the location of the UAV-RIS, and the IoTD transmission scheduling for large-scale Internet of Things wireless networks. Analysis of loss functions and extensive simulations is performed to compare the stability and convergence performance of the two algorithms. The results reveal the superiority of the On-Policy approach, PPO, over the Off-Policy approach, DQN, in terms of stability, convergence speed, and under diverse environment settings.

Free space optics (FSO), which uses light as the carrier to transmit data in free space, has been demonstrated as a secure and high-speed solution for long distance and line-of-sight wireless communications. Applying FSO as fronthaul/backhaul communications between base stations (BSs) and the gateway can significantly increase the fronthaul/backhaul link capacity. Traditionally, the gateway has to be equipped with multiple FSO transceivers, each of which is used to communicate with a BS by establishing a dedicated FSO. In this paper, we propose to use a steerable FSO system, where the gateway is equipped with a steerable FSO transceiver to communicate with multiple FSO transceivers at different BSs in a time division multiplexing manner. Applying the steerable FSO system can reduce the number of FSO transceivers at the gateway, and thus reduce the capital cost of implementing an FSO based fronthaul/backhaul network. We formulate the transmission scheduling problem in the steerable FSO system to optimize the active time for each FSO link associated with the steerable FSO transceiver such that the overall delay of transmitting a packet from geo-distributed BSs to the steerable FSO transceiver at the gateway is minimized, while guaranteeing the latency requirements of the BSs. We propose the laTency aWare transmIssionScheduling for sTeerable FSO (TWIST) algorithm, which is designed based on Sequential Quadratic Programming, to efficiently solve the proposed problem. The performance of TWIST is validated via extensive simulations.

Owing to the resource-constrained feature of Internet of Things (IoT) devices, offloading tasks from IoT devices to the nearby mobile edge computing (MEC) servers can not only save the energy of IoT devices but also reduce the response time of executing the tasks. However, offloading a task to the nearest MEC server may not be the optimal solution due to the limited computing resources of the MEC server. Thus, jointly optimizing the offloading decision and resource management is critical, but yet to be explored. Here, offloading decision refers to where to offload a task and resource management implies how much computing resource in an MEC server is allocated to a task. By considering the waiting time of a task in the communication and computing queues (which are ignored by most of the existing works) as well as tasks priorities, we propose the Deep reinforcement lEarning based offloading deCision and rEsource managemeNT (DECENT) algorithm, which leverages the advantage actor critic method to optimize the offloading decision and computing resource allocation for each arriving task in real-time such that the cumulative weighted response time can be minimized. The performance of DECENT is demonstrated via different experiments.

Federated learning (FL) is a collaborative machine learning framework to enable different clients such as Internet of Things (IoT) devices to participate in a machine learning model training process, while preserving data privacy. Client selection is critical to determine the performance of FL. Most of the existing client selection methods aim to maximize the number of selected clients, who can upload their local models before the deadline, in each global iteration, thus potentially accelerating the model convergence rate. However, these methods ignore the fact that most of the IoT devices are powered by on-board batteries and harvested green energy from the environment to prolong battery life. Hence, clients selected by these methods may not have sufficient energy to upload their local models in a global iteration or are unable to participate in the training process in the near future due to battery drainage. In this paper, we propose a novel client selection, entitled “EnerGy-AwaRe CliEnt SElection for Green FeDerated Learning (GREED)”, to optimize the trade-off between maximizing the number of selected clients and minimizing the energy drawn from batteries for the selected clients, while ensuring that all the selected clients have sufficient energy to upload their local models before the deadline. The performance of GREED is validated via extensive simulations.

