Title: Multi Agent DeepRL based Joint Power and Subchannel Allocation in IAB networks Show Chinese title

Title: 基于多智能体深度强化学习的IAB网络联合功率和子信道分配

Abstract: Integrated Access and Backhauling (IAB) is a viable approach for meeting the unprecedented need for higher data rates of future generations, acting as a cost-effective alternative to dense fiber-wired links. The design of such networks with constraints usually results in an optimization problem of non-convex and combinatorial nature. Under those situations, it is challenging to obtain an optimal strategy for the joint Subchannel Allocation and Power Allocation (SAPA) problem. In this paper, we develop a multi-agent Deep Reinforcement Learning (DeepRL) based framework for joint optimization of power and subchannel allocation in an IAB network to maximize the downlink data rate. SAPA using DDQN (Double Deep Q-Learning Network) can handle computationally expensive problems with huge action spaces associated with multiple users and nodes. Unlike the conventional methods such as game theory, fractional programming, and convex optimization, which in practice demand more and more accurate network information, the multi-agent DeepRL approach requires less environment network information. Simulation results show the proposed scheme's promising performance when compared with baseline (Deep Q-Learning Network and Random) schemes. Show Chinese abstract

Abstract: 集成接入和回传（IAB）是一种可行的方法，用于满足未来一代对更高数据速率的前所未有的需求，作为密集光纤有线链路的成本效益替代方案。 在约束条件下设计此类网络通常会导致非凸和组合性质的优化问题。 在这些情况下，获得联合子信道分配和功率分配（SAPA）问题的最优策略具有挑战性。 在本文中，我们开发了一个基于多智能体深度强化学习（DeepRL）的框架，用于IAB网络中功率和子信道分配的联合优化，以最大化下行数据速率。 使用DDQN（双深度Q学习网络）的SAPA可以处理与多个用户和节点相关的计算成本高昂、动作空间庞大的问题。 与传统方法如博弈论、分数规划和凸优化不同，这些方法在实践中需要越来越多的准确网络信息，而多智能体DeepRL方法所需的环境网络信息较少。 仿真结果表明，与基线（深度Q学习网络和随机）方案相比，所提出的方案表现出良好的性能。