Pub Date : 2025-02-01DOI: 10.1016/j.vehcom.2024.100862
Widhi Yahya , Ying-Dar Lin , Faysal Marzuk , Piotr Chołda , Yuan-Cheng Lai
Traffic offloading is crucial for reducing computing latency in distributed edge systems such as vehicle-to-everything (V2X) networks, which use roadside units (RSUs) and access network mobile edge computing (AN-MEC) with ML agents. Traffic offloading is part of the control plane problem, which requires fast decision-making in complex V2X systems. This study presents a novel ratio-based offloading strategy using the twin delayed deep deterministic policy gradient (TD3) algorithm to optimize offloading ratios in a two-tier V2X system, enabling computation at both RSUs and the edge. The offloading optimization covers both vertical and horizontal offloading, introducing a continuous search space that needs fast decision-making to accommodate fluctuating traffic in complex V2X systems. We developed a V2X environment to evaluate the performance of the offloading agent, incorporating latency models, state and action definitions, and reward structures. A comparative analysis with metaheuristic simulated annealing (SA) is conducted, and the impact of single versus multiple offloading agents with deployment options at a centralized central office (CO) is examined. Evaluation results indicate that TD3's decision time is five orders of magnitude faster than SA. For 10 and 50 sites, SA takes 602 and 20,421 seconds, respectively, while single-agent TD3 requires 4 to 24 milliseconds and multi-agent TD3 takes 1 to 3 milliseconds. The average latency for SA ranges from 0.18 to 0.32 milliseconds, single-agent TD3 from 0.26 to 0.5 milliseconds, and multi-agent TD3 from 0.22 to 0.45 milliseconds, demonstrating that TD3 approximates SA performance with initial training.
{"title":"Offloading in V2X with road side units: Deep reinforcement learning","authors":"Widhi Yahya , Ying-Dar Lin , Faysal Marzuk , Piotr Chołda , Yuan-Cheng Lai","doi":"10.1016/j.vehcom.2024.100862","DOIUrl":"10.1016/j.vehcom.2024.100862","url":null,"abstract":"<div><div>Traffic offloading is crucial for reducing computing latency in distributed edge systems such as vehicle-to-everything (V2X) networks, which use roadside units (RSUs) and access network mobile edge computing (AN-MEC) with ML agents. Traffic offloading is part of the control plane problem, which requires fast decision-making in complex V2X systems. This study presents a novel ratio-based offloading strategy using the twin delayed deep deterministic policy gradient (TD3) algorithm to optimize offloading ratios in a two-tier V2X system, enabling computation at both RSUs and the edge. The offloading optimization covers both vertical and horizontal offloading, introducing a continuous search space that needs fast decision-making to accommodate fluctuating traffic in complex V2X systems. We developed a V2X environment to evaluate the performance of the offloading agent, incorporating latency models, state and action definitions, and reward structures. A comparative analysis with metaheuristic simulated annealing (SA) is conducted, and the impact of single versus multiple offloading agents with deployment options at a centralized central office (CO) is examined. Evaluation results indicate that TD3's decision time is five orders of magnitude faster than SA. For 10 and 50 sites, SA takes 602 and 20,421 seconds, respectively, while single-agent TD3 requires 4 to 24 milliseconds and multi-agent TD3 takes 1 to 3 milliseconds. The average latency for SA ranges from 0.18 to 0.32 milliseconds, single-agent TD3 from 0.26 to 0.5 milliseconds, and multi-agent TD3 from 0.22 to 0.45 milliseconds, demonstrating that TD3 approximates SA performance with initial training.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"51 ","pages":"Article 100862"},"PeriodicalIF":5.8,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142825408","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-01DOI: 10.1016/j.vehcom.2024.100867
Yaoxin Duan , Wendi Nie , Victor C.S. Lee , Kai Liu
With the rise and development of intelligent vehicles, the computation capability of vehicles has increased rapidly and considerably. Vehicle-to-Vehicle (V2V) offloading, in which computation-intensive tasks are offloaded to underutilized vehicles, has been proposed. However, V2V offloading faces the challenges of task transmission reliability and task computation reliability. In V2V offloading, tasks are transmitted via V2V communication, which is volatile and spotty because of rapidly changing network topology and channel conditions between vehicles, resulting in time-varying delays of task transmission and even loss of connectivity. Thus, it is challenging to complete V2V offloading within a given delay constraint. In addition, the realistic diverse vehicular environment always comes with malicious vehicles, which can cause irreparable harm to V2V offloading. Therefore, in this paper, we propose a V2V task offloading scheme called Redundant Task Offloading with Dual-Reliability (RTODR), aiming to minimize task offloading costs while ensuring both task transmission reliability and task computation reliability in a Mobile Edge Computing (MEC)-assisted vehicular network. Specifically, for a computation task, a V2V connection is considered reliable only if the task can be successfully transmitted via the V2V connection within the deadline of the task. To ensure task computation reliability, task computation results from a trusty service vehicle are considered to be reliable. Then we formally model a Minimizing Task Offloading Cost with Dual-reliability (MTOCD) problem, which is mathematically formulated as a multi-objective optimization problem. Afterward, we propose a heuristic redundant task offloading algorithm, named Dual-Reliability Offloading (DRO), to solve the problem. Finally, comprehensive experiments have been conducted to demonstrate that RTODR achieves lower costs compared with other approaches.
