The arrival of 5G era brings great convenience and unprecedented opportunities to human society under the new technical features of ultra-high speed, ultra-low latency, and mega-connectivity. As the management hub and "brain" of the 5G network architecture, the security of the core network is paramount. However, its evolution towards cloudification has led to frequent cybersecurity incidents; an attack on the 5G core network could trigger severe consequences. To address this, this paper first systematically analyzes potential vulnerabilities and attack paths within the 5G core network from an attacker’s perspective. Subsequently, an endogenous proactive defense architecture, grounded in the principles of dynamism, heterogeneity, and redundancy, is proposed. The working mechanism of this architecture is elaborated in detail, alongside a theoretical analysis of its security properties. To validate the efficacy of the proposed architecture, defensive performance was first compared across different schemes via simulation. The results demonstrate that our architecture achieves optimal defensive performance while significantly increasing the attacker’s cost. Furthermore, the proposed architecture was implemented on the Unified Data Management (UDM) network function within the open-source free5gc platform, and its enhanced protective capability was assessed through real-world attack testing. Experimental results indicate that the architecture effectively elevates the overall security posture of the 5G core network, introducing only marginal performance overhead.

As a key intelligent component of core networks, NWDAF (Network Data Analytics Function) provides network intelligence capabilities while facing increasingly data security challenges. Current research on NWDAF data security primarily adopts federated learning framework, which trains models locally to avoid privacy leakage risks. However, existing NWDAF federated learning studies mostly rely on idealized assumptions, failing to adequately address two critical issues in real-world network deployments: 1) uneven data distribution across different regions, and 2) differentiated optimization requirements for high priority security regions. To solve these problems, we propose a region dual-factor aware federated learning framework for NWDAF (NWDAF-FedRDA), specifically designed to handle regional data imbalance and security priority differences in core networks. The framework comprises three core components: 1) Model parameter distribution: enabling allocation of global model parameters to regional nodes; 2) Local training: performing regional data preprocessing and model training; 3) Region dual-factor aware aggregation: computing regional weights through awareness factors and executing global model aggregation. Experimental evaluations across four network scenarios demonstrate the framework’s performance in anomaly detection tasks. Results show that compared to baseline methods, our framework exhibits significant advantages in uneven data distribution environments while effectively enhancing model performance for high-priority security regions.

As the brain of B5G networks, the core networks enable more ubiquitous intelligent connectivity over previous generations of mobile networks, thanks to the decentralized user plane close to edges. As mitigation against abnormal signaling attacks on edge core networks, signal protection mechanisms for the user planes at the N4 interface are widely investigated. However, the prior art fails to adequately address the distribution characteristics of abnormal signaling at this interface, where single-point defences are insufficient for the complexities of beyond 5G (B5G) distributed architecture. This article proposes a decentralized federated anomaly signaling protection framework, called B5GCASP, based on a functionally layered anomalous signaling detection model (FLAD). Mainly, B5GCASP analyses abnormal signaling distribution under the packet forwarding control protocol (PFCP) at the N4 interface, distinguishing significant and nonsignificant anomalies. Coupled with a decentralized, federated detection mechanism, B5GCASP creates a comprehensive point-and-area detection architecture. Extensive experiments on the 5GC PFCP dataset show that B5GCASP achieves higher accuracy and faster detection of abnormal signaling compared to single-point defending baselines, which offer robust anomaly signaling protection for the B5G core network.

As containerization technology progresses, the network functions within the 5G core are abstracted into Virtual Network Functions (VNFs), operating within a cloud-based ecosystem. To bolster network resilience, autoscaling strategies dynamically allocate resources in response to workloads, ensuring the Quality of Service (QoS) is maintained. However, the inherent openness of 5G networks exposes them to signaling storms and unexpected disturbances, leading to substantial workload fluctuations and bursts. Current reactive autoscaling strategies are slow to respond to these challenges, falling short of user expectations for low-latency, high-quality internet services. To address this challenge, we propose a novel auto-scaling framework specifically designed for the 5G cloud core network system, dubbed 5G-PPDE. Our framework harnesses the Crossformer model for accurate workload forecasting and employs a Multi-Layer Perceptron (MLP) classification model to offer four distinct scaling strategies. The scaling-down decision algorithm implements a Gradual Decrease Strategy (GDS) for resource management, while the burst scaling strategy adapts to sudden workload changes by referencing the maximum service quantities observed during historical window periods. We have constructed a validation system leveraging the open-source Free5GC and Free5GMANO, using the wrk tool to simulate burst traffic scenarios. Prometheus is utilized for real-time data collection, thereby validating the efficacy of our proposed framework. The findings demonstrate that our method effectively mitigates the impact of burst traffic, outperforming current solutions.

