AoI-Aware Pilot Sequence Construction for Active User Detection in MTC Networks

Many Internet of Things (IoT) applications require fresh information for proper operation, and the age of information (AoI) metric serves as a key indicator of freshness. To guarantee the delivery of fresh information, transmitting users must be correctly identified so that their packets can be decoded, making active user detection (AUD) a prerequisite for establishing communication. The AUD becomes particularly challenging in scenarios with multiple active users and limited orthogonal resources. While compressed sensing (CS) techniques have shown promise for AUD by enabling simultaneous communication in resource-scarce environments, most research has focused on improving detection precision rather than optimizing AoI. This work introduces a novel modification to CS-based AUD to prioritize high-AoI transmissions, reducing the network’s average AoI. Specifically, we propose a pilot sequence construction method where users generate pilots as linear combinations of an orthogonal sequence basis, with unique weight sets assigned to each user. High-AoI users are allocated exclusive pilot sequences, ensuring orthogonality between high- and low-AoI transmissions and enhancing the detection of high-AoI users, even in congested settings. Simulations reveal that optimizing the AoI threshold and the number of sequences dedicated to high- and low-AoI transmissions leads to significant reductions in average AoI, especially as the number of active users increases. Interestingly, this comes at the cost of reduced detection precision compared to standard setups. A special case dedicates all pilots to high-AoI users, further enhancing AoI reduction at the expense of discarding low-AoI packets.