{"title":"Redundant task offloading with dual-reliability in MEC-assisted vehicular networks","authors":"Yaoxin Duan , Wendi Nie , Victor C.S. Lee , Kai Liu","doi":"10.1016/j.vehcom.2024.100867","DOIUrl":"10.1016/j.vehcom.2024.100867","url":null,"abstract":"<div><div>With the rise and development of intelligent vehicles, the computation capability of vehicles has increased rapidly and considerably. Vehicle-to-Vehicle (V2V) offloading, in which computation-intensive tasks are offloaded to underutilized vehicles, has been proposed. However, V2V offloading faces the challenges of task transmission reliability and task computation reliability. In V2V offloading, tasks are transmitted via V2V communication, which is volatile and spotty because of rapidly changing network topology and channel conditions between vehicles, resulting in time-varying delays of task transmission and even loss of connectivity. Thus, it is challenging to complete V2V offloading within a given delay constraint. In addition, the realistic diverse vehicular environment always comes with malicious vehicles, which can cause irreparable harm to V2V offloading. Therefore, in this paper, we propose a V2V task offloading scheme called Redundant Task Offloading with Dual-Reliability (RTODR), aiming to minimize task offloading costs while ensuring both task transmission reliability and task computation reliability in a Mobile Edge Computing (MEC)-assisted vehicular network. Specifically, for a computation task, a V2V connection is considered reliable only if the task can be successfully transmitted via the V2V connection within the deadline of the task. To ensure task computation reliability, task computation results from a trusty service vehicle are considered to be reliable. Then we formally model a Minimizing Task Offloading Cost with Dual-reliability (MTOCD) problem, which is mathematically formulated as a multi-objective optimization problem. Afterward, we propose a heuristic redundant task offloading algorithm, named Dual-Reliability Offloading (DRO), to solve the problem. Finally, comprehensive experiments have been conducted to demonstrate that RTODR achieves lower costs compared with other approaches.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"51 ","pages":"Article 100867"},"PeriodicalIF":5.8,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142825409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-01DOI: 10.1016/j.vehcom.2024.100871
Brooke Kidmose
The humble, mechanical automobile has gradually evolved into our modern connected and autonomous vehicles (CAVs)—also known as “smart vehicles.” Similarly, our cities are gradually developing into “smart cities,” where municipal services from transportation networks to utilities to recycling to law enforcement are integrated. The idea, with both smart vehicles and smart cities, is that more data leads to better, more informed decisions. Smart vehicles and smart cities would acquire data from their own equipment (e.g., cameras, sensors) and from their connections—e.g., connections to fellow smart vehicles, to road-side infrastructure, to smart transportation systems (STSs), etc.
Unfortunately, the paradigm of smart vehicles in smart cities is rife with danger and ripe for misuse. One vulnerable system or service could become an attacker's entry point, facilitating access to every connected vehicle, device, etc. Worse, smart vehicles and smart cities are inherently cyber-physical; a cyberattack can have physical consequences, including destruction of infrastructure and loss of life. Lastly, to leverage all the benefits of smart vehicles in smart cities, we would need to accept exorbitant levels of data collection and surveillance, which, in the absence of ironclad privacy protections, could lead to total lack of privacy.
In this work, we define the automotive context—i.e., smart vehicles—within the larger context of smart cities as our threat landscape. Then, we enumerate and describe all of the (1) threats, (2) attack surfaces & targets, (3) areas of concern (indirect vulnerabilities & threats), and (4) impacts of smart vehicles in smart cities. Our objective is to demonstrate that the dangers are real and imminent—in the hope that they will be addressed before an attack on the “smart vehicles in smart cities” paradigm results in loss of life.
