Article
Computer Science, Information Systems
Muhammad Sheraz, Shahryar Shafique, Sohail Imran, Muhammad Asif, Rizwan Ullah, Muhammad Ibrar, Andrzej Bartoszewicz, Saleh Mobayen
Summary: In this study, we propose a mobility-aware data caching approach for offloading data through device-to-device (D2D) communication. By observing users' connectivity patterns and considering users' data preferences, our proposed mechanism achieves better performance in terms of data offloading gain and cache hit rate compared to other commonly used caching schemes.
Article
Computer Science, Information Systems
Minjoong Rim, Chung G. Kang
Summary: The increasing amount of traffic in wireless networks can be mitigated using device-to-device caching technology, but in the early stages, limited number of devices and small storage might pose a challenge. Common content popularity makes it difficult to achieve satisfactory performance with small caches, while individual users may have diverse content preferences.
Article
Engineering, Electrical & Electronic
Su Yao, Mu Wang, Qiang Qu, Ziyi Zhang, Yi-Feng Zhang, Ke Xu, Mingwei Xu
Summary: This paper proposes a blockchain-empowered collaborative task offloading method for Cloud-Edge-Device (CED) computing. The blockchain plays a central role in task offloading, resource brokerage, and incentives. By modifying the blockchain consensus process, participants can reach an agreement by solving the task offloading problem. A truthful incentive mechanism is also proposed to encourage resource contributions and honesty in the system.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2022)
Article
Computer Science, Information Systems
Steffi Jayakumar, S. Nandakumar
Summary: Device to Device communication (D2D) is a promising technology in future wireless communication systems, and this paper proposes an RL-based resource allocation scheme to optimize system throughput and fairness. The RL method allows the system to learn from trial-and-error without prior knowledge, and expanding the observation space improves the accuracy of the learning algorithm. Simulation results demonstrate improvements in throughput, energy efficiency, spectrum efficiency, and fairness, compared to conventional methods.
Article
Computer Science, Information Systems
Ziyu Peng, Gaocai Wang, Wang Nong, Yu Qiu, Shuqiang Huang
Summary: Multiple-Services Mobile Edge Computing allows dynamic updates of cached services in edge servers, enabling task offloading to improve system performance. However, the dynamic nature of service requirements, computing demands, and data transfer poses a challenge in adapting the subset of service types and making resource allocation decisions. To solve this, a deep reinforcement learning-based algorithm called DSOR is proposed, which converts the problem into a Markov decision process and jointly determines service caching, task offloading, bandwidth allocation, and computing resource allocation.
COMPUTER COMMUNICATIONS
(2023)
Article
Computer Science, Information Systems
Zhu Xiao, Jinmei Shu, Hongbo Jiang, John C. S. Lui, Geyong Min, Jiangchuan Liu, Schahram Dustdar
Summary: This paper investigates the optimization problem of content caching and parallel task offloading in device-to-device aided mobile edge computing networks. An enhanced binary particle swarm optimization algorithm and an improved multi-objective bat algorithm are proposed to solve the problem. Experimental results show that the algorithms can significantly reduce delay and energy consumption, and maintain a high parallel task offloading ratio even with a large number of mobile devices.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2023)
Article
Computer Science, Information Systems
Razie Roostaei, Marzieh Sheikhi, Zeinab Movahedi
Summary: This paper proposes a Device-to-Device enabled mobile cloud computing framework, with a credit and reputation-based incentive mechanism, using second-price reverse auction to price resources, monitoring user behavior through activity logs for reputation value, reducing NP-hard decision algorithm complexity, with polynomial time complexity, and satisfying individual rationality, computational efficiency, truthfulness, and Pareto optimality. Extensive experiments demonstrate the effectiveness of the proposed scheme.
