To address the substantial measurement challenges in characterizing self-interference (SI) intensity for full-duplex (FD) large-scale arrays, we propose an innovative prediction method using Vision Transformer (ViT) and transfer learning. The proposed model integrates array multi-modal features and employs multi-head attention mechanisms to accurately predict SI intensity in FD array systems with limited measured data. The proposed transfer learning strategy is composed of a pre-training phase based on the spherical-wave model and a fine-tuning phase utilizing limited measured data, which further enhances the predictive performance of the proposed model. Our experimental evaluation in the upper 6 GHz (U6G) band (6675 ∼ 6875 MHz) shows that, compared to traditional methods, the proposed method achieves approximately 16% and 10% improvements in Structural Similarity Index Measure (SSIM) and the proportion of predictions with errors below 3 dB, respectively. These results confirm its effectiveness for SI intensity prediction in FD arrays and indicate its potential for supporting beam management and optimization in B5G/6G FD systems.

Space-ground integration is a key component of 6G with diverse topologies, time-varying topologies, broad geographical and temporal scales, and limited spatial node resources, bringing new development opportunities and scope for mobile communications. Before designing the 3D 6G space-ground integrated network architecture, we first examine the main difficulties of space-ground integrated networks, and then examine the common use cases and requirements of 6G-oriented space-ground integrated networks. Finally, we propose a space-ground integrated network architecture based on a decentralized core network and a ground-centralized core network. For the former, the core network is placed on the ground while the airborne platform performs certain access network activities. To meet the needs of various application scenarios, the latter airborne platform performs access network operations and loads a portion of the core network components. Dynamic deployment of key network functions becomes possible as the network functions are reconfigured to meet the changing application scenario needs and network operation requirements.

The High Altitude Platform Station (HAPS), as a as an NTN (Non-Terrestrial Network) network node, has evolved as an important component of the 6G wireless network. Acting as an IMT base station (HIBS) in mobile services, HAPS is a promising way to achieve vertical heterogeneous coverage enhancement. HIBS can operate in the stratosphere up to 20 km high to provide more coverage, however, it is limited by the ability to accommodate more users in scarce spectrum resources. To improve the SE of the uplink, we propose a MIMO-assisted HIBS technique in this study. A Monte Carlo method is used to calculate the CDF curves of the system SE for each configuration, which makes the simulation results closer to the actual deployment. HIBS is configured with three different combinations of MIMO antenna arrays, including single antenna, 2x2MIMO+ 2x4MIMO, and 4x4MIMO+ 4x8MIMO. The simulation results show that appropriately sized MIMO antenna panels can significantly improve the HIBS uplink link SE.

Reconfigurable Intelligent Surface (RIS), is a technology that uses a meta-surface to reflect signals to the user. the use of RIS as a passive full-duplex relay to support wireless communications is a powerful means of providing a potential 6G technology with the unique qualities of low cost, low energy consumption, good reliability and high capacity. Active fullduplex relaying is a similar technology that has evolved from the half-duplex relaying known today and is also a potential 6G technology. both RIS and full/half-duplex relaying act as nodes between source and destination, how can we pick between the two to improve system performance? In this study, we examine the performance of these two technologies from the operator’s perspective. To provide a deterministic calculation for the comparison, we derive closed expressions for the reach rates of RIS and active relaying and propose an evaluation model to make the comparison fair. The results of the study show that RIS and different relaying methods, especially RIS and full-duplex relaying, two newly developed techniques, are candidates for 6G and have significant advantages in overcoming the deterioration of NLOS routing.

The High Altitude Platform Station (HAPS), as a network node in the Non-Terrestrial Network (NTN), has evolved as an important component of the 6G wireless network. Acting as an IMT base station (HIBS) in mobile services, HAPS is a promising approach to achieve vertical heterogeneous coverage enhancement. HIBS can operate in the stratosphere up to 20 km high to provide more coverage, however, it is limited by the ability to accommodate more users in scarce spectrum resources. In our study, a MIMO-assisted HIBS approach based on a 7-cell network structure is proposed to improve the spectral efficiency (SE) of the downlink. HIBS is configured with three different combinations of MIMO antenna arrays, including single antenna, 2x2MIMO+2x4MIMO, and 4x4MIMO+4x8MIMO. A Monte Carlo method is used to calculate the system SE for each configuration of the CDF curve, which brings the simulation results closer to the actual deployment. The simulation results show that the downlink SE of the system may be significantly improved by using MIMO-assisted HIBS based on a 7-cell scheme, and that the downlink SE can be further improved by scaling up the MIMO array.

we present a semi-intelligent antenna based on pattern reconfigurable dual-polarized Dielectric Resonator Antenna (DRA) for indoor WLAN base station. The DRA is the air-gap DRA which feed by a traditional 3dB hybrid coupler with 1/4λ delay line. Tow of this network are utilized to realize this antenna, and selecting one of the four ports can be implemented a boresight or cardioid radiation pattern on the vertical or horizontal direction of polarization independently. It is the first known antenna of one element to be reconfigurable two patterns in two polarization directions simultaneously. The pattern reconfigurable DRA operate around 5-6GHz, reconfigurable radiation patterns can be achieved in dual-polarizations. The relative impedance bandwidth (S11<-10dB) of the four radiation patterns is 30.9%, 31.4%, 14.8% and 18.2% respectively. This antenna has moderate gain between 6–8 dBi. The configuration of multi-port switching multi-mode makes the proposed antenna a suitable candidate for the semi-intelligent antenna of IEEE 802.11ad. Due to the easy switching way, the proposed antenna can change the beam according the hotspots prediction through the machine learning algorithm to achieve semi-intelligence.