Adaptive Retransmission Time Out in Flying Ad-Hoc Network By LSTM Machine Learning: Round Trip Time Prediction

Mirzaeinia, Amir; Mirzaeinia, Mehdi; Shekaramiz, Mohammad; Hassanalian, Mostafa

doi:10.2514/6.2020-0053

Cited by 2 publications

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recurrent Neural Networks (RNN) are able to perform their tasks based on some old information recorded in their memories. They are similar to a network that has a loop in them in order to keep track of old information [4].LSTM have the sufficient ability to solve the problem of long-term dependencies which general RNNs (Vanilla RNNs) cannot learn for the prediction [5]. The Gated Recurrent Unit (GRU), another type of RNN, offers a simpler structure compared to LSTM.…”

Section: Introductionmentioning

confidence: 99%

Predicting Future Aircraft Emissions in Indonesia: A Comparative Analysis using LSTM and GRU Methods

Mahendra,

Hannim

et al. 2024

Proceedings of the 2nd International Conference on Aviation Industry, Education, and Regulation, AVINER 2023, 8-9 November 2023

View full text Add to dashboard Cite

The aviation industry in Indonesia continues to expand resulting in concerns over the environmental impact of aircraft emissions have become paramount. This article presents a comprehensive analysis of future emission predictions using two advanced time series forecasting methods, Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU), applied to historical aircraft emission data (Q3 2022 to Q3 2023). Leveraging the sequential nature of the data, LSTM and GRU networks are harnessed to model the intricate temporal dependencies and inherent seasonality present in the emission time series. The evaluation encompasses various performance metrics, including Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R-squared (R²), to gauge the models' predictive capabilities across different forecasting horizons. Results indicate that both LSTM and GRU methods demonstrate promising forecasting capabilities, outperforming traditional time series models. However, subtle distinctions emerge in their predictive efficiency. LSTM exhibits superior performance in capturing long-term dependencies and handling complex emission patterns, whereas GRU showcases efficiency in shorter forecasting horizons. Remarkably, the research uncovers the profound impact of ML, with the Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) methods emerging as the most potent tools with an accuracy reached up to 87%.

show abstract

Section: Introductionmentioning

confidence: 99%

Predicting Future Aircraft Emissions in Indonesia: A Comparative Analysis using LSTM and GRU Methods

Mahendra,

Hannim

et al. 2024

Proceedings of the 2nd International Conference on Aviation Industry, Education, and Regulation, AVINER 2023, 8-9 November 2023

View full text Add to dashboard Cite

show abstract

“…Furthermore, it requires a thorough understanding of the application limitations such as the low resources of UAVs in terms of computation and energy [16] . Additionally, RL provides resilience towards various environmental hazards, such as weather changes or storms, when the design of elements is hazards-aware [17].…”

Section: Introductionmentioning

confidence: 99%

Reinforcement-Based Clustering in Flying Ad-Hoc Networks for Serving Vertical and Horizontal Routing

Abdulhae,

Mandeep,

Islam

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Unmanned aerial vehicles (UAVs) are receiving increasing attention due to their wide range of applications. Flying ad-hoc networks (FANETs) enable inter-communication between UAVs, but their dynamics and fast changes in mobility and topology make it challenging to achieve stable routing. Clustering based FANETs provide a self-organizing approach for routing, but clustering is a decentralized process based on different utilities that can affect performance. In this work, we propose a centralized agent placed in a remote ground station to adjust the role of each utility in the cluster-head arbitration. Our approach uses six utilities, including node centrality, residual energy, link holding time, velocity similarity, buffer occupancy, and diversity. The developed reinforcement learning (RL)-clustering approach enables routing over multiple planes, with each plane representing a network cell with a different coverage radius of nodes, including Femto, Pico, and Micro planes. We propose a novel reward formulation for effective learning of agents that includes indicators of stability based on role change, energy consumption, and confirmation message through the selected node. For evaluation, the developed RL based agent was evaluated on three types of agents, namely, Q-learning, DQN, and DDPG and it was compared with random agent. Results have shown the superiority of our developed RL based clustering in terms of stability, energy consumption and most network metrics.

show abstract

Adaptive Retransmission Time Out in Flying Ad-Hoc Network By LSTM Machine Learning: Round Trip Time Prediction

Cited by 2 publications

References 19 publications

Predicting Future Aircraft Emissions in Indonesia: A Comparative Analysis using LSTM and GRU Methods

Predicting Future Aircraft Emissions in Indonesia: A Comparative Analysis using LSTM and GRU Methods

Reinforcement-Based Clustering in Flying Ad-Hoc Networks for Serving Vertical and Horizontal Routing

Contact Info

Product

Resources

About