A Fast Surrogate Model-Based Algorithm Using Multilayer Perceptron Neural Networks for Microwave Circuit Design

Question

1. What techniques have been used to suppress undesirable harmonics in microwave circuit design?

2. How does machine learning aid in microwave circuit design?

3. What is the purpose of designing a preliminary resonator?

4. What is the transition zero frequency of the proposed single polygonal resonator?

Accepted Answer

Several techniques have been employed to suppress undesirable harmonics in microwave circuit design. These include EBG and DGS techniques, open stubs, resonators, and lumped components. EBG and DGS techniques have been used to decrease circuit size and suppress harmonics. Open stubs have been utilized for harmonics suppression and size reduction, but only a few harmonics can be suppressed using this method. Resonators, such as rectangular and T-shaped resonators, have been incorporated in the divider structure to suppress harmonics. Lumped components, including L and C, have been used to create compact dividers with harmonics suppression. Additionally, a new design algorithm has been proposed to aid in the design of complicated devices, such as filters and power dividers, with the suppression of more than 12 unwanted harmonics and a wide suppression band from 2 GHz to 20 GHz.

Accepted Answer

Machine learning, particularly neural networks, plays a crucial role in microwave circuit design by addressing complex design challenges and optimizing the design process. In this context, the application of multilayer perceptron neural networks as a fast surrogate model-based algorithm enables efficient and accurate modeling techniques. By incorporating machine learning techniques, the algorithm facilitates the design of complicated devices within microwave and communication systems, such as filters and power dividers. The algorithm follows a seven-step process, starting with the preliminary low-pass filter (LPF) design, followed by the proposed ANN structure, training using electromagnetic simulation, finalizing dimensions based on desired parameters, applying the final LPFs to the divider branches, and achieving the final wide-band power divider (WPD). This approach leverages extensive datasets and learning patterns to predict structural responses, fatigue life, and failure modes, ultimately enhancing the design and performance of microwave circuits.

Accepted Answer

The purpose of designing a preliminary resonator is to achieve desired performance and features, such as small dimensions, wide stop-band bandwidth, and sharp response in the transient region. It serves as a starting point for improving the resonator's design by adding additional resonators and suppressors. Simulations are carried out using ADS software to optimize the resonator's performance.

Accepted Answer

The transition zero frequency of the proposed single polygonal resonator is 2.75 GHz. This resonator was designed on an RT/Duroid 5880 substrate with a dielectric coefficient of 2.2 and a dielectric thickness of 0.508 mm. The structure of the resonator is demonstrated in Figure 4, which shows a folded transmission line. The resonator's design and characteristics make it suitable for applications requiring a specific transition zero frequency, such as in RF and microwave circuits. The transition zero frequency is an essential parameter in resonator design, as it determines the resonant frequency at which the device exhibits minimal impedance. In this case, the resonator's transition zero frequency of 2.75 GHz indicates its suitability for applications operating within the GHz range. The design and fabrication of resonators with precise transition zero frequencies are crucial for achieving desired performance in various electronic and communication systems.

Accepted Answer

The proposed structure of the preliminary dual polygonal resonator is a symmetrical design with Port 1, Port 2, W1, W2, W3, W4, and L1, L1. This design is illustrated in Figure 5, which shows the resonator loaded with folded transmission lines. The symmetrical structure aims to improve the results by enhancing the resonator's performance. The dual polygonal resonator design is expected to provide better resonance characteristics and improved signal transmission. The symmetrical structure allows for balanced resonance and reduces unwanted harmonics, leading to a more efficient and reliable resonator. Overall, the proposed design offers potential advancements in resonator technology, making it an interesting area of research for further exploration.

Accepted Answer

In the presented ANN model, 31 samples are used for training the network. This number is chosen to ensure that the model has enough data to learn from and generalize well to new, unseen data. The training process involves adjusting the weights and biases of the neural network to minimize the error between the predicted output and the actual output. By using a sufficient number of samples, the model can capture the underlying patterns and relationships in the data, leading to more accurate predictions. Additionally, using a larger training dataset can help prevent overfitting, where the model becomes too specialized to the training data and performs poorly on new data. Therefore, using 31 samples for training the ANN model is a crucial step in developing a reliable and effective surrogate model.

