Beamwidth and Peak Radiation Analysis

Create AI-Based and Standard Horn Antennas

Use the design function to create an AI-based horn antenna model operating at 10 GHz. Then use the exportAntenna function to generate a standard horn antenna from the AI-based model.

f = 10e9;
antAI = design(horn,f,ForAI=true);
antEm = exportAntenna(antAI);

Analyze Beamwidth

AI‑based beamwidth analysis determines the half‑power beamwidth (HPBW) and the corresponding angles in the elevation and azimuth planes, which intersect along the major lobe of the antenna’s radiation pattern. HPBW is defined as the angular width of the major lobe between the points where the radiated power drops to half (−3 dB) of its maximum value. It is a key parameter used to characterize the directional performance of an antenna.

tic
[bw,ang,plane] = beamwidth(antAI,f)

bw = 2×1

   27.5029
   30.4996

ang = 2×2

  346.7615   14.2644
  344.7485   15.2481

plane=2×3 table
       Plane       Slice    SliceValue
    ___________    _____    __________

    "Elevation"    "Az"         0     
    "Azimuth"      "El"         0     

tAIBw = toc;

The plane output specifies the two principal planes to calculate the beamwidth. The first element in bw corresponds to the beamwidth in the elevation plane, obtained by slicing the radiation pattern at an azimuth angle (Az) of 0 degrees. The second element in bw represents the beamwidth in the azimuth plane, obtained by slicing the pattern at an elevation angle (El) of 0 degrees. The ang output provides the lower and upper angular bounds of the beamwidth in each respective plane.

Compute beamwidth using full-wave EM simulation along the elevation and azimuth planes. Quantize the results with a 1 degree resolution.

tic
[bwpEl,angpEl] = beamwidth(antEm,f,0,0:1:360);
[bwpAz,angpAz] = beamwidth(antEm,f,0:1:360,0);
tEmBw = toc;

Evaluate the accuracy of the AI-predicted beamwidth values against those obtained from the full-wave simulation. Compute the relative error for the beamwidth and for the corresponding lower and upper angular bounds in both planes using the calcerr function defined in this section.

errbwp1 = calcerr(bwpEl,bw(1));
errLowerAngp1 = calcerr(angpEl(1),ang(1,1));
errUpperAngp1 = calcerr(angpEl(2),ang(1,2));

errbwp2 = calcerr(bwpAz,bw(2));
errLowerAngp2 = calcerr(angpAz(1),ang(2,1));
errUpperAngp2 = calcerr(angpAz(2),ang(2,2));

function err = calcerr(act,pred)
err = (abs(act-pred)/act)*100;
end

Tabulate the results to compare the AI and EM analysis outputs along with the computed errors and analysis time.

predValues = [bwpEl,angpEl(1),angpEl(2), bwpAz,angpAz(1),angpAz(2)]';
actualValues = [bw(1),ang(1,1),ang(1,2),bw(2),ang(2,1),ang(2,2)]';
errors = [errbwp1,errLowerAngp1,errUpperAngp1,errbwp2,errLowerAngp2,errUpperAngp2]';

tab_bw = table(predValues,actualValues,errors,...
    VariableNames=["EM","AI","Relative Error (%)"],RowNames=["Beamwidth Az (deg)",...
    "Lower Angle Az (deg)","Upper Angle Az (deg)","Beamwidth El (deg)",...
    "Lower Angle El (deg)","Upper Angle El (deg)"])

tab_bw=6×3 table
                            EM       AI      Relative Error (%)
                            ___    ______    __________________

    Beamwidth Az (deg)       28    27.503          1.7754      
    Lower Angle Az (deg)    346    346.76          0.2201      
    Upper Angle Az (deg)     14    14.264          1.8888      
    Beamwidth El (deg)       32      30.5          4.6886      
    Lower Angle El (deg)    344    344.75         0.21759      
    Upper Angle El (deg)     16    15.248          4.6991      

tab_bw_time = table(tEmBw,tAIBw,tEmBw/tAIBw,VariableNames=["EM","AI","Speed Up"],RowNames="Time (sec)")

tab_bw_time=1×3 table
                    EM        AI      Speed Up
                  ______    ______    ________

    Time (sec)    7.6804    1.2628     6.0823 

The AI-based beamwidth analysis shows strong agreement with the EM results, with all relative errors within 5%. Additionally, the significant reduction in computation time highlights the efficiency of the AI approach.

Analyze Peak Radiation

Analyze the peak radiation of the AI-model-based horn antenna using the peakRadiation function. This function returns the maximum gain of the antenna along with the corresponding azimuth and elevation angles.

tic
[pgAI,azAngAI,elAngAI] = peakRadiation(antAI,f);
tAIPg = toc;

Analyze peak radiation of the regular horn antenna.

tic
[pgEm,azAngEm,elAngEm] = peakRadiation(antEm,f,0:1:360,-90:1:90);
tEmPg = toc;

Calculate the error for three outputs.

errPg = calcerr(pgEm,pgAI);
errAz = calcerr(azAngEm,azAngAI);
errEl = calcerr(elAngEm,elAngAI);

The AI-based analysis shows strong agreement with the EM simulation results while requiring significantly less computation time. Refer to the table for a comparison of the time taken by the AI and EM analyses.

predValues = [pgEm,azAngEm,elAngEm]';
actualValues = [pgAI,azAngAI,elAngAI]';
errors = [errPg,errAz,errEl]';
tab_pg = table(predValues,actualValues,errors,...
    VariableNames=["EM","AI","Relative Error  (%)"],RowNames=["Peak Gain (dBi)",...
    "Azimuth Angle (deg)","Elevation Angle (deg)"])

tab_pg=3×3 table
                               EM         AI       Relative Error  (%)
                             ______    ________    ___________________

    Peak Gain (dBi)          15.472       15.51          0.24546      
    Azimuth Angle (deg)           0           0              NaN      
    Elevation Angle (deg)         0    0.042063              Inf      

tab_pg_time = table(tEmPg,tAIPg,tEmPg/tAIPg,VariableNames=["EM","AI","Speed Up"],RowNames="Time (sec)")

tab_pg_time=1×3 table
                    EM         AI       Speed Up
                  ______    ________    ________

    Time (sec)    70.692    0.061116     1156.7