Antenna Layout Optimization Method Based on MATLAB and CST Interactive Simulation

Chengliang Zhong1, a, *, Xiaodong Mu1, b, and Xiangchen He2, c 1Research Institute of Hi-Tech, Xi’an 710025, China

2Beijing Institute of Remote Sensing Technology, Beijing 100039, China ajhxl2@163.com, bmxd402@163.com, chithxc@aliyun.com.

*Corresponding author

Keywords: layout optimization, genetic algorithm, electromagnetic simulation, interactive simulation

Abstract: Due to the limited space of the carrier, the multi-antenna system on the carrier is prone to electromagnetic interference. Most of the researchers used genetic algorithms to optimize the antenna layout. On this basis, this paper proposes a method of interactive simulation of MATLAB and electromagnetic simulation software CST. Firstly, the position information of the antenna and the carrier platform is initialized in MATLAB, and then the program of the genetic algorithm is run, and the layout scheme of each generation is automatically transmitted to the CST through the ActiveX. The CST solves the antenna coupling degree and automatically returns it to MATLAB as the fitness value of the scheme. After dozens of generations of genetic algorithms, a better layout scheme can be found. Compared with the previous optimization methods, the method of this paper does not need to build complex electromagnetic numerical calculation model, and it has strong operability and high degree of visualization. Therefore, it has certain reference for the layout optimization of multi-antenna systems of other carriers.

1. Introduction

In order to meet different communication needs, a variety of communication devices are often integrated in the same carrier. However, due to the limited space of the carrier, most of the antennas of the communication device are densely mounted on the surface of the carrier, such as communication vehicles, aircraft, ships, etc., which leads to electromagnetic compatibility problems between antennas of different devices. Therefore, in a limited space range, it is necessary to optimize the layout of the antenna without affecting the performance of each antenna.

The layout of the antenna is a complex and nonlinear optimization problem. To solve it, the researchers mostly use the genetic algorithm which is suitable for solving complex problems such as global optimization and nonlinearity. In the solution process, since the degree of coupling between antennas is an important index for evaluating electromagnetic compatibility between antennas, the

2018 3rd International Conference on Computer Science and Information Engineering (ICCSIE 2018)

Published by CSP © 2018 the Authors 258

fitness value in genetic algorithms is generally measured by the degree of coupling, but the calculation of coupling is a more complicated problem in the electromagnetic field. Some authors used the methods of computational electromagnetic to program to calculate the coupling degree, such as method of moments (MoM) in [1], but this is very difficult and time consuming. With the development of commercial electromagnetic computing software, more and more researchers use CST, FEKO and other electromagnetic software to analyze the coupling degree of the antenna [2,3]. The author wrote the genetic algorithm through the VBA macro language inside the CST software in [4], and passed the calculation of the coupling degree to the CST, and automatically searched through the global layout scope to obtain a better scheme, but there is still a lack of visualization and data processing. In this paper, considering the advantages of MATLAB in numerical calculation, data analysis and visualization, the automatic control of CST by MATLAB is realized through Activex, and the optimization of antenna layout is completed by genetic algorithm, which has good practicability.

2. Antenna Layout Optimization Model

2.1 Coupling Degree.

A multi-antenna system can be regarded as a generalized multi-port network [2], each antenna corresponding to one port of a generalized network, and the feeder line is equivalent to a port transmission line. According to the generalized network definition, the coupling degree between antenna feeders can be represented by parameter S of multi-port network. The relationship between the normalized incident wave and the reflected wave of each port of the n-port network can be represented by an S-parameter matrix, shown in Equation (1).

1 11 12 1 1

2 21 22 2 2

1 2

n

n

n n nn nn

b S S S ab S S S a

S S S ab

=

, (1)

Where iiS represents the reflection coefficient of port i when all ports except port i are connected to match load. ijS indicates the transmission coefficient of port j to port i when all ports except port j are connected to match load. Thus, the coupling degree between the antennas can be obtained by

the Equation (2).

( ) 20lg ijL w S= . (2)

Therefore, S21 solved by CST is used to measure the coupling degree of the two antennas.

2.2 Optimization Model.

As the objective function to be optimized, there are many factors affecting the coupling between the antennas, but the most important one is the relationship between the geometric positions of the antennas. Since a pair of antennas have a coupling degree, as a multi-antenna system, it is necessary to reduce the overall coupling degree in the system. Therefore, based on the geometric position of the antenna as the independent variable and the total coupling degree of the antenna system as the dependent variable, an optimization model is established in this paper. If there are m fixed antennas

259

and n movable antennas ( 3, 1m n n+ ≥ > ) in a certain layout range, when the position of the movable antenna is adjusted, the coupling degree between the fixed antennas is constant, so the number of antenna pairs to be examined is 2 2

m n nN C C+= − . Since the importance of different antennas may be different, a weighting factor is added before each coupling degree. The final optimization model is shown in Equation (3).

1 21 1

1min 1 1max

2min 2 2max

min max

min ( , , )

.

n m

m ij iji j i

m m m

f X X X w S

s t X X XX X X

X X X

= = +

=

≤ ≤≤ ≤

≤ ≤

∑ ∑

, (3)

Where ijw is the weighting coefficient of the antenna i and j and iX is the coordinate vector of the antenna i, which has a certain range limitation.

The essence of antenna layout optimization is a numerical optimization problem with upper and lower bound constraints. The genetic algorithm is adopted in this paper, and the specific ideas are as follows:

Step1: Encoding. The position of the movable antenna is encoded. An individual is a layout scheme, that is, every encoded sequence contains the position information of all the movable antennas.