Accepted Answer

The predicted circuit parameters using the proposed ANN model show acceptable accuracy. The real and predicted comparison values of the f c (GHZ), z (dB/GHz), BW (MHZ), and RL (dB) parameters for the training and testing data are shown in Figure 6. The test errors are higher compared to the training errors, but the values of error are acceptable, and the worst-case error for both the testing and the training process is below than 0.2 MRE. Additionally, the obtained results of the proposed filter and power divider, predicted using the proposed ANN, show that the small difference between the real and predicted data in the testing process did not affect the filter and power divider output parameters, and the desired results were obtained for the designed filter and divider. The real and predicted testing and training values of the f c (GHZ), z (dB/GHz), BW (MHZ), and RL (dB) parameters versus the number of data samples in the proposed model are shown in Figure 7. Two samples were chosen for validation, and these are shown in Figure 7. The validation samples validate the accuracy of the proposed model, and help in designing the proposed resonator with the desired parameters. The final results of the proposed ANN model are listed in Table 2, which shows the high accuracy of the predictions of the proposed model. The errors reported in this table correspond to the denormalized data. According to this table, the model was trained perfectly using the training data, and the validation sample was accurately predicted by the proposed model, which was then used to form the structure of the proposed resonator.

Accepted Answer

The frequency response of the proposed preliminary dual polygonal resonator, as shown in Figure 8, indicates a transition zero with more than a 60 dB attenuation level. This means that the resonator effectively blocks or attenuates signals at a specific frequency. Additionally, the resonator has an insertion loss of about 0.5 dB in the pass band, which refers to the loss of signal power in the frequency range where the resonator allows signals to pass through. Overall, the frequency response demonstrates the resonator's ability to selectively filter and manipulate signals based on their frequency components.

Accepted Answer

The dimensions of the proposed dual-side resonators are as follows: L1 = 1.2 mm, W1 = 1.56 mm, L5 = 3 mm, W5 = 0.1 mm, L6 = 1.8 mm, W6 = 1.4 mm, L7 = 0.8 mm, W7 = 9.9 mm, L8 = 0.8 mm. These dimensions were designed to improve the results and obtain a microstrip low-pass filter with smaller dimensions. The resonators were loaded at the main transmission line of the filter, as depicted in Figure 9. The frequency response of these resonators, shown in Figure 10, demonstrates the creation of two transmission zeros with more than a 60 dB attenuation level, enhancing the filter's performance.

Accepted Answer

The frequency response of the proposed combined resonator, as shown in Figure 12, creates three transmission zeros and provides a wide stop band bandwidth from approximately 1.9 GHz to 8.9 GHz, with an attenuation level of more than 20 dB. The proposed combined resonator has a sharp transition band of about 0.294 GHz from -3 to -20 dB, which provides a 57.8 dB/GHz roll of the rate parameter, resulting in a desirable frequency response.

Accepted Answer

Suppressor cells, as shown in Figure 13, are designed with specific dimensions (L1, W1, L2, W2, L3, W3, L4, W4, W9) to provide suppression at higher frequencies. The frequency response of these cells, depicted in Figure 14, demonstrates their ability to attenuate frequencies beyond the combined resonator's performance at low frequencies. By adding suppressor cells to the combined resonator, the overall system achieves better performance across a wider frequency range, making it suitable for applications requiring suppression at higher frequencies.

Accepted Answer

The proposed LPF has a cutoff frequency of 1.8 GHz and passes signals below this frequency with a very low insertion loss that is close to zero. The frequency response (S11 and S21 parameters) of the proposed LPF is depicted in Figure 16, showing a -3 dB value at 1.8 GHz and a -40 dB value at 2 GHz. This indicates that the proposed LPF effectively filters out signals above the cutoff frequency while maintaining a minimal insertion loss.

