Abstract
Cyber-physical systems in a smart city environment offer secure computations in addition to robust resources and secure information exchanges. From the security aspect, the selection of legitimate computing resources is considered the most trustworthy measure for preserving user and data privacy. However, the outlier information results in false computations and time-consuming privacy measures for smart city users. This article introduces the Computation Annealed Selection Process (CASP) for combating outlier information in cyber-physical system networks. This process is assimilated with different infrastructure units in the smart city in an adaptable fashion. The adaptive measure administers information and user privacy through mutual time-key-based authentication. Information privacy is endured until the smart city provider provides a valid service interval. W2The service interval is evaluated for its consistency in maintaining the privacy and is validated using decision-tree learning. The decision tree performs interval classification and key assignment recommendations. Based on the decisions, the privacy is either prolonged or withdrawn for the information exchanged between the users and service providers. This process is repeated until the end of the service interval, identifying the outlier information in legitimate intervals. The proposed process improves detection, legitimacy rate, interval time, and interval rate by 7.57%, 10.4%, 13.15%, and 7.59%, respectively.