Pre-Registration Seminar Electronics Communication ppt

Electronics

1. Pre-Registration Seminar of Pawan Kumar Jaiswal (195EC20) Under the guidance of Dr. Rajarshi Bhattacharya Department of Electronics and Communication Engineering National Institute of Technology, Patna

2. OBJECTIVE OF RESEARCH WORK My research work comes under the field of Microwave Imaging with major objective beings: To design a suitable directive antenna for imaging. To optimize the gain of directive antenna using machine learning . To apply the designed antenna to collect scattered signal for image construction. To detect presence of scatter. To identify the type of scatter.

3. My research work comes under the field of : MICROWAVE IMAGING Application areas: Medical Imaging 2. Nondestructive Testing 3. Concealed weapon detection Through – the-wall imaging Food Contamination detection Microwave imaging Quantitative Imaging Techniques Qualitative Imaging Techniques

4. Area or research Signal Processing / Image Processing Area of Research MICROWAVE IMAGING Application areas- Medical imaging Concealed object detection Other civil work applications Food contamination detection Scattered signal = Total Signal – Incident Signal Processing and Analysis of scattered signal P rediction about Object Under Test (OUT). Incident wave Scattered wave Problem Description Object under test Antenna Design Selection of Application Area

5. key challenges : Selection of Application Area. Antenna Design ( as per application need) . Methodology and algorithms for the analysis of scattered data . Simulation and real time based measurement.

6. Application Area: Microwave Techniques for Food Contamination Detection in Packed Food Jar

7. Literature Review SL NO. TITLE OF PAPER PUBLICATION 1. Noninvasive Inline Food Inspection via Microwave Imaging Technology: An Application Example in the Food Industry IEEE ANTENNAS & PROPAGATION MAGAZINE (October 2020) 2. Microwave Imaging Technology for In-line Food Contamination Monitoring Antennas and Propagation Society International Symposium (2019) 3. Monitoring of Food Contamination via Microwave Imaging ACES 2019 4. Microwave tomography for Food contamination Monitoring EuCAP 2021 5. A Machine learning based Microwave sensing Approach to Food contamination Detection ISCAS 2020 6. A Low Cost and Portable Microwave Imaging System for Breast Tumor Detection Using UWB Directional Antenna array. Scientific Report (NATURE) OCTOBER 2019

8. Problem Statement-1 To design suitable antenna for microwave Imaging

9. Problem statement-2 Optimization of antenna for high gain using machine learning.

10. Problem Statement-3 Detection of presence of impurity in food jar with simulated data and measured data using image processing.

11. Problem statement-4 Identification of types of impurity using a machine learning algorithm.

12. First step of work is to design a suitable Antenna for Microwave Imaging (Problem statement-1) Requirements of Antenna: Highly Directional High Efficiency High Gain Ultra Wideband Less Group Delay Compact size Less Beam Tilting Stable Radiation pattern Horn Antenna Vivaldi Antenna Antipodal Vivaldi Antenna

13. Antipodal Vivaldi Antenna L   L   W W   Top Buttom Fig. 1

14. Antipodal Vivaldi Antenna Design Overview: It behaves as Resonating antenna at lower frequencies and as Travelling wave antenna at higher frequencies. It has two flares: Inner flare & Outer flare. simple feeding structure. Two conductors on same substrate, mirror imaged to each other. L ower cutoff frequency depends on,   =  

15. Design for Basic Antipodal Vivaldi Antenna (BAVA) Stage-I We have calculated the value of flare width at outer end to get lower cutoff frequency near to 2.45 GHz ,so that it could cover the ISM band by using this formula; C is the velocity of light in free space , is the relative permittivity of substrate material (RO4003C-3.38) is the maximum width of the elliptical flare opening =67.08 mm Upper cutoff frequency is designed by width of feed line:   M. T. Islam, M. Z. Mahmud, M. T. Islam, S. Kibria , and M. Samsuzzaman , “A Low Cost and Portable Microwave Imaging System for Breast Tumor Detection Using UWB Directional Antenna array,” Sci. Rep. , vol. 9, no. 1, pp. 1–13, 2019, doi : 10.1038/s41598-019-51620-z =  

16. Design equation of elliptical flare: are the initial and final points of exponential curve. is opening rate of exponential flare.  

