A DNA sequencing based approach to fault diagnosis in automotive electronic networks

The increasing number of Electronic Control Units within the network of a vehicle is increasing the level of complexity of these networks. Thus, fault diagnosis of these sophisticated systems becomes more complex. The aim of this paper is to provide a technique to enable CAN-based fault detection in a premium vehicle network or any other network which uses the Controller Area Network (CAN) protocol for communication. The fault detection technique described here is based on a sequential behaviour of the CAN network of a vehicle and by using signal processing methods commonly used in DNA sequencing analysis, fault detection was achieved and data were classified in clusters of normal scenarios and fault scenario.