Enhanced Resource and Mobility Management for Non-Terrestrial Networks
- 주제(키워드) resource management , mobility management , NTN
- 주제(DDC) 004.6
- 발행기관 아주대학교 일반대학원
- 지도교수 Jae-Hyun Kim
- 발행년도 2025
- 학위수여년월 2025. 2
- 학위명 박사
- 학과 및 전공 일반대학원 AI융합네트워크학과
- 실제URI http://www.dcollection.net/handler/ajou/000000034437
- 본문언어 영어
- 저작권 아주대학교 논문은 저작권에 의해 보호받습니다.
초록/요약
Non-terrestrial networks (NTN) are regarded as a key en- abler for enhancing the performance of next-generation wire- less communication networks. It can expand coverage, en- hance frequency efficiency, and reduce digital divide. Although the NTN has many advantages, it faces challenges such as propagation delay, path loss, channel estimation, Doppler ef- fect, resource management, and mobility management. Also, the existing resource and mobility management schemes have been developed to suit terrestrial networks (TN) environ- ment. Therefore, this dissertation proposed resource and mo- bility management to address these challenges for the NTN. First, a group contention-based (GC) medium access control (MAC) protocol for guaranteeing full-duplex (FD) pair in unmanned aerial vehicle (UAV) relay network is proposed. This protocol is modified based on the carrier sense multi- ple access with collision avoidance (CSMA/CA) protocol to support FD communication. If the UAV informs destination node, the nodes capable of forming an FD pair with it con- tend for transmission opportunities. The performance results confirm that the proposed MAC protocol has better perfor- mance compared to conventional MAC protocols. Second, a HO scheme using multi-agent deep reinforcement learning to decide the optimal HO for air-to-ground (ATG) network is proposed. When the aircraft has fast speed compared with the TN platforms, the channel condition is fast changed. It is difficult to decide the basic handover (BHO). Therefore, to address this issue, the novel HO protocol is needed. It consid- ers not only signal strength but also additional factors such as the number of HO, unnecessary HO (UHO), radio link fail- ure (RLF), and fairness. As the training epochs progress, the base station (BS) conducts optimal HO decision. The perfor- mance results show that the proposed HO enhances the HO performance. Lastly, the novel HO scheme that combines ran- dom access channel (RACH)-less HO (RHO) and conditional HO (CHO) is proposed for low Earth orbit (LEO) satellite. If the execution timing of node and the uplink (UL) grant timing of cell is uncertain, the signaling overhead increases. Therefore, to address this issue, the prediction of UL grant start timing algorithm is proposed. It predicts UL grant start timing using antenna gain of the serving and the potential target cells without additional message. The performance re- sults show that the proposed HO improves HO performance without degradation. Furthermore, the prediction algorithm demonstrates that the prediction of exact timing is possible.
more목차
1 Introduction 1
1.1 Background and motivation 1
1.2 Contributions 4
1.3 Overview of dissertation 7
2 Related work 8
2.1 UAV communication techniques 8
2.2 HO schemes for high speed vehicles 11
2.3 LEO satellite HO schemes 13
3 Group Contention-based MAC Protocol for Guaranteeing FD Pair in UAV Relay Network 16
3.1 GC MAC protocol 20
3.1.1 System model 20
3.1.2 Establishment of contention group table 20
3.1.3 Novel UACK packet for GC 23
3.1.4 Procedure of GC MAC protocol 23
3.2 Design of analytical model 26
3.3 Performance evaluation 29
3.4 Concluding remarks 35
4 Attention-based Multi Agent Deep Q-Network Handover for ATG Network 36
4.1 ATG network 38
4.1.1 System model 38
4.2 A-MADQN HO for ATG network 42
4.2.1 Problem formulation 42
4.2.2 MDP definition 43
4.2.3 MADQN algorithm for HO decision 45
4.2.4 Attention mechanism 48
4.2.5 Procedure of A-MADQN HO scheme 50
4.3 Performance evaluation 51
4.4 Concluding remarks 59
5 RACH-less Conditional Handover with Timing Prediction for Intra-LEO Satellite Network 60
5.1 LEO satellite network 62
5.1.1 System model 62
5.1.2 Satellite channel model 62
5.2 HO enhancements in 3GPP 65
5.2.1 Overview of RHO 65
5.2.2 Overview of CHO 66
5.3 Proposed RCHO for LEO satellite 66
5.3.1 Procedure of RCHO 66
5.3.2 UL grant start timing information 70
5.3.3 Prediction of UL grant start timing 70
5.3.4 Prediction algorithm of UL grant start timing 71
5.4 Performance evaluation 72
5.5 Concluding remarks 80
6 Conclusion 82
References 85
