A Cooperative Scheme Against Cache Pollution Attack in Vehicular Content Centric Network
- 0 Collaborators
We design a scheme based on the cooperation of Vehilces and RSUs in Vehicular Content Centric Network to detect and defense Cache Pullution Attack. ...learn more
Project status: Concept
Overview / Usage
Vehicular Content Centric Network (VCCN) has been proposed to address user mobility and sporadic connectivity issues in Vehicular Ad hoc Network (VANET). In VCCN, in-network data caching is frequently used to store frequently accessed data on vehicles. Data caching, unfortunately, suffers from Cache Pollution Attack (CPA) that send out fabricated requests to pollute data caches. In this work, we propose a cooperative detection and defense scheme of CPA in VCCN. In our scheme, each device maintains a suspicious list by recording those users who have repeatedly submitted interests to contents that are way out of the normal fluctuation of requests to popular contents in the network. By cooperating with all vehicles, each RSU computes a final attack table based on the suspicious lists from the direct observation and indirect feedbacks from those that have come into range. As a line of defense, requests from suspected attackers will be dropped in order to protect VCCN from CPA. Simulation evaluations demonstrate that our scheme is very effective in detecting and defending CPA with higher cache hit, higher detecting ratio, and higher cache accuracy compared to other state-of-the-art schemes.
Methodology / Approach
As soon as an Interest is received, each node makes the following statistics: the number of Interests for the same content, request interval between two consecutive Interests for the same content and the corresponding requester.
Periodically, each node computes the popularity increment of each content based on the above statistics. If the increment is more than the threshold, this node will add a record to its suspicious list including the content name, the requester Identifier (ID) and the number of suspicious packets sent by the requester.
Each vehicle sends its suspicious list to its RSU periodically if there is any update in its list. Each RSU can calculate a final attack table based on its direct observation and feedbacks from other vehicles. Then, the attack table is broadcast within its communication range.
Each node will discard the packets from the malicious requesters contained in the attack table.
Technologies Used
The Opportunistic Network Environment (ONE) simulator .
Java programming skills.