SCIENCE@home

DAG-AD-NAVs: Novel Attacks and Defenses on Decentralized Networks of Autonomous Vehicles With Directed Acyclic Graph Structures

Sazid Nazat,Walaa M. Alayed,Mustafa Abdallah

Published 2025 in IEEE Access

ABSTRACT

Autonomous vehicles (AVs) provide major enhancements to transportation networks, but securing decentralized AV networks against attacks remains a critical challenge. This paper introduces a novel framework for AV connectivity using Directed Acyclic Graph (DAG) structures. We propose four different attack scenarios on DAG connectivity framework of AVs (including insider attacks like Attack-A and Attack-B, and isolation attacks like Attack-C and Attack-D). The goal of these attacks is to demonstrate the vulnerabilities of these DAG-based AV networks. We also suggest different defense strategies against each of the attack scenarios (including adaptive DAG admission control, RSU restoration protocols, a centralized Regional Transport Supervision Agency (RTSA), and a modified PRoPHET routing algorithm incorporating authentication scores). We evaluate our proposed attacks and defenses on five different structures that represent different ways for connecting AVs. These DAG structures are: binary tree, partial mesh, star, along with two other structures from prior works. We also provide baseline comparisons against Trust-Based Admission and Sybil-Resistant Join defenses. In our evaluation, we test the effect of our different attacks and defenses on the evolution of number of anomalous and benign AVs over time steps for all of these five DAG structures. Our defense strategies reduce anomalous AVs by an average of 5.8x across attack scenarios, including 10x for binary tree, 3x for partial mesh, and 4.5x for attack Graph-A. It increases successful message delivery between benign AVs by 8.2x across all time steps and ensures 100% connection retention among benign AVs through RSU restoration. These findings offer critical insights for designing secure, scalable, and resilient decentralized AV systems.

PUBLICATION RECORD

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-53 of 53 references · Page 1 of 1

CITED BY

  • No citing papers are available for this paper.

Showing 0-0 of 0 citing papers · Page 1 of 1