{"title":"A review of smart vehicles in smart cities: Dangers, impacts, and the threat landscape","authors":"Brooke Kidmose","doi":"10.1016/j.vehcom.2024.100871","DOIUrl":"10.1016/j.vehcom.2024.100871","url":null,"abstract":"<div><div>The humble, mechanical automobile has gradually evolved into our modern connected and autonomous vehicles (CAVs)—also known as “smart vehicles.” Similarly, our cities are gradually developing into “smart cities,” where municipal services from transportation networks to utilities to recycling to law enforcement are integrated. The idea, with both smart vehicles and smart cities, is that more data leads to better, more informed decisions. Smart vehicles and smart cities would acquire data from their own equipment (e.g., cameras, sensors) and from their connections—e.g., connections to fellow smart vehicles, to road-side infrastructure, to smart transportation systems (STSs), etc.</div><div>Unfortunately, the paradigm of smart vehicles in smart cities is rife with danger and ripe for misuse. One vulnerable system or service could become an attacker's entry point, facilitating access to every connected vehicle, device, etc. Worse, smart vehicles and smart cities are inherently cyber-physical; a cyberattack can have physical consequences, including destruction of infrastructure and loss of life. Lastly, to leverage all the benefits of smart vehicles in smart cities, we would need to accept exorbitant levels of data collection and surveillance, which, in the absence of ironclad privacy protections, could lead to total lack of privacy.</div><div>In this work, we define the automotive context—i.e., smart vehicles—within the larger context of smart cities as our threat landscape. Then, we enumerate and describe all of the (1) threats, (2) attack surfaces & targets, (3) areas of concern (indirect vulnerabilities & threats), and (4) impacts of smart vehicles in smart cities. Our objective is to demonstrate that the dangers are real and imminent—in the hope that they will be addressed before an attack on the “smart vehicles in smart cities” paradigm results in loss of life.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"51 ","pages":"Article 100871"},"PeriodicalIF":5.8,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142825355","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-31DOI: 10.1016/j.vehcom.2025.100891
Parmila Devi, Manoranjan Rai Bharti, Dikshant Gautam
The surge in wireless network attacks has intensified the focus on physical layer security (PLS) within academia and industry. As PLS provides security solutions by leveraging the randomness of wireless channels without the need for encryption/decryption keys, fading channels play a major role in PLS solutions. This survey aims to understand the effect of fading on PLS for 5G/6G communications by utilizing various PLS techniques such as beamforming, artificial noise injection, cooperative and opportunistic relaying, physical authentication, and intelligent reflective surface-based PLS over various fading channels. Initially, the role of PLS in 5G/6G communications, its fundamentals, and various techniques available for 5G/6G communications are examined. Since PLS for 5G communications has been extensively studied in the literature, we categorize it into two cases, direct and indirect communications, and provide a comprehensive survey on PLS for 5G communications over various fading channels. Thereafter, we survey the PLS for 6G communications over various fading channels, noting that the work available for PLS in 6G communications is limited and in its early stages. Given the increasing attention on artificial intelligence and machine learning (AI/ML) for wireless communications, this survey also explores PLS based on AI/ML techniques over various fading channels. Finally, the survey concludes with observations on challenges and future directions.
{"title":"A survey on physical layer security for 5G/6G communications over different fading channels: Approaches, challenges, and future directions","authors":"Parmila Devi, Manoranjan Rai Bharti, Dikshant Gautam","doi":"10.1016/j.vehcom.2025.100891","DOIUrl":"10.1016/j.vehcom.2025.100891","url":null,"abstract":"<div><div>The surge in wireless network attacks has intensified the focus on physical layer security (PLS) within academia and industry. As PLS provides security solutions by leveraging the randomness of wireless channels without the need for encryption/decryption keys, fading channels play a major role in PLS solutions. This survey aims to understand the effect of fading on PLS for 5G/6G communications by utilizing various PLS techniques such as beamforming, artificial noise injection, cooperative and opportunistic relaying, physical authentication, and intelligent reflective surface-based PLS over various fading channels. Initially, the role of PLS in 5G/6G communications, its fundamentals, and various techniques available for 5G/6G communications are examined. Since PLS for 5G communications has been extensively studied in the literature, we categorize it into two cases, direct and indirect communications, and provide a comprehensive survey on PLS for 5G communications over various fading channels. Thereafter, we survey the PLS for 6G communications over various fading channels, noting that the work available for PLS in 6G communications is limited and in its early stages. Given the increasing attention on artificial intelligence and machine learning (AI/ML) for wireless communications, this survey also explores PLS based on AI/ML techniques over various fading channels. Finally, the survey concludes with observations on challenges and future directions.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"53 ","pages":"Article 100891"},"PeriodicalIF":5.8,"publicationDate":"2025-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Vehicular digital twin network is partitioned into multiple networks either due to the geographical differences or their accelerating expansion, which necessitates a secure and incessant transition of cross-regional vehicles. Therefore, in this dynamic topology, the handover process for cross-regional vehicles becomes imperative. The literature encompasses an abundance of blockchain-based handover mechanisms, specifically designed for vehicle and the roadside units. Unfortunately, some of these are not feasible for vehicular digital twin networks due to their high computational overhead and susceptibility to security threats. Therefore, this paper presents a handover authentication protocol for the blockchain-based vehicular digital twin networks, leveraging the smart contract. It entirely depends on digital twin, which reduces the burden of the vehicle and enhances the efficiency and security of the handover process. Security strengths and competency against attacks like sybil and impersonation attacks are investigated through a real-or-random oracle model (ROR) and non-mathematical analysis. The operational analysis evaluates the proposed mechanism with pertinent works based on security functionalities, computation, and communication overhead. Moreover, to illustrate suggested smart contract's viability and the reasonable cost of blockchain consumption, it is implemented via the Ethereum test network. Hence, obtained results indicate the relevancy of the mechanism for vehicular digital twin networks.
