AI based sniffing anomaly detector
artea.com developed for a car part supplier an IoT Big Data system that can detect data from the car’s control unit and identify anomalies in real-time.
Data is carried over the network in packets (TCP) through seven physical layers. Each of these layers could have a throughput of 100gbit/s: the ability to process data in real-time allows the identification of hacker attacks or network failure. Denial-of-service (DoS) is a type of cyber-attack where hackers seek to make a resource (or the entire network) unavailable by flooding them with unnecessary connection requests in order to overload systems and block communication devices.
AI allows the monitoring of network traffic from physical to application layer. Real-time processed data is able to detect causes of a control unit anomaly. Once intercepted through an AI engine, anomalies are then reported to a call center. Data is then quickly analyzed by the call center and instructions are seamlessly sent to the driver. This service is also offered on subscription.
- Feature extraction A hardware data dump device at OSI level-1 can reduce processor load by making traffic information available to the user.
- On-field anomaly detection FPGA hardware allows the processing of ML algorithms on physical appliances to detect anomalies as soon as possible.
- Cloud Training Collected data is copied to the Cloud for storage and advanced analytics are applied to improve effectiveness of ML algorithms through training.