Realistically simulating spatial–temporal correlated complex sea clutter using statistical model-based methods is very challenging because the intricate interplay of various physical mechanisms in both space and time poses significant obstacles for clutter modeling. In this article, to overcome this challenge, we propose a data-driven method based on the 2-D amplitude and phase matching optimization (APMO) with two steps. First, the simulated clutter amplitudes with desired distribution and correlation characteristics are generated by performing the frequency-domain inverse transform and correlation transfer techniques on the measured clutter. Second, an optimization strategy is developed to acquire the well-matched phases for the simulated clutter amplitudes by constraining the phases to the simulated clutter amplitudes and the desired Doppler and range spectra. Two different algorithms are separately adopted to implement the optimization strategy, and their advantages and disadvantages are compared theoretically and experimentally. The APMO method is direct in the analysis of the measured clutter’s statistical characteristics and universal to different radar and environmental conditions. It is shown that the simulation results can reproduce the distribution and spatial–temporal correlation features of the real-world X-band complex sea clutter. The relative mean-squared deviation (RMSD) between the measured and simulated clutter’s Doppler and range spectra can be notably reduced to 0.0499 and 0.1237, respectively.

Due to the unequispaced pulse repetition interval (PRI), the low-oversampling ratio and the range-variant blockage, the echo of the low-oversampled staggered SAR (LS-SAR) is nonuniformly sampled with sub-Nyquist and range-variant rate. However, the existing LS-SAR processing methods lack robustness with regards to the scenario type and the PRI variation mode. In this article, a compressive-sensing-based image reconstruction method for the LS-SAR is proposed. First, a hybrid-domain model (HDM) of the LS-SAR echo is presented. In the HDM, the coupled range cell migration (RCM), the unequispaced PRI, and the conflict blockage are formulated as the matrix multiplications with a 3-D tensor, a 2-D matrix, and a Hadamard product, respectively. Based on the HDM, the image reconstruction is realized through the 2-D fast iterative shrinkage thresholding algorithm (ISTA), in which the gradient is derived by exploiting the properties of the tensor and matrix trace. The fast Fourier transform (FFT) and the nonuniform FFT are implemented to accelerate the computation. Due to good accommodation of the RCM and the LS-SAR sampling characteristics, the proposed method can work well for various PRI variation modes and scenario types. Simulations using the point scatter and the distributed target with wide-swath extension demonstrate the effectiveness as well as the robustness of the proposed method.

In this paper, the concept of quantile is introduced and elaborately related to the truncation depth, based on which quantile truncated statistics (QTS) is put forward. The QTS gives a reasonable explanation of truncation depth and makes the selection of truncation depth well-founded and controllable. In addition, maximum likelihood estimation based on QTS (QTS-MLE) for the probability density function (PDF) parameters is derived. We start the analysis from Weibull background assuming that the shape parameter is known, and then extend it to the case where the shape parameter is unknown. By analyzing the variance and mean square error (MSE) of the estimated parameters in Weibull background, it is found that QTS-MLE has better estimation performance than the MLE based on truncated statistics (TS-MLE). On this basis, the constant false alarm rate (CFAR) detector based on QTS-MLE, i.e. QTS-CFAR, is proposed. The analytic expressions of the false alarm rate and detection probability of QTS-CFAR are derived under the Weibull background with known shape parameter. The full CFAR characteristics of TS- and QTS-CFAR detectors in Weibull background with unknown shape parameter are proved by invariant theory. Monte Carlo simulations show that QTS-CFAR detector has better anti-interference performance and false alarm control ability in multiple-target environment. Furthermore, the superiority of QTS-CFAR detector is verified by the real data collected by skywave over-the-horizon radar. Finally, we present the expression of QTS-MLE for the scale parameter in the Gamma background.

This paper focuses on processing low oversampling echo data of staggered synthetic aperture radar (SAR). In staggered mode, the non-uniformly sampling and echo data loss cause severe azimuth ambiguity. To solve this problem, we propose a method combining the compressed sensing (CS) tool and the conformal Fourier transform (CFT) algorithm. First, the 2-D-fast iterative shrinkage thresholding algorithm (FISTA) is used to efficiently solve the optimization problem based on CS theory to recovery the missing data. Second, the CFT algorithm is used to accurately compute the spectrum of the non-uniformly sampled echo data. Unlike other methods suiting fast pulse repetition interval (PRI) change only, this method performs well for both fast and slow PRI change. Plus, the proposed method can be well applied to both point and distributed targets. Simulation results demonstrate that the proposed method can effectively and robustly suppress the azimuth ambiguity for staggered SAR data.