{"title":"SC-VDTwinAuth: Smart-contract Assisted Handover Authentication Protocol for Vehicular Digital Twin Network","authors":"Deepika Gautam, Garima Thakur, Sunil Prajapat, Pankaj Kumar","doi":"10.1016/j.vehcom.2025.100890","DOIUrl":"10.1016/j.vehcom.2025.100890","url":null,"abstract":"<div><div>Vehicular digital twin network is partitioned into multiple networks either due to the geographical differences or their accelerating expansion, which necessitates a secure and incessant transition of cross-regional vehicles. Therefore, in this dynamic topology, the handover process for cross-regional vehicles becomes imperative. The literature encompasses an abundance of blockchain-based handover mechanisms, specifically designed for vehicle and the roadside units. Unfortunately, some of these are not feasible for vehicular digital twin networks due to their high computational overhead and susceptibility to security threats. Therefore, this paper presents a handover authentication protocol for the blockchain-based vehicular digital twin networks, leveraging the smart contract. It entirely depends on digital twin, which reduces the burden of the vehicle and enhances the efficiency and security of the handover process. Security strengths and competency against attacks like sybil and impersonation attacks are investigated through a real-or-random oracle model (ROR) and non-mathematical analysis. The operational analysis evaluates the proposed mechanism with pertinent works based on security functionalities, computation, and communication overhead. Moreover, to illustrate suggested smart contract's viability and the reasonable cost of blockchain consumption, it is implemented via the Ethereum test network. Hence, obtained results indicate the relevancy of the mechanism for vehicular digital twin networks.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"53 ","pages":"Article 100890"},"PeriodicalIF":5.8,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143083303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-29DOI: 10.1016/j.vehcom.2025.100889
Siva Sai , Sudhanshu Mishra , Vinay Chamola
Currently, resource allocation in Unmanned Aerial Vehicles (UAVs) is a major topic of discussion among industrialists and researchers. Considering the different emerging applications of UAVs, if the resource allocation problem is not addressed effectively, the upcoming UAV applications will not serve their proposed purpose. Although there are numerous and diverse research works addressing the resource allocation in UAVs, there is an evident lack of a comprehensive survey describing and analyzing the existing methods. Addressing this research gap, we present an extensive review of the resource allocation in UAVs. In this work, we classify the existing research works based on four criteria - optimization goal-based classification, mathematical model-based classification, game theory framework-based classification, and machine learning model-based classification. Our findings revealed that the mathematical models are relatively more explored to solve the resource allocation problem in UAVs. Researchers have explored a variety of game theory techniques, like the Stackelberg model, mean-field game theory, cooperative games, etc., for optimized resource allocation in UAVs. The optimization of energy and throughput factors is more seen in the literature compared to the other optimization goals. We also observed that the reinforcement learning technique is a heavily exploited technique for resource allocation in UAVs compared to all other machine learning-based methods. We have also presented several challenges and future works in the field of resource allocation in UAVs.