17. Substrate selection The sensitivity to substrate thickness is one of the main reasons for performance decrement at higher frequencies. Unwanted modes are excited when the electrical thickness grows. These modes change the phase of waves as they pass along the flare section, causing pattern distortion and increased cross-polarization . To avoid these issues, thin, low-dielectric-constant substrate materials are recommended. Considering this aspect we have chosen (RO4003C ) having a dielectric constant of 3.38 and loss tangent 0.0027 of thickness of 0.508 mm for antenna fabrication.

18. Antenna Layout Stage- I (BAVA)     L 0rigin   L   W W   Fig. 2

19. Calculated Dimensions of BAVA (in mm): L = 135, W= 92 = 1.18, =35   The above design has been simulated in CST (Computer Simulation Tool) for different frequency ranges: (1-22)GHz, (22-35)GHz, (35-45)GHz (45-55) GHz with strict boundary conditions.

20. Lower cutoff frequency at 2.1 GHz and band notches seen near to 3,4,7 & 8 GHz. Fig. 3

21. Antenna Layout Stage-II To reduce the Lower cutoff frequency from 2.1 GHz to cover the DCS band of 1.7 GHz and BLL (Back Lobe Label) . we have used the concept of slot cutting on the both conductor. An elliptical slot has been cut on both conductors .

22. L1 L2 L3 3 0.04 -6 1.5 0.06 L1 L2 L3 31.31 14 17.13 3 0.04 -6 1.5 0.06 L1 L2 L3 31.31 14 17.13 Optimized dimension of parameters Slot Slot Fig. 4

23. Lower Cutoff frequency shifted to 1.3 GHz from 2.1 GHz due to slot and BLL reduces by 2 dB. Fig. 5

24. Antenna Layout Stage-III To make Antenna more Directive; a parasitic patch has been introduced between the flares. In existing literature there are so many different types of parasitic structure has been proposed by various authors. We have adopted elliptical shape of parasitic patch of optimized dimension to get more directive waves; hence gain improved.

25. PATCH + = 1   20 4.5 20 4.5 Fig. 6

26. Gain improvement of approx 2 dB has seen for the frequency range 26-33 GHz. Fig. 7

27. Antenna Layout stage-IV Our intension is to increase the gain of antenna up to maximum level; to do so we have used the dielectric lens of same substrate of elliptical shape at the end of inner flare. The main work of lens is to make wave more converging so that directivity of antenna increases; hence gain improves. There are different types of dielectric lens exits in literature; we have adopted elliptical lens.

28. + = 1   Dielectric lens 40 20 40 20 Fig. 8

29. Peak gain of 2 dB is obtained at the higher frequencies 27-30 GHz Fig. 9

30. Antenna Layout Stage-V Antenna with Novel Metamaterial Metamaterial structure plays a vital role in gain improvement of antenna. In existing literature there are so many different types of metamaterial (in respect of shape, size & placing) has been implemented by the authors to improve the gain. In this work we have designed a novel metamaterial of dimension λg /8 at 22 GHz(8.3mm). The optimized dimension of length/width is 1.2 mm.

31. Parameters M1 M2 M3 M4 M5 M6 Dimensions (mm) 2 2 1.2 0.5 0.4 0.2 Parameters M7 M8 M9 M10 M11 M12 Dimensions (mm) 0.55 0.15 0.1 0.1 0.25 0.1 M1×M2 (2×2 mm 2 ) M3×M3 ( 1.2×1.2 mm 2 ) Fig. 10

32. This metamaterial behaves as zero index metamaterial having both permittivity and permeability near to zero for 1-22 GHz. H k E x z y Fig. 11

33. We have reached to this particular arrangement of meta surface after so many iteration of metamaterial array to make more converging wave in the end fire direction; h ence directivity of We have reached to this particular arrangement of meta surface after so many iteration of metamaterial array to make more converging wave in the end fire direction; h ence directivity of antenna enhanced . vantenna enhanced Metamaterial surface We have reached to this particular arrangement of meta surface after so many iteration of metamaterial array to make the wave more converging in the end fire direction; hence directivity of antenna enhanced . Fig. 12

34. Gain at 1.3 GHz is ;which is not significant to consider. Gain at 1.5 is ;hence lower cut off frequency is 1.5 GHz. Peak gain improvement by 1.92 dB and highest gain obtained is 15.12 dB at 15 GHz. This metamaterial shows negative impact on gain at higher frequencies above 22 GHz particularly at 24 GHz. Fig. 13

35. With Metasurface (b) Without Metasurface Electric Field distribution at 15 GHz With Metasurface (b) Without Metasurface Fig. 9: Electric Field distribution at 24 GHz Fig. 14

