Miniaturised antenna

Contents

We show next an example of matching an antenna’s reflection in the band 870-900 MHz. Before starting, the reflection parameter of the antenna has been saved to a .s1p (touchstone) file called superdirective.s1p.

Matching problem set up

The first step is loading the antenna data and creating the matching object. For that we use the function MultiPort.read.
Additionally we interpolate the data obtaining 1000 frequency points from 800MHz to 1GHz. This step is important since a good frequency resolution will be required for the rational approximation

1
2

data = MultiPort.read('superdirective.s1p');
data = resample(data,linspace(800e6,1e9,1000));

Define frequency band object (from 870MHz to 900MHz)

3

BAND = FreqBand('Band',[870e6,900e6]);

Pick rational approximation order and filter order. We choose A=2 for the order of the rational approximation and M=5 for the order of the matching filter. Thus the global system is of order N=A+M=7.

4
5

A=2;
M=5;

Create the matching object specifying the order of the global system, the order of the rational approximation and the desired passband. Also we can provide additional options as parameter-value pairs.

6
7
8
9
10

MATCHING = matching(data,...
                    'SystemOrder',A+M,...
                    'ApproximationOrder',A,...
                    'Passband',BAND,...
                    'Fixpoint',true);

Rational approximation

Rational approximation of the load is required. The plot function can be used to check the quality of the approximation at this stage. 

11
12
13
14
15

MATCHING.rationalApproximation;
figure
plot(MATCHING)
title('Rational approximation');
legend('show','Location','Best');

Solve the matching problem

It can be seen that a good fit is obtained in the band of interest with the rational approximation. Therefore we can proceed to solve the problem.

16

MATCHING.solve;

Extract the results

Once the solve method finished we can extract the result from the matching object. The property MATCHING.Results is a structure containing the rational model of the computed matching filter as well as the obtained global system. Moreover the structure MATCHING.ReflectionLevel contains the reflection level corresponding to: a) the lower hard bound, b) the reflection level obtained after the simplification of the blaschke product (fixpoint) if applicable, c) the reflection level obtained after a local optimisation (if it is performed)

17
18

Results = MATCHING.Results;
L = MATCHING.ReflectionLevel
L = struct with fields:
     optimal: -8.6620
    fixpoint: -8.5652

Verify the results by plotting the scattering parameters of the matching filter and the global system. Also parameter-value pairs can be used to provide additional options. In this case we decided not to plot the rationall approximation again (we already saw it was a very good approximation) and also to plot the filter response with dash (–) lines.

19
20
21
22
23
24
25
26

figure
plot(MATCHING,'PlotApprox',false,...
              'Scale','MHz',...
              'FilterArgs',{'LineStyle','--'});
legend('show','Location','Best');
ylim([-15 0]);
xlim([800 950]);
title('Matching of superdirective antenna');

Comments are closed.