In the IoT space CSPs can observe anomalies like unexpectedly high (or low) device activity in a specific location, unexpected mobility changes, deviations from expected service usage (such as sudden peaks in voice call usage) or unexpectedly high activity of devices in a specific location (which may indicate that several devices of the same type have been stolen).
The main problem is that a lot of critical information needed for decision-making is delivered to the telecom operators too late. Traditional solutions based on business intelligence/data warehouse platforms enable you to perform post-mortem, offline data mining and analysis, and provide actionable results no sooner than several days or even weeks after the event took place. This is no longer fast enough in a world where everything needs to happen in real time.
Real-time big data processing can, with the application of various types of algorithms, be very effective in anomaly detection. Technology based on machine learning allows the identification of previously unknown issues that may be responsible for quality problems and security threats. This enables service providers to eliminate existing anomalies and prevent future ones, and to detect problems more rapidly and solve them pro-actively.
Anomalies can be detected by analyzing device behavior, networks, groups of devices owned by one customer in any segment, or location.
By analyzing mobile subscriptions, it’s possible to detect numerous frauds, including the use of duplicated SIM cards and abusive roaming cases. This is extremely useful in the IoT space where large assets need to be supervised. The AI/ML engine continuously checks if the place and nature of the service usage is suspicious.
Usage Analytics (examples):
Movement Control (examples):
The engine is complimented by a graphic representation of the device movement on a map. Potential fraud becomes clearly visible (“why are my devices running in Australia?”).
Tell us about your business needs. We will find the perfect solution.