Article
Computer Science, Information Systems
Fariba Majidi, Mohammad Reza Khayyambashi, Behrang Barekatain
Summary: The rapid development of the Internet of Things has led to a significant increase in the number of terminals and network traffic. This article proposes a hierarchical federated deep reinforcement learning (HFDRL) method to predict user's future requests and optimize content transfer while avoiding network overload and learning difficulties. By categorizing edge devices hierarchically, the method improves the performance of both local base station networks and the global network by reducing content redundancy and latency.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Hardware & Architecture
Guanhua Qiao, Supeng Leng, Yan Zhang
Summary: This paper introduces a framework of device-to-device edge computing and networks (D2D-ECN), which enables collaborative optimization between communication and computation by utilizing resource-rich devices. To address the issue of computation interruption in task intensive applications, the D2D-ECN is equipped with energy harvesting technology to provide a green computation network and ensure service continuity. Reinforcement learning and Lyapunov optimization techniques are employed to overcome the challenges posed by renewable energy, channel state, and task generation rates.
MOBILE NETWORKS & APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Sepideh Malektaji, Amin Ebrahimzadeh, Halima Elbiaze, Roch H. Glitho, Somayeh Kianpisheh
Summary: With the increasing demands for data, content delivery networks are facing challenges in meeting end-users' quality-of-experience requirements, especially in terms of delay. This paper proposes a content migration strategy using deep reinforcement learning to minimize costs and reduce content access delay, achieving up to a 70% reduction compared to conventional strategies. However, selecting which contents to migrate and to which neighboring cache to migrate remains a complex problem.
IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT
(2021)
Article
Telecommunications
Yingchun Wang, Jingyi Wang, Weizhan Zhang, Yufeng Zhan, Song Guo, Qinghua Zheng, Xuanyu Wang
Summary: With the rapid development of mobile devices and deep learning, mobile smart applications using deep learning technology have emerged as a main research focus. Although deep learning has achieved tremendous success in various research fields, deploying such applications on resource-restricted mobile devices remains a challenge.
DIGITAL COMMUNICATIONS AND NETWORKS
(2022)
Article
Computer Science, Hardware & Architecture
Junna Zhang, Jiawei Chen, Xiang Bao, Chunhong Liu, Peiyan Yuan, Xinglin Zhang, Shangguang Wang
Summary: Edge computing provides abundant computing and storage resources to meet the growing requirements of delay-sensitive mobile applications. This paper proposes an efficient dependent task offloading mechanism that can optimize the overall task completion time in scenarios where edge servers have limited service caches and computing power.
JOURNAL OF NETWORK AND COMPUTER APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Dongjin Yu, Tong Wu, Chengfei Liu, Dongjing Wang
Summary: Edge caching has become a hot research topic in Mobile Edge Computing (MEC) as an effective way to reduce traffic burden in cellular networks. By placing contents at the edge and utilizing Device-to-Device (D2D) links, caching and recommendation can improve efficiency and reduce transmission cost in edge caching. This article proposes a Joint Content Caching and Recommender System (JCCRS) that uses a Collaborative Filtering algorithm and a Deep Deterministic Policy Gradient (DDPG) based method to solve the content caching and recommendation problem.
IEEE TRANSACTIONS ON SERVICES COMPUTING
(2023)
Article
Engineering, Electrical & Electronic
Jian Xiong, Yuzhe Fang, Peng Cheng, Zhiping Shi, Wei Zhang
Summary: Content caching plays an important role in converged networks but is limited by the storage capacity of terminals. Research shows that caching popular services in router nodes close to the terminators can improve energy efficiency, yet faces challenges with service scheduling.
IEEE TRANSACTIONS ON BROADCASTING
(2021)
Article
Computer Science, Information Systems
Fayez Alqahtani, Mohammed Al-Maitah, Osama Elshakankiry
Summary: The mobile edge computing (MEC) paradigm provides cloud and application services, but heterogeneous services can increase delay, requiring caching and offloading features. A proactive caching technique with offloading (PCTO) can meet the needs of parallel user services, reducing response time through demand-aware offloading. Network-level caching and deep learning are used to streamline failed service distribution intervals and improve performance.
COMPUTER COMMUNICATIONS
(2022)
Article
Computer Science, Theory & Methods
Huan Zhou, Mingze Li, Ning Wang, Geyong Min, Jie Wu
Summary: With the rapid development of IoT and deep learning, there is an urgent need to enable deep learning inference on IoT devices in MEC. To address the computation limitation, computation offloading is proposed as a promising approach. This paper proposes a novel FL-based DNN model parallelism method to accelerate inference by converting a DNN layer into several smaller layers for increased offloading flexibility.
