最新刊期
- 摘要：Conventional single-atom hybrid beamforming has limited representation capability in near-field extremely large-scale multiple-input multiple-output (XL-MIMO) systems because of spherical-wave propagation and spatial non-stationarity. A dual-resolution rate-oriented hybrid beamforming (DR²-HBF) algorithm is proposed. The algorithm constructs local objectives from the spectral-efficiency gradient, jointly represents each beam using two atoms and a constrained complex coefficient, and performs each update through two-stage Top-K candidate screening, hardware projection, and validation based on the original rate objective. Simulation results show that the proposed algorithm improves spectral efficiency under identical constant-modulus hardware constraints and enhances statistical robustness under imperfect channel state information. Results obtained with amplitude-and-phase-adjustable architectures further indicate that performance gains should be evaluated together with hardware cost.关键词：near-field communications;extremely large-scale multiple-input multiple-output;hybrid beamforming;dual-resolution representation;rate-oriented optimization21|3|0更新时间：2026-10-12
- 摘要：To address the issue that precise OSNR simulation in the design phase of 400G backbone networks involves numerous parameters and complex calculations, which hinders rapid evaluation of site placement schemes, this paper derives the explicit relationship between optical amplifier span length and the cumulative equivalent span count ∑Eeq based on the standard OSNR estimation model (58-model), and proposes two evaluation models: a precise calculation model and an engineering simplified model. The simplified model normalizes standard spans, enabling designers to perform preliminary screening through manual calculation using only span lengths. Validation with field data from five existing trunk lines shows that the precise model is fully consistent with exact OSNR calculations; the simplified model exhibits systematic over-conservatism across all samples, with ∑Eeq overestimated by approximately 8–20 equivalent spans compared to the precise model, causing multiple lines that could actually operate without regeneration to be misjudged as infeasible. Based on the error calibration results, a three-level criterion is established: "green zone for direct approval, yellow zone for initiating precise verification, and red zone for direct rejection" (for G.652.D: simplified ∑Eeq ≤ 22 for approval, 22–30 for verification, >30 for rejection; for G.654.E: ≤26 for approval, 26–30 for verification, >30 for rejection), with the yellow-red boundary value of 30 empirically derived from field samples. Designers can quickly identify scheme risks by calculating only the simplified ∑Eeq, avoiding investment waste caused by misjudgment. The method requires few parameters and is easy to calculate, making it suitable for rapid scheme comparison and selection of G.652.D/G.654.E fiber-based 400G PM-QPSK backbone projects in the feasibility study and preliminary design stages.关键词：400G backbone networks;optical amplifier site placement schemes;rapid assessment;equivalent span count;misjudgment avoidance;three-level criterion7|2|0更新时间：2026-10-10
-
Coherent-Source DOA Estimation in Hybrid Arrays via Spatial Covariance Structure Reorganization AI导读
摘要：To address the difficulty of directly acquiring the complete array covariance information in hybrid analog–digital (HAD) arrays with a limited number of radio-frequency (RF) chains, as well as the rank deficiency of the signal covariance matrix caused by coherent sources, this paper develops a spatial covariance structure reorganization (SCSR) method for HAD arrays. The proposed method first reconstructs the sample covariance matrix of the physical array in the element domain using multiple groups of switch–phase-shifter combining observations. Subsequently, a Toeplitz diagonal diffusion mechanism is introduced into the reconstructed covariance domain. Through cross-diagonal extraction, diffusion along the main-diagonal direction, and count-based fusion, a high-dimensional equivalent covariance matrix suitable for subspace decomposition is constructed. The multiple signal classification (MUSIC) algorithm is then employed for direction-of-arrival (DOA) estimation. Simulation results demonstrate that, for a 24-element array, SCSR preserves a well-defined spatial spectral structure in scenarios with 21 mixed coherent sources and with 11 fully coherent sources. Across different signal-to-noise ratio (SNR) levels, the proposed method generally achieves lower root-mean-square error (RMSE) than the benchmark methods and exhibits a lower performance breakdown threshold in the low-SNR region. These results indicate that the proposed SCSR method provides a practical covariance-domain processing approach for coherent multi-source DOA estimation in HAD arrays with limited RF chains.