36. ( b) 50 GHz 3D-Radiation Pattern with Metasurface at 49 and 50 GHz (a) 49 GHz At higher frequencies greater 49 GHz and 50 GHz the metasurface behaves as reflector ; hence completely reflecting the wave in opposite direction. Fig. 15

37. Antenna Layout: Stage-VI Antenna with notches without Meta surface This meta material is suitable for lower frequencies up to 22 GHz above to this it shows negative impact on antenna gain . To improve the antenna gain at higher frequencies; the concept of notching on inner flares of both conductor has been used . T he two notches work as radiator and it enhance the gain of 5G band frequency for 24-35 GHz. Notches are designed for the 5G frequency (24-27 GHz); width and depth is taken in terms of λg /8 & λg /4 respectively; the optimized dimension are 1 mm & 1.96 mm and 1 mm & 1.47 mm for larger and smaller notch respectively. J. Eichenberger , E. Yetisir , and N. Ghalichechian , “High-Gain Antipodal Vivaldi Antenna With Pseudoelement and Notched Tapered Slot Operating at (2.5 to 57) GHz,” IEEE Trans. Antennas Propag . , vol. 67, no. 7, pp. 4357–4366, Jul. 2019, doi : 10.1109/TAP.2019.2906008.

38. z y Stage-VI Fig. 16

39. By using the two notches on inner flare of each conductor, peak gain of 17.52 dB is obtained at 24 GHz and 27 GHz respectively. The enhancement of gain by 1.8 dB and reduction of SLL by 3.5 dB is obtained for frequency range 24-27 GHz.

40. E-Plane_24 GHz H-Plane_24 GHz E-Plane_27 GHz H-Plane_27 GHz Normalized radiation pattern of Stage IV and VI at 24 & 27 GHz Fig. 17

41. 24 GHz (Stage-VI) with notches 24 GHz (Stage-IV) 27 GHz (Stage-IV) 27 GHz (Stage-VI) with notches E-Field Distribution Fig. 18

42. In this work we have gone for fabrication of antenna with Meta surface (Stage-V) for the frequency range 1-22 GHz. Due to the limitations of Anechoic chamber and measurement facility , we have done our measurement up to 18 GHz at 2.45 GHz,5.5 GHz,15 GHz and 18 GHz for gain and E & H co and cross polarization. Reflection coefficient (S 11 ) is measured using Agilent N5247A Vector Network Analyzer. The measured result is very much similar to simulated result.

43. Top View Bottom View FABRICATION Fig. 19

44. S-Parameter Analysis E-Plane Measurement H-Plane Measurement MEASUREMENT Fig. 20

45. RADIATION PATTERN 2.45 GHz 2.45 GHz 5.5 GHz 5.5 GHz

46. 15 GHz 15 GHz 18 GHz 2.45 GHz 18 GHz

47. 24 GHz 24 GHz 27 GHz 27 GHz Fig. 21

48. Fig. 22

49. Freq. 3-dB Beamwidth (deg.) Max. Gain (dB) SLL (dB) BLL (dB)/-FBR (dB) E-Plane H-Plane E-Plane H-Plane E-Plane H-Plane E-Plane H-Plane 2.45 108.3 158.8 4.24 4.24 -5.1 Nil -21.24 -21.24 5.5 29.8 65.5 10.9 10.9 -9.8 -12.5 -22.01 -22.01 15 23.5 28.0 15.1 15.1 -12.2 -12.0 -16.84 -16.84 18 17.8 34.0 15.0 15.0 -11.4 -6.9 -20.17 -20.17 24 12.7 24.1 17.4 17.4 -12.9 -7.8 -20.24 -20.24 27 11.5 22.2 17.52 17.52 -10.3 -7.3 -26.65 -26.65 28 12.0 23.8 17.2 17.2 -9.3 -6.7 -23.24 -23.24 29 12.4 24.0 17.1 17.1 -9.0 -6.6 -30.83 -30.83 30 13.1 25.6 16.5 16.5 -8.5 -5.0 -23.18 -23.18 37 10.2 31.6 15.0 15.0 -6.7 -7.1 -27.96 -27.96 40 08.9 39.8 13.8 13.8 -6.3 -8.4 -20.07 -20.07 A comparison chart of the co-polarization radiation patterns in terms of Gain, Beam width , SLL, and BLL (2.45-18 GHz-Stage-V; 24-40 GHz-Stage-VI)