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS
(2023)
Article
Computer Science, Theory & Methods
Xin Li, Junsong Zhou, Xin Wei, Dawei Li, Zhuzhong Qian, Jie Wu, Xiaolin Qin, Sanglu Lu
Summary: Loosely coupled and highly cohesive microservices running in containers have become the new paradigm for application development. Compared to monolithic applications, microservices architecture allows for independent deployment and scaling, promising to simplify software development and operation. However, the increase in microservices scale and east-west network traffic in data centers has made cluster management more complex. This paper proposes a Microservice-Oriented Topology-Aware Scheduling Framework (MOTAS) that optimizes the network overhead of microservice applications by effectively utilizing microservices and cluster topologies through a heuristic graph mapping algorithm. The framework also guarantees cluster resource utilization and incorporates a mechanism for detecting and handling QoS violations in dynamic microservice environments.
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS
(2023)
Article
Engineering, Electrical & Electronic
Wenyu Zhang, Haijun Zhang, Hui Ma, Hua Shao, Ning Wang, Victor C. M. Leung
Summary: This paper proposes a predictive and adaptive deep coding (PADC) framework that achieves flexible code rate optimization with a given target transmission quality requirement. By using a variable code length enabled DeepJSCC model, an Oracle Network model, and a CR optimizer, PADC can minimize bandwidth consumption while guaranteeing the PSNR constraint for each image data in wireless image transmission tasks.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Wenbo Du, Tao Tan, Haijun Zhang, Xianbin Cao, Gang Yan, Osvaldo Simeone
Summary: A set of low-cost sensors is used to infer the network topology of a self-organizing wireless network by extracting timing information from data packets and acknowledgment (ACK) packets. A new EM-based algorithm, called EM-CDA, is introduced to handle the impact of packet losses on causality metrics. Extensive experiments on the NS-3 simulation platform validate the effectiveness of the method.
IEEE TRANSACTIONS ON COMMUNICATIONS
(2023)
Article
Computer Science, Hardware & Architecture
En Wang, Mijia Zhang, Wenbin Liu, Haoyi Xiong, Bo Yang, Yongjian Yang, Jie Wu
Summary: Mobile CrowdSensing (MCS) is a popular data collection paradigm that often faces the issue of sparse sensed data. To address this problem, sparse MCS recruits users to sense important areas and completes the data through data completion. However, in real-world scenarios, there are various types of data that can complement each other, including important outliers. Detecting and recovering these outliers poses challenges due to their infrequency and the complex spatiotemporal relations among the data.
IEEE-ACM TRANSACTIONS ON NETWORKING
(2023)
Article
Computer Science, Information Systems
Sijie Huang, He Huang, Guoju Gao, Yu-E Sun, Yang Du, Jie Wu
Summary: Blockchain is a distributed ledger system used in Bitcoin to protect transaction histories. In the mining process, high computing power is required, making it hard to implement on mobile devices. This article proposes a more realistic scenario where edge/cloud service providers have different propagation delays, and analyzes the pricing and scheduling problem in a three-stage multi-leader multi-follower Stackelberg game to achieve equilibrium. Extensive simulations demonstrate the effectiveness of the proposed solution.
IEEE TRANSACTIONS ON SERVICES COMPUTING
(2023)
Article
Computer Science, Information Systems
Huan Zhou, Zhenning Wang, Geyong Min, Haijun Zhang
Summary: This article investigates a UAV-aided mobile-edge computing network for computation offloading to provide additional computation capability and wide coverage for mobile users. It proposes a game model and a gradient-based algorithm to achieve the maximization of system utility.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Information Systems
Yunhui Qin, Zhongshan Zhang, Huangfu Wei, Haijun Zhang, Keping Long
Summary: This letter investigates the cooperative resource allocation of cellular networks with simultaneous wireless information and power transfer in the time-varying channel environment. The soft actor-critic (SAC) algorithm is exploited to tackle the optimization problem which aims to find a feasible resource allocation policy to maximize the data rate and system fairness while minimizing the channel switching penalty. Considering the costly agent-to-environment interactions and the restricted empirical dataset of the SAC algorithm, this letter explores the permutation equivalence of the optimization objective, and designs two data augmentation schemes for the experience replay buffer of SAC. The cumulative discount reward shows that data augmentation assisted algorithms outperform the baseline in the learning speed. The simulation results referring to the average data rate and system fairness show that the proposed schemes benefit to the training model and effectively improve the performance of algorithms.