关键词：Hybrid analog–digital array;spatial covariance structure reorganization;coherent sources;direction-of-arrival estimation;MUSIC18|1|0更新时间：2026-09-30 - 摘要：Aiming at the problem that the EoSoO (Ethernet over SDH over ODU) services carried by the optical transport network (OTN) within Nanjing metropolitan area rapidly consumed the low-order cross-connection capacity of core network elements, the causes of capacity shortage were firstly analyzed. Then, based on the overall architecture of OTN digital twin, a digital twin functional model suitable for the optimization of low-order virtual container (VC) bearer resources was proposed. Under given topology and capacity constraints, the model was used to adjust unreasonable routes of existing services and to allocate bearer resources for incremental services according to optimization strategies. Meanwhile, an AI small model was introduced to perform time-series prediction based on the optimized resource consumption trend, so as to identify potential bottlenecks in advance. Finally, the twin capabilities were technically self-evaluated by combining standardization achievements such as OTN digital twin intelligence grading and evaluation methods. The evaluation results show that the model can effectively support the intelligent requirements of OTN operation and maintenance.关键词：OTN;low-order cross-connection capacity;digital twin;resource optimization;Intelligent Operation and Maintenance8|2|0更新时间：2026-09-30
-
Efficient 3D Antenna Radiation Pattern Reconstruction via Attention-Enhanced Feature Modulation AI导读
摘要：To address the difficulty of acquiring complete three-dimensional (3D) antenna radiation patterns for wireless simulation and network planning, an efficient reconstruction method based on attention-enhanced feature modulation was proposed. A lightweight robust loss function was designed, where structural consistency constraints were used to replace external autoencoder-based supervision, thereby reducing additional training overhead. Meanwhile, bidirectional cross-attention and feature modulation mechanisms were introduced to dynamically fuse horizontal and vertical slice features, thereby improving global reconstruction accuracy. Furthermore, the standard U-Net architecture was refined to achieve faster error convergence and higher reconstruction accuracy under a comparable parameter budget. Extensive simulations on a diverse antenna dataset demonstrate that the proposed method consistently outperforms both traditional analytical methods and existing artificial intelligence (AI) baselines in root mean square error (RMSE) and structural similarity index (SSIM) metrics, confirming its effectiveness for antenna radiation pattern reconstruction.关键词：antenna radiation pattern;deep learning;attention mechanism;U-Net11|0|0更新时间：2026-09-30 - 摘要：In optical-electrical decoupled architectures, OSNR estimation methods suffer from inadequate accuracy, poor cross-vendor scenario adaptability, and heavy reliance on manual maintenance. This paper investigates precise OSNR assessment techniques for autonomous optical networks and proposes a corresponding intelligent management and control scheme. Drawing on AI technologies, we first decouple the optical-layer and electrical-layer parameters, and introduce cross-domain adaptive alignment to mitigate heterogeneous shifts across equipment from different vendors. A hybrid OSNR estimation model combining Transformer and CatBoost is then constructed to achieve high-accuracy estimation. By embedding this model into an autonomous closed-loop mechanism, we establish a full-process autonomous system, covering data acquisition, OSNR prediction, anomaly identification, configuration optimization, iterative convergence, and report generation. Experimental results from a live network environment show that the model achieves an MAE of 0.381 dB in OSNR estimation, corresponding to a 78.99% improvement in estimation accuracy over the conventional coupled model, thus verifying the feasibility of this approach for supporting intelligent management and control of autonomous optical networks under an optical-electrical decoupled architecture.关键词：autonomous optical networks;optical-electrical decoupling;OSNR estimation;intelligent control13|0|0更新时间：2026-09-30
- 摘要：To address the degradation of historical failure identification performance caused by model updating when soft failure classes continuously appear in coherent optical networks, a continual learning-based incremental soft failure identification method was proposed. First, according to the digital signal processing process of coherent receivers, a physics-guided feature representation was constructed using the received electrical power spectral density, equalizer frequency response, and receiver-side performance indicators. The features were used to characterize typical soft failures, including filter shift, filter tightening, erbium-doped fiber amplifier (EDFA) gain degradation, and amplified spontaneous emission (ASE) noise increase. Then, soft failure identification was formulated as a class-incremental learning problem. When new failure classes were continuously introduced, dark experience replay++ (DER++) was adopted to store a small number of historical samples and their classification scores, and catastrophic forgetting was alleviated by combining historical-sample supervised constraints with classification-score preservation constraints. Simulation results showed that, when only 0.95% of the training samples of each class were cached, the proposed method achieved a final average accuracy of 93.93% and reduced the average forgetting rate to 7.95%. The method can learn new failure classes while retaining the identification capability for historical failure classes.关键词：autonomous optical network;soft failure identification;continual learning;class-incremental learning;experience replay6|0|0更新时间：2026-09-30
