Online Parameter Extraction and Control of Reconfigurable High-Frequency Filters

• Applicants: Professor Dr.-Ing. Michael Höft; Professor Dr.-Ing. Holger Maune
• Subject Area: Communication Technology and Networks, High-Frequency Technology and Photonic Systems, Signal Processing and Machine Learning for Information Technology
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 519897406

The primary goal of this proposal is to research the real-time reconfiguration of high-frequency filters. For this purpose, the reconfiguration of a microwave filter is to be performed automatically after the input of a new target characteristic and various effects, such as temperature drift, are to be compensated. The reconfiguration should take place without interrupting the signal path. The aim is to develop tunable microwave filters in planar and non-planar technology and to influence the transmission characteristics with the help of a tuning system. The different technologies for tuning have different non-linearities and quality factors. Among others, it can be a planar microstrip filter with varactors as tuning elements or a coaxial resonator filter with mechanical tuning screws. In these cases, the control is realized with variable voltage sources or stepper motors. An essential aspect of this project is the development of a measuring system to determine the transmission characteristics of the microwave filter during operation. Here, the filter structure should not be disconnected, so that a possible transmission is not interrupted. Potential approaches are presented in the application, which must be examined in detail for their advantages and disadvantages. For example, an investigation of the signal in the time domain would be possible, so that the transfer function can be concluded by transformation. Subsequently, different tuning algorithms will be investigated with which the transmission characteristics of the filter structures can be adjusted. The algorithms will be compared based on various aspects such as runtime and quality of optimization. Coupling matrix extraction provides an efficient method of tuning, but how to provide reliable and robust extraction needs to be investigated. Reconfiguration algorithms based on scattering parameters and time domain analysis are to be developed so that tuning is possible without the need to construct an equivalent circuit. As a last aspect, artificial neural networks are to be investigated, which can be used to reproduce unknown, even very complex functions. Finally, the overall system consisting of microwave filter, control electronics, measurement system and tuning algorithm is to be set up and tested under various aspects. On the one hand, the temperature drift that occurs when the structure heats up is to be compensated and, on the other hand, the filter is to be tuned to a new transmission characteristic consisting of a new center frequency and bandwidth.

• DFG Programme: Research Grants