betaFeedback

WorksEngineering

Pattern Synthesizing of the Concentric Ring Arrays Using Recursive Least Square Method

Abstract

This letter presents a recursive technique to synthesize the array factor (AF) of a concentric ring array. In this method, first, the problem is modeled using the traditional least square method (LSM). In the second step, a recursive technique is applied to the defined problem. It is shown that for many practical ring arrays, only a few iteration number is sufficient to obtain a solution with high accuracy. To evaluate the performance and benefits of the proposed method, various concentric ring arrays with different array factor, including equi-ripple pattern, pattern with deep nulls, flat-top array factor, are examined. It is shown that the introduced procedure is a very noble candidate for this purpose.

Introduction

In electrical engineering, especially in the communication field, single element antenna is widely utilized for numerous applications [1, 2, 3, 4, 5, 6]. A single element antenna is simple, cheap, and low profile. Although a single element antenna it is a very good choice, especially for the industrial projects, but its gain is typically low. To increase the gain, several single element antennas should be arranged in a linear, planar, or ring geometry. This arrangement is called the antenna array. An array of single element antennas are extensively used in the various applications such as 5G/6G networks, seekers, and electromagnetic compatibility instruments [7, 8, 9, 10].

All of the antenna arrays have high gain. Also, the radiation pattern of them is easily controlled by changing the magnitude and phase of excitation coefficients of each elements. To control the radiation pattern of an array, a beam-forming network (BFN) is necessary. Additionally, the beam-forming network can control the side lobe level (SLL) of the radiation pattern, increase or decrease the half-power beam-width (HPBW), directivity, etc. However, it is not easy, and there are several challenge for this purpose [11, 12, 13, 14]. So far, several types of beam-forming networks, including BFN based on the substrate integrated waveguide (SIW), microstrip BFN, BFN with non-uniform profile, BFN on a perforated substrate, gap-waveguide BFN, and coplanar BFN [15, 16, 17, 18, 19, 20, 21, 22].

The total radiation pattern of a linear, planar and ring arrays is shaped by picking the appropriate values of the magnetic and phase of array weights. Hereafter, synthesizing the array factor is a chief contest in electrical engineering, especially in antenna engineering. Nowadays, analytical, iterative, and algorithm-based tactics are announced for this problem, which some of them can moderate the number of array components [23, 24, 25, 26, 27, 28]. Similar to the linear and planar arrays, concentric ring arrays are employed in several applications such as radars and satellites. Concentric ring arrays have a low volume occupation. Also, the beam-forming networks of them is simpler than the planar arrays.

So far, a smaller number of techniques are presented for synthesizing the radiation pattern of a concentric ring array. In [29] and [30], the algorithm-based approaches are hired to find the excitation coefficients of the array components of a non-uniform circular arrays. In these works, it is tried to suppress the maximum side lobe levels of the array factor. In [31], a new method is proposed for this type of array, which reduce the number of array rings by the simple try-error technique. In this work, conventional least square method, Richardson technique, and QR factorization are simultaneously used. The synthesis of the array factor of a concentric ring array is a complicated problem. To this end, evolutionary techniques such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO) can also be employed. These algorithms offer a solution with high accuracy. However, the computational cost of these algorithms is very high. This can be a serious problem, especially for the large concentric arrays.

In this letter, and in the first step, the array factor of a concentric ring array is defined. In the second step, the traditional least square method is used to establish a system of linear equations. It is shown that using the LSM method in the traditional formation leads to a solution with low accuracy. So, a recursive technique have to be employed. For practical arrays, only a few iteration is sufficient to reach an acceptable accuracy. To assess the advantages of the proposed recursive method, several concentric arrays are considered. Numerical results approve that the proposed recursive techniques is a promising method. In the next section, theory of the method will be shown. In the following, numerical results will report. The last section includes some conclusions.

Mathematical Theory

Figure 1 shows the geometry of a concentric ring array. In a concentric ring array, the radius of all rings are not small. So, the array factor of it is independent of the azimuth angle in the spherical coordinate. If rnr_n, NrN_r, InI_n and NnN_n represent the radius of ring nn, the number of rings, elements weights for ring nn and the number of elements in ring nn, respectively, the array factor is defined as follows [31].

