Distributed Deep Learning-based Task Offloading for UAV-enabled Mobile Edge Computing

Mukherjee, Mithun; Kumar, Vikas; Lat, Ankit; Guo, Mian; Matam, Rakesh; Lv, Yunrong

doi:10.1109/infocomwkshps50562.2020.9162899

Cited by 36 publications

(13 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For large-scale systems with high-dimensional state and action spaces, the learning process for RL may take a long time to converge to the best policy. To overcome the limitations of RL, DRL, which leverage the advantages of deep neural networks (DNNs) [151,152] to train the learning process, has been proposed. A DNN which is referred to as an artificial neural network (ANN) with two or more hidden layers is widely used for optimization problems to find the solutions by manipulating mathematics appropriately [141].…”

Section: Deep Reinforcement Learningmentioning

confidence: 99%

A comprehensive survey on aerial mobile edge computing: Challenges, state-of-the-art, and future directions

Song

Qin

Hao

et al. 2022

Computer Communications

View full text Add to dashboard Cite

Section: Deep Reinforcement Learningmentioning

confidence: 99%

A comprehensive survey on aerial mobile edge computing: Challenges, state-of-the-art, and future directions

Song

Qin

Hao

et al. 2022

Computer Communications

View full text Add to dashboard Cite

“…The authors suggest that their approach suppresses the motivation of selfish behavior and faked service record attacks by malicious edge servers. Also in 2020, the authors Mukherjee et al study a task offloading model for energy-constrained mobile devices in UAV-enabled MEC networks [56]. In their framework, end users can process tasks themselves or offload them to UAVs that act as access points.…”

Section: Edge Servers Consortium Blockchainmentioning

confidence: 99%

Exploring the Intersection of Consortium Blockchain Technologies and Multi-Access Edge Computing: Chronicles of a Proof of Concept Demo

Vera-Rivera

Hossain

Refaey

2022

IEEE Open J. Commun. Soc.

View full text Add to dashboard Cite

The efficient administration of network resources in Multi-Access Edge Computing (MEC) is an active research topic these days. Task sharing, in particular, is one of the fundamental problems regarding MEC architectures although the existing literature approaches the topic mostly from the mobile user point of view. In this article, we present the chronicles of EdgeChain network, a proof of concept demo of a blockchain-based model for secure and private task sharing collaboration specially designed for edge computing servers. EdgeChain operates on a decentralized approach that counts on the blockchain services provided by the Hyperledger Fabric platform. The network offers enhanced security features that leverage the permissioned nature of Fabric, more specifically, it relies on Fabric's membership service to validate identities and allowed behavior of participant nodes in the network. This design choice restrains external attackers from interfering with the normal operation of the task sharing scheme. The network also offers enhanced privacy features powered by a smart contract design that makes use of multiple decoupled Fabric channels with separate operation rules, ledgers, and peerto-peer communication networks. This design choice guarantees that the computational tasks circulating in the network are only exposed to servers participating in task sharing services. The trait prevents other servers in the network to have access to private tasks and their data. The article goes through the multiple stages of the design and construction process of the proof of concept demo, from the blockchain-based task sharing framework and system model, all the way to the implementation details of the network.

show abstract

“…Many efforts have proposed DML, such as FL and partitioned learning techniques, to allow wireless devices to acquire a global model with limited data exchange or based on partial models and datasets. Mukherjee et al [29] introduced a distributed deep neural network-based offloading strategy to minimize the weighted sum of the delay and the energy consumption in UAV-assisted MEC networks. McMahan et al [30] presented the Federated Averaging (FedAvg) algorithm, which trains an aggregate model and does not require uploading client data to a server.…”

Section: Distributed Machine Learningmentioning

confidence: 99%

Predictive UAV Base Station Deployment and Service Offloading With Distributed Edge Learning

Zhao

Pacheco

Santos

et al. 2021

IEEE Trans. Netw. Serv. Manage.

View full text Add to dashboard Cite

In modern networks, edge computing will be responsible for processing and learning from the critical networkand user-generated data, such as wireless link usage, mobility information, application requests, and many others. The presence of Artificial Intelligence-based (AI) applications at the edge of the network will enable the network to predict necessary user behavior and its impact on network infrastructure, such as base station overloading. One of the main strategies for offloading users and base stations is to deploy UAV base stations, or flying base stations, which can dynamically provide service and connectivity. In this article, we introduce a framework for distributed learning over Multi-access Edge Computing (MEC), which manages data applications in a fully distributed setting across edge servers, thus reducing the cost of collecting user information in a centralized server. We couple the proposed distributed learning with a novel similarity metric for user trajectories, which can aggregate neural network models with similar costs as other model aggregation techniques. However, the aggregation technique can achieve much higher accuracy. Furthermore, we apply the proposed distributed learning scheme to manage and deploy flying base stations to areas that experience high demand or poor user connectivity, thus optimizing connectivity in terms of user satisfaction, delay, and network throughput.

show abstract

Distributed Deep Learning-based Task Offloading for UAV-enabled Mobile Edge Computing

Cited by 36 publications

References 22 publications

A comprehensive survey on aerial mobile edge computing: Challenges, state-of-the-art, and future directions

A comprehensive survey on aerial mobile edge computing: Challenges, state-of-the-art, and future directions

Exploring the Intersection of Consortium Blockchain Technologies and Multi-Access Edge Computing: Chronicles of a Proof of Concept Demo

Predictive UAV Base Station Deployment and Service Offloading With Distributed Edge Learning

Contact Info

Product

Resources

About