Age of Information Optimization by Deep Reinforcement Learning for Random Access in Machine Type Communication

Jeong, Minseok; Seo, Giup; Hwang, Euiseok

doi:10.1109/bigdata55660.2022.10020660

Cited by 2 publications

(1 citation statement)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It was demonstrated that compared with FCFS strategy, the LCFS strategy has less average AoI because it always chooses to transmit the latest status updates. The work of Jeong et al 25 proposed a novel deep reinforcement learning scheme to directly optimize AoI for machine‐type communication in a slotted ALOHA RA channel. By taking into account urgency and packet age as extra local information for a reward, it was achieved that AoI could be improved while preserving the throughput.…”

Section: Introductionmentioning

confidence: 99%

Age of error information and throughput for truncated and layer‐coded HARQ‐based satellite‐IoT systems

Zhao,

Jia,

Xie

et al. 2023

Int J Communication

View full text Add to dashboard Cite

SummaryThe work focuses on the simultaneous improvement on the throughput and information freshness of the promising satellite‐based Internet of Things (S‐IoT). To this end, we first propose a method based on the adaptive modulation coding (AMC) and truncated layer‐coded hybrid automatic repeat request (L‐HARQ) strategies by especially considering the difference in the channel state information (CSI) between the transmitter and receiver. On account of the large signal delay of the S‐IoT systems, this paper establishes the channel model based on the correlation of the signal‐to‐noise ratio (SNR) under the outdated channel and the ideal channel. Secondly, considering the fact that the improved backtrack decoding is used in the L‐HARQ, the classical definition of age of information (AoI) is insufficient to quantify the status update freshness, we propose the age of error information (AoEI), which is defined as the elapsed time since the generation of the lastly successfully backtrack‐decoded status update. Thirdly, we characterize the statistical descriptions of the maximum wait time of the successfully recovered status updates and the interdeparture time between the two consecutive status update packets that are lastly successfully received. Then, the closed‐form expression of average AoEI is derived under the assumption of independent and identically distributed fading channels. Finally, the Markov decision process (MDP) framework is used to optimize the packet parameters in each round of L‐HARQ process under the general fading scenario. Thereby, the optimized transmission scheme that minimizes AoI with throughput constraint is solved. The simulations investigate the effect of system parameters on the average AoEI.

show abstract

Section: Introductionmentioning

confidence: 99%