Step2: Calculating coupling degree. Through Activex, MATLAB introduces the decoded position information and simulation parameters into CST, and runs the CST time domain solver to calculate the coupling degree between the antennas and save it as a ‘data’ file. After MATLAB reads the coupling degree data of each antenna, according to Equation (3), the total coupling degree in the scheme is calculated.

In order to accelerate the simulation, the layout plan and the coupling degree calculated by the CST are stored in an external file each time. Before calculating the coupling degree of a certain scheme, it is first determined whether the scheme exists in this file. If it exists, it is read directly, without calling CST, because each simulation takes much more time than the query.

Step3: Selection, crossover and mutation. With the degree of coupling as the fitness of the individual, if the convergence condition is not met, the corresponding selection, crossover, and mutation operations are performed according to the principle of the genetic algorithm to generate the next generation of new individuals, and then the new coupling degree is calculated by entering Step 2.

Step4: Model correction. By analyzing the optimal coupling degree of each generation and genetic algebra, the related images are drawn, and the crossover probability and mutation probability of the genetic algorithm are adjusted to prevent the local optimal search.

3. Implementation Process

3.1 MATLAB Controls CST.

Through Activex, the ‘invoke()’ function in MATLAB is used to control CST. For example, (1) Start CST and open the target file: cst = actxserver(' CSTStudio.application ');

260

mws = invoke(cst, 'OpenFile', 'AntennaEMC1.cst'); (2) Modify the model parameters: invoke( mws, 'StoreParameter', 'x', x); (3) Start CST and open the target file: r21 = invoke(mws, 'Result1D', 'a2(1)1(1)').

3.2 Antenna Model Establishment.

This paper takes the creation of 300MHZ and 625MHZ monopoles as an example to calculate the coupling degree of two antennas, as shown in Fig. 1 and Fig. 2. The two antennas are mounted on a 0.2m*0.2m square metal plate with an antenna radius of 0.006m. The height of 300MHZ antenna is 0.25m, and the 625MHZ is set to 0.12m. A Cartesian coordinate system is established with the center of the platform as the origin. Each unit represents 1 mm, and the two antennas are arranged at (-50, 0), (50, 0). To simplify the model, the complex coaxial model is replaced with discrete ports.

Figure. 1. Antenna layout.

Figure. 2. The relationship between the coupling degree S21 of two antennas and frequency.

It can be seen from Fig. 2 that two antennas with similar distances and frequencies have higher coupling, especially near the respective frequencies.

4. Simulation Result

The position of the two antennas is fixed in Fig.1. The existing two movable antennas need to be arranged on the above platform, numbered І, Ⅱ, and the frequencies are 750MHZ and 500MHZ, respectively.

The layout is optimized according to the model established in the above section. The parameters set in the solution are as follows: The antenna coordinates are binary coded, and the code length is 28 bits. The first 14 bits represent the antenna І, and the last 14 bits represent the antenna Ⅱ. The initial population is 8, and the code is randomly generated. The crossover probability is set to 0.8 and the

261

mutation probability is set to 0.1. This paper adopts the optimal preservation strategy. All antennas are equally important (weighting factor is 1) and the simulation will exit after 120 generations.

The solution results are as follows: The optimal layout of the first generation is that the coordinate of antenna І is (-28, 60.5), the coordinate of antenna Ⅱ is (3, 57.5), and the total coupling degree is -55.0159 db. After 50 generations of simulation analysis, the relationship between the average coupling degree and the optimal coupling degree is shown in Fig. 3. The optimal layout position generated by the algorithm is the coordinates of antenna І (-52, -7), antenna Ⅱ (-49, -28), and the total coupling degree is -71.9017 db, which is reduced by nearly 20 db, shown in Fig. 4.

Figure. 3. Average coupling degree and optimal coupling degree of each generation.

Figure. 4. Optimize the coupling degree after simulation.

5. Summary

In this paper, the genetic algorithm is used to optimize the layout between the antennas to reduce the coupling degree. The co-simulation method is adopted, which makes full use of the data processing capability of MATLAB and the electromagnetic computing capability of CST. The method in this paper is applicable to the antenna layout optimization problem of other carriers. If the carrier structure and the objective function change, researchers only need to re-create the new model in the CST and write the corresponding fitness function.

References

[1] Yuan, J., Qiu, Y., Liu, Q.Z., Tian, J. and Xie, Y.J.. (2006) Position optimal design of vehicular antenna via space mapping and genetic algorithm. Chinese Journal of Radio Science, 21-1, 26.

262

[2] Yu, J.H., Ma, X.L. and Zheng, Y.L. (2008) Simulation of Electromagnetic Character and Coupling Degree of Vehicle Whip Antennas. Journal of System Simulation, 20-6, 1603. [3] Wang, W., Wang, X.T., Wang, W., Song, S., Li, M.Y. and Chen, L. (2015) Antenna coupling between electronic information equipment in flat ground situation. Chinese Journal of Radio Science, 30-4, 814. [4] Bai, X. (2013) CST-based system-level electromagnetic compatibility analysis and software design. M.S. thesis, National University of Defense Technology, Changsha, China. [5] Uthansakul, P., Assanuk, D. and Uthansakul, M. (2011) An Optimal Design of Multiple Antenna Positions on Mobile Devices Based on Mutual Coupling Analysis. International Journal of Antennas and Propagation, 1687-5869, 235. [6] Koper, E.M., Wood, W.D. and Schneider, S.W. (2004) Aircraft antenna coupling minimization using genetic algorithms and approximation. IEEE Transactions on Aerospace and Electronic Systems,40-2, 42.

263