SortAlgo-Metrics: Identification of Cloud-Based Server Via a Simple Algorithmic Analysis

This paper introduces a novel technique to detect spoof or fake software systems via the generation of a unique digital signature based on a direct analysis of the construction of the system. Specifically, we model a novel mechanism referred to as SortAlgo-Metrics analysis to identify cloud-based servers. Experimentally, we deployed four cloud-based servers to run four sorting algorithms in order to extract features that are employed to perform statistical analysis upon with the aim to obtain their metrics which has further underpin the investigation of their behaviours. The model has been validated by comparing training data and unknown data, and the result has shown server 2-4 have a strong identification with 96% probability, while server 1 with 55%, it is surmised that is could be as the result of insufficient sample data. However, if such a simple model can produce a result with this high probability, this shows that with more complex features and sufficient data pulled from cloud-based servers, SortAlgo-Metrics model could generate a higher degree of basis numbers for ICMetrics technology entropy key generation and other complex systems.