The geometry of a concentric ring array.

Figure 1. The geometry of a concentric ring array.

F=I0+∑n=1NrInNnJ0(krnu)(1)F = I_0 + \sum_{n=1}^{N_r} I_n N_n J_0\left(k r_n u\right) \tag*{(1)}
u=sin⁡θ(2)u = \sin\theta \tag*{(2)}
Nn=round⁡(2πrndn)(3)N_n = \operatorname{round}\left(\frac{2\pi r_n}{d_n}\right) \tag*{(3)}
dn=2rnsin⁡(πNn)(4)d_n = 2r_n \sin\left(\frac{\pi}{N_n}\right) \tag*{(4)}

where θ\theta, I0I_0 and dnd_n are the elevation angle in spherical coordinate, the weight of center element, and the arc distance between adjacent elements, respectively. It should be noted that equation (1) is used to calculate the array factor of a concentric ring array with a very good approximation. The array factor can be rewritten from its samples [32]. According to the Nyquist theorem, only MM samples is sufficient. So, we have.

Fr=∑m=1MF(um)S(u−um)(5)F_r = \sum_{m=1}^{M} F(u_m)S(u-u_m) \tag*{(5)}

In above equation, FrF_r and S(.)S(.) are the rewritten array factor and the sampling function as follows.

S(u)=sin⁡(Mu/2)Msin⁡(u/2)(6)S(u) = \frac{\sin(Mu/2)}{M\sin(u/2)} \tag*{(6)}

The smallest value of MM is determined as.

M≥2(Nr−1)∣rn+1−rn∣λ(7)M \ge\frac{2(N_r-1)|r_{n+1}-r_n|}{\lambda} \tag*{(7)}

where λ\lambda is the free space wave-number. Using the sampling function, a system of linear equation is established as follows.

AM×(Nr+1)I(Nr+1)×1=BM×1(8)\mathbf{A}_{M \times(N_r+1)}\mathbf{I}_{(N_r+1)\times1}=\mathbf{B}_{M\times1} \tag*{(8)}
A=[Amn]→Amn={NnJ0(krnum)n=1,…,Nr1n=Nr+1(9)\mathbf{A}=[A_{mn}] \rightarrow A_{mn} = \begin{cases} N_n J_0(kr_nu_m) & n=1,\ldots,N_r \\ 1 & n=N_r+1 \end{cases} \tag*{(9)}
B=[Fr(um)];m=1,2,…,M(10)\mathbf{B}=[F_r(u_m)];\quad m=1,2,\ldots,M \tag*{(10)}
I=[I0I1⋯INr]T(11)\mathbf{I}=[I_0\quad I_1\quad\cdots\quad I_{N_r}]^T \tag*{(11)}

In above, A\mathbf{A} is the coefficient matrix of the system of equation, B\mathbf{B} is the vector including the array factor samples, and I\mathbf{I} is a column vector includes the complex excitation coefficients of array. According to the traditional least square method, the solution is found as follows.

I=(ATA)−1ATB(12)\mathbf{I}=(\mathbf{A}^T\mathbf{A})^{-1}\mathbf{A}^T\mathbf{B} \tag*{(12)}

In the traditional LSM, cost of all equations is equal. So, this issue leads to the self-biasing of the problem. As a results, the accuracy of the solution will decrease. To overcome that, recursive technique is proposed.

It is assumed that in the first stage, only MM samples are employed to initial guess of the solution. However, in the recursive least square method, M0M_0 samples should be regarded. Our studies show that the smallest value of M0M_0 is determined as.

M0≃2M≥8(Nr−1)∣rn+1−rn∣λ(13)M_0 \simeq2M \ge\frac{8(N_r-1)|r_{n+1}-r_n|}{\lambda} \tag*{(13)}

The previous equations can be rewritten as follows.

