Antenna Designing And Optimization Using Machine Learning For ISM Band Application

Project Title

Antenna Designing And Optimization Using Machine Learning For ISM Band Application

Project Area of Specialization Artificial IntelligenceProject Summary

INTRODUCTION: 

Microstrip Antenna: In telecommunication, a microstrip antenna is a kind of internal antenna. They are mostly used at microwave frequencies. Antenna optimization aims at creating advanced and complex electromagnetic devices that must be competitive in terms of performance, serviceability, and cost-effectiveness.

ISM BAND: The industrial, scientific, and medical radio band (ISM band) refers to a group of radio bands reserved for the use of Radio Frequency (RF) for scientific, medical, and industrial applications. ISM bands are generally open license bands, used for scientific, medical, and industrial applications.

PROBLEM STATEMENT: The upcoming era of the Internet of Things (IoT) has enabled an immense growth in the demand for application-specific antennas, which are needed for almost all electronic devices. Hence, the requirement for a smart and efficient way of antenna designing has become inevitable. Current antenna design heavily relies on the designer’s empirical experiences and EM simulations. Traditional methods are inherently inefficient and computationally intensive, making them impractical when there are a large number of antenna design parameters to be optimized such as for 3-D printed antennas.

OUR APPROACH TO THE PROBLEM: Our idea is that artificial Intelligence and  Machine Learning (ML) is a promising choice to provide automated, computational feasible, and practically effective approaches for antenna design. The ultimate goal of this communication is to further extend the proposed ideas to the more complex design of antennas and develop scalable and efficient algorithms to tackle computational challenges, by handling a large number of design parameters.

OUR CONTRIBUTION: The main contribution of this communication is to fill the gap by presenting new classes of ML-based methods for automated antenna design optimization, evaluating their performance in terms of prediction accuracy and robustness, and making comparisons with EM simulations.

Our work shows that ML is a promising choice to provide automated, computational feasible, and practically effective approaches for antenna design for the industrial, scientific, and medical  (ISM) bands. The ultimate goal of this communication is to further extend the proposed ideas to the more complex design of antennas and develop scalable and efficient algorithms to tackle computational challenges, by handling a large number of design parameters.

Project Objectives

PROJECT OBJECTIVES:

 To address challenges in designing complex 3-D structures, machine learning (ML) techniques may be highly beneficial. ML has been widely used as indispensable data analysis and decision-making tool in a broad range of applications, ranging from hand-written digit recognition to human genomics.

Our work shows that artificial Intelligence and  Machine Learning (ML) is a promising choice to provide automated, computational feasible, and practically effective approaches for the industrial, scientific, and medical  (ISM) band antenna design.

Our ultimate goal is to develop a machine-learning algorithm to solve and predict the different parameters of the complex design of antennas and develop scalable and efficient algorithms to tackle computational challenges, by handling a large number of design parameters.

our model will make an automated way to design a microstrip antenna for the ISM band applications.

Project Implementation Method

Designing Sample Antenna in CST Software:

first, we designed an antenna in CST  software and simulated it as just a random design having a turning frequency is near to 2.38GHz.  

CST Simulations:

We changed the antenna parameters one by one and run the simulation for different values of the parameters we did a large number of simulations more than 400 sets. the purpose of this simulation is to extract data from CST that we will use for the training of our machine learning model. 

Data Collection and pre-Preprocessing:

from CST we extract all the simulated data as a . CSV file the data contain the different values of length width frequency etc for every simulated result.  the data are pre-processed and cleaned.

Machine Learning Algorithm:

Machine Learning algorithms are developed using support vector regression (SVR) using linear and Radial Basis Function (RBF) and artificial neural networks in the model designing we did data pre-processing, and data scaling, and then we fit the model on the data that we extract from CST software. 

Split Data into Train and Test Sets:

the data are divided into two sets one set is used for training the model while the other one is used for the testing of the model 

New antenna requirements:

once the model gets trained then the machine will predict the new desired feature for the new antenna  based on the requirements 

Predicted Features By ML model:

the machine learning model learns from data that we extract from simulated software and gets trained for a new feature to be predicted our machine learning model shows a good result by testing a new unable feature the result ml model predicts the feature with  99.4% accuracy.

