带通滤波器
拓扑(电路)
滤波器(信号处理)
谐振器
节点(物理)
电子工程
物理
波导滤波器
计算机科学
原型滤波器
传输(电信)
章节(排版)
网络拓扑
模块化设计
通带
电子滤波器拓扑
网络综合滤波器
零极点图
业务流程重组
联轴节(管道)
滤波器设计
m-导出滤波器
滤波理论
工程类
控制理论(社会学)
作者
Lin Yan,Jing‐Yu Lin,Hong-Ji Li,Sai‐Wai Wong,Yi Wang
标识
DOI:10.1109/tmtt.2026.3686917
摘要
Two novel low-pass prototype topologies for realizing quasi-elliptic bandpass filter (BPF) responses are discussed here. Each topology can generate a pair of transmission zeros (TZs), configurable at either a resonating or nonresonating node (NRN). In the case where the TZ pair is associated with a resonating node, the corresponding filter section is composed of one bandpass resonator (BPR) and two bandstop resonators (BSRs), denoted as BPR = 1 and BSR = 2. Alternatively, when the TZ pair is assigned to an NRN, the section employs two BSRs, represented as BPR = 0 and BSR = 2. These modular sections can be cascaded to create compact filters with multiple TZs. Meanwhile, to overcome the computational complexity of traditional synthesis on the configuration of proposed two low-pass prototypes, a filter synthesis framework based on surrogate-assisted optimization is proposed. This framework is capable of designing filters integrating different resonant sections, and each section can carry a TZ pair at either a resonant or nonresonant node. To validate the proposed approach, the synthesis and design of two BPFs with distinct topological arrangements are demonstrated: one is a 5th-order inline BPF constructed using three BPR = 1 and BSR = 2 sections and two BPR = 0 and BSR= 2 sections; the other is a 5th-order filter that incorporates one BPR = 1 and BSR = 2 section along with two BPR = 0 and BSR= 2 sections. Additionally, a manifold quadruplexer is implemented, utilizing two BPR = 1 and BSR = 2 sections and one BPR = 0 and BSR = 2 section to achieve the third-order channel filtering response. The synthesized, simulated, and measured results exhibit excellent agreement, confirming the effectiveness and accuracy of the proposed methodologies.
科研通智能强力驱动
Strongly Powered by AbleSci AI