PROJECT TITLE :

Traffic Prediction and Fast Uplink for Hidden Markov IoT Models

ABSTRACT:

In this work, we present a novel framework for the traffic prediction and fast uplink (FU) capabilities of Internet of Things (IoT) networks that are controlled by binary Markovian events. First, we apply the forward algorithm with hidden Markov models (HMMs) in order to schedule the available resources to the devices with maximum likelihood activation probabilities using the FU grant. This is done so that we can maximize the likelihood that an activation will occur. In order to evaluate the accuracy of the prediction, we also take into account the regret metric by counting the number of transmission slots that were unused. The next step is to construct an optimization problem for fairness, with the goal of reducing the Age of Information (AoI) while maintaining as little regret as is humanly possible. Finally, we present an iterative algorithm for estimating the model hyperparameters (activation probabilities) in a real-time application and apply an online-learning version of the proposed traffic prediction scheme. Both of these are presented as a conclusion to this paper. The results of the simulations show that the proposed algorithms outperform baseline models such as time-division multiple access (TDMA) and grant-free (GF) random-access in terms of the amount of regret, the efficiency of system usage, and the area of influence (AoI).