Quesada: A Framework for Reliable and Trustworthy Data Acquisition in 6G-IoT

The convergence of Artificial Intelligence (AI) and the Internet of Things (IoT) in future 6G networks (6G-IoT) promises to unlock unprecedented capabilities. However, the continuous collection and analysis of large-scale, low-density data pose significant threats to the reliability and trustworthiness of these systems, leading to high energy consumption and potential decision-making based on stale information. To address these critical challenges, this paper proposes a novel architecture for building reliable and trustworthy 6G-IoT services. Our approach involves three key contributions: 1) We leverage the multi-access edge computing (MEC) paradigm to locally process raw data, filtering redundancy and thereby ensuring that AI models operate on more meaningful information. 2) We design a decoupled, two-level (edge-cloud) decision-making mechanism that explicitly manages the trade-off between data trustworthiness (quantified by information freshness) and system energy consumption, a cornerstone of long-term reliability. 3) We implement these principles in a new, distributed end-edge-cloud framework named Quesada (Query-control-based safe and data-trustworthy acquisition), which coordinates edge and cloud decisions to enhance overall system performance. To validate our approach, we conduct a series of comparative experiments. The results demonstrate that the Quesada framework significantly improves both system reliability and data trustworthiness, making it a viable architecture for future 6G-IoT applications.