Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Emerging Science Journal
Nana Rachmana Syambas
School of Electrical Engineering and Informatics, Bandung Institute of Technology (ITB), Bandung 40132
Environmental Science

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Dual-Agent Q-Learning for Cross-Layer IEEE 802.11bd Optimization in Dense VANETs Galih Nugraha Nurkahfi; Suyoto; Agus Subekti; Budi Prawara; Ratna Mayasari; Andy Triwinarko; Nasrullah Armi; Eueung Mulyana; Nana Rachmana Syambas
Emerging Science Journal Vol. 10 No. 3 (2026): June
Publisher : Ital Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.28991/ESJ-2026-010-03-019

Abstract

Dense vehicular ad hoc networks face critical challenges in reliably delivering safety messages due to channel congestion, packet collisions, and interference. This study develops a dual-agent Q-learning framework for cross-layer IEEE 802.11bd optimization to improve latency and power efficiency while maintaining acceptable packet delivery ratios in dense traffic. We propose a decomposed architecture separating PHY-layer power control and MAC-layer beacon rate adaptation, with deterministic SINR-based MCS selection ensuring IEEE 802.11bd compliance. The framework is evaluated using a Python-based VANET simulator implementing the IEEE 802.11bd PHY/MAC stack with realistic SUMO mobility, multi-class background traffic, and omnidirectional/sectoral antennas across 20-90 vehicles/km densities. Results show dual-agent Q-learning reduces average latency by 44.6% (31.1ms to 17.2ms) and transmission power by 55% (15-20dBm to 9dBm) compared to static baselines, with acceptable 5-11% PDR reduction (94.2% to 88.6%). The approach converges within 8,500 episodes, significantly faster than single-agent Q-learning (12,500) and dual-agent DQN (14,000-35,000). This work introduces the first dual-agent tabular Q-learning for joint power-rate-MCS optimization in IEEE 802.11bd VANETs, demonstrating that agent decomposition reduces state-action complexity while enabling interpretable, fast-converging control suitable for sub-100ms vehicular applications.
Imbalanced Learning for Multiclass Fault Prediction in Optical Networks Using Enhanced Hybrid Sampling Rohmat Tulloh; Nana Rachmana Syambas; Eueung Mulyana
Emerging Science Journal Vol. 10 No. 5 (2026): October
Publisher : Ital Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.28991/ESJ-2026-010-05-08

Abstract

Fault prediction in optical networks faces the problem of class imbalance, where failure event data is much less than normal event data. Conventional classification models typically used in balanced or low-imbalance conditions tend to be biased toward the normal class. Sampling techniques such as oversampling tend to increase overlap. This causes high performance values in imbalanced classes not always to reflect the model's true ability to detect minority classes. This paper proposes a new framework called SMOTE-ENN-CatBoost-Optuna. This approach combines a hybrid sampling technique using synthetic minority oversampling (SMOTE) and edited nearest neighbors (ENN), controlled by Categorical Boosting (CatBoost), with hyperparameter tuning using Optuna. The proposed method dynamically adjusts the oversampling rate based on model error, followed by noise filtering to refine class boundaries. An iterative training strategy with Matthews correlation coefficient (MCC)-based convergence is further introduced to stabilize performance. To assess the proposed model's ability to adapt to varying levels of imbalance, imbalanced classes were grouped into three categories (medium, high, and extreme) based on the Imbalance Ratio (IR). Experimental results show that the proposed model consistently outperforms baseline methods. It achieves improvements of up to 0.9944 in MCC and 0.9951 in Macro F1-score. These results indicate the framework effectively enhances fault-detection reliability and maintains robust performance across most classes.
Co-Authors Agus Subekti Andy Triwinarko Budi Prawara Eueung Mulyana Galih Nugraha Nurkahfi Nasrullah Armi Ratna Mayasari Rohmat Tulloh Suyoto