Internet users operate lots of different applications at various network endpoints, which generate different traffic patterns. It would be practical to analyze these patterns based on an unified standpoint, which is universally applicable in all the cases. To solve this problem we should turn to the field of Mathematics, because it does not regard the differences deriving from distributed networks and multiple usage.
In my work, I research algorithms for comparing time series and implement these algorithms in JAVA to solve the formerly described problem. Several gigabytes of traffic data was collected with different bandwidth settings and by different network applications. In this way I managed to simulate the internet habits of the average user. I had to find a solution for managing big data in reasonable time, so in the work I implement a sliding window solution. The results of the implemented algorithms are evaluated by runtime and by specific factors. Conclusions are drawn and suggestions are made for further development.