50. Simulated Group Delay Stage-V Fig. 23

51. Fig. 24

52. Ref. Dimensions (mm 3 , ) Substrate–Dielectric Constant Peak Realized Gain (dB/ dBi ) Frequency (GHz) – Gain (dB/ dBi ) Operational Bandwidth [12] 96×100×0.51, 0.67×0.7=0.47 RO4003C substrate-3.38 12 4–7.5 20–12 2.1–27 [39] 77×186×0.55, 0.64×1.55=0.99 RT/Duroid 6002 substrate-2.94 16 4–10, 20–12 40–14 2.5–57 [26] 120×260×0.762, 0.4×0.87=0.35 Dielectric Constant–3 14.6 (1–5.3), (8–11.6) (8–28) >11.6 1-28 [40] 40×90×0.51, 0.45×1.022 =0.46 RO4003C substrate-3.38 15 3.4–8, 19.5–12.8 40–14 3.4–40 [41] 40×100×1, 0.37×0.93=0.34 Dielectric constant-2.55 10.55 5 ≥ 7 20 __ 9.5, 40 ≥ 5 2.8–40 [42] ----- Rogers Arlon Iso 917–2.2 16 1–5, 4–14 6–16 1–7 [43] 30×55×0.51, 0.45×0.82=0.37 RT 4003C–3.38 12.7 (without array) 5–4, 15–10.5 30–11, 35–12.7 4.5–50 [44] 70×166×0.76, 1.09×2.60=2.83 RO4350–3.48, 14.8 5–10.3 16–14.8 4.7–20 [45] 30×55×0.254, 2×3.67=7.33 RT/Duroid 5880 substrate–2.2 14.22 20–11 37–14.22 20-40 [46] --, 0.53×0.4=0.21 Fully Metal 12 10–9 45–12 2–50 [47] 21×37×1.6, 0.25×0.45=0.11   FR4 substrate–4.3 4.5 3.5– -0.5 5.5–4.5, 8.5–4 3.6–12 Prop. 95×175×0.51, 0.48×0.88=0.42 RO4003C–3.38 17.52 1.5–3.22, 4–8 15–15.45, 27–17.52 37–15.0, 55–12.4 1.5–55 Ref. Substrate–Dielectric Constant Peak Realized Gain (dB/ dBi ) Frequency (GHz) – Gain (dB/ dBi ) Operational Bandwidth [12] 96×100×0.51, 0.67×0.7=0.47 RO4003C substrate-3.38 12 4–7.5 20–12 2.1–27 [39] 77×186×0.55, 0.64×1.55=0.99 RT/Duroid 6002 substrate-2.94 16 4–10, 20–12 40–14 2.5–57 [26] 120×260×0.762, 0.4×0.87=0.35 Dielectric Constant–3 14.6 (1–5.3), (8–11.6) (8–28) >11.6 1-28 [40] 40×90×0.51, 0.45×1.022 =0.46 RO4003C substrate-3.38 15 3.4–8, 19.5–12.8 40–14 3.4–40 [41] 40×100×1, 0.37×0.93=0.34 Dielectric constant-2.55 10.55 5 ≥ 7 20 __ 9.5, 40 ≥ 5 2.8–40 [42] ----- Rogers Arlon Iso 917–2.2 16 1–5, 4–14 6–16 1–7 [43] 30×55×0.51, 0.45×0.82=0.37 RT 4003C–3.38 12.7 (without array) 5–4, 15–10.5 30–11, 35–12.7 4.5–50 [44] 70×166×0.76, 1.09×2.60=2.83 RO4350–3.48, 14.8 5–10.3 16–14.8 4.7–20 [45] 30×55×0.254, 2×3.67=7.33 RT/Duroid 5880 substrate–2.2 14.22 20–11 37–14.22 20-40 [46] --, 0.53×0.4=0.21 Fully Metal 12 10–9 45–12 2–50 [47] 21×37×1.6, 0.25×0.45=0.11   FR4 substrate–4.3 4.5 3.5– -0.5 5.5–4.5, 8.5–4 3.6–12 Prop. 95×175×0.51, 0.48×0.88=0.42 RO4003C–3.38 17.52 1.5–3.22, 4–8 15–15.45, 27–17.52 37–15.0, 55–12.4 1.5–55 Comparison Table