{"title":"Resource allocation in unmanned aerial vehicle networks: A review","authors":"Siva Sai , Sudhanshu Mishra , Vinay Chamola","doi":"10.1016/j.vehcom.2025.100889","DOIUrl":"10.1016/j.vehcom.2025.100889","url":null,"abstract":"<div><div>Currently, resource allocation in Unmanned Aerial Vehicles (UAVs) is a major topic of discussion among industrialists and researchers. Considering the different emerging applications of UAVs, if the resource allocation problem is not addressed effectively, the upcoming UAV applications will not serve their proposed purpose. Although there are numerous and diverse research works addressing the resource allocation in UAVs, there is an evident lack of a comprehensive survey describing and analyzing the existing methods. Addressing this research gap, we present an extensive review of the resource allocation in UAVs. In this work, we classify the existing research works based on four criteria - optimization goal-based classification, mathematical model-based classification, game theory framework-based classification, and machine learning model-based classification. Our findings revealed that the mathematical models are relatively more explored to solve the resource allocation problem in UAVs. Researchers have explored a variety of game theory techniques, like the Stackelberg model, mean-field game theory, cooperative games, etc., for optimized resource allocation in UAVs. The optimization of energy and throughput factors is more seen in the literature compared to the other optimization goals. We also observed that the reinforcement learning technique is a heavily exploited technique for resource allocation in UAVs compared to all other machine learning-based methods. We have also presented several challenges and future works in the field of resource allocation in UAVs.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"52 ","pages":"Article 100889"},"PeriodicalIF":5.8,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143083366","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-21DOI: 10.1016/j.vehcom.2025.100888
Murali Krishna Tanati , Manimaran Ponnusamy
The vehicle ad hoc network, or VANET, is a fantastic tool for smart transport since it improves efficiency, management, traffic safety, and comfort. Distributed Denial of Service (DDoS) attacks on VANET infrastructure have the potential to compromise traffic safety by causing collisions and fatalities. Therefore, while integrating VANETs into intelligent transport networks, the pertinent security issues must be addressed. This paper provides an efficient routing optimization as well as a deep learning-based attack detection approach. The input data are first collected from publically accessible datasets. After that, a unique Dense Capsule Stacked Auto Encoder (DCSAE) network is developed to detect the presence of DDoS attacks in the inputs. Here, the detection method is enabled by the hybridization of the Capsule Network with a Stacked Auto Encoder. Moreover, the Improved Fire Hawks Optimization Algorithm (IFHOA) is employed to refine the proposed detection technique. Once assaults have been discovered, bandwidth is allocated using the Hybrid Remora Whale Optimization (HRWO) approach. Finally, an Improved Osprey Optimization (IOO) method is utilized to identify a better routing path by taking into account aspects such as energy usage, delay, and drop. The DDoS SDN dataset is employed to implement the proposed method. In the results section, the suggested technique is compared to existing methods in terms of recall, accuracy, precision, F1 score, Mean Absolute Error (MAE), Packet Delivery Ratio (PDR), Packet Loss Ratio (PLR), and consumption of energy. The proposed model achieved an accuracy of 94.07 % while achieving the precision, recall, and F1-score of 94.2 %, 93.33 %, and 93.88 %, respectively. The model achieved the MAE of 0.132, delay of 4812.976, energy consumption of 40.13 %, PDR of 95.1805, and PLR of 3.6816 %, respectively.
{"title":"Dense capsule stacked auto-encoder model based DDoS attack detection and hybrid optimal bandwidth allocation with routing in VANET environment","authors":"Murali Krishna Tanati , Manimaran Ponnusamy","doi":"10.1016/j.vehcom.2025.100888","DOIUrl":"10.1016/j.vehcom.2025.100888","url":null,"abstract":"<div><div>The vehicle ad hoc network, or VANET, is a fantastic tool for smart transport since it improves efficiency, management, traffic safety, and comfort. Distributed Denial of Service (DDoS) attacks on VANET infrastructure have the potential to compromise traffic safety by causing collisions and fatalities. Therefore, while integrating VANETs into intelligent transport networks, the pertinent security issues must be addressed. This paper provides an efficient routing optimization as well as a deep learning-based attack detection approach. The input data are first collected from publically accessible datasets. After that, a unique Dense Capsule Stacked Auto Encoder (DCSAE) network is developed to detect the presence of DDoS attacks in the inputs. Here, the detection method is enabled by the hybridization of the Capsule Network with a Stacked Auto Encoder. Moreover, the Improved Fire Hawks Optimization Algorithm (IFHOA) is employed to refine the proposed detection technique. Once assaults have been discovered, bandwidth is allocated using the Hybrid Remora Whale Optimization (HRWO) approach. Finally, an Improved Osprey Optimization (IOO) method is utilized to identify a better routing path by taking into account aspects such as energy usage, delay, and drop. The DDoS SDN dataset is employed to implement the proposed method. In the results section, the suggested technique is compared to existing methods in terms of recall, accuracy, precision, F1 score, Mean Absolute Error (MAE), Packet Delivery Ratio (PDR), Packet Loss Ratio (PLR), and consumption of energy. The proposed model achieved an accuracy of 94.07 % while achieving the precision, recall, and F1-score of 94.2 %, 93.33 %, and 93.88 %, respectively. The model achieved the MAE of 0.132, delay of 4812.976, energy consumption of 40.13 %, PDR of 95.1805, and PLR of 3.6816 %, respectively.