Dynamic Selfish Node Detection With Link Quality Consideration in Vehicular Networks

With the emergence of various applications in intelligent transportation systems (ITS), the cooperative communication among the nodes on roads is becoming more and more important in vehicular networks. However, selfish nodes, which are not willing to relay data packets for others, drastically decrease the communication performance of the network in various aspects including packet delivery ratio, end-to-end delay, communication bandwidth, etc. It is critical to detect the selfish nodes in the network for avoiding the decrement of network performance. Nevertheless, it is extremely challenging to correctly and fairly detect selfish nodes in vehicular networks where the network topology is highly dynamic and link quality between nodes is lossy. In order to address this issue, a novel detection scheme is proposed in this paper. The proposed scheme jointly considers the link quality, mobility, and vehicle behaviors at the network layer and MAC layer with fuzzy logic in a fully decentralized manner. Furthermore, an innovative perception module is deployed based on probabilistic calculations to refine the detection decisions of the scheme. The scheme is validated by simulation platforms which integrate Veins, SUMO, and INET Framework under IEEE 802.11p and AODV protocols. The computer simulations are conducted with both highway and urban traffic scenarios of Xi'an city in China. As the results shown, the proposed scheme demonstrates better detection capabilities than the conventional schemes in terms of
Precision
,
Recall
, and
F1-score
with an acceptable communication overhead.