# Traffic Prediction Based Fast Uplink Grant for Massive IoT

arXiv: https://arxiv.org/abs/2008.02207 · HTML: https://mlsanity.com/paper/2008.02207/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Aug 2020.

## Built on by (most cited first, 4 of 4)

- [A Learning-Based Fast Uplink Grant for Massive IoT via Support Vector Machines and Long Short-Term Memory](https://mlsanity.com/paper/2108.10070/) (2021)
- [Energy-Efficient Wake-Up Signalling for Machine-Type Devices Based on Traffic-Aware Long-Short Term Memory Prediction](https://mlsanity.com/paper/2206.06058/) (2022)
- [Coordinated Random Access for Industrial IoT With Correlated Traffic By Reinforcement-Learning](https://mlsanity.com/paper/2109.08389/) (2021)
- [RACH Traffic Prediction in Massive Machine Type Communications](https://mlsanity.com/paper/2405.05235/) (2024)

## Abstract (arXiv)

This paper presents a novel framework for traffic prediction of IoT devices activated by binary Markovian events. First, we consider a massive set of IoT devices whose activation events are modeled by an On-Off Markov process with known transition probabilities. Next, we exploit the temporal correlation of the traffic events and apply the forward algorithm in the context of hidden Markov models (HMM) in order to predict the activation likelihood of each IoT device. Finally, we apply the fast uplink grant scheme in order to allocate resources to the IoT devices that have the maximal likelihood for transmission. In order to evaluate the performance of the proposed scheme, we define the regret metric as the number of missed resource allocation opportunities. The proposed fast uplink scheme based on traffic prediction outperforms both conventional random access and time division duplex in terms of regret and efficiency of system usage, while it maintains its superiority over random access in terms of average age of information for massive deployments.