- 摘要：Aiming at the development of the low-altitude aerial economy, the requirements of network coverage height, width, range, and other aspects under various low-altitude scenarios were analyzed in this paper. A single-station model for millimeter-wave (mmWave) sensing in low-altitude scenarios was proposed accordingly. Based on this model, the frequency-division cellular networking scheme, area networking scheme, and linear networking scheme for mmWave sensing were analyzed in detail. Considering practical engineering conditions, reasonable inter-site distance ranges, site height selections, azimuth angle ranges, and tilt adjustments were proposed. Furthermore, through actual field tests, the overall performance of the mmWave sensing cellular networking scheme was effectively validated. Key performance indicators, including sensing altitude, velocity, accuracy, false alarm rate, miss detection rate, and multi-target detection capability, all achieved the expected results.关键词：low-altitude aerial intelligent network;integrated sensing and communication;sensing networking;millimeter-wave4|0|0更新时间：2026-09-30
- 摘要：With the evolution of quantum key distribution (QKD) technology from single-protocol point-to-point transmission to multi-protocol large-scale networking, multi-protocol QKD optical networks have become a research hotspot. From the perspective of the collaborative evolution of the physical layer and the network layer, the research progress of physical-layer protocol design and performance optimization, as well as the development trend of network-layer field verification and heterogeneous interconnection, are systematically reviewed in this paper. Based on the typical characteristics of multi-protocol QKD optical networks, three core technologies, including the multi-protocol integration node model, the multi-protocol cooperative control scheme, and the multi-protocol key supply strategy, are mainly discussed, laying a foundation for the future construction of efficient, flexible, elastic, adaptive, secure and controllable multi-protocol QKD optical networks.关键词：optical networks;quantum key distribution;multiple protocols6|0|0更新时间：2026-09-30
- 摘要：The Rayleigh distance of extremely large-scale reconfigurable intelligent surface (XL-RIS) increases with its aperture size. Consequently, conventional far-field discrete Fourier transform (DFT) angular-domain codebooks cannot accurately match near-field spherical-wave channels, resulting in angular-domain energy leakage, multibeam focal mismatch, and reflection-gain loss. To address these issues, a wavenumber-domain projection-based multibeam design method was proposed for near-field XL-RIS multiuser communications. First, continuous-aperture near-field channel and constant-modulus reflection models were established to characterize spherical-wave propagation and the basis mismatch induced by the far-field assumption. Second, a finite-dimensional wavenumber-domain equivalent matrix was constructed through projection onto two-dimensional Fourier wavenumber bases. The dominant propagation modes were retained via low-wavenumber truncation, and an upper bound on the truncation error was derived. Finally, under the passive constant-modulus constraint, the minimum user spectral efficiency was maximized, and a block coordinate descent and majorization-minimization (BCD-MM) algorithm was employed to achieve joint focusing for multiple users. The system performance improved with the signal-to-noise ratio, decreased as the number of users increased, first increased and then approached saturation as the wavenumber truncation radius increased, and improved with the XL-RIS aperture area. The proposed method provides a design approach that balances performance, low-dimensional representation, and physical interpretability for adaptive XL-RIS multibeam reconfiguration based on channel and location information in 6G endogenous-intelligence networks.关键词：extremely large-scale reconfigurable intelligent surface (XL-RIS);near field;multi-beam;wavenumber domain;constant-modulus constraint7|0|0更新时间：2026-09-30