Federated learning (FL) has been proposed to efficiently and privacy-preserving distributed machine learning architecture for the Internet of Things (IoT). In a wireless FL system, clients in IoT devices train their local models over the local data sets. The derived local models are uploaded to an FL server to generate a global model, broadcasted to the clients in the next global iteration for further training. Owing to the heterogeneous feature of the clients, client selection is critical to determine the overall training time. Traditionally, the objective of client selection is to select the maximum number of clients who can derive and upload their local models before the deadline in each global iteration. However, selecting more clients increases the energy consumption of the clients. Moreover, selecting the maximum number of clients is unnecessary as having fewer clients in early global iterations and more clients in later global iterations have been proved to achieve higher model accuracy. Hence, this article proposes to dynamically adjust and optimize the tradeoff between maximizing the number of selected clients and minimizing the total energy consumption of the clients by selecting suitable clients and allocating appropriate resources in terms of CPU frequency and transmission power. We formulate the joint client selection and resource management problem and design the energy and latency-aware resource management and client selection (ELASTIC) algorithm to efficiently solve the problem. Extensive simulations are conducted to demonstrate the performance of ELASTIC.

In a free space optics (FSO)-based drone-assisted mobile network, a drone-mounted base station (DBS) can be rapidly deployed over a place of interest to relay traffic between Internet of Things (IoT) devices and a macro base station (MBS), and FSO is applied as the fronthaul solution between the DBS and the MBS to provide a high link capacity at a long distance. However, due to inevitable optical beam misalignment, having a smaller divergence angle of the optical beam may increase the outage probability of the FSO-based fronthaul link. On the other hand, having a larger divergence angle may reduce the capacity of the FSO-based fronthaul link. So, it is critical but challenging to determine the divergence angle in order to optimize the tradeoff between minimizing the outage probability and maximizing the capacity for the FSO-based fronthaul link in the context of the FSO-based drone-assisted mobile network. In this article, we formulate an optimization problem to determine the optimal divergence angle that can minimize the link outage probability, while guaranteeing the capacity of the FSO-based fronthaul link no less than the threshold. The link outage and capacity aware divergence angle (LEARN) algorithm is designed to efficiently solve the problem. The performance of LEARN is demonstrated via extensive simulations.

Random interleavers are widely employed in digital communication systems to combat channel fading and burst errors. In applications such as grant-free access by Internet of Things (IoT) devices, accurately identifying a specific irregular interleaving pattern within an interleaver period is vital to both terminal recognition and data recovery. In this work, we investigate effective approaches for random interleaving pattern identification in Reed-Solomon (RS) coded data streams. We first propose an algorithm of low computational complexity to detect positions of code symbols belonging to the same RS codeword group (RSCG) under modest bit error rate. We further develop another low-complexity algorithm to successfully identify random interleaving patterns for RS code symbols within each RS codeword under moderate to high error rate applications. Our theoretical analysis and simulation results corroborate to demonstrate the effectiveness of our algorithms.

Federated learning (FL) is a distributed machine learning technique to address the data privacy issue. Participant selection is critical to determine the latency of the training process in a heterogeneous FL architecture, where users with different hardware setups and wireless channel conditions communicate with their base station to participate in the FL training process. Many solutions have been designed to consider computational and uploading latency of different users to select suitable participants such that the straggler problem can be avoided. However, none of these solutions consider the waiting time of a participant, which refers to the latency of a participant waiting for the wireless channel to be available, and the waiting time could significantly affect the latency of the training process, especially when a huge number of participants are involved in the training process and share the wireless channel in the time-division duplexing manner to upload their local FL models. In this paper, we consider not only the computational and uploading latency but also the waiting time (which is estimated based on an M/G/1 queueing model) of a participant to select suitable participants. We formulate an optimization problem to maximize the number of selected participants, who can upload their local models before the deadline in a global iteration. The Latency awarE pARticipant selectioN (LEARN) algorithm is proposed to solve the problem and the performance of LEARN is validated via simulations.

