6G and Beyond Dense Network Deployment: A Deep Reinforcement Learning Approach

Integrated Access and Backhaul (IAB) networksoffer a versatile and scalable solution for expanding broadbandcoverage in urban environments. However, optimizing the deploy-ment of IAB nodes to ensure reliable coverage while minimizingcosts poses significant challenges, particularly given the locationconstraints and the highly dynamic nature of urban settings. Thiswork introduces a novel Deep Reinforcement Learning (DRL)approach for IAB network planning, considering urban con-straints and dynamics. We employ Deep Q-Networks (DQNs) withaction elimination to learn optimal node placement strategies.Our framework incorporates DQN, Double DQN, and DuelingDQN architectures to handle large state and action spaces ef-fectively. Simulations across various initial donor configurations,including five-dice, vertical, and pentagon patterns, demonstratethe superiority of our DRL approach. The Dueling DQN achievesthe most efficient deployment, reducing node count by an averageof 12.3% compared to a heuristic method. This study highlightsthe potential of advanced DRL techniques in addressing complexnetwork planning challenges, offering an efficient and adaptivesolution for IAB deployment in diverse urban environments.