RL-based Distributed Parametric Resource Allocation Scheme for Multi-Hop IAB Networks

As the communication demand increases and base stations (BSs) are deployed more densely, the traditional fiber backhaul will increase the cost of network construction and limit the scalability of the network. Especially for mountainous areas and suburban rural areas, using wireless backhaul links instead of the traditional fiber backhaul is an attractive scheme. Compared with traditional integrated access and backhaul (IAB) networks, this paper studies resource allocation in two-hop IAB networks based on layer division multiplexing (LDM) and multi-connection (MC). IAB-donor obtains data from the core network (CN) and transmits it to the corresponding UE groups through the two-hop IAB network. Since the resource allocation problem is intractable by conventional algorithms, a distributed and parameterized framework is proposed. Simulation results show that the proposed two-hop IAB resource allocation algorithm based on LDM and MC can achieve higher capacity than traditional IAB networks.