Article
Computer Science, Hardware & Architecture
Preeti Rani, Rohit Sharma
Summary: This paper proposes an intelligent transport system for IOVs-based vehicular network traffic in a smart city scenario, using tree-based Decision Tree, Random Forest, Extra Tree, and XGBoost machine learning models. Simulation results show that the proposed system achieves high detection accuracy and low computational costs through ensemble learning and feature selection. The tree-based ML techniques with feature selection outperform those without feature selection for IOV-based vehicular network traffic.
COMPUTERS & ELECTRICAL ENGINEERING
(2023)
Article
Engineering, Civil
Zhigao Zheng, Ali Kashif Bashir
Summary: This paper discusses the challenges of data processing in intelligent vehicular networks (IVN) and the motivation to address these challenges using graph processing technologies. The characteristics of widely used graph algorithms and graph processing frameworks on GPU are explored, and several graph-based optimization technologies for IVN data processing are proposed. Experimental results demonstrate that graph processing technologies on GPU can achieve excellent performance on IVN data.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Review
Computer Science, Information Systems
Qimei Cui, Xingxing Hu, Wei Ni, Xiaofeng Tao, Ping Zhang, Tao Chen, Kwang-Cheng Chen, Martin Haenggi
Summary: This study surveys the current state and recent advances in vehicular mobility models, categorizes them into vehicular distribution, vehicular traffic, and driving behavior models, and analyzes their application in various scenarios. The study also highlights research opportunities in deep learning platform design, evaluating the representativeness and completeness of models, and developing hybrid models.
SCIENCE CHINA-INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Salvador Balkus, Honggang Wang, Brian D. Cornet, Chinmay Mahabal, Hieu Ngo, Hua Fang
Summary: This paper provides a comprehensive literature survey on the intersection between machine learning and vehicular communications in autonomous driving. It explains how shared data can improve the performance of autonomous vehicles and addresses five major questions related to the use of vehicle-to-vehicle and vehicle-to-everything communications.
IEEE COMMUNICATIONS SURVEYS AND TUTORIALS
(2022)
Article
Computer Science, Information Systems
Muhammad Haris, Munam Ali Shah, Carsten Maple
Summary: This paper proposes an Intelligent Volunteer Computing-based VANETs architecture to fulfill the computational requirements of vehicles applications intelligently. It uses machine learning approaches to predict the capability of vehicles for task execution and compares different regression techniques to reduce errors.
Article
Engineering, Civil
Mahdi Abbasi, Mina Yaghoobikia, Milad Rafiee, Mohammad R. Khosravi, Varun G. Menon
Summary: With the rapid growth of Internet of Vehicles (IoV) and the increased data volume and computational loads, traditional cloud computing solutions pose substantial delays in handling workloads. By utilizing Genetic Algorithm to optimize power consumption at edge systems, a more even workload distribution can be achieved to reduce processing delays significantly.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2021)
Article
Engineering, Civil
Mohsin Kamal, Muhammad Tariq, Gautam Srivastava, Lukas Malina
Summary: With the growth of IoV in IATS, the volume of data exchanged between vehicles has increased, leading to an increased need for secure data transfer. This paper proposes optimized security algorithms using symmetric encryption for secure multimedia data transfer between vehicles and validates their performance through experimental analysis.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Engineering, Civil
Wei Duan, Xiaohui Gu, Miaowen Wen, Yancheng Ji, Jianhua Ge, Guoan Zhang
Summary: A hierarchical model with QoS-aware and power-aware resource management is proposed for the intelligent vehicular network, optimizing system latency and energy efficiency. The Minimum Latency with Migration Loads scheme is developed for workload balance among multiple MECSs, while workload redistribution and dynamic VM reconfiguration optimize energy efficiency.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Chemistry, Analytical
Nazik Alturki, Turki Aljrees, Muhammad Umer, Abid Ishaq, Shtwai Alsubai, Oumaima Saidani, Sirojiddin Djuraev, Imran Ashraf
Summary: This research paper investigates the latest trends in safety, security, and privacy related to drones and highlights the importance of secure drone networks. The proposed framework incorporates intelligent machine learning models into the design and structure of IoT-aided drones, rendering adaptable and secure technology to mitigate cyber-security threats.
Article
Engineering, Electrical & Electronic
Jinkai Zheng, Tom H. Luan, Yilong Hui, Zhisheng Yin, Nan Cheng, Longxiang Gao, Lin X. Cai
Summary: This paper investigates the joint network selection and power level allocation problem in data synchronization between vehicles and digital twins in heterogeneous access networks. A learning-based heterogeneous network selection scheme is proposed, and transfer learning is applied to improve learning efficiency and decrease data synchronization latency. Extensive experiments show that the proposed scheme achieves optimal network selection and better performance compared to existing approaches.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2022)
Article
Computer Science, Theory & Methods
Clayson Celes, Azzedine Boukerche, Antonio A. F. Loureiro
Summary: This article provides an overview of publicly available vehicular mobility traces, discusses preprocessing issues, introduces methods for characterizing and modeling mobility data, and reviews existing proposals for applying hidden knowledge extracted from mobility traces in vehicular networks. It also discusses open research problems and provides directions for further work.
