Forming a Two-Tier Heterogeneous Air-Network via Combination of High and Low Altitude Platforms

$Q$-Learning Aided Intelligent Routing With Maximum Utility in Cognitive UAV Swarm for Emergency Communications

“…More recently, there has been increasing interest in fusing HAPs, UAVs, and ground networks for setting up the threelayer aerial-ground integrated network architecture with different applications [21]- [25]. Ahmadinejad and Falahati in [21] designed an aerial heterogeneous wireless with an HAP and multiple UAVs serving as the quasi-stationary ABSs to provide radio access services to the GDs via the downlink orthogonal division multiple access.…”

“…More recently, there has been increasing interest in fusing HAPs, UAVs, and ground networks for setting up the threelayer aerial-ground integrated network architecture with different applications [21]- [25]. Ahmadinejad and Falahati in [21] designed an aerial heterogeneous wireless with an HAP and multiple UAVs serving as the quasi-stationary ABSs to provide radio access services to the GDs via the downlink orthogonal division multiple access. A non-orthogonal multiple access (NOMA)-enabled airborne access vehicular ad hoc networks (VANETs) architecture was proposed in [22] to provide reliable downlink communication services to vehicles by an HAP and several UAV relays via the decode-and-forward protocol.…”

“…A non-orthogonal multiple access (NOMA)-enabled airborne access vehicular ad hoc networks (VANETs) architecture was proposed in [22] to provide reliable downlink communication services to vehicles by an HAP and several UAV relays via the decode-and-forward protocol. Apart from the downlink access services brought by the hierarchical aerial-ground network architecture in the research efforts [21], [22], some related works [23]- [25] have studied the uplink transmission or offloading as well. In [23], Qin et al adopted an XAPS model with an HAP serving as a macro ABS and several LAPs serving as small ABSs to empower the clustered-NOMA systems in 6G heterogeneous Internet of Things (IoTs).…”

“…The above solutions [21]- [25] have laid a solid foundation on the three-layer aerial-ground integrated network architecture. To our best knowledge, the use of UAV self-organization and collaboration to create the UAV swarm for distributing sensing tasks has not been addressed in HAP-UAV hierarchical architecture and remains an appealing study.…”

Self-Evolving Integrated Vertical Heterogeneous Networks

“…In particular, intelligent network management seems to have the most dominant role in envisioning an integrated VHetNet architecture with self-evolving capability, i.e., SEI-VHetNet. Note that: 1) joint network optimization problems as described in Section IV are highly complex and coupled, so they are computationally prohibitive to be solved optimally; 2) almost all proposed analytical solutions are offline resulting in static network management; 3) the highly dynamic nature of integrated VHetNet architecture, which demands a realtime integration and coordination process between the tiers; 4) the life-cycle cost of running a mobile vertical network, which requires eliminating manual configuration of network elements at the time of deployment through dynamic and intelligent optimization and troubleshooting during operations; 5) the need for adaptive and real-time responses to novel and dynamic user service requests and to improve network Throughput ANN [180] Energy-efficiency MAB [159] Radio coverage Q-learning [181] Throughput and data traffic balance DL [160] QoS Q-learning [183] The number of offloaded tasks and resource allocation Multi-agant DDPG [161] Spectral efficiency DQN [184] Throughput and trajectory design DRL [162] Coordination of multiple UAVs Q-learning [185] Total energy consumption and path planning MDP [163], [165] Radio coverage Decentralized DRL [186] Sum power consumption and resource management FL [164] QoS Double Q-learning [187] QoE and path planning DDPG [166] Energy efficiency SMGD [188] Energy consumption and trajectory design DDPG performance and customer experience [144]. In Fig.…”