Work

3次元noma無線網のためのaiを活用した物理層とmac層の最適化法

Year: 2023
Type: dissertation
Abstract: On the other hand, to unlock the full potential of NOMA and reap its benefits, MAC layer-level resource management operation optimization is crucial.Specifically, proper allocation of transmission power and selection of candidate users for pairing over the same resource block are critical for an efficient v utilization of the available resources.Reinforcement learning deploying double deep-Q networks (DDQN) is a promising framework that can be adopted for tackling the problem.In this thesis, an RL-based DDQN scheme is proposed for opportunistic user pairing to access downlink NOMA systems with capacity-limited backhaul link connections.The proposed algorithm relies on proactive data caching to alleviate the throttling caused by backhaul bottlenecks, and optimized UE selection and power allocation are accomplished through the continuous interaction between an RL agent and the NOMA environment to increase the overall system throughput.Moreover, for non-terrestrial 3D-NOMA operation, this thesis proposes two energy-efficient RL-based algorithms for millimeter wave (mmWave)-enabled unmanned aerial vehicle (UAV) communications toward 6G.This can be especially useful in ad-hoc communication scenarios within a neighborhood with main-network connectivity problems such as in areas affected by natural disasters.This is important since 6G mobile network design must accommodate use cases for mission-critical communications during emergencies.Dynamic selection of suitable hovering spots within the target zone where the batteryconstrained UAV needs to be positioned as well as calibrated NOMA power control with proper device pairing are critical for optimized performance.We propose cost-subsidized multiarmed bandit (CS-MAB) and DDQN-based solutions to address these problems jointly.For 3D-NOMA PHY layer, the proposed AI-aided OFDM PAPR reduction scheme operates at 16.7% faster signal processing speed along with about 12% improvement in the cubic metric measure of the OFDM signal compared to conventional clipping and filtering (SCF) for a 256-subcarrier OFDM PHY layer with QPSK modulation.On the other hand, at the MAC layer level, an 18.4% faster data transfer rates can be accommodated by the proposed DDQN RL agent for cache-enabled NOMA downlink resource management over a capacity-limited backhaul link when operating with 38-GHz mmWaves and 500 MHz system bandwidth.Moreover, significant UAV battery energy savings exceeding 90% can be accomplished for mission-critical, non-terrestrial NOMA downlink transmissions.vi
Institution: Kyushu University
Language: English
Open Access status: green
OpenAlex ID: W7146265100
Cites: 143
Cited by:
Related to: 10