ACM COMPUTING SURVEYS
(2022)
Article
Automation & Control Systems
Uche Onyekpe, Vasile Palade, Anuradha Herath, Stratis Kanarachos, Michael E. Fitzpatrick
Summary: This paper proposes a deep learning approach to accurately position wheeled vehicles in GNSS deprived environments using wheel speed information. Experimental results show that the method effectively reduces positioning errors in various driving scenarios and long-term GNSS outage situations.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2021)
Article
Telecommunications
Shiyi Wang, Yong Liao
Summary: The paper proposes a CSI intelligent feedback network model named residual mix-net (RM-Net) for V2I scenarios, which shows fast training speed, fewer training samples required, and superior performance in learning channel characteristics in high-speed V2I scenarios.
CHINA COMMUNICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Muhammad Saleem, Sagheer Abbas, Taher M. Ghazal, Muhammad Adnan Khan, Nizar Sahawneh, Munir Ahmad
Summary: Smart cities have developed rapidly, with traffic congestion being a major concern. Communication delays, traffic flow, and road safety are challenges for Intelligent Transportation Systems. Vehicular Networks propose novel approaches, including vehicle communication, navigation, and traffic control. This research presents a fusion-based intelligent traffic congestion control system for VNs using machine learning, achieving improved traffic flow and accuracy in alleviating congestion.
EGYPTIAN INFORMATICS JOURNAL
(2022)
Article
Automation & Control Systems
M. Ramya Devi, I. Jasmine Selvakumari Jeya
Summary: Internet of Vehicles (IoV) is an intelligent vehicular technology that enables vehicles to communicate with each other via internet. This study proposes a hybrid Cooperative, Vehicular Communication Management Framework called CAMINO (CA) to improve communication delay and throughput, and enhance security using Common Vulnerability Scoring System (CVSS) methodology.
INTELLIGENT AUTOMATION AND SOFT COMPUTING
(2023)
Article
Telecommunications
Farhad Mirkarimi, Chintha Tellambura, Geoffrey Ye Li
Summary: This letter proposes the development of minimum mean-squared error (MMSE) estimators based on deep neural networks for data detection. To overcome the performance degradation caused by linear MMSE approximations, a near-optimal estimator is developed using the Donsker-Varadhan representation of mutual information (MI) and the derivative relationship between MI and MMSE. This near-optimal MMSE estimator can be computed using a deep neural network, which is trained using mini-batches of input and output samples. Several examples are provided to demonstrate the effectiveness of the proposed estimator, and its application in an end-to-end communication system shows promising performance compared to conventional techniques.
IEEE COMMUNICATIONS LETTERS
(2023)
Article
Engineering, Electrical & Electronic
Peiwen Jiang, Chao-Kai Wen, Shi Jin, Geoffrey Ye Li
Summary: Video conferencing is a popular mode of meeting, but it consumes considerable communication resources. Conventional video compression reduces resolution under limited bandwidth. Semantic video conferencing (SVC) maintains high resolution by transmitting keypoints to represent motions. However, the study on transmission errors' influence on keypoints is limited. In this paper, an SVC network based on keypoint transmission is established to reduce transmission resources and prevent detailed expression loss.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Bowen Zhang, Houssem Sifaou, Geoffrey Ye Li
Summary: This paper presents a new localization system that utilizes a novel attention-augmented residual convolutional neural network for indoor positioning and a denoising task for tracking. The proposed methods outperform existing approaches in performance and generality improvement.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Jiabao Gao, Caijun Zhong, Geoffrey Ye Li, Joseph B. Soriaga, Arash Behboodi
Summary: Hybrid analog-digital (HAD) architecture is widely used in millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) systems to reduce hardware cost and energy consumption. However, channel estimation in HAD systems is challenging due to limited RF chains. Existing compressive sensing (CS) algorithms may suffer from practical effects and high complexity. To address these issues, a deep learning-based channel estimation approach is proposed, where the sparse Bayesian learning (SBL) algorithm is unfolded into a deep neural network (DNN). Simulation results show that the proposed approach outperforms existing approaches in terms of performance and complexity.