In this paper, we propose to cache popular Internet of Things (IoT) resources in the brokers (which can be considered as application layer middlewares) by applying the CoAP Publish/Subscribe protocol in order to reduce the energy consumption of the servers (e.g., IoT devices), which host these resources. If an IoT resource is cached in a broker, all the requests to retrieve the content of the IoT resource will be delivered to the broker, which responses to the requests by sending related contents, thus increasing the power consumption of the broker. In order to reduce the operational expenditure of the broker provider, each broker is powered by green energy and uses on-grid energy as a backup. On-gird energy consumption of the brokers may be different. That is, some brokers with low green energy generation and more cached IoT resources may consume more on-grid energy consumption than brokers with high green energy generation and less cached IoT resources. In order to minimize the total on-grid energy consumption of the brokers, the Green Energy Aware Resource caching (GEAR) algorithm is proposed to balance energy demands by re-allocating/re-caching the popular IoT resources among the brokers. The performance of GEAR is validated via simulations.

In the Green Cloudlet Network (GCN) architecture, each User Equipment (UE) is associated with an Avatar (a private virtual machine for executing its UE's offloaded tasks) in a cloudlet located at the network edge. In order to reduce the operational expenditure for maintaining the distributed cloudlets, each cloudlet is powered by green energy and uses on-grid power as a backup. Owing to the spatial dynamics of energy demands and green energy generations, the energy gap (i.e., energy demand minus green energy generation) among different cloudlets in the network is unbalanced, i.e., some cloudlets' energy demands can be fully provisioned by their green energy generations but others need to utilize on-grid power to meet their energy demands. The unbalanced energy gap increases the on-grid power consumption of the cloudlets. In this paper, we propose the Green-energy aware Avatar Placement (GAP) strategy to minimize the total on-grid power consumption of the cloudlets by migrating Avatars among the cloudlets according to the cloudlets' residual green energy, while guaranteeing the service level agreement (the End-to-End (E2E) delay requirement between a UE and its Avatar). Simulation results show that GAP can save 57.1 and 57.6 percent of on-grid power consumption as compared to the two other Avatar placement strategies, i.e., Static Avatar Placement and Follow me AvataR, respectively.

Multiple drone-mounted base stations (DBSs) are used to be deployed over a disaster struck area to help mobile users (MUs) communicate with working BSs, which are located beyond the disaster-struck area. DBSs are considered as relay nodes between MUs and working BSs. In order to relax the bottleneck in wireless backhaul links, we propose a cooperative drone assisted mobile access network architecture by enabling DBSs (whose backhaul links are congested) to offload their traffic to other DBSs (whose backhaul links are not congested) via DBS-to-DBS communications. We formulate the DBS placement and channel allocation problem in the context of the cooperative drone assisted mobile access network architecture, and design a COoperative DBS plAcement and CHannel allocation (COACH) algorithm to solve the problem. The performance of COACH is demonstrated via extensive simulations.

In a traditional big data network, data streams generated by User Equipments (UEs) are uploaded to the remote cloud (for further processing) via the Internet. However, moving a huge amount of data via the Internet may lead to a long End-to-End (E2E) delay between a UE and its computing resources (in the remote cloud) as well as severe traffic jams in the Internet. To overcome this drawback, we propose a cloudlet network to bring the computing and storage resources from the cloud to the mobile edge. Each base station is attached to one cloudlet and each UE is associated with its Avatar in the cloudlet to process its data locally. Thus, the E2E delay between a UE and its computing resources in its Avatars is reduced as compared to that in the traditional big data network. However, in order to maintain the low E2E delay when UEs roam away, it is necessary to hand off Avatars accordingly-it is not practical to hand off the Avatars' virtual disks during roaming as this will incur unbearable migration time and network congestion. We propose the LatEncy Aware Replica placemeNt (LEARN) algorithm to place a number of replicas of each Avatar's virtual disk into suitable cloudlets. Thus, the Avatar can be handed off among its cloudlets (which contain one of its replicas) without migrating its virtual disk. Simulations demonstrate that LEARN reduces the average E2E delay. Meanwhile, by considering the capacity limitation of each cloudlet, we propose the LatEncy aware Avatar hanDoff (LEAD) algorithm to place UEs' Avatars among the cloudlets such that the average E2E delay is minimized. Simulations demonstrate that LEAD maintains the low average E2E delay.