Am+1=[Amam+1T],m=1,2,…,(M0−M)(14)\mathbf{A}_{m+1} = \begin{bmatrix} \mathbf{A}_m \\ a_{m+1}^T \end{bmatrix},\quad m=1,2,\ldots,(M_0-M) \tag*{(14)}
am+1=[am+1]→am+1={Nn0J0(krn+m+Mu)n=1,…,Nr1n=Nr+1(15)\mathbf{a}_{m+1}=[a_{m+1}] \rightarrow a_{m+1}= \begin{cases} N_{n_0}J_0(kr_{n+m+M}u) & n=1,\ldots,N_r \\ 1 & n=N_r+1 \end{cases} \tag*{(15)}
Bm+1=[BBm]T(16)\mathbf{B}_{m+1}=[\mathbf{B}\quad B_m]^T \tag*{(16)}
Bm=Fr(um+M)(17)\mathbf{B}_m=F_r(u_{m+M}) \tag*{(17)}

Finally, the excitation coefficients are calculated in a recursive manner as the following equations.

Im+1=Im+Pm+1am+1{Bm+1−am+1TIm}(18)\mathbf{I}_{m+1}=\mathbf{I}_m+\mathbf{P}_{m+1}\mathbf{a}_{m+1}\{\mathbf{B}_{m+1}-\mathbf{a}_{m+1}^T\mathbf{I}_m\} \tag*{(18)}
Pm+1=Pm−Km+1am+1TPm(19)\mathbf{P}_{m+1}=\mathbf{P}_m-\mathbf{K}_{m+1}\mathbf{a}_{m+1}^T\mathbf{P}_m \tag*{(19)}
Pm=(AmTAm)−1(20)\mathbf{P}_m=(\mathbf{A}_m^T\mathbf{A}_m)^{-1} \tag*{(20)}
Km+1=Pmam+1(am+1TPmam+1+1)−1(21)\mathbf{K}_{m+1}=\mathbf{P}_m\mathbf{a}_{m+1}\left(\mathbf{a}_{m+1}^T\mathbf{P}_m\mathbf{a}_{m+1}+1\right)^{-1} \tag*{(21)}

The studied shown that only three iterations is sufficient. To reduce the number of rings, the similar procedure shown in [31] can be considered.

Numerical Examples and Discussion

In the following, various concentric ring arrays with different array factor are examined. To this end, MATLAB-based program is written.

Flat-Top Array Factor

A flat-top array factor is considered as the first example. This pattern has non-zero values at −0.4≤u≤+0.4-0.4\le u\le+0.4. During the synthesis process, it is assumed that Nr=9N_r=9, rn=nλ/2r_n=n\lambda/2 and Nn=round⁡(2πn)N_n=\operatorname{round}(2\pi n). Obtained two-dimensional array factor and the target array factor are plotted in Figure 2. It is seen that the accuracy is acceptable. The obtained ring weights are displayed in Figure 3. Also, Figure 4 shows the three-dimensional array factor. The array factor is symmetric.

2D array factor of the first example

Figure 2. The 2D array factor of the first example.

Excitation coefficients of the first example

Figure 3. The excitation coefficients of the first example.

3D array factor of the first example

Figure 4. The 3D array factor of the first example.

Difference Array Factor

In the second example, a difference array factor with maximum side lobe level about -25 dB and Nr=11N_r=11 is regarded. After applying the synthesis process, the obtained two-dimensional array factor and the target array factor are depicted in Figure 5. It is seen that the accuracy is reasonable. The obtained ring weights are displayed in Figure 6. Also, Figure 7 shows the three-dimensional array factor. Similar to previous example, the array factor is symmetric.

2D array factor of the second example

Figure 5. The 2D array factor of the second example.

Excitation coefficients of the second example

Figure 6. The excitation coefficients of the second example.

3D array factor of the second example

Figure 7. The 3D array factor of the second example.

Equi-Ripple Array Factor

A concentric ring array with equi-ripple array factor with side lobe level around -30 dB and Nr=10N_r=10 is considered as the third example. The obtained two-dimensional array factor and the target array factor are depicted in Figure 8. It is seen that the accuracy is good. The obtained ring weights are displayed in Figure 9. Also, Figure 10 shows the three-dimensional array factor. Similar to previous examples, the array factor is symmetric.

2D array factor of the third example

Figure 8. The 2D array factor of the third example.

Excitation coefficients of the third example

Figure 9. The excitation coefficients of the third example.

