کنفرانس مهندسی برق ایران

صفحه اصلی / سی و سومین کنفرانس بین المللی مهندسی برق

Agglomerative Hierarchical Clustering Based on Q-learning for D2D Communication in Public Safety Communication Networks

نویسندگان :

Sahel Alipour¹ Mohammad Mansour Kesargheh² Abdulhamid Zahedi³ Ghasem Mirjalily⁴ Jamshid Abouei⁵

1- Yazd university 2- Iran University of Science and Technology (IUST) 3- Kermanshah Universiy of Technology 4- Yazd university 5- Yazd university

کلمات کلیدی :

Public safety communication network (PSCN)،D2D communication،Power control،Agglomerative hierarchical clustering (AHC)،Q-learning algorithm

چکیده :

When a disaster situation happens, communication networks often fail due to damaged base stations (BSs), significantly impacting emergency services. This paper presents an efficient machine learning-based hierarchical clustering method to enhance device-to-device (D2D) communication within public safety communication networks (PSCNs). By deploying relay stations at the edges of adjacent functional cells, the proposed architecture ensures robust connectivity between functional and disaster regions, enabling efficient communication for rescue teams and survivors. This approach dynamically forms clusters based on the agglomerative hierarchical clustering (AHC) method where using the Q-learning algorithm, the power control approach is presented to tackle the interference. Simulation results demonstrate the effectiveness of the proposed clustering scheme which improves the scalability and energy efficiency by reducing user power consumption and enhancing communication reliability in emergency scenarios.