Optimum topology for power handling in dual bandpass filters
Lead Research Organisation:
University of Leeds
Department Name: Electronic and Electrical Engineering
Abstract
Due to the ever-increasing need for capacity, the typical filtering requirements for RF filters deployed at cellular base stations are particularly stringent. Such high performance filters require estimation of their power handling capacity to avoid breakdown. Investigation of the stored energy distribution in network topologies of RF filters is one approach to estimating their power handling capacity.
The total time-averaged stored energy for a given transfer function is constant. It's distribution, however, is strongly influenced by the synthesised network topology. The aim of this research is to use a pattern recognition algorithm to find similarities in the stored energy distribution within filter networks realising single and dual band transfer functions. By finding such similarities, guidelines for estimating the optimum network topology for a given transfer function are to be deduced.
The total time-averaged stored energy for a given transfer function is constant. It's distribution, however, is strongly influenced by the synthesised network topology. The aim of this research is to use a pattern recognition algorithm to find similarities in the stored energy distribution within filter networks realising single and dual band transfer functions. By finding such similarities, guidelines for estimating the optimum network topology for a given transfer function are to be deduced.
Organisations
Publications
Musonda E
(2019)
Synthesis of Multiband Filters by Linear Optimization
in IEEE Transactions on Microwave Theory and Techniques
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/N509681/1 | 30/09/2016 | 29/09/2021 | |||
| 1803835 | Studentship | EP/N509681/1 | 31/08/2016 | 31/10/2019 | Rucha Paradkar Smith |
| Description | Microwave filter design begins by determining an equivalent lumped or distributed filter network that approximates the desired filter response. The process of determining the filter network, given the excitation and the desired response (specifications), is known as synthesis. The technique for the synthesis of single band filters is well-established. With the advent of advanced wireless systems, the demand for capacity is increasing continually. This has led to the advent of multi-band filters that are to be employed at the RF front-end of cellular base stations. Therefore, a novel multi-band synthesis technique was developed in this work and was validated by designing dual and triple band filters. The power handling capacity of microwave filters is limited by the internal peak electric fields that can lead to its breakdown. The peak internal electric fields in the filter can be lowered by choosing a filter topology (arrangement of filter resonators) that lowers the peak time-averaged stored energy (t.a.s.e) in the equivalent filter network. The computed peak stored energy in various practically realisable filter topologies for numerous single band filter responses were investigated using k-means clustering algorithm. By analysing the results, patterns or "guidelines" were established to predict a topology for any given transfer function that would yield the lowest peak t.a.s.e. This topology would constitute the optimum power handling topology for the particular transfer function. Similarly, practically realisable filter topologies for dual band responses were investigated using the clustering algorithm to establish patterns that aid in the optimum topology for power handling in dual band transfer functions. These patterns were validated by comparing the EM simulations of two different filter topologies. |
| Exploitation Route | The novel synthesis technique can be implemented to synthesise high-order, complex multi-band transfer functions, from which the practical multi-band filters can be designed. The patterns established using the clustering algorithm, for single and dual bandpass transfer functions, can be exploited to predict the optimum topology. This eradicates the need to investigate all possible filter topologies for a given transfer function, which can be time-consuming and tedious. The predicted filter topology lowers the peak internal electric fields in a filter structure, thus, eliminating, if not minimising the risk of a breakdown. |
| Sectors | Aerospace Defence and Marine Electronics |