Trust evaluation mechanism for data collection in smart cities

Abstract

Driven by Artificial Intelligence (AI) and Internet of Things (IoT) technologies, smart cities are becoming more sophisticated. In smart cities, intelligent applications such as traffic optimization, environmental monitoring and emergency response develop intelligent decisions with AI models. The core of smart decision-making lies in the intelligent analysis and processing of real-time data collected by IoT devices, which supports the intelligent and precise operation of cities. However, the credibility of data sources directly affects the quality of decision-making and the effectiveness of AI models. Issues such as tampering, falsification, or failure of data sources not only lead to biased decisions but may also jeopardize the stability and security of urban systems. Therefore, conducting trust evaluations on data collection terminals in intelligent scenarios is crucial for ensuring the reliable operation of smart cities. This paper proposes a multidimensional trust evaluation method for data collection in intelligent applications of smart cities. Moreover, this method can effectively detect sophisticated attacks, such as data poisoning and insider threats. By incorporating static, dynamic, and energy trust assessments, and applying the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) to generate an integrated trust value, this approach ensures data credibility from the source and provides reliable data support for smart decision-making.