Asynchronous Federated Deep Reinforcement Learning-Based URLLC-Aware Computation Offloading in Space-Assisted Vehicular Networks

Pan, Chao; Zhao, Wang; Liao, Haijun; Zhou, Zhenyu; Wang, Xiaoyan; Tariq, Muhammad; Otaibi, Sattam Al

doi:10.1109/tits.2022.3150756

Cited by 22 publications

(12 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Pan et al [67] proposed an asynchronous federated DQN-based and URLLC-aware computation offloading algorithm (ASTEROID) to maximize throughput under long-term URLLC constraints. In this paper, a URLLC constraint model, based on extreme value theory, is established.…”

Section: Deep Reinforcement Learning Algorithm Based On Value Functionmentioning

confidence: 99%

Deep Learning for Video Application in Cooperative Vehicle-Infrastructure System: A Comprehensive Survey

Dai

2022

Applied Sciences

View full text Add to dashboard Cite

Video application is a research hotspot in cooperative vehicle-infrastructure systems (CVIS) which is greatly related to traffic safety and the quality of user experience. Dealing with large datasets of feedback from complex environments is a challenge when using traditional video application approaches. However, the in-depth structure of deep learning has the ability to deal with high-dimensional data sets, which shows better performance in video application problems. Therefore, the research value and significance of video applications over CVIS can be better reflected through deep learning. Firstly, the research status of traditional video application methods and deep learning methods over CVIS were introduced; the existing video application methods based on deep learning were classified according to generative and discriminative deep architecture. Then, we summarized the main methods of deep learning and deep reinforcement learning algorithms for video applications over CVIS, and made a comparative study of their performances. Finally, the challenges and development trends of deep learning in the field were explored and discussed.

show abstract

Section: Deep Reinforcement Learning Algorithm Based On Value Functionmentioning

confidence: 99%

Deep Learning for Video Application in Cooperative Vehicle-Infrastructure System: A Comprehensive Survey

Dai

2022

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…It also has the ability to predict the unknowable popularity of cached contents. With the joining and leaving of vehicles, the merging and splitting of the cluster, and the change of communication security level, the group leader needs to update the group key of the intracluster communication in real-time [ 30 , 31 ]. The dynamic optimization process of the required length and update frequency of the group key, the security level of the group communication, and the calculation delay of communication encryption and decryption can be regarded as a Markov process.…”

Section: Key Management Technology Based On Reinforcement Learningmentioning

confidence: 99%

“…When λ = 3, compared with the GKA scheme, the proposed technique's encryption and decryption calculation delay are reduced by 18.1%, and the security level is improved by 24.1%. This is because, compared with the fixed vital length and update frequency of the GKA scheme, the RLKA technology proposed in this paper can be based on the number of vehicles in the current cluster, the number of vehicles joining and leaving the group, and the safety feedback of cluster vehicle intrusion detection [ 31 ]. Prescan simulation software uses a basic vehicle dynamics model, which is unable to drive the intelligent vehicle precisely in the vertical direction or to effectively reflect the vehicle's dynamic properties in that direction.…”

Section: Simulation Experimentsmentioning

confidence: 99%

Minimum Latency-Secure Key Transmission for Cloud-Based Internet of Vehicles Using Reinforcement Learning

Akilandeswari¹,

Kumar

Thilagamani³

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

The Internet of Vehicles (IoV) communication key management level controls the confidentiality and security of its data, which may withstand user identity-based attacks such as electronic spoofing. The IoV group’s key is updated with a defined frequency under the current key management method, which lengthens the time between crucial changes and encryption. The cluster key distribution management is used as the study object in this paper, which is based on the communication security on the Internet of Vehicles cluster. When vehicles enter and exit the cluster, the Internet of Vehicles must update the group key in real-time to ensure its forward and backward security. A low-latency IoV group key distribution management technology based on reinforcement learning is proposed to optimize the group owner vehicle according to factors such as changes in the number of surrounding vehicles and essential update records and the update frequency and the key length of its group key. The technology does not require the group leader vehicle to predict the nearby traffic flow model. The access-driven cache attack model reduces the delay of encryption and decryption and is verified in the simulation of the IoV based on advanced encryption standards. The simulation results show that, compared with the benchmark group key management scheme, this technology reduces the transmission delay of key updates, the calculation delay of encryption and decryption of the IoV, and improves the group key confidentiality.

show abstract

“…Besides, for heterogeneous data, the transmission latency of each synchronous model aggregation mechanism is unacceptable for time-sensitive devices. In this way, several works have proposed asynchronous model aggregation methods, where only one participant device would update the global model each time [19]- [21]. Meanwhile, when one device uploads its model, the others continue to complete their training.…”

Section: Introductionmentioning

confidence: 99%

Semi-Asynchronous Hierarchical Federated Learning Over Mobile Edge Networks

Chen

You

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Mobile edge network has been recognized as a promising technology for future wireless communications. However, mobile edge networks usually gathering large amounts of data, which makes it difficult to explore data science efficiently. Currently, federated learning has been proposed as an appealing approach to allow users to cooperatively reap the benefits from trained participants. In this paper, we propose a novel Semi-Asynchronous Hierarchical Federated Learning (SAHFL) framework for mobile edge networks that enables elastic edge to cloud model aggregation from data sensing. We further formulate a joint edge node association and resource allocation problem under the proposed SAHFL framework to prevent personalities of heterogeneous devices and achieve communication-efficiency. To deal with our proposed Mixed integer nonlinear programming (MINLP) problem, we introduce a distributed Alternating Direction Method of Multipliers (ADMM)-Block Coordinate Update (BCU) algorithm. With this algorithm, a tradeoff between training accuracy and transmission latency has been derived. Numerical results demonstrate the advantages of the proposed algorithm in terms of training overhead and model performance.INDEX TERMS Edge association, federated learning, mobile edge network, resource allocation, semiasynchronous.

show abstract

Asynchronous Federated Deep Reinforcement Learning-Based URLLC-Aware Computation Offloading in Space-Assisted Vehicular Networks

Cited by 22 publications

References 33 publications

Deep Learning for Video Application in Cooperative Vehicle-Infrastructure System: A Comprehensive Survey

Deep Learning for Video Application in Cooperative Vehicle-Infrastructure System: A Comprehensive Survey

Minimum Latency-Secure Key Transmission for Cloud-Based Internet of Vehicles Using Reinforcement Learning

Semi-Asynchronous Hierarchical Federated Learning Over Mobile Edge Networks

Contact Info

Product

Resources

About