53. [1] P. J. Gibson, “VIVALDI AERIAL.,” in Conference Proceedings - European Microwave Conference , 1979, pp. 101–105, doi : 10.1109/euma.1979.332681. [2] E. Gazit , “Improved design of the Vivaldi antenna,” IEE Proc. H Microwaves, Antennas Propag . , vol. 135, no. 2, p. 89, 1988, doi : 10.1049/ip-h-2.1988.0020. [3] J. Fisher, “Design and Performance Analysis of a 1-40GHz Ultra-Wideband Antipodal Vivaldi Antenna,” Proc. Ger. Radar Symp . GRS , vol. 6002, 2000, [Online]. Available: http://www.roke.co.uk/resources/papers/Analysis-of-a-Ultra-wideband-Antipodal-Vivaldi-Antenna.pdf. [4] J. Y. Siddiqui , Y. M. M. Antar , A. P. Freundorfer , E. C. Smith, G. A. Morin, and T. Thayaparan , “Design of an ultrawideband antipodal tapered slot antenna using elliptical strip conductors,” IEEE Antennas Wirel . Propag . Lett. , vol. 10, pp. 251–254, 2011, doi : 10.1109/LAWP.2011.2128296. [5] R. Natarajan, J. V. George, M. Kanagasabai , and A. Kumar Shrivastav , “A Compact Antipodal Vivaldi Antenna for UWB Applications,” IEEE Antennas Wirel . Propag . Lett. , vol. 14, pp. 1557–1560, 2015, doi : 10.1109/LAWP.2015.2412255 . References:

54. [6] M. Sun, X. Qing, and Z. N. Chen, “60-GHz end-fire fan-like antennas with wide beamwidth ,” IEEE Trans. Antennas Propag . , vol. 61, no. 4, pp. 1616–1622, 2013, doi : 10.1109/TAP.2012.2237153. [7] A. M. De Oliveira, M. B. Perotoni , S. T. Kofuji , and J. F. Justo, “A palm tree Antipodal Vivaldi Antenna with exponential slot edge for improved radiation pattern,” IEEE Antennas Wirel . Propag . Lett. , vol. 14, pp. 1334–1337, 2015, doi : 10.1109/LAWP.2015.2404875. [8] P. Fei , Y. C. Jiao, W. Hu, and F. S. Zhang, “A miniaturized antipodal vivaldi antenna with improved radiation characteristics,” IEEE Antennas Wirel . Propag . Lett. , vol. 10, pp. 127–130, 2011, doi : 10.1109/LAWP.2011.2112329. [9] J. Bai, S. Shi, and D. W. Prather, “Modified compact antipodal Vivaldi antenna for 4-50-GHz UWB application,” IEEE Trans. Microw . Theory Tech. , vol. 59, no. 4 PART 2, pp. 1051–1057, Apr. 2011, doi : 10.1109/TMTT.2011.2113970. [10] W. Mazhar , D. Klymyshyn , and A. Qureshi, “Log periodic slot-loaded circular vivaldi antenna for 5–40 GHz UWB applications,” Microw . Opt. Technol. Lett. , vol. 59, no. 1, pp. 159–163, Jan. 2017, doi : 10.1002/mop.30252.

55. [ 11] M. Abbak , M. N. Akinci , M. Çayören , and I. Akduman , “Experimental microwave imaging with a novel corrugated vivaldi antenna,” IEEE Trans. Antennas Propag . , vol. 65, no. 6, pp. 3302–3307, 2017, doi : 10.1109/TAP.2017.2670228. [12] M. Moosazadeh and S. Kharkovsky , “Design of Ultra-Wideband Antipodal Vivaldi Antenna for Microwave Imaging Applications,” Nov. 2015, doi : 10.1109/ICUWB.2015.7324435. [13] M. H. H. Agahi , H. Abiri , and F. Mohajeri , “Investigation of a New Idea for Antipodal Vivaldi Antenna Design,” Int. J. Comput . Electr . Eng. , vol. 3, no. 2, pp. 277–281, 2011, doi : 10.7763/ijcee.2011.v3.327. [14] I. T. Nassar and T. M. Weller, “A Novel Method for Improving Antipodal Vivaldi Antenna Performance,” IEEE Trans. Antennas Propag . , vol. 63, no. 7, pp. 3321–3324, Jul. 2015, doi : 10.1109/TAP.2015.2429749. [15] M. Moosazadeh and S. Kharkovsky , “A Compact High-Gain and Front-to-Back Ratio Elliptically Tapered Antipodal Vivaldi Antenna With Trapezoid-Shaped Dielectric Lens,” IEEE Antennas Wirel . Propag . Lett. , vol. 15, pp. 552–555, 2016, doi : 10.1109/LAWP.2015.2457919.

