DEEP LEARNING AIDED FLOW CONTROL MECHANISM SELECTION

Authors : Merih Leblebici, Ali Çalhan

Pages : 10-20

Doi:10.22531/muglajsci.1707304

View : 74 | Download : 174

Publication Date : 2025-12-31

Article Type : Research Paper

Abstract :The rapid growth in global mobile data traffic demands the development of advanced communication systems such as the sixth generation (6G). Integrated sensing and communication (ISAC) is recognized as a pivotal technology within the 6G, enabling simultaneous information transmission and environmental awareness to enhance spectrum efficiency and fulfill the requirements of future applications. This study presents a thorough cross-layer framework for ISAC that integrates physical and medium access control (MAC) layer processes, overcoming the limitations of traditional, isolated approaches. The framework incorporates deep learning models for real-time signal-to-noise ratio (SNR) prediction and flow control mechanism, leveraging fuzzy logic to dynamically adjust automatic repeat request (ARQ) mechanisms based on SNR, bandwidth, and delay. Through detailed analyses of fuzzy control surfaces, the study proves the ability of this system in optimizations of resource allocation, adaptation in a dynamic environment, and achieving a balance in reliability and efficiency. Results confirm that SNR dominates ARQ decision-making, while bandwidth and delay significantly influence performance under certain conditions. These findings validate capability of the fuzzy inference system to enable intelligent communication systems and establish ISAC as a foundational component for 6G networks.
Keywords : Akış Kontrolü, Derin Öğrenme, 6G, ISAC

ORIGINAL ARTICLE URL