Dynamic relay selection for AoI minimization in short packets based full-duplex energy harvesting-IoT networks

The burgeoning demand for real-time status updates in the Internet of Things (IoT) underscores the importance of the age of information (AoI). While full-duplex (FD) relaying and energy harvesting (EH) are promising techniques for extending network lifetime and improving spectral efficiency, their joint impact on information freshness remains poorly understood, especially in the presence of residual self-interference, stochastic energy supply, and short-packet transmissions. Unlike existing AoI-related works that focus on half-duplex relaying, fixed relay participation, or infinite blocklength assumptions, this paper investigates the AoI performance of a multi-relay EH-IoT network operating in FD mode under the finite blocklength regime. To overcome the excessive waiting delay inherent in conventional multi-relay cooperation, we propose a novel earliest- partial relay cooperation scheme, in which only the first relays that successfully decode a status update participate in forwarding. This design fundamentally reshapes the delay-reliability trade-off in AoI-oriented systems. Building on this scheme, we develop a unified analytical framework that jointly captures stochastic energy arrivals, imperfect self-interference cancellation in FD relays, co-channel interference, and short-packet communications. By leveraging order statistics, we derive a closed-form expression for the average AoI, explicitly revealing how packet length, relay selection size, and residual interference interact to determine information freshness. Numerical results validate the analysis and demonstrate that the average AoI exhibits non-monotonic behavior with respect to both packet length and. Moreover, a key insight is that FD relaying may be inferior to half-duplex relaying unless self-interference is sufficiently suppressed, providing important design guidelines for practical status update systems.