56. [16] M. Moosazadeh , “High-gain antipodal vivaldi antenna surrounded by dielectric for wideband applications,” IEEE Trans. Antennas Propag . , vol. 66, no. 8, pp. 4349–4352, Aug. 2018, doi : 10.1109/TAP.2018.2840839. [17] M. Amiri , F. Tofigh , A. Ghafoorzadeh-Yazdi , and M. Abolhasan , “Exponential Antipodal Vivaldi Antenna with Exponential Dielectric Lens,” IEEE Antennas Wirel . Propag . Lett. , vol. 16, pp. 1792–1795, 2017, doi : 10.1109/LAWP.2017.2679125. [18] M. Bhaskar , E. Johari, Z. Akhter, and M. J. Akhtar, “Gain enhancement of the Vivaldi antenna with band notch characteristics using zero-index metamaterial,” Microw . Opt. Technol. Lett. , vol. 58, no. 1, pp. 233–238, Jan. 2016, doi : 10.1002/mop.29534. [19] P. Usha and C. Krishnan, “Epsilon near zero metasurface for ultrawideband antenna gain enhancement and radar cross section reduction,” AEU - Int. J. Electron. Commun . , vol. 119, p. 153167, 2020, doi : 10.1016/j.aeue.2020.153167. [20] M. Sun, Z. N. Chen, and X. Qing, “Gain enhancement of 60-GHz antipodal tapered slot antenna using zero-index metamaterial,” IEEE Trans. Antennas Propag . , vol. 61, no. 4, pp. 1741–1746, 2013, doi : 10.1109/TAP.2012.2237154 . [21] S. Operating, “High-Gain Antipodal Vivaldi Antenna With Pseudoelement and Notched Tapered,” IEEE Trans. Antennas Propag . , vol. 67, no. 7, pp. 4357–4366, 2019, doi : 10.1109/TAP.2019.2906008. [22] M. T. Islam, M. Z. Mahmud, M. T. Islam, S. Kibria , and M. Samsuzzaman , “A Low Cost and Portable Microwave Imaging System for Breast Tumor Detection Using UWB Directional Antenna array,” Sci. Rep. , vol. 9, no. 1, pp. 1–13, 2019, doi : 10.1038/s41598-019-51620-z.

57. [ 23] 2015. Balanis , Constantine A. Antenna theory: analysis and design. John wiley & sons, Book . . [24] M. Moosazadeh , Antipodal Vivaldi Antennas for Microwave Imaging of Construction Materials and Structures . Cham: Springer International Publishing, 2019. [25] S. Das and D. Mitra , “A compact wideband flexible implantable slot antenna design with enhanced gain,” IEEE Trans. Antennas Propag . , vol. 66, no. 8, pp. 4309–4314, 2018, doi : 10.1109/TAP.2018.2836463 . [26] X. Shi, Y. Cao, Y. Hu, X. Luo, H. Yang, and L. H. Ye, “A High-Gain Antipodal Vivaldi Antenna With Director and Metamaterial at 1–28 GHz,” IEEE Antennas Wirel . Propag . Lett. , vol. 20, no. 12, pp. 2432–2436, Dec. 2021, doi : 10.1109/LAWP.2021.3114061. [27] S. El- Nady , H. M. Zamel , M. Hendy, A. H. A. Zekry , and A. M. Attiya , “Gain enhancement of a millimeter wave antipodal vivaldi antenna by epsilon-near-zero metamaterial,” Prog . Electromagn . Res. C , vol. 85, no. July, pp. 105–116, 2018, doi : 10.2528/pierc18050302. [28] S. Geetha Priyadharisini and E. Rufus, “A double negative metamaterial inspired miniaturized rectangular patch antenna with improved gain and bandwidth,” Prog . Electromagn . Res. Symp . , vol. 2017-Novem, pp. 2907–2913, 2017, doi : 10.1109/PIERS-FALL.2017.8293629. [29] A. Kumar, A. Q. Ansari, B. K. Kanaujia , J. Kishor , and L. Matekovits , “A Review on Different Techniques of Mutual Coupling Reduction Between Elements of Any MIMO Antenna. Part 2: Metamaterials and Many More,” Radio Sci. , vol. 56, no. 3, pp. 1–22, 2021, doi : 10.1029/2020rs007222. [30] S. Enoch, G. Tayeb , P. Sabouroux , N. Guérin , and P. Vincent, “A Metamaterial for Directive Emission,” Phys. Rev. Lett. , vol. 89, no. 21, pp. 1–4, 2002, doi : 10.1103/PhysRevLett.89.213902