3D array factor of the third example

Figure 10. The 3D array factor of the third example.

Array Factor with Deep Nulls

An equi-ripple array factor with SLL=−16\mathrm{SLL}=-16 dB, two −40-40 dB deep nulls, and Nr=14N_r=14 is considered as the fourth example. The obtained two-dimensional array factor and the target array factor are depicted in Figure 11. It is seen that the accuracy is good. The obtained ring weights are displayed in Figure 12. Also, Figure 13 shows the three-dimensional array factor.

2D array factor of the fourth example

Figure 11. The 2D array factor of the fourth example.

Excitation coefficients of the fourth example

Figure 12. The excitation coefficients of the fourth example.

3D array factor of the fourth example

Figure 13. The 3D array factor of the fourth example.

CONCLUSION

A recursive least square method is proposed to synthesize the array factor of a concentric ring array. First, the problem is modeled by the classical least square technique. Then, a recursive procedure is applied to the described problem. It is shown that for the practical concentric ring arrays, a few iteration number is enough to obtain a solution with high precision. To estimate the performance and profits of the method, several concentric ring arrays with dissimilar array factor, including equi-ripple pattern, pattern with deep nulls, flat-top array factor, are examined.

References

  1. [1]M. Boozari, M. G. H. Alijani, “Investigation on Transient Response of Linear Dipole Antennas in the Presence of HEMP Based on Equivalent Circuit,” Progress in Electromagnetics Research Letters, vol. 66, pp. 39-43, 2017, doi: 10.2528/PIERL16123006.DOI
  2. [2]D. Qin, B. Sun and R. Zhang, “VHF/UHF Ultrawideband Slim Monopole Antenna With Parasitic Loadings,” IEEE Antennas and Wireless Propagation Letters, vol. 21, no. 10, pp. 2050-2054, Oct. 2022, doi: 10.1109/LAWP.2022.3190015.DOI
  3. [3]M. G. H. Alijani, S. Sheikh and A. Kishk, “Development of Closed-Form Formula for Quick Estimation of Antenna Factor,” 2021 15th European Conference on Antennas and Propagation (EuCAP), Dusseldorf, Germany, 2021, pp. 1-5, doi: 10.23919/EuCAP51087.2021.9411008.DOI
  4. [4]Z. Chen et al., “Compact Wideband Circularly Polarized Loop Antenna Based on Dual Common and Differential Modes,” IEEE Antennas and Wireless Propagation Letters, vol. 21, no. 8, pp. 1567-1571, Aug. 2022, doi: 10.1109/LAWP.2022.3174400.DOI
  5. [5]M. Boozari, R. Bayderkhani, M. G. Alijani and M. H. Neshati, “Closed form formula for longitudinal slot of LWAs with minimum side lobe level,” 2017 Iranian Conference on Electrical Engineering (ICEE), Tehran, Iran, 2017, pp. 1820-1824, doi: 10.1109/IranianC.2017.7985348.DOI
  6. [6]C. Chen, “A Compact Wideband Endfire Filtering Antenna Inspired by a Uniplanar Microstrip Antenna,” IEEE Antennas and Wireless Propagation Letters, vol. 21, no. 4, pp. 853-857, April 2022, doi: 10.1109/LAWP.2022.3151800.DOI
  7. [7]A. Tamandani and M. G. H. Alijani, “Development of an Analytical Method for Pattern Synthesizing of Linear and Planar Arrays with Optimal Parameters,” AEU-International Journal of Electronics and Communications, vol. 146, pp. 1-7, March 2022, doi: 10.1016/j.aeue.2022.154135.DOI
  8. [8]K. Marák, J. Kráček and S. Bilicz, “Antenna Array Pattern Synthesis Using an Iterative Method,” IEEE Transactions on Magnetics, vol. 56, no. 2, pp. 1-4, Feb. 2020, Art no. 7508304, doi: 10.1109/TMAG.2019.2952809.DOI