Validate the Features Accuracy using  CST  Software:

the feature that we get from the machine learning model was validated using simulated software CST the result shows that the feature values that our ML model predict is almost similar to the original value that we get from CST software the accuracy of the ML model is approximately equal to 99.4%

Fabrications and Testing:

once the microstrip antenna is simulated based on the user requirements like its resonance frequency, reflection coefficient, directivity, radiation pattern, etc.

the antenna is fabricated using the commercial service of a printed circuit board manufacturing company the antenna is sent for testing and measurements 

Final Designed Antenna:

finally, once the required results are achieved the antenna is now ready and can be used in ISM  band applications.

Benefits of the Project

Benefits Of The Project:

The upcoming era of the Internet of Things (IoT) has enabled an immense growth in the demand for application-specific antennas, which are needed for almost all electronic devices. Hence, the use of machine learning for a smart and efficient way of antenna designing has become inevitable.

Current antenna design heavily relies on the designer’s empirical experiences and EM simulations. our machine learning will make it easy as our machine learning algorithms will learn from the previous data and will predict new design parameters based on the requirements of the designer. 

Traditional methods are inherently inefficient and computationally intensive, making them impractical when there are a large number of antenna design parameters to be optimized such as for 3-D printed antennas. by using a machine learning algorithm it will make the design process fast and accurate.

To address challenges in designing complex 3-D structures, machine learning (ML) techniques are highly beneficial. ML has been widely used as indispensable data analysis and decision-making tool in a broad range of applications, ranging from hand-written digit recognition to human genomics.

The main contribution of this communication is to fill the gap by presenting new classes of ML-based methods for automated antenna design optimization, evaluating their performance in terms of prediction accuracy and robustness, and making comparisons with EM simulations.

Our model suggests that ML is a promising choice to provide automated, computational feasible, and practically effective approaches for antenna design. The ultimate goal of this communication is to further extend the proposed ideas to the more complex design of antennas and develop scalable and efficient algorithms to tackle computational challenges, by handling a large number of design parameters.

Technical Details of Final Deliverable

Technical Details:

for industrial, scientific, and medical  (ISM) bands.  application-specific antenna designs are required, which are needed for almost all electronic devices. Hence, the use of machine learning for a smart and efficient way of antenna designing has become inevitable.  the benefit to use our model will make the design fast, accurate, and automated.

Current antenna design heavily relies on the designer’s empirical experiences and EM simulations. our machine learning will make it easy as our machine learning algorithms will learn from the previous data and will predict new design parameters based on the requirements of the designer. 

once the designer specifies has requirements for a custom design of the antenna then our model takes these requirements as input and predicts the antenna design parameters fast as compared to electromagnetic simulation software which takes a larger time to simulate the antenna for the desired applications.

Traditional methods are inherently inefficient and computationally intensive, making them impractical when there are a large number of antenna design parameters to be optimized such as for 3-D printed antennas. by using a machine learning algorithm it will make the design process fast and accurate.

To address challenges in designing complex 3-D structures, machine learning (ML) techniques are highly beneficial. ML has been widely used as indispensable data analysis and decision-making tool in a broad range of applications, ranging from hand-written digit recognition to human genomics.

The main contribution of this communication is to fill the gap by presenting new classes of ML-based methods for automated antenna design optimization, evaluating their performance in terms of prediction accuracy and robustness, and making comparisons with EM simulations.

Our ML model is a promising choice to provide automated, computational feasible, and practically effective approaches for antenna design. The ultimate goal of this communication is to further extend the proposed ideas to the more complex design of antennas and develop scalable and efficient algorithms to tackle computational challenges, by handling a large number of design parameters. 

Final Deliverable of the Project HW/SW integrated systemCore Industry TelecommunicationOther Industries IT Core Technology Artificial Intelligence(AI)Other Technologies OthersSustainable Development Goals Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	52000
machine learning algorithm development	Equipment	1	30000	30000
fabrication cost of the antenna	Equipment	1	9000	9000
Arduino UNO	Equipment	2	3000	6000
testing facility and measurements	Miscellaneous	1	5000	5000
adapters	Miscellaneous	2	600	1200
connector cable	Equipment	2	400	800