- 摘要：Due to the scarcity of labeled data for physical layer impairments in optical transport networks (OTNs) and the inadequate credibility of model training data under few-shot scenarios, an urgent need was identified to improve the quality of impairment data generation and the reliability of pseudo-labels in such scenarios to support the engineering implementation of intelligent processing technologies for OTN physical layer impairments. A collaborative method integrating generative data augmentation and semi-supervised labeling with optical transmission physical constraints was proposed in this paper. Specifically, a conditional diffusion generation model with physical constraints was constructed, and a joint loss function incorporating physical consistency, distribution matching, and boundary constraints was designed. Furthermore, a generation-labeling collaborative semi-supervised framework was established, which suppressed pseudo-label noise through a dual mechanism of Monte Carlo uncertainty estimation and physical verification, thereby realizing the bidirectional iterative optimization of the data augmentation and labeling models. Experimental results demonstrated that the proposed method could significantly enhance the distribution credibility, physical credibility, and labeling credibility of OTN physical layer impairment data, effectively addressing the key issue of insufficient and unreliable model training data in few-shot scenarios. Consequently, this method exhibited strong engineering applicability and promotion value.关键词：OTN;physical layer impairment;few-shot scenario;generative data augmentation;semi-supervised labeling3|0|0更新时间：2026-09-30
- 摘要：To address the performance analysis and resource allocation challenges of index modulation-based frequency hopping spread spectrum (IM-FHSS) systems under mixed jamming, independent jamming parameters were introduced to model different jamming types. Complete bit error rate (
) expressions for both additive Gaussian white noise (AWGN) and Rayleigh fading channels were derived based on hypergeometric distribution probability models. By defining 3 core performance metrics, outage probability, energy efficiency, and resource occupancy rate, two complementary optimization models, namely resource minimization and minimization, were established. The nonlinear randomly reuse-based mutated whale optimization algorithm (NRRMWOA) and the improved composite particle swarm optimization (ICPSO) algorithm were employed to solve the two optimization problems, respectively, addressing their mixed-integer nonlinear, non-convex, and multi-modal characteristics. Simulation results demonstrated that the theoretical derivations closely matched Monte Carlo simulations, and the proposed algorithms significantly outperformed traditional methods in terms of solution accuracy, convergence speed, and stability. 关键词：index modulation;frequency hopping;mixed jamming;anti-jamming strategies6|0|0更新时间：2026-09-30 - 摘要：In high-mobility vehicular networks, communication links are frequently blocked by surrounding obstacles, sensing accuracy degrades considerably as vehicle speed increases, and onboard energy storage is strictly limited—these three factors constitute the primary challenge in current system design. To address this issue, a simultaneously transmitting and reflecting surface (STARS) is mounted on the vehicle roof. By virtue of its independent tunability over transmitted and reflected signals, the STARS enables natural spatial separation between in-cabin communication coverage and out-vehicle environmental sensing. The Cramér-Rao bound (CRB) is adopted to quantify sensing accuracy. Under the dual constraints of a CRB upper bound and per-user signal-to-interference-plus-noise ratio (SINR) lower bounds, an optimization problem is formulated to minimize the base station transmit power. Auxiliary variables are introduced and Schur complement properties are invoked to handle the non-convex CRB term. Block coordinate descent (BCD) and fractional programming (FP) are employed in an alternating fashion to jointly obtain the beamforming vectors and reflection coefficients. Simulation results confirm that the proposed approach effectively reduces the base station output power while maintaining satisfactory communication quality and sensing accuracy, demonstrating favorable applicability to energy-constrained onboard scenarios.关键词：simultaneously transmitting and reflecting surface;integrated sensing and communication;vehicular networks;power minimization;Cramér-Rao bound;joint optimization3|0|0更新时间：2026-09-30
- 摘要：Network operation logs record key status information of devices, and effectively analyzing them enables early fault warning. To address the problems of traditional fault prediction methods being black-box and lacking interpretability, as well as existing reflection models suffering from continuous memory bloat and getting stuck in local optima, an internet operation and maintenance fault prediction method based on multi-agent context learning is proposed. Taking large language model agents as the core computational units, the method works as follows. First, agents parse log semantics and perform structured extraction based on error codes and stack traces, constructing a multi-dimensional fault labeling system and a hierarchical sample set. Second, a tripartite semantic game architecture composed of two differentiated generators and one discriminator is constructed, replacing parametric numerical fitting with critical interaction, and achieving dynamic calibration of prediction strategies and evaluation criteria without gradient updates. Furthermore, inspired by the dynamic optimization idea of particle swarm optimization (PSO), a heuristic memory filtering mechanism is designed to represent interaction experiences as structured memories stored in a memory pool, track individual and global best experiences from the interaction memory, and dynamically inject them into the context, forming an “adversarial-adaptive-evolution” loop that drives agents to escape local optima under non-parametric conditions. Experiments on the GAIA (Generic AIOps Atlas) dataset show that the model achieved an average accuracy of 92.6%, a stable F1 score of 91.1%, and missed alarm and false alarm rates controlled at 10.2% and 7.4%, respectively, significantly outperforming same-paradigm baselines such as single-agent chain-of-thought (CoT) reasoning, Reflexion, and RCAgent. Ablation experiments confirmed the indispensability of the multi-agent game and the dual-filter mechanism, and the more stringent triple-matching experiment further verified the model’s ability to output structured root-cause texts. Engineering metric analysis indicates that during testing, the model trades a time cost on the order of hundreds of seconds and relatively low token consumption for highly reliable predictions, making it highly practical in operational scenarios where the cost of missed alarms far outweighs computing costs.关键词：log text;fault prediction;multi-agent;in-context learning9|0|0更新时间：2026-09-30
- 摘要：To address the difficulties of multi-source positioning coordination, delayed cross-scenario handover, and insufficient service continuity in complex indoor environments such as large commercial complexes, a large language model assisted scheduling and continuity control method for indoor multi-source positioning was proposed inspired by the 3GPP positioning framework and extended at the platform level. Cellular and Wi-Fi measurements were primarily aggregated on the location service platform, with extensible adaptation to Bluetooth/UWB and inertial sensing data. Dynamic source selection was achieved through spatio-temporal alignment, signal semantic encoding, scenario recognition, constrained scheduling, and fusion calculation. The method was evaluated with field data collected from eight floors of a 400 000 m² commercial complex. The results showed that the weighted scenario recognition accuracy reached 94.3%, the root mean square error in open LOS atrium scenarios was 0.75 m, the cross-scenario positioning continuity reached 88%, and the average positioning error was reduced by 49.5% compared with the fixed-weight fusion method. The proposed method improves the continuity, stability, and cross-scenario adaptability of indoor positioning services.关键词：indoor positioning;multi-source perception;large language model;scenario recognition;continuity control;5G-advanced4|0|0更新时间：2026-09-30
- 摘要：To address the trade-off among communication security, sensing accuracy, and harvested energy in a simultaneously transmitting and reflecting surface (STARS)-assisted integrated sensing, communication, and power transfer (ISCAP) system, a robust resource allocation scheme based on a stackelberg game-based meta soft actor-critic (ST-Meta-SAC) algorithm is proposed. First, closed-form expressions for the secrecy rate (SR) and harvested energy are derived under the sensing channel state information (CSI) errors constraints. Subsequently, we formulate a multi-objective optimization problem that maximizes the communication secrecy rate while minimizing the determinant of the sensing Cramer-Rao bound (CRB) matrix subject to the minimum harvested energy constraint, by jointly optimizing the STARS reflection and transmission coefficient matrices, the sensing signal covariance matrix, and the signal beamforming vector. Finally, since the high-dimensional, multi-level, and non-convex nature of the problem makes it intractable for conventional methods, we propose a ST-Meta-SAC algorithm to obtain an efficient solution. Simulation results show that our proposed method outperforms benchmark schemes in terms of convergence speed, demonstrating its effectiveness.关键词：simultaneously transmitting and reflecting surface;Cramér–Rao bound;secrecy rate;harvested energy3|1|0更新时间：2026-09-30
-
Intelligent covert communication optimization via temporal feature extraction from jamming feedback AI导读
摘要：To address the problem that existing studies on reactive jamming mainly focused on passive avoidance after jamming occurred and rarely utilized the detection feedback of the warden to guide covert strategy design, a dynamic multi-channel covert communication joint optimization method was investigated for a scenario where the warden was equipped with both eavesdropping detection and reactive jamming capabilities. A closed-loop time-slot interaction mechanism of “sensing-learning-transmission-feedback” was constructed. The discrete channel selection and continuous power control were modeled as a Markov decision process with a hybrid action space. Moreover, a joint optimization algorithm based on a long short-term memory temporal encoder and a hybrid soft actor-critic framework was proposed, in which decisions were made by combining historical channel states and jamming feedback information. Thus, the communication strategy was transformed from “passive avoidance” to “active prevention and control”. Simulation results showed that the proposed algorithm could learn an effective covert transmission strategy without prior information about the warden. Its performance approached the ideal upper bound of the KL-constrained optimization algorithm under perfect channel state information, and a better balance between covertness and effectiveness was achieved compared with the deep Q-network. The proposed method provides an effective solution for covert communication optimization in reactive jamming environments.