MATLAB Answers

0

How can I use replicated subarrays and still set individual taper and phases?

Asked by Garrett Newell on 21 Aug 2019
Latest activity Commented on by Garrett Newell on 22 Aug 2019
Hi, I am trying to create a URA from sub arrays and steer the entire array by setting the wieghts and phases of each individaul element in each subarray. When I try to use 'SubarraySteering set to custom it yeilds an error asking for a differnet size input for ws. When I give it the size it wants it asks for another size. Is this a bug or am I appling the weighting wrong?
fc = 10e9;
c = physconst('lightspeed');
lambda = c/fc;
ele = phased.CosineAntennaElement;
antenna_spacing = lambda/2;
element_per_sub = [2 2];
sub_array = phased.URA('Element', ele, 'ElementSpacing', lambda/2, 'Size', element_per_sub);
num_sub_arrays = [2 4];
array = phased.ReplicatedSubarray('Subarray', sub_array, 'Layout', 'Custom', 'SubarraySteering', 'custom');
sub_pos =[];
sub_norm = [];
for ii = 1:num_sub_arrays(1)
for jj = 1:num_sub_arrays(2)
sub_pos(:,end+1) = [mod((ii-1),4)*antenna_spacing*2; mod((jj-1), 80)*2*antenna_spacing; 0];
sub_norm(:,end+1) = [0;-90];
end
end
array.SubarrayPosition=sub_pos;
array.SubarrayNormal = sub_norm;
taper = taylorwin(sub_array.Size(1)*num_sub_arrays(1)).*taylorwin(sub_array.Size(2)*num_sub_arrays(2))';
sub_array.Taper=taper;
ws = [];
for ii =1:num_sub_arrays(1)
for jj =1:num_sub_arrays(2)
temp_val = taper(element_per_sub(1)*(ii-1)+1:element_per_sub(1)*(ii), element_per_sub(2)*(jj-1)+1:element_per_sub(2)*(jj));
ws(:,end+1) = temp_val(:);
end
end
sv = phased.SteeringVector('SensorArray', array, 'IncludeElementResponse', true);
step(array, fc, c, [30;30], ws)
pattern(array, fc, 'Weights', ws(:))

  0 Comments

Sign in to comment.

1 Answer

Answer by Honglei Chen
on 22 Aug 2019
 Accepted Answer

If you set SubarraySteering to Custom, it means you want to control each element individually, thus you need to provide a weights that matches the number of elements. This is different than the SteeringVector since for a subarray, the steering vector is only computed at subarray level. You can specify the weights for each element using ElementWeights option when calling pattern(), like
fc = 10e9;
c = physconst('lightspeed');
lambda = c/fc;
ele = phased.CosineAntennaElement;
antenna_spacing = lambda/2;
element_per_sub = [2 2];
sub_array = phased.URA('Element', ele, 'ElementSpacing', lambda/2, 'Size', element_per_sub);
num_sub_arrays = [2 4];
array = phased.ReplicatedSubarray('Subarray', sub_array, 'Layout', 'Custom', 'SubarraySteering', 'custom');
sub_pos =[];
sub_norm = [];
for ii = 1:num_sub_arrays(1)
for jj = 1:num_sub_arrays(2)
sub_pos(:,end+1) = [mod((ii-1),4)*antenna_spacing*2; mod((jj-1), 80)*2*antenna_spacing; 0];
sub_norm(:,end+1) = [0;-90];
end
end
array.SubarrayPosition=sub_pos;
array.SubarrayNormal = sub_norm;
taper = taylorwin(sub_array.Size(1)*num_sub_arrays(1)).*taylorwin(sub_array.Size(2)*num_sub_arrays(2))';
% sub_array.Taper=taper;
ws = [];
for ii =1:num_sub_arrays(1)
for jj =1:num_sub_arrays(2)
temp_val = taper(element_per_sub(1)*(ii-1)+1:element_per_sub(1)*(ii), element_per_sub(2)*(jj-1)+1:element_per_sub(2)*(jj));
ws(:,end+1) = temp_val(:);
end
end
sv = phased.SteeringVector('SensorArray', array, 'IncludeElementResponse', true);
step(array, fc, [30;30], c, ws)
pattern(array, fc, 'ElementWeights', ws)
HTH

  1 Comment

Great, that worked. Although it appears to not actaully change the array response with the elementWeights arguments.
% function [array] = MakeArray()
% This will create the entire array and pass it back
fc=10e9;
rad_ele = phased.CosineAntennaElement('CosinePower', [1.5 1.5], 'FrequencyRange', [0 20e9]);
antenna_spacing = .58*.0254; % in meters
array_size = [64 1280];
sub_array_size = [16 16];
% num_sub_arrays = array_size./sub_array_size;
num_sub_arrays = [2 4];
sub_array = phased.URA('Element', rad_ele, 'ElementSpacing', antenna_spacing, 'Size', sub_array_size, 'ArrayNormal', 'x');
array = phased.ReplicatedSubarray('Subarray', sub_array, 'Layout', 'Custom', 'SubarraySteering', 'Phase');
sub_pos =[];
sub_norm = [];
for ii = 1:num_sub_arrays(1)
for jj = 1:num_sub_arrays(2)
sub_pos(:,end+1) = [ 0 ;mod((ii-1),4)*antenna_spacing*16; mod((jj-1), 80)*16*antenna_spacing];
sub_norm(:,end+1) = [0;0];
end
end
array.SubarrayPosition = sub_pos;
array.SubarrayNormal= sub_norm;
taper = taylorwin(sub_array.Size(1)*num_sub_arrays(1)).*taylorwin(sub_array.Size(2)*num_sub_arrays(2))';
ws=[];
for x = 1:num_sub_arrays(2)
for y = 1:num_sub_arrays(1)
temp_taper = taper(sub_array_size(2)*(y-1)+1:sub_array_size(2)*(y),sub_array_size(1)*(x-1)+1:sub_array_size(1)*(x));
temp_taper = flipud((temp_taper)');
temp = ~temp;
ws(:,end+1) = temp_taper(:);
end
end
p=pattern(array, fc, 0, -10:.1:10, 'ElementWeights', ws, 'CoordinateSystem', 'Rectangular', 'normalize', true, 'type' , 'powerdb');
%% Refernce
temp_array = phased.URA('Element', rad_ele, 'ElementSpacing', antenna_spacing, 'Size', num_sub_arrays.*sub_array_size)
p_ref1 = pattern(temp_array, fc, -10:.1:10,0, 'CoordinateSystem', 'Rectangular', 'normalize', true, 'type' , 'powerdb');
temp_array.Taper = taper;
p_ref = pattern(temp_array, fc, -10:.1:10,0, 'CoordinateSystem', 'Rectangular', 'normalize', true, 'type' , 'powerdb');
figure; hold;
plot( -10:.1:10, p, '-xb', 'DisplayName', 'SubArray Response With Taper')
plot( -10:.1:10, p_ref, 'DisplayName', 'URA Response With Taper')
plot( -10:.1:10, p_ref1, 'DisplayName', 'URA Response Without Taper')
legend
Results.jpg

Sign in to comment.