Applying Drone-mounted Base Station (DBS) to assist Macro Base Station (MBS) can potentially increase the throughput of the mobile access network. In this paper, we first derive the spectral efficiency of delivering traffic from the MBS to a Ground User (GU) via the DBS, which is operated in the half-duplex in-band mode, upon which we formulate the problem by jointly optimizing the DBS placement (i.e., the altitude of the DBS) and user association in order to maximize the spectral efficiency of the hotspot area. We design the Spectral efficienT Aware DBS pLacement and usEr association (STABLE) algorithm to solve the proposed problem and demonstrate the performance of STABLE via simulations.

We propose to apply constrained application protocol publish/subscribe to cache popular Internet of Things (IoT) resources in a broker to reduce the energy consumption of servers (which host these popular resources) and the average delay for delivering the IoT resources' contents to the clients. We provide the smart parking application in smart cities as an example to demonstrate the benefit for conducting popular IoT resource caching. However, caching popular IoT resources in the broker may not always be the optimal choice, i.e., the broker may be congested by caching too many IoT resources, and so the average delay for enabling the broker to deliver contents of the IoT resources may be unbearable. Thus, we propose a novel energy aware and latency guaranteed dynamic resource caching (EASE) strategy to enable the broker to cache suitable popular resources such that the energy savings from the servers are maximized, while the average delay for publishing the contents of the resources to the corresponding clients is minimized. We demonstrate the performances of EASE via simulations as compared to other two baseline IoT resource caching strategies.

At the Internet of Things (IoT) application layer, a physical phenomenon, which is sensed by a server (i.e., an IoT device), is defined as an IoT resource. In this paper, we propose to cache popular IoT resources in brokers, which are considered as the application layer middleware nodes. Caching popular resources in the brokers is to move the traffic loads (for delivering the up-to-date contents of the resources) from the servers (which host these popular resources) to the brokers, thus reducing the energy consumption of the servers. However, many brokers may be geographically distributed in the network and caching popular resources in nearby brokers may result in unbalanced traffic loads among the brokers, and may thus dramatically increase the average delay of the brokers in delivering the contents of their cached popular resources to clients. To reduce the average delay among the brokers, we propose to re-cache/re-allocate the popular resources from heavily loaded brokers into lightly loaded brokers in order to balance the traffic loads among brokers. We formulate the popular resource re-caching problem as an optimization problem, which is proven to be NP-hard. We design the latency aware popular resource re-caching (LEARN) algorithm to efficiently solve the problem, and demonstrate the performance of LEARN via simulations.

Different from traditional static small cells, Drone Base Stations (DBSs) exhibit their own advantages, i.e., faster and cheaper to deploy, more flexibly reconfigured, and likely to have better communications channels owing to the presence of short-range line-of-sight links. Thus, applying DBSs into the cellular network has great potential to increase the throughput of the network and improve Quality of Service (QoS) of Mobile Users (MUs). In this paper, we focus on how to place the DBS (i.e., jointly determining the location and the association coverage of a DBS) in order to improve the QoS in terms of minimizing the total average latency ratio of MUs by considering the energy capacity limitation of the DBS. We formulate the DBS placement problem as an optimization problem and design a Latency aware dronE bAse station Placement (LEAP) algorithm to solve it efficiently. The performance of LEAP is demonstrated via simulations as compared to other two baseline methods.