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"52 ","pages":"Article 100888"},"PeriodicalIF":5.8,"publicationDate":"2025-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143083304","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-21DOI: 10.1016/j.vehcom.2025.100887
Bingfeng Xu , Jincheng Zhao , Bo Wang , Gaofeng He
Detecting zero-day attacks is a critical challenge in the Internet of Vehicles (IoV). Due to the limited availability of labeled attack data, anomaly-based methods are predominantly employed. However, the variability in the driving environment and behavioral patterns of vehicles introduces significant fluctuations in normal behavior, which in turn leads to high false positive rates when using these methods. In this work, we propose a novel detection method for zero-day attacks in IoV through sample augmentation. We first analyze the similarities between known and zero-day attacks in IoV. Based on the analysis, a Few-shot Learning Conditional Generative Adversarial Network (FLCGAN) model with multiple generators and discriminators is developed. Within this framework, an attack sample augmentation algorithm is designed to enhance input data by expanding the known attack dataset, thereby reducing false positives. To address the data imbalance caused by the limited number of input attack samples, an ensemble focal loss function is incorporated into the generator to ensure diversity and dispersion of the generated samples. Additionally, a collaborative focal loss function is introduced into the discriminator to improve the classification of difficult-to-classify data. A theoretical analysis is also conducted on the coverage of samples generated by the model. Extensive experiments conducted on the IoV simulation tool Framework For Misbehavior Detection (F2MD) demonstrate that the proposed method surpasses existing approaches in both detection effect and detection delay for zero-day attacks.
{"title":"Detection of zero-day attacks via sample augmentation for the Internet of Vehicles","authors":"Bingfeng Xu , Jincheng Zhao , Bo Wang , Gaofeng He","doi":"10.1016/j.vehcom.2025.100887","DOIUrl":"10.1016/j.vehcom.2025.100887","url":null,"abstract":"<div><div>Detecting zero-day attacks is a critical challenge in the Internet of Vehicles (IoV). Due to the limited availability of labeled attack data, anomaly-based methods are predominantly employed. However, the variability in the driving environment and behavioral patterns of vehicles introduces significant fluctuations in normal behavior, which in turn leads to high false positive rates when using these methods. In this work, we propose a novel detection method for zero-day attacks in IoV through sample augmentation. We first analyze the similarities between known and zero-day attacks in IoV. Based on the analysis, a Few-shot Learning Conditional Generative Adversarial Network (FLCGAN) model with multiple generators and discriminators is developed. Within this framework, an attack sample augmentation algorithm is designed to enhance input data by expanding the known attack dataset, thereby reducing false positives. To address the data imbalance caused by the limited number of input attack samples, an ensemble focal loss function is incorporated into the generator to ensure diversity and dispersion of the generated samples. Additionally, a collaborative focal loss function is introduced into the discriminator to improve the classification of difficult-to-classify data. A theoretical analysis is also conducted on the coverage of samples generated by the model. Extensive experiments conducted on the IoV simulation tool Framework For Misbehavior Detection (F2MD) demonstrate that the proposed method surpasses existing approaches in both detection effect and detection delay for zero-day attacks.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"52 ","pages":"Article 100887"},"PeriodicalIF":5.8,"publicationDate":"2025-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049923","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-20DOI: 10.1016/j.vehcom.2025.100879
Amir Masoud Rahmani , Amir Haider , Khursheed Aurangzeb , May Altulyan , Entesar Gemeay , Mohammad Sadegh Yousefpoor , Efat Yousefpoor , Parisa Khoshvaght , Mehdi Hosseinzadeh
