Deep Reinforcement Learning for Cooperative Edge Caching in Vehicular Networks

Xing, Yuping; Sun, Yanhua; Lan, Qiao; Wang, Zhuwei; Si, Pengbo; Zhang, Yanhua

doi:10.1109/iccsn52437.2021.9463666

Cited by 11 publications

(4 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Vehicles construct clusters and receive requested content from either their CHs or RSUs; part of the mathematical model is solvable as a knapsack problem [6] Collaborative caching Vehicles broadcast the names of cached content [17] Collaborative caching Cache policy is determined with deep reinforcement learning [18] Collaborative caching The problem of which content to store and where to store it is determined by machine learning [19] Collaborative caching Data carrier node is selected by reinforcement learning [20] Cache place determination Cache placement is treated as an MWVCP problem [21] Cache place determination Node n (n < k) only stores content c, where c mod k = n [22] Relay vehicle determination Vehicles that lie in the common communication area are preferentially selected…”

Section: Referencementioning

confidence: 99%

See 1 more Smart Citation

Efficient V2V Communications by Clustering-Based Collaborative Caching

Tokunaga,

Tang

2024

Electronics

View full text Add to dashboard Cite

Vehicle-to-vehicle (V2V) communication plays an important role in enabling autonomous driving. However, when multiple vehicles request the same content, like road conditions, delivering it individually by V2V communication can significantly increase traffic volume, potentially causing congestion in the wireless channel. To address this issue, Content-Centric Network (CCN) technology is applied to V2V communication, which improves communication efficiency by exploiting content cached at vehicles. However, previous methods faced the following challenges: (i) vehicles could not use content stored in nearby vehicles outside the communication path, and (ii) redundant caching of the same content occurred at nearby vehicles. To tackle these challenges, this paper proposes a collaborative caching method in which vehicles are grouped into clusters and each cluster has a designated head responsible for managing caches across all vehicles within the cluster. In this way, this method enables vehicles to use the content cached at adjacent vehicles that are not directly on a communication path. In addition, it eliminates redundant caches, allowing a more diverse range of content storage. Extensive simulation results demonstrate that the proposed approach effectively reduces content delivery latency by 33% compared to the method using clusters without cooperative caching and by 19% compared to the ECV+ method.

show abstract

Section: Referencementioning

confidence: 99%

“…With the emergence of AI, some studies suggest combining AI and vehicular networks: Ref. [17] calculates the cache policy with deep reinforcement learning, Ref. [18] uses machine learning to decide which content to store and where to store it, and Ref.…”

Section: Collaborative Cachingmentioning

confidence: 99%

Efficient V2V Communications by Clustering-Based Collaborative Caching

Tokunaga,

Tang

2024

Electronics

View full text Add to dashboard Cite

show abstract

“…Vehicles travel at a uniform speed, and the distance d n between adjacent CAVPs in the same lane follows a uniform distribution U. The frame length is T. We consider the spatial location and environment around the vehicle to be semi-static, i.e., not changing significantly within frame T but changing dynamically over multiple frames [33]. The detailed scenario parameters are set as shown in Table 1.…”

Section: Simulation Setupmentioning

confidence: 99%

Reinforcement Learning-Based Resource Allocation for Multiple Vehicles with Communication-Assisted Sensing Mechanism

Fan,

Fei,

Huang

et al. 2024

Electronics

View full text Add to dashboard Cite

Autonomous vehicles (AVs) can be equipped with Integrated sensing and communications (ISAC) devices to realize sensing and communication functions simultaneously. Time-division ISAC (TD-ISAC) is advantageous due to its ease of implementation, efficient deployment and integration into any system. TD-ISAC greatly enhances spectrum efficiency and equipment utilization and reduces system energy consumption. In this paper, we propose a communication-assisted sensing mechanism based on TD-ISAC to support multi-vehicle collaborative sensing. However, there are some challenges in applying TD-ISAC to AVs. First, AVs should allocate resources for sensing and communication in a dynamically changing environment. Second, the limited spectrum resources bring the problem of mutual interference of multi-vehicle signals. To address these issues, we construct a multi-vehicle signal interference model, formulate an optimization problem based on the partially observable Markov decision process (POMDP) framework and design a decentralized dynamic allocation scheme for multi-vehicle time–frequency resources based on a deep reinforcement learning (DRL) algorithm. Simulation results show that the proposed scheme performs better in miss detection probability and average system interference power compared to the DRQN algorithm without the communication-assisted sensing mechanism and the random algorithm without reinforcement learning. We can conclude that the proposed scheme can effectively allocate the resources of the TD-ISAC system and reduce interference between multiple vehicles.

show abstract

“…The installation of a mobile edge computing (MEC) server in the RSU creates space for the caching content, further enabling the delivery of the content from a location closer than that of the remote server. Caching to the RSUs decreases the latency and relieves the backhaul burden [ 4 ]. However, because the vehicle is constantly moving, the residence time in the RSU range is relatively shorter and the network topology continues to vary [ 5 ].…”

Section: Introductionmentioning

confidence: 99%

Deep Reinforcement Learning for Edge Caching with Mobility Prediction in Vehicular Networks

Choi

Lim

2023

Sensors

View full text Add to dashboard Cite

As vehicles are connected to the Internet, various services can be provided to users. However, if the requests of vehicle users are concentrated on the remote server, the transmission delay increases, and there is a high possibility that the delay constraint cannot be satisfied. To solve this problem, caching can be performed at a closer proximity to the user which in turn would reduce the latency by distributing requests. The road side unit (RSU) and vehicle can serve as caching nodes by providing storage space closer to users through a mobile edge computing (MEC) server and an on-board unit (OBU), respectively. In this paper, we propose a caching strategy for both RSUs and vehicles with the goal of maximizing the caching node throughput. The vehicles move at a greater speed; thus, if positions of the vehicles are predictable in advance, this helps to determine the location and type of content that has to be cached. By using the temporal and spatial characteristics of vehicles, we adopted a long short-term memory (LSTM) to predict the locations of the vehicles. To respond to time-varying content popularity, a deep deterministic policy gradient (DDPG) was used to determine the size of each piece of content to be stored in the caching nodes. Experiments in various environments have proven that the proposed algorithm performs better when compared to other caching methods in terms of the throughput of caching nodes, delay constraint satisfaction, and update cost.

show abstract

Deep Reinforcement Learning for Cooperative Edge Caching in Vehicular Networks

Cited by 11 publications

References 13 publications

Efficient V2V Communications by Clustering-Based Collaborative Caching

Efficient V2V Communications by Clustering-Based Collaborative Caching

Reinforcement Learning-Based Resource Allocation for Multiple Vehicles with Communication-Assisted Sensing Mechanism

Deep Reinforcement Learning for Edge Caching with Mobility Prediction in Vehicular Networks

Contact Info

Product

Resources

About