58. Problem statement-2 Optimization of antenna with notches for high gain using machine learning.

59. z y Fig. 25

60. In the first work : Two notches are cut on inner flares of AVA to improve the gain at the frequencies 24 GHz and 27 GHz .The width and depth of notches has been taken in terms of the respective frequencies of interest. The spacing between the notches is taken near to to optimize the gain . [ J. Eichenberger , E. Yetisir , and N. Ghalichechian , “High-Gain Antipodal Vivaldi Antenna With Pseudoelement and Notched Tapered Slot Operating at (2.5 to 57) GHz,” IEEE Trans. Antennas Propag . , vol. 67, no. 7, pp. 4357–4366, Jul. 2019, doi : 10.1109/TAP.2019.2906008.] In case of more number of notches (more than two) it is very difficult to identify which parameters among width, depth and spacing is going to dominant to improve the gain . In literature very few papers are available on Antenna using Notches. In those papers the author has optimized the dimension of parameters either by CST optimizer tool or by Hit and Trail method to improve the gain but no one has proposed the concept to identify : which parameters of notch dimension is playing the key role to improve the Gain of antenna for a particular frequency band.  

61. Top view Bottom view Fig. 26

62. P1 P2 P3 P4 Q R S T U V W 1 2 3 4 Fig. 27

63. 1 1 Q Q Q Q 2 3 4 P3 P1 P2 P4 1 Q S U W R T V Fig. 28

64. Problem Statement: To identify the parameters which are responsible for overall gain improvement for a frequency band in case of more number of notches on the inner flare of AVA. P1 : T : P4 : Q : U : R : P2 : V : S : W : Notches 1,2,3 & 4 corresponds to the frequency 24,27,30,40 GHz respectively. Frequency band (22-40 GHz) The different parameters of Notches are:

65. The Base value of parameters P1, Q, R, S, P2, T, U, P3,V, W, P4 are respectively: 0.94, 1.89, 7.59, 1.68, 0.84, 6.74, 1.51, 0.75, 6.07, 1.14, 0.57 In this problem statement: Decision has to be taken regarding the Parameters, which will improve the overall gain. ML Algorithms: 1. Principle Component Analysis (PCA) 2.Decision Tree Algorithm(CART) 3. Random Forest Algorithm 4. Support V ector Machine (SVM) 5. ANN

66. Antenna Design equations + = 1, for i=6 Notch Design:    

67. Fig. 29

68. TABLE I SPECIFICATIONS FOR THE PROPOSED AVA

69. TABLE II OPTIMIZED DIMENSIONS OF NOTCHES

70. Ideal characteristics of the band-limited approach Simulated realized gain due to single notches and with all notches Fig. 30

71. MACHINE LEARNING APPROACH AND MODEL CONSTRUCTION Process diagram for data set creation Process diagram for Train and Test data-set creation Fig. 31

72. Process diagram for antenna gain classification Fig. 32

73. Data Set Creation for Machine Learning Parameters values are created with 5%,10% and 15% variations in base value Parameters Q R S T U V W Base Data 1.89 7.59 1.68 6.74 1.51 6.07 1.14 10% Incremented Data Logic: YES 2.08 8.35 1.85 7.41 1.66 6.67 1.25 10% Decremented Data Logic :NO 1.7 6.83 1.51 6.07 1.36 5.47 1.03

74. Q R S T U V W Average Gain 1.7 6.83 1.51 6.07 1.36 5.47 1.03   1.7 6.83 1.51 6.07 1.36 5.47 1.25   1.7 6.83 1.51 6.07 1.36 6.67 1.03   1.7 6.83 1.51 6.07 1.36 6.67 1.25   1.7 6.83 1.51 6.07 1.66 5.47 1.03   1.7 6.83 1.51 6.07 1.66 5.47 1.25   1.7 6.83 1.51 6.07 1.66 6.67 1.03   1.7 6.83 1.51 6.07 1.66 6.67 1.25   1.7 6.83 1.51 7.41 1.36 5.47 1.03   7 parameters forms total 128 combinations. Here binary classification is used (YES/NO) for ML algorithms.