  9. [9]H. Kähkönen, J. Ala-Laurinaho and V. Viikari, “A Modular Dual-Polarized Ka-Band Vivali Antenna Array,” IEEE Access, vol. 10, pp. 36362-36372, 2022, doi: 10.1109/ACCESS.2022.3164201.DOI
  10. [10]J. Tian, S. Lei, Z. Lin, Y. Gao, H. Hu and B. Chen, “Robust Pencil Beam Pattern Synthesis With Array Position Uncertainty,” IEEE Antennas and Wireless Propagation Letters, vol. 20, no. 8, pp. 1483-1487, Aug. 2021, doi: 10.1109/LAWP.2021.3088095.DOI
  11. [11]M. Sharifi, M. Boozari, M. G. Alijani and M. H. Neshati, “Development a New Algorithm to Reduce SLL of an Equally Spaced Linear Array,” Electrical Engineering (ICEE), Iranian Conference on, Mashhad, Iran, 2018, pp. 554-557, doi: 10.1109/ICEE.2018.8472414.DOI
  12. [12]L. Chen, W. Ma, Y. Wang and H. Zhou, “Array Factor Forming With Regularization for Aperture Synthesis Radiometric Imaging With an Irregularly Distributed Array,” IEEE Geoscience and Remote Sensing Letters, vol. 17, no. 1, pp. 97-101, Jan. 2020, doi: 10.1109/LGRS.2019.2914926.DOI
  13. [13]M. G. H. Alijani, M. H. Neshati and M. Boozari, “Side Lobe Level Reduction of any Type of Linear Equally Spaced Array Using the Method of Convolution,” Progress In Electromagnetics Research Letters, vol. 66, pp. 79-84, 2017, doi: 10.2528/PIERL16121608.DOI
  14. [14]L. F. Yepes, D. H. Covarrubias, M. A. Alonso and R. Ferrús, “Hybrid Sparse Linear Array Synthesis Applied to Phased Antenna Arrays,” IEEE Antennas and Wireless Propagation Letters, vol. 13, pp. 185-188, 2014, doi: 10.1109/LAWP.2014.2301012.DOI
  15. [15]H. Karimian-Sarkesh, M. H. Neshati, M. G. H. Alijani, “Development of an analytical method to determine the propagation characteristics of microstrip line on artificial perforated substrates,” AEU-International Journal of Electronics and Communications, vol. 140, pp. 1-8, October 2021, doi: 10.1016/j.aeue.2021.153951.
  16. [16]G. F. Hamberger, S. Trummer, U. Siart and T. F. Eibert, “A Planar Dual-Polarized Microstrip 1-D-Beamforming Antenna Array for the 24-GHz Band,” IEEE Transactions on Antennas and Propagation, vol. 65, no. 1, pp. 142-149, Jan. 2017, doi: 10.1109/TAP.2016.2618847.DOI
  17. [17]M. Alijani-Ghadikolae and M. H. Neshati, “Developing an accurate and simple dispersion analysis of TE10 mode of substrate integrated waveguides,” 2013 21st Iranian Conference on Electrical Engineering (ICEE), Mashhad, Iran, 2013, pp. 1-4, doi: 10.1109/IranianC.2013.6599879.DOI
  18. [18]T. Djerafi and K. Wu, “A Low-Cost Wideband 77-GHz Planar Butler Matrix in SIW Technology,” IEEE Transactions on Antennas and Propagation, vol. 60, no. 10, pp. 4949-4954, Oct. 2012, doi: 10.1109/TAP.2012.2207309.
  19. [19]N. Yasi, M. Boozari, M. G. H. Alijani, M. H. Neshati, “A new method for analyzing non-uniform guiding structures,” Scientific Reports, vol. 13, pp. 1-11, February 2023, doi: 10.1038/s41598-023-30145-6.DOI
  20. [20]M. G. H. Alijani, S. Sheikh, A. Kishk, “Analytical method for single and coupled nonuniform guiding structures,” AEU-International Journal of Electronics and Communications, vol. 162, pp. 1-8, April 2023, doi: 10.1016/j.aeue.2023.154590.DOI
  21. [21]M. G. H. Alijani, S. Sheikh and A. Kishk, “New Coupling Analysis of Transmission Lines by Port Solver Method,” 2022 IEEE USNC-URSI Radio Science Meeting (Joint with AP-S Symposium), Denver, CO, USA, 2022, pp. 36-37, doi: 10.23919/USNC-URSI52669.2022.9887509.DOI
