Modelling and Analysis of Age of Information in IoT Networks

Abstract

Rapid rise in the number of communication devices for diverse applications with the advancement in network efficiency, has transformed the lifestyle of humans. The fifth-generation network endeavor to provide better capac ity, better mobility support, lower latency, and increased coverage area as compared to the services provided by the predecessor. These advancements provoke development of more applications for human ease including intelli gent transportation systems, smart agriculture, smart healthcare to name a few. Most of these applications related to the status update systems demand lower delays in delivering the information from the generation to the monitor for their smooth functioning. The time elapsed since the generation of an information till its delivery to the monitor is termed as Age of information (AoI), which is a critical metric in status update systems as the delivery of stale information in such systems is futile. Existing works attempted various techniques to preserve information freshness in internet of thing (IoT) sys tems exploiting various queuing techniques, access schemes and scheduling algorithms. However, to fill the room, in this thesis AoI in a smart agricul ture area is analyzed by optimizing the trajectory of the mobile reader. As the path taken by the reader while collecting information from the deployed sensors in the area is directly proportional to the propagation time of infor mation travelling from the nodes to the reader. Similarly, AoI and packet viii ix drops probability are optimized by managing the queuing and scheduling policy in transmission of information from the nodes to the monitor. For this purpose, the queuing in managed according to the arrival probability at nodes and the priority of each node using Markov decision process (MDP). MDP optimizes the scheduling of nodes by scheduling the right node at the right time according to their priorities. NOMA on the other hand can trans mit multiple packets in a single time slot optimizing AoI and probability of packet drops at the cost of increased outage probability as compared to OMA. Therefore, a hybrid OMA and NOMA approach is used to further optimize the AoI and packet drops in an IoT network. The results show that the proposed technique outperforms previous techniques in minimizing AoI and packet drops in IoT systems.