Along with the rapid evolution of mobile communication technologies, such as 5G, there has been a significant increase in telecom fraud, which severely dissipates individual fortune and social wealth. In recent years, graph mining techniques are gradually becoming a mainstream solution for detecting telecom fraud. However, the graph imbalance problem, caused by the Pareto principle, brings severe challenges to graph data mining. This emerging and complex issue has received limited attention in prior research. In this paper, we propose a G raph AT tention network with CO st-sensitive BO osting (GAT-COBO) for the graph imbalance problem. First, we design a GAT-based base classifier to learn the embeddings of all nodes in the graph. Then, we feed the embeddings into a well-designed cost-sensitive learner for imbalanced learning. Next, we update the weights according to the misclassification cost to make the model focus more on the minority class. Finally, we sum the node embeddings obtained by multiple cost-sensitive learners to obtain a comprehensive node representation, which is used for the downstream anomaly detection task. Extensive experiments on two real-world telecom fraud detection datasets demonstrate that our proposed method is effective for the graph imbalance problem, outperforming the state-of-the-art GNNs and GNN-based fraud detectors. In addition, our model is also helpful for solving the widespread over-smoothing problem in GNNs. The GAT-COBO code and datasets are available at https://github.com/xxhu94/GAT-COBO.

This study reports a multi‐layered dual‐mode substrate integrated waveguide (SIW) bandpass filter (BPF), in which the input and output ports are located on the same substrate layer (i.e. the upper one). The dual‐mode SIW cavity in the lower substrate layer is coupled to the upper dual‐mode SIW cavity by two crossed slots, which can independently control the inter‐couplings. The proposed filter can be modelled by a dual‐extended doublet coupling scheme with considering bypass couplings between source and load. Based on the proposed structure, single‐ and dual‐band filtering responses with four controllable finite‐transmission zeros (FTZs) are realised and analysed. For the demonstration, a single‐band filter and a dual‐band filter were fabricated and measured. The measured results agree well with simulated ones.

With the rapid evolution of mobile communication networks, the number of subscribers and their communication practices is increasing dramatically worldwide. However, fraudsters are also sniffing out the benefits. Detecting fraudsters from the massive volume of call detail records (CDR) in mobile communication networks has become an important yet challenging topic. Fortunately, Graph neural network (GNN) brings new possibilities for telecom fraud detection. However, the presence of the graph imbalance and GNN oversmoothing problems makes fraudster detection unsatisfactory. To address these problems, we propose a new fraud detector. First, we transform the user features with the help of a multilayer perceptron. Then, a reinforcement learning-based neighbor sampling strategy is designed to balance the number of neighbors of different classes of users. Next, we perform user feature aggregation using GNN. Finally, we innovatively treat the above augmented GNN as weak classifier and integrate multiple weak classifiers using the AdaBoost algorithm. A balanced focal loss function is also used to monitor the model training error. Extensive experiments are conducted on two open real-world telecom fraud datasets, and the results show that the proposed method is significantly effective for the graph imbalance problem and the oversmoothing problem in telecom fraud detection.

An unbalanced microstrip bandpass filter (BPF) in dual‐extended doublet topology with source‐load (S‐L) coupling is first presented and researched based on dual‐layer capacitive‐loaded dual‐mode loop resonators. The input and output ports are located on the upper substrate layer. The lower dual‐mode loop resonator is coupled to the upper loop resonator using four coupling apertures in the middle metal layer. Single‐ and dual‐band responses with four controllable finite‐transmission zeros (FTZs) are analysed. Then, a balanced BPF in dual‐extended doublet topology without S‐L coupling on the basics of the proposed unbalanced filter is presented and researched. Single‐ and dual‐band differential mode responses with three controllable FTZs can be realised with good common mode suppression. Some design examples are provided to illustrate the performances of the realised single‐ and dual‐band unbalanced and balanced filters with controllable FTZs. The measured results agree well with simulated ones.

This paper presents improved quad-mode resonators based on the open/short-end load of the center stub, which introduces the self-coupled structure. Compared with the traditional stub-loaded stepped impedance resonator (SLSIR), the proposed quad-mode resonators can significantly reduce the circuit size and further increase the design freedom. First of all, a comprehensive theoretical analysis is performed to infer the resonant characteristics of the proposed resonator. Then, two design prototypes based on different center loads of the resonator were demonstrated to verify the feasibility of the design method. First, a prototype of a wideband high temperature superconducting (HTS) filter with an FBW of 89% is proposed, whose center frequency (CF) can be flexibly controlled. Next, a quad-band HTS filter using different central stub load is proposed, which centered at 1.20, 2.52, 3.51, 4.44 GHz with the 3-dB FBW of 4.8%, 3.96%, 3.97% and 2.2%, respectively. The internal and external coupling mechanisms are also carefully studied to meet the specified requirements. The two prototypes proposed in this paper both have the advantages of compact size, ultra-low insertion loss and steep suppressed skirts. In addition, the measured results are in good agreement with the EM full-wave simulated results.

Herein, a rigorous derivation of an isosceles right‐angled triangular (IRT) metallic cavity is analysed. Then, the transverse field components in the IRT substrate‐integrated waveguide (SIW) cavity are derived, and the expression of the resonant frequencies can also be obtained. Then, a novel perturbed IRT SIW cavity is proposed, which is based on the characteristics of resonant modes in IRT SIW cavity and the capacitive‐loading technique. Three prototypes were designed to verify the validity of the derivation and analysis in this study. First, a dual‐mode single passband filter was implemented using the perturbed modes TE ₁₀₂ and TE ₂₀₂ . Next, two tri‐mode triple‐band bandpass filters (BPFs) using the perturbed modes TE ₁₀₁ , TE ₁₀₂ and TE ₂₀₂ are proposed, synthesised, verified and measured for demonstration, including a second‐order direct‐coupled one which operates at 8.02, 11.44 and 12.52 GHz, and a fourth‐order direct‐coupled one which operates at 7.7, 11.0 and 12.1 GHZ. The measured curves agree well with the simulated ones. Very compact circuit sizes and excellent filtering performances have been achieved for the three prototypes.