Secure device identification using multidimensional mapping

In this paper we investigate several potential hardware features from multiple devices for suitability during the employment of a device identification. The generation of stable and unique digital identity from features is challenging in device identification because of the unstable operation environments that implies the features employed are likely to vary under normal operating conditions. To address this, we introduce a novel multi-dimensional key generation technology which maps from multi-dimensional feature space directly to a key space. Furthermore, normalized distributions of features give the necessary data to model the characteristics, from which we derive intra-sample device feature distributions, and correlate the distinct features to generate a secure key to identify the device. Furthermore, to evaluate our experiment, we considerably carried out measurement using the mathematical & statistical modelling.