75. P Q R S T U V W GAIN NO NO NO NO NO NO NO YES   NO NO NO NO NO NO NO YES   NO NO NO NO NO NO YES YES   NO NO NO NO NO NO YES YES   NO NO NO NO NO YES NO YES   NO NO NO NO NO YES NO YES   NO NO NO NO NO YES YES YES   NO NO NO NO NO YES YES YES   NO NO NO NO YES NO NO YES   NO NO NO NO YES NO NO YES   NO NO NO NO YES NO YES YES   NO NO NO NO YES NO YES YES   NO NO NO NO YES YES NO YES   NO NO NO NO YES YES NO YES   NO NO NO NO YES YES YES YES   NO NO NO NO YES YES YES YES   NO NO NO 2.18 NO NO NO YES   NO NO NO 2.18 NO NO NO YES   NO NO NO 2.18 NO NO YES YES  

76. Steps to be carried out : Creation of 128 data sets using 5%,10% and 15% variation in base data . Simulate the Antenna for all data set and find the realized gain for all frequencies (22-40)GHz . Convert the data set in binary data set (YES/NO) . Apply the Decision Tree Algorithm and Random forest Algorithm one by one to identify the root parameters ; which are responsible for Gain improvement . 5. Validate the result obtained from both algorithms.

77. Result Analysis

78. FABRICATION AND MEASUREMENT OF ANTENNA Fig. 33

79. Fig. 34

80. Radiation Pattern

81. Fig. 35

82. E-Field Plot Fig. 36

83. COMPARISON AND RESULT ANALYSIS Fig. 37

84. PROPOSED ANTENNA COMPARATIVE ANALYSIS WITH EXISTING WORKS

85. References [1] P. J. Gibson, “VIVALDI AERIAL.,” in Conference Proceedings - European Microwave Conference, 1979, pp. 101–105, doi: 10.1109/euma.1979.332681. [2] E. Gazit, “Improved Design of the Vivaldi Antenna.,” IEE Proc. H Microwaves, Antennas Propag., vol. 135, no. 2, pp. 89–92, 1988, doi: 10.1049/ip-h-2.1988.0020. [3] J. Fisher, “Design and Performance Analysis of a 1-40GHz Ultra-Wideband Antipodal Vivaldi Antenna,” Proc. Ger. Radar Symp. GRS, vol. 6002, pp. 1–5, 2000. [4] R. Natarajan, J. V. George, M. Kanagasabai, and A. Kumar Shrivastav, “A Compact Antipodal Vivaldi Antenna for UWB Applications,” IEEE Antennas Wirel. Propag. Lett., vol. 14, pp. 1557–1560, 2015, doi: 10.1109/LAWP.2015.2412255. [5] M. Sun, X. Qing, and Z. N. Chen, “60-GHz end-fire fan-like antennas with wide beamwidth,” IEEE Trans. Antennas Propag., vol. 61, no. 4, pp. 1616–1622, 2013, doi: 10.1109/TAP.2012.2237153.

86. [6] A. M. De Oliveira, M. B. Perotoni, S. T. Kofuji, and J. F. Justo, “A palm tree Antipodal Vivaldi Antenna with exponential slot edge for improved radiation pattern,” IEEE Antennas Wirel. Propag. Lett., vol. 14, pp. 1334–1337, 2015, doi: 10.1109/LAWP.2015.2404875. [7] P. Fei, Y. C. Jiao, W. Hu, and F. S. Zhang, “A miniaturized antipodal vivaldi antenna with improved radiation characteristics,” IEEE Antennas Wirel. Propag. Lett., vol. 10, pp. 127–130, 2011, doi: 10.1109/LAWP.2011.2112329. [8] J. Bai, S. Shi, and D. W. Prather, “Modified compact antipodal Vivaldi antenna for 4- 50-GHz UWB application,” IEEE Trans. Microw. Theory Tech., vol. 59, no. 4 PART 2, pp. 1051–1057, Apr. 2011, doi: 10.1109/TMTT.2011.2113970. [9] W. Mazhar, D. Klymyshyn, and A. Qureshi, “Log periodic slot-loaded circular vivaldi antenna for 5–40 GHz UWB applications,” Microw. Opt. Technol. Lett., vol. 59, no. 1, pp. 159–163, Jan. 2017, doi: 10.1002/mop.30252. [10] M. Abbak, M. N. Akinci, M. Çayören, and I. Akduman, “Experimental microwave imaging with a novel corrugated vivaldi antenna,” IEEE Trans. Antennas Propag., vol. 65, no. 6, pp. 3302–3307, 2017, doi: 10.1109/TAP.2017.2670228.

87. Problem Statement-3 Detection of presence of impurity in food jar with simulated data using image processing.