We propose a cloudlet network architecture to bring the computing resources from the centralized cloud to the edge. Thus, each User Equipment (UE) can communicate with its Avatar, a software clone located in a cloudlet, and can thus lower the end-to-end (E2E) delay. However, UEs are moving over time, and so the low E2E delay may not be maintained if UEs' Avatars stay in their original cloudlets. Thus, live Avatar migration (i.e., migrating a UE's Avatar to a suitable cloudlet based on the UE's location) is enabled to maintain the low E2E delay between each UE and its Avatar. On the other hand, the migration itself incurs extra overheads in terms of resources of the Avatar, which compromise the performance of applications running in the Avatar. By considering the gain (i.e., the E2E delay reduction) and the cost (i.e., the migration overheads) of the live Avatar migration, we propose a PRofIt Maximization Avatar pLacement (PRIMAL) strategy for the cloudlet network in order to optimize the tradeoff between the migration gain and the migration cost by selectively migrating the Avatars to their optimal locations. Simulation results demonstrate that as compared to the other two strategies (i.e., Follow Me Avatar and Static), PRIMAL maximizes the profit in terms of maintaining the low average E2E delay between UEs and their Avatars and minimizing the migration cost simultaneously.

We propose a Green Cloudlet Network (GCN) architecture to provide seamless Mobile Cloud Computing (MCC) services to User Equipments (UEs) with low latency in which each cloudlet is powered by both green and brown energy. Fully utilizing green energy can significantly reduce the operational cost of cloudlet providers. However, owing to the spatial dynamics of energy demand and green energy generation, the energy gap among different cloudlets in the network is unbalanced, i.e., some cloudlets' energy demands can be fully provided by green energy but others need to utilize on-grid energy (i.e., brown energy) to satisfy their energy demands. We propose a Green-energy awarE Avatar migRation (GEAR) strategy to minimize the on-grid energy consumption in GCN by redistributing the energy demands via Avatar migration among cloudlets according to cloudlets' green energy generation. Furthermore, GEAR ensures the Service Level Agreement (SLA) in terms of the maximum Avatar propagation delay by avoiding Avatars hosted in the remote cloudlets. We formulate the GEAR strategy as a mixed integer linear programming problem, which is NP-hard, and thus apply the Branch and Bound search to find its sub-optimal solution. Simulation results demonstrate that GEAR can save on-grid energy consumption significantly as compared to the Follow me AvataR (FAR) migration strategy, which aims to minimize the propagation delay between an UE and its Avatar.

This paper presents a mobile cloud computing application model and addresses how to minimize the energy consumption for uploading L size of data load within the T delay constraint. We propose a bandwidth allocation strategy for a LTE network with homogeneous sub-channel condition. Our objective is to allocate more bandwidth to each UE when its relative channel condition becomes better. We formulate the UE's objective function as the sum of two penalty functions: channel condition penalty function which incentivizes base stations to minimize the energy consumption for every UE and Service Level Agreement (SLA) demand penalty function which guarantees L size of data load that can be uploaded in time. In the network scenario, we formulate the EnerGy Optimized (EGO) bandwidth allocation strategy as a linear programming model and solve it by the Simplex Method. Simulation results show that EGO can save energy of up to 60% for each UE and decrease the SLA violation rate in the network of up to 30% in comparison with the existing bandwidth allocation strategy in the uplink of the LTE network.

Bandwidth is a key resource in cloud networks. Every tenant wants to be assigned the bandwidth which is proportional to the price they have paid. At the cloud vender side, the link bandwidth utilization could be enhanced to support more clients. In this paper, we show that the traditional PS-N (Proportional Sharing at Network level) bandwidth allocation algorithm cannot achieve the network proportionality fairness when the network is over-subscribed. PPSN (Persistence Proportional Sharing at Network level) is proposed to solve the unfairness issue. However, the bandwidth utilization of both algorithms is not good enough to meet venders' demands. BEPPS-N (Bandwidth Efficiency Persistence Proportional Sharing at Network level) is thus proposed to enhance the bandwidth utilization by assigning more bandwidth to the communication pairs that are not passing through bottleneck links, and at the same time, keep proportionality fairness per different tenant. Finally, simulations and performance analysis have been conducted to substantiate the viability of our proposed approach.