A model of a secure information system for cognitive data processing in IoT sensor networks for laboratory climatic testing

Information protection during data transmission in IoT-based sensor networks used in laboratory testing is an important task in the context of growing demand for laboratory climatic tests. The use of cognitive artificial intelligence in the process of managing laboratory climatic tests requires taking into account not only factors affecting the test object, but also factors of possible network-based attacks on the cyberphysical system. That is why the paper presents a model of a secure information system for cognitive data processing in IoT sensor networks. When cognitive information processing is performed, situational factors of the first kind (data on climatic conditions) and the second kind (potential network-based attacks) are taken into account. Data protection is implemented through cryptographic information transformation based on the generation of two-operand asymmetric CET-operations with an accuracy of up to the permutation of the second operand when transmitting information from sensors to the cognitive data processing unit (Algorithm I). To protect the information transmitted from the database to the cognitive processing unit, a synthesis of direct and inverse operations is used based on the generation of two-operand asymmetric CET-operations with an accuracy of up to the permutation of the first operand (Algorithm II). To protect the information transmitted from the cognitive data processing unit to the actuators, a synthesis of direct and inverse operations is used based on the generation of two-operand asymmetric CET-operations with an accuracy of up to the permutation of the transformation result (Algorithm III). Situational factors of the first kind affect the information, due to which encryption is performed according to algorithm I. The presence of algorithms II and III is due to situational factors of the second kind. A graphical and tabular representation of the information system model was formed, which determined the role and place of protection algorithms in the functional structure of an IoT-based sensor network. The analysis of the development in modeling managerial decision-making based on cognitive data processing, taking into account situational factors of the first and second kind under conditions of network-based attack simulation, showed improved efficiency. The development is 7% more efficient compared to the results of managerial decision-making by a person and 49% more efficient compared to the use of the LIMS considered in the work.