Towards a Fog Based Computing Conceptual Framework for Biometric Digital Identification Systems.

The contribution posited in this paper is the provision of a conceptual framework for the use of Biometric facial recognition camera systems as a form of digital identification within your internet of things (IOT) integrated - data cloud environments, and in of itself an overlay to a data fog network layer. The authors conduct quantitative quasi experimental assessments that seek to answer the following research questions (i) To what extent can the characteristics of an Internet of things (IOT) biometric facial recognition camera system be used to generate false alerts. The false alerts seek to determine veracity of the collected and enrolled image templates stored within our test Microsoft Azure cloud environments and compared with non-cloud deployments to determine the veracity. The characteristics referred to in question # 1 are treated as the independent variables within the study and the false alerts as the dependent variable. We assume a threshold on all alerts to be a maximum of 0.1. Our experiments show the delta change between the non-cloud and the cloud based setup was identical for matching templates that were collected. The second research question seek to determine (ii) what is a suitable unifying framework to address question # 1 raised. In case of question # 2, the overlay of using the NIST biometric standards, NIST, Cyber-Security, and the NIST digital identity provided a strong terms of reference of this work such that a unified conceptual framework based on all three were integrated. The overall results of our study based on the two research questions provide a comprehensive baseline for which benchmark testing on biometric facial recognition identification technologies located within data clouds, and situated on edge fog networks can be further studied and our work represents one of the first anywhere in literature that speaks to these application study approaches described in this paper. While we grounded our benchmark simulation studies within the University environment, future work seeks to explore enterprise wide operational use cases for supporting performance evaluations within small to medium size production environments.