Flying ad hoc networks (FANETs) are a new example of ad hoc networks, which arrange unmanned aerial vehicles (UAVs) in an ad hoc form. The features of these networks, such as the movement of UAVs in a 3D space, high speed of UAVs, dynamic topology, limited resources, and low density, have created vital challenges for communication reliability, especially when designing routing methods in FANETs. In this paper, a novel cylindrical filtering-based greedy perimeter stateless routing scheme (CF-GPSR) is suggested in FANETs. In CF-GPSR, cylindrical filtering reduces the size of the initial candidate set to accelerate the selection of the next-hop node. In this phase, the formulation of the cylindrical filtering construction process is expressed in the cylindrical coordinate system because the filtered area is a cylinder enclosed within the communication range of flying nodes. The cylindrical filtering construction process includes three steps, namely transferring coordinate axes, rotating coordinate axes, and cylinder construction. When selecting the next-hop node, CF-GPSR first uses this cylindrical filtering to limit the candidate set of each flying node. Then, CF-GPSR decides on the best next-hop UAV based on a merit function, which includes four criteria, namely velocity factor, ideal distance, residual energy, and movement angle, and selects a candidate node with the highest merit value as the next-hop UAV. Finally, the simulation process is performed using the NS 3.23 simulator, and four simulation scenarios are defined based on the number of UAVs, the communication area of nodes, network connections, and the size of packets to evaluate CF-GPSR. In the simulation process, CF-GPSR is compared with the three GPSR-based routing schemes, namely UF-GPSR, GPSR-PPU, and GPSR in terms of delay, data delivery ratio, data loss ratio, and throughput. In the first scenario, namely the change in the number of flying nodes, CF-GPSR improves delay, PDR, PLR, and throughput by 17.34%, 4.83%, 16%, and 7.05%, respectively. Also, in the second scenario, namely the change in communication range, the proposed method optimizes delay, PDR, PLR, and throughput by 4.91%, 5.71%, 6.12%, and 8.45%, respectively. In the third scenario, namely the change in the number of connections, CF-GPSR improves EED, PDR, PLR, and throughput by 18.41%, 9.09%, 9.52%, and 7.03%, respectively. In the fourth simulation scenario, namely the change in the packet size, CF-GPSR improves delay, PDR, PLR, and throughput by 14.81%, 19.39%, 7.19%, and 0.39%, respectively.
{"title":"A novel cylindrical filtering-based greedy perimeter stateless routing scheme in flying ad hoc networks","authors":"Amir Masoud Rahmani , Amir Haider , Khursheed Aurangzeb , May Altulyan , Entesar Gemeay , Mohammad Sadegh Yousefpoor , Efat Yousefpoor , Parisa Khoshvaght , Mehdi Hosseinzadeh","doi":"10.1016/j.vehcom.2025.100879","DOIUrl":"10.1016/j.vehcom.2025.100879","url":null,"abstract":"<div><div>Flying ad hoc networks (FANETs) are a new example of ad hoc networks, which arrange unmanned aerial vehicles (UAVs) in an ad hoc form. The features of these networks, such as the movement of UAVs in a 3D space, high speed of UAVs, dynamic topology, limited resources, and low density, have created vital challenges for communication reliability, especially when designing routing methods in FANETs. In this paper, a novel cylindrical filtering-based greedy perimeter stateless routing scheme (CF-GPSR) is suggested in FANETs. In CF-GPSR, cylindrical filtering reduces the size of the initial candidate set to accelerate the selection of the next-hop node. In this phase, the formulation of the cylindrical filtering construction process is expressed in the cylindrical coordinate system because the filtered area is a cylinder enclosed within the communication range of flying nodes. The cylindrical filtering construction process includes three steps, namely transferring coordinate axes, rotating coordinate axes, and cylinder construction. When selecting the next-hop node, CF-GPSR first uses this cylindrical filtering to limit the candidate set of each flying node. Then, CF-GPSR decides on the best next-hop UAV based on a merit function, which includes four criteria, namely velocity factor, ideal distance, residual energy, and movement angle, and selects a candidate node with the highest merit value as the next-hop UAV. Finally, the simulation process is performed using the NS 3.23 simulator, and four simulation scenarios are defined based on the number of UAVs, the communication area of nodes, network connections, and the size of packets to evaluate CF-GPSR. In the simulation process, CF-GPSR is compared with the three GPSR-based routing schemes, namely UF-GPSR, GPSR-PPU, and GPSR in terms of delay, data delivery ratio, data loss ratio, and throughput. In the first scenario, namely the change in the number of flying nodes, CF-GPSR improves delay, PDR, PLR, and throughput by 17.34%, 4.83%, 16%, and 7.05%, respectively. Also, in the second scenario, namely the change in communication range, the proposed method optimizes delay, PDR, PLR, and throughput by 4.91%, 5.71%, 6.12%, and 8.45%, respectively. In the third scenario, namely the change in the number of connections, CF-GPSR improves EED, PDR, PLR, and throughput by 18.41%, 9.09%, 9.52%, and 7.03%, respectively. In the fourth simulation scenario, namely the change in the packet size, CF-GPSR improves delay, PDR, PLR, and throughput by 14.81%, 19.39%, 7.19%, and 0.39%, respectively.