关键词：reactive jamming;covert communication;long short-term memory temporal encoder;deep reinforcement learning10|1|0更新时间：2026-09-30 - 摘要：This paper proposes an active UAV-RIS-assisted secure beamforming method for integrated sensing and communication (ISAC) in maritime environments. A shore based dual-functional base station simultaneously serves legitimate maritime users and senses a potential eavesdropper. The dual-functional beamformers, active RIS reflection coefficients, and horizontal UAV position are jointly optimized to maximize the sum secrecy rate subject to sensing SNR, transmit power, RIS output power, and user QoS constraints. To address the resulting non-convex problem with strongly coupled variables, an alternating-optimization algorithm integrating weighted minimum mean-square error (WMMSE), successive convex approximation (SCA), and projected gradient ascent is developed. At a maximum BS transmit power of 30 dBm, the proposed method achieves a sum secrecy rate of approximately 9.5 bps/Hz, providing gains of about 14.5% over the active RIS scheme with a fixed UAV and 39.7% over the passive RIS scheme with optimized UAV placement. The algorithm also converges within a small number of outer iterations. These results demonstrate the benefit of jointly exploiting active-RIS amplification and UAV spatial deployment for secure maritime ISAC transmission.关键词：maritime communication;integrated sensing and communication;Active Reconfigurable Intelligent Surface;unmanned aerial vehicle;physical layer security7|0|0更新时间：2026-09-30
- 摘要：To address the low detection probability and high inference latency of spectrum sensing models for tactical edge networks under low signal-to-noise ratio (SNR) conditions, a lightweight spectrum sensing method based on dual-domain physical priors and confidence-aware decoupled self-distillation (CDSD) is proposed. First, a dual-domain physical-prior input is constructed by fusing time-domain conjugate differential features with frequency-domain magnitude-spectrum features, thereby enhancing the ability of shallow models to characterize the coherent structures of weak signals and their spectral energy distributions. Furthermore, a self-distillation training framework is incorporated into a multi-scale attention backbone network. A confidence-aware mask is employed to decouple and control decision-level distillation and representation-level distillation, reducing the negative transfer caused by erroneous soft labels generated by the teacher branch under low-SNR conditions. Finally, attention transfer and contrastive distillation constraints are jointly introduced to improve the feature separability of the student model. Experimental results show that, during deployment, the number of parameters in the student model is reduced from 14.36 M to 1.01 M, corresponding to a reduction of approximately 93%, while the inference speed is increased by a factor of 5.45. At a false-alarm probability of 10%, the proposed method achieves a detection probability of 75.6% at an SNR of −16 dB, outperforming baseline methods including VT-CNN2, USKD, FRSDK, MixSKD, and ZiLS by more than 26.6%. The proposed method achieves a favorable balance among detection performance, model complexity, and inference efficiency, providing effective support for low-SNR spectrum sensing at tactical edge nodes.关键词：spectrum sensing;self-distillation;dual-domain fusion;model compression;tactical edge networks3|0|0更新时间：2026-09-30
- 摘要：With the continuous deepening of research on 5G non-terrestrial networks (NTN) and the proposal of the 6G integrated space-air-ground-sea vision, the direct-to-satellite technology for legacy smartphones, featuring advantages such as low cost, easy deployment, and rapid implementation, was regarded as one of the important technical approaches to achieve global seamless coverage. Firstly, the core challenges faced by the 5G legacy smartphone direct-to-satellite scenario were analyzed, including time and frequency synchronization failure, random access window overrun and insufficient link budget. Secondly, to address the above issues, a time synchronization scheme based on dynamic reference points, a frequency synchronization scheme based on static reference points, and a random access enhancement scheme for legacy smartphone direct-to-satellite communication were proposed. Finally, through simulations and laboratory measurements, the feasibility of the proposed technical schemes was fully verified, and it was also demonstrated that the proposed time-frequency synchronization schemes can provide key technical support for terminal access and synchronization in global navigation satellite system (GNSS)-free scenarios of 6G NTN.关键词：satellite-direct-to-cell;legacy smartphone;non-terrestrial network;time synchronization;frequency synchronization8|0|0更新时间：2026-09-30
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
- Total Visits：1155036 Visits Today：719 Online Users：6
0