IEEE TRANSACTIONS ON COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Linghui Ge, Hua Zhang, Jun-Bo Wang, Geoffrey Ye Li
Summary: In this paper, a reconfigurable wireless relaying system aided by multiple unmanned aerial vehicle (UAV)-carried intelligent reflecting surfaces (IRSs) is proposed to enhance multiuser downlink transmission. The joint optimization of UAVs' placement, active beamforming and power allocation at the BS, and passive beamforming at the UAV-IRSs is conducted to maximize the weighted sum rate. Simulation results show significant performance gains compared to benchmark schemes.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Hengtao He, Rui Wang, Weijie Jin, Shi Jin, Chao-Kai Wen, Geoffrey Ye Li
Summary: Millimeter-wave (mmWave) communications are promising for future wireless networks, but channel estimation is challenging in wideband mmWave systems. To address this, a lens-based beamspace massive MIMO system is considered, and a model-driven unsupervised learning network, LDGEC, is proposed for channel estimation. LDGEC can be trained with limited measurements and real channel data using a denoiser-by-denoiser approach, and outperforms state-of-the-art algorithms when the receiver has limited RF chains and low-resolution ADCs.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Ding Shi, Linfeng Song, Wenqi Zhou, Xiqi Gao, Cheng-Xiang Wang, Geoffrey Ye Li
Summary: In this paper, the authors investigate channel acquisition for high frequency skywave MIMO communications with OFDM modulation. They propose the concept of triple beams in the SFT domain and establish a channel model based on sampled triple steering vectors. The authors also study optimal channel estimation and pilot design, and propose a channel prediction method for data segments. Simulation results show the superior performance of the proposed channel acquisition approach.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Telecommunications
Ouya Wang, Jiabao Gao, Geoffrey Ye Li
Summary: In recent years, deep learning has been widely used in communications and achieved remarkable performance improvement. Most existing works rely on data-driven deep learning, which requires a large amount of training data and computing resources. This paper introduces few-shot learning to reduce the training data requirement for new environments by leveraging the learning experience from known environments. By embedding an attention network into the deep learning-based communication model, different environments can be learned together, allowing the communication model to perform well in new environments with only a few pilot blocks. Through an example of deep-learning-based channel estimation, this novel design method achieves better performance than the existing data-driven approach for few-shot learning.
IEEE TRANSACTIONS ON COGNITIVE COMMUNICATIONS AND NETWORKING
(2023)
Article
Engineering, Electrical & Electronic
Huiqiang Xie, Zhijin Qin, Geoffrey Ye Li
Summary: This paper introduces a deep learning-based semantic communication system with memory, called Mem-DeepSC, to mimic human-like communication. By using a universal Transformer-based transceiver to extract semantic information and a memory module to process context information, the reliability and efficiency of communication are improved. Experimental results show that Mem-DeepSC outperforms benchmarks in terms of answer accuracy and transmission efficiency.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Junchao Shi, An-An Lu, Wen Zhong, Xiqi Gao, Geoffrey Ye Li
Summary: In this paper, the authors investigate the downlink robust precoding for massive MIMO communications with imperfect CSI. They propose a robust WMMSE precoder that maximizes the ergodic sum rate. The precoding vectors are characterized by low-dimensional parameters learned from available CSI through a neural network. The deep learning design significantly reduces computational complexity while achieving near-optimal performance.
IEEE TRANSACTIONS ON COMMUNICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Shenglong Zhou, Geoffrey Ye Li
Summary: The researchers have developed an efficient optimization algorithm for federated learning, which is capable of combating the stragglers' effect, is computationally and communication-efficient, and has good convergence under mild conditions. Moreover, it outperforms several state-of-the-art algorithms in terms of numerical performance.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Engineering, Electrical & Electronic
Xianglong Yu, Xiqi Gao, An-An Lu, Jinlin Zhang, Hebing Wu, Geoffrey Ye Li
Summary: This paper investigates the robust precoding for high frequency skywave massive MIMO communications with imperfect channel state information (CSI). It is proven that the robust precoder for ergodic sum-rate maximization can be designed by optimizing the beam domain robust precoder (BDRP) without any loss of optimality. Furthermore, a low-complexity BDRP design with an ergodic sum-rate upper bound is developed, simplifying the iterative algorithm.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Computer Science, Hardware & Architecture
Peiwen Jiang, Chao-Kai Wen, Shi Jin, Geoffrey Ye Li
Summary: This study discusses recent advancements in semantic communication, utilizing conventional modules in wireless systems to improve performance. It also demonstrates the importance of utilizing traditional hybrid automatic repeat request and modulation methods for novel semantic coding and metrics to enhance wireless semantic communication.
IEEE WIRELESS COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Shenglong Zhou, Geoffrey Ye Li
Summary: In this paper, a hybrid federated learning algorithm (FedGiA) that combines gradient descent and the inexact alternating direction method of multipliers was proposed to address the challenges of saving communication resources, reducing computational costs, and achieving convergence. The proposed algorithm was proven to be theoretically and numerically more efficient in terms of communication and computation than other state-of-the-art algorithms, and it also achieves global convergence under mild conditions.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2023)
Proceedings Paper
Computer Science, Information Systems
Lei Yan, Zhijin Qin, Rui Zhang, Yongzhao Li, Geoffrey Ye Li
Summary: This paper investigates the problem of semantic-aware resource allocation in a multi-cell multi-task network. A novel measure of semantic entropy is proposed to quantify the semantic information for different tasks. A quality-of-experience (QoE) model is also proposed, and the resource allocation problem is formulated in terms of the number of transmitted semantic symbols, channel assignment, and power allocation. Simulation results demonstrate the effectiveness and superiority of the proposed method on the overall QoE.
2022 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM 2022)
(2022)