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"52 ","pages":"Article 100879"},"PeriodicalIF":5.8,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049926","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The next generation (5G/B5G) vehicular cognitive radio networks (VCRNs) flag the track to intelligence-based autonomous driving in the initiation of future wireless networking and make daily vehicular operation more convenient, greener, efficient, and safer. However, with the continuous evolution of vehicles, the vehicular network becomes large-scale, dynamic, and heterogeneous, making it tough to fulfill the strict necessities, such as high security, resource allocation, massive connectivity, and ultralow latency. The combination of cognitive radio (CR) networks (different network coexistence) and machine learning (ML) has arisen as an influential artificial intelligence (AI) approach to make both the communication system and vehicle more adaptable and efficient. Naturally, applying ML to VCRNs has become an active research area and is being extensively considered in industry and academia. In this work, a reinforcement learning (RL) based optimal resource allocation (RORA) technique is proposed to solve the myopic decision-making problem by an autonomous vehicle (RL agent) takes its action to select the power level and optimal sub-band and maximize long-term rewards with a maximum payoff in VCRNs. The aim of this work is to design and implement an intelligent, resource allocation framework that ensures efficient and adaptive spectrum utilization while minimizing communication latency, energy consumption, and transmission cost in VCRNs. As a schema for the realization and capabilities evaluations, the CR networks consisting of LTE cellular network inter-working with Wi-Fi network with constant inter-space between Wi-Fi access points (APs) installed along the pathway is analysed. This framework is further analysed with variable inter-space between Wi-Fi APs. The key research problem addressed in this work is the challenge of optimizing spectrum and power allocation in highly dynamic vehicular environments characterized by rapid mobility, fluctuating network conditions, and interference from multiple vehicular CR nodes. The results show that the proposed RORA technique is more operative and outperforms other resource allocation schemes in terms of prediction accuracy and throughput.
Impact Statement
In this work, we proposed a machine learning-based technique applied to vehicular networks and opportunistic spectrum access parts in cognitive radio networks. The proposed technique envisions ways of enabling AI toward a future intelligent transportation system (ITS), including network intelligentization and the development of intelligent radio (IR).
{"title":"RORA: Reinforcement learning based optimal distributed resource allocation strategies in vehicular cognitive radio networks for 6G","authors":"Mani Shekhar Gupta , Akanksha Srivastava , Krishan Kumar","doi":"10.1016/j.vehcom.2025.100882","DOIUrl":"10.1016/j.vehcom.2025.100882","url":null,"abstract":"<div><div>The next generation (5G/B5G) vehicular cognitive radio networks (VCRNs) flag the track to intelligence-based autonomous driving in the initiation of future wireless networking and make daily vehicular operation more convenient, greener, efficient, and safer. However, with the continuous evolution of vehicles, the vehicular network becomes large-scale, dynamic, and heterogeneous, making it tough to fulfill the strict necessities, such as high security, resource allocation, massive connectivity, and ultralow latency. The combination of cognitive radio (CR) networks (different network coexistence) and machine learning (ML) has arisen as an influential artificial intelligence (AI) approach to make both the communication system and vehicle more adaptable and efficient. Naturally, applying ML to VCRNs has become an active research area and is being extensively considered in industry and academia. In this work, a reinforcement learning (RL) based optimal resource allocation (RORA) technique is proposed to solve the myopic decision-making problem by an autonomous vehicle (RL agent) takes its action to select the power level and optimal sub-band and maximize long-term rewards with a maximum payoff in VCRNs. The aim of this work is to design and implement an intelligent, resource allocation framework that ensures efficient and adaptive spectrum utilization while minimizing communication latency, energy consumption, and transmission cost in VCRNs. As a schema for the realization and capabilities evaluations, the CR networks consisting of LTE cellular network inter-working with Wi-Fi network with constant inter-space between Wi-Fi access points (APs) installed along the pathway is analysed. This framework is further analysed with variable inter-space between Wi-Fi APs. The key research problem addressed in this work is the challenge of optimizing spectrum and power allocation in highly dynamic vehicular environments characterized by rapid mobility, fluctuating network conditions, and interference from multiple vehicular CR nodes. The results show that the proposed RORA technique is more operative and outperforms other resource allocation schemes in terms of prediction accuracy and throughput.</div></div><div><h3>Impact Statement</h3><div>In this work, we proposed a machine learning-based technique applied to vehicular networks and opportunistic spectrum access parts in cognitive radio networks. The proposed technique envisions ways of enabling AI toward a future intelligent transportation system (ITS), including network intelligentization and the development of intelligent radio (IR).</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"52 ","pages":"Article 100882"},"PeriodicalIF":5.8,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049925","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号