We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.NI

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Networking and Internet Architecture

Title: Optimizing Vehicle-to-Edge Mapping with Load Balancing for Attack-Resilience in IoV

Abstract: Attack-resilience is essential to maintain continuous service availability in Internet of Vehicles (IoV) where critical tasks are carried out. In this paper, we address the problem of service outage due to attacks on the edge network and propose an attack-resilient mapping of vehicles to edge nodes that host different types of service instances considering resource efficiency and delay. The distribution of service requests (of an attack-affected edge node) to multiple attack-free edge nodes is performed with an optimal vehicle-to-edge (V2E) mapping. The optimal mapping aims to improve the user experience with minimal delay while considering fair usage of edge capacities and balanced load upon a failure over different edge nodes. The proposed mapping solution is used within a deep reinforcement learning (DRL) based framework to effectively deal with the dynamism in service requests and vehicle mobility. We demonstrate the effectiveness of the proposed mapping approach through extensive simulation results using real-world vehicle mobility datasets from three cities.
Comments: 7 pages, 9 figures, 1 table, Accepted in IEEE CCNC 2023
Subjects: Networking and Internet Architecture (cs.NI)
Cite as: arXiv:2210.01375 [cs.NI]
  (or arXiv:2210.01375v1 [cs.NI] for this version)

Submission history

From: Anum Talpur [view email]
[v1] Tue, 4 Oct 2022 04:35:55 GMT (8568kb,D)

Link back to: arXiv, form interface, contact.