Statistical AoI Minimization: Sampling and Transmission Adaption to Wireless Fading Channels

Wireless status update aims to keep destination node realizing real-time status at source node as timely as possible by continuously sampling, packaging, and transmitting the status information from the source to the destination. Statistical age-of-information (AoI) is a novel metric in a long-term manner to evaluate the freshness of information, which bridges the gap between average peak AoI and maximum peak AoI. Both sampling interval and the transmission time of each status packet have a significant impact on statistical AoI. Different from the existing works where the transmission time for one status packet is fixed or equal to the sampling interval, we in this paper consider the case where both sampling interval and transmission time can be jointly optimized to lower statistical AoI. Toward this end, we formulate the statistical AoI minimization problem over wireless fading channels, where the average power constraint is considered. To tackle this nonconvex problem, the Dinkelbach's transform nested with block coordinate descent (BCD) method is adopted. Then, an adaptive-sampling-interval and adaptive-transmission-time (ASIATT) scheme is obtained, which suggests that the transmission time should be smaller than sampling interval and then equal to it as channel power gain increases. Therefore, sometimes the channel enters into the idle state. Numerical results validate the superiority of the ASIATT scheme.