Accepted Answer

The normalized circuit size (NCS) is defined in Equation (6). It is a measure of the size of a circuit relative to a reference circuit. The formula takes into account the dimensions of the circuit and the architecture factor (AF) parameter. The AF parameter corresponds to the planar or 3D structure of the circuit, with a value of 1 for 2D circuits and 2 for 3D circuits. The NCS formula helps in comparing the size of different circuits and optimizing the design for efficiency and performance.

Accepted Answer

The proposed Wilkinson Divider Design Process suppresses unwanted harmonics by incorporating one open-ended stub and two filters into the conventional WPD structure. These filters are designed to target and suppress specific harmonics, resulting in improved performance. In the provided section, it is mentioned that the proposed divider has a wide rejection band from 2 GHz to 20 GHz, with an attenuation level of more than 20 dB, effectively suppressing the 2nd-13th unwanted harmonics. This design approach ensures that the divider has good performance at the operating frequency of 1.5 GHz and at higher frequencies, as demonstrated by the frequency response depicted in Figure 18. The proposed divider's ability to suppress unwanted harmonics contributes to its overall effectiveness and superiority compared to other similar works, as shown in Table 4.

Accepted Answer

The LPF designed using the proposed algorithm achieves a sharpness parameter of approximately 185 dB/GHz. This indicates a sharp response within the transmission band, demonstrating excellent performance in the stopband with an attenuation level of more than 20 dB from 2 GHz to 18.2 GHz. The compact size of the LPF (19.3 mm x 9.9 mm, 0.15 lg x 0.081 lg) enables its integration within the divider structure for harmonics suppression, making it a suitable choice for modern communication systems.

A Fast Surrogate Model-Based Algorithm Using Multilayer Perceptron Neural Networks for Microwave Circuit Design

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What techniques have been used to suppress undesirable harmonics in microwave circuit design?

2. How does machine learning aid in microwave circuit design?

3. What is the purpose of designing a preliminary resonator?

4. What is the transition zero frequency of the proposed single polygonal resonator?

5. What is the proposed structure of the preliminary dual polygonal resonator?

6. How many samples are used for training the ANN model?

7. How accurate are the predicted circuit parameters using the proposed ANN model?

8. What is the frequency response of the proposed preliminary dual polygonal resonator?

9. What are the dimensions of the proposed dual-side resonators?

10. What is the frequency response of the proposed combined resonator?

11. How do suppressor cells enhance frequency response?

12. What is the cutoff frequency and insertion loss of the proposed low-pass filter (LPF)?

13. What is the normalized circuit size (NCS) formula?

14. How does the proposed Wilkinson Divider Design Process suppress unwanted harmonics?

15. What is the sharpness parameter of the LPF designed using the proposed algorithm?

Citations

Spoof Surface Plasmon Polariton Based on Stepped Grooves and Its Application in Compact Low-Pass Filter Design

Design of a Microwave Quadrature Hybrid Coupler with Harmonic Suppression Using Artificial Neural Networks

A neural networks approach for designing compact all-optical photonic crystal based AND logic gate

Design of a Microstrip Wilkinson Power Divider Using a Low Pass Filter with the Particle Swarm Optimization Algorithm

Virtual worlds, real opportunities: A review of marketing in the metaverse

References

Suppression of harmonics in Wilkinson power divider using dual-band rejection by asymmetric DGS

Miniaturized microstrip lowpass filter with wide stopband using double equilateral U-shaped defected ground structure

Precise Void Fraction Measurement in Two-phase Flows Independent of the Flow Regime Using Gamma-ray Attenuation

Wilkinson power divider incorporating quasi-elliptic filters for improved out-of-band rejection

Power divider with microstrip electromagnetic bandgap element for miniaturisation and harmonic rejection

Related Papers (5)

Performance Analysis of Resonators for microwave Applications

A miniaturized microstrip Wilkinson power divider with harmonics suppression using radial/rectangular-shaped resonators

Novel Half-Ring Microstrip Resonator

Microstrip bandpass filters using cross-shaped triple-mode resonator

Miniaturized Microstrip Wilkinson Power Divider With Harmonic Suppression