Priority Aggregation Network With Integrated Computation and Sensation for Ultra Dense Artificial Intelligence of Things

In the realm of ultra dense Intelligence of Things, efficient utilization of computation and sensation (CAS) resources remains a challenge due to their uneven temporal and spatial distribution. Priority Aggregation Network (PAN) is introduced to employ priority as an intermediary to restore, combine, and exchange idle CAS resources' permissions for cross-temporal and cross-modal resource leasing. PAN awards priority to devices that lease their sensing and computing resources to others, facilitating efficient resource allocation. Numerical results indicate PAN maintains high cross-temporal task completion rates while conserving equipment resources, with increased willingness to pay for resources.