IEEE WIRELESS COMMUNICATIONS LETTERS
(2023)
Article
Computer Science, Information Systems
Wenyu Zhang, Sherali Zeadally, Wei Li, Haijun Zhang, Jingyi Hou, Victor C. M. Leung
Summary: The breakthrough of AI techniques has accelerated their applications in various industries, including security protection, transportation, agriculture, and medical care. With the support of edge computing environments, providing AIaaS with latency guarantee can speed up the deployment of data-intensive and computation-intensive AI applications and reduce customers' investment cost. However, existing studies have not addressed the specific deployment architecture, working mechanism design, and performance optimization problems for AIaaS with configurable data quality and model complexity. To tackle this, we propose a configurable model deployment architecture (CMDA) for edge AIaaS and a flexible working mechanism that allows joint configuration of data quality ratios (DQRs) and model complexity ratios (MCRs) for AI tasks.
IEEE TRANSACTIONS ON CLOUD COMPUTING
(2023)
Article
Engineering, Electrical & Electronic
Jie Zheng, Haijun Zhang, Jiawen Kang, Ling Gao, Jie Ren, Dusit Niyato
Summary: In this paper, we propose a method to support covert over-the-air computation (OAC) using intelligent reflecting surfaces (IRS). By optimizing the joint problem and designing a covert difference-of-convex-functions program (CDC), we can select the maximum covert devices participating in model aggregation while satisfying the mean squared error (MSE) requirement. Simulation results demonstrate that using the IRS in covert OAC can achieve significant performance gain compared to baseline algorithms.
IEEE TRANSACTIONS ON COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Haijun Zhang, Hongyu Wang, Yabo Li, Keping Long, Arumugam Nallanathan
Summary: This paper proposes a dynamic resource allocation scheme for task-oriented semantic communication networks based on deep reinforcement learning, which allows data with richer semantic information to preferentially occupy limited communication resources to improve long-term transmission efficiency.
IEEE TRANSACTIONS ON COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Heng Wang, Haijun Zhang, Xiangnan Liu, Keping Long, Arumugam Nallanathan
Summary: Due to limited computation capacity of wireless user devices, multi-access edge computing (MEC) has become an effective way to meet the real-time demands. To increase system capacity, a UAV-assisted computation offloading architecture in the terahertz (THz) band is proposed. Deep reinforcement learning (DRL) based approaches, such as DDQN and DDPG, are used to solve the non-convex optimization problem of minimizing latency.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Xiaonan Li, Haijun Zhang, Huan Zhou, Ning Wang, Keping Long, Saba Al-Rubaye, George K. Karagiannidis
Summary: This paper proposes a framework for resource allocation in the terrestrial-satellite network based on non-orthogonal multiple access (NOMA). A deployment method of local cache pools is also given to achieve lower time delay and maximize energy efficiency. The proposed method, which utilizes multi-agent deep deterministic policy gradient (MADDPG), shows better performance compared to traditional single-agent deep reinforcement learning algorithm in optimizing resource allocation and cache design in the integrated terrestrial-satellite network.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Computer Science, Theory & Methods
Yubin Duan, Jie Wu
Summary: This paper proposes a method to optimize the communication cost of the parameter server framework in distributed training by compressing the model and optimizing data and parameter allocation. Experimental results show that this compression and allocation scheme can efficiently reduce the communication overhead for both linear and deep neural network models.
JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING
(2023)
Article
Computer Science, Hardware & Architecture
Ning Chen, Sheng Zhang, Zhi Ma, Yu Chen, Yibo Jin, Jie Wu, Zhuzhong Qian, Yu Liang, Sanglu Lu
Summary: The usage of live streaming services has led to a substantial increase in live video traffic. However, the perceived quality of experience of users is frequently limited by variations in the upstream bandwidth of streamers. To address this issue, we propose ViChaser, a block-oriented super-resolution approach that performs neural super-resolution on potential blocks of interest in the media server and uses online learning to adapt to the dynamic content of the video. ViChaser achieves higher video quality and faster processing speed compared to existing methods.
IEEE-ACM TRANSACTIONS ON NETWORKING
(2023)