  22. [22]A. Tamandani and M. G. H. Alijani, “Analysis of a nonuniform guiding structure by the adaptive finite-difference and singular value decomposition methods,” ETRI journal, vol. 45, no. 24, pp. 1-9, March 2023, doi: 10.4218/etrij.2022-0076.DOI
  23. [23]J. Tian, S. Lei, Z. Lin, Y. Gao, H. Hu and B. Chen, “Robust Pencil Beam Pattern Synthesis With Array Position Uncertainty,” IEEE Antennas and Wireless Propagation Letters, vol. 20, no. 8, pp. 1483-1487, Aug. 2021, doi: 10.1109/LAWP.2021.3088095.DOI
  24. [24]M. G. H. Alijani and M. H. Neshati, “A New Noniterative Method for Pattern Synthesis of Unequally Spaced Linear Arrays,” International Journal of RF and Microwave Computer-Aided Engineering, vol. 29, no. 11, pp. 1-9, August 2019, doi: 10.1002/mmce.21921.DOI
  25. [25]C. Cui, W. T. Li, X. T. Ye, Y. Q. Hei, P. Rocca and X. W. Shi, “Synthesis of Mask-Constrained Pattern-Recnfigurable Nonuniformly Spaced Linear Arrays Using Artificial Neural Networks,” IEEE Transactions on Antennas and Propagation, vol. 70, no. 6, pp. 4355-4368, June 2022, doi: 10.1109/TAP.2022.3140214.DOI
  26. [26]M. G. H. Alijani and M. H. Neshati, “Development of a New Method for Pattern Synthesizing of Linear and Planar Arrays Using Legendre Transform With Minimum Number of Elements,” IEEE Transactions on Antennas and Propagation, vol. 70, no. 4, pp. 2779-2789, April 2022, doi: 10.1109/TAP.2021.3137200.DOI
  27. [27]M. G. H. Alijani and M. H. Neshati, “Development a New Array Factor Synthesizing Technique by Pattern Integration and Least Square Method,” IEEE Transactions on Antennas and Propagation, vol. 66, no. 12, pp. 6869-6874, Dec. 2018, doi: 10.1109/TAP.2018.2871715.DOI
  28. [28]F. Yang, Y. Ma, Y. Chen, S. Qu and S. Yang, “A Novel Method for Maximum Directivity Synthesis of Irregular Phased Arrays,” IEEE Transactions on Antennas and Propagation, vol. 70, no. 6, pp. 4426-4438, June 2022, doi: 10.1109/TAP.2021.3138543.DOI
  29. [29]M. A. Panduro, C. A. Brizuela, R. Dominguez, G. Romero, “Design of non-uniform circular antenna arrays for side lobe reduction using the method of genetic algorithms,” AEU-International Journal of Electronics and Communications, vol. 60, no. 10, pp. 713-717, 2006, doi: 10.1016/j.aeue.2006.03.006.DOI
  30. [30]M. A. Panduro, C. A. Brizuela, “Evolutionary multi-objective design of non-uniform circular phased arrays,” The International Journal for Computation and Mathematics in Electrical and Electronic Engineering, vol. 27, no. 2, pp. 551-566, 2008, doi: 10.1108/03321640810847823.DOI
  31. [31]A. Tamandani and M. G. H. Alijani, “Development of a new method for pattern synthesizing of the concentric ring arrays with minimum number of rings,” AEU-International Journal of Electronics and Communications, vol. 152, pp. 1-7, July 2022, doi: 10.1016/j.aeue.2022.154262.DOI
  32. [32]M. G. H. Alijani and M. H. Neshati, “Development a New Technique Based on Least Square Method to Synthesize the Pattern of Equally Space Linear Arrays,” International Journal of Engineering, Transactions B: Applications, vol. 32, no. 11, pp. 1620-1626, November 2019, doi: 10.5829/ije.2019.32.11b.13.DOI

Paper details

vlnc:W_wdqnqsjj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 10

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Engineering

Not recorded for this paper

View statement
Looking for related papers…
Contents