Thanks for your response. symbol_book is 1x4 matrix which has the QPSK reference quadrants {0.7071+0.7071i, -0.7071+0.7071i, -0.7071-0.7071i, 0.7071-0.7071i}. Now, the x_hat is the estimated data with 50000x1 matrix. What i am trying to do is find the min distance between the QPSK reference and the actual received data. Not sure if maybe when i switched to 2015 matlab would have an effect! please advise
I get an error in knnsearch
6 views (last 30 days)
Show older comments
rec_syms = knnsearch([real(symbol_book) imag(symbol_book)], [real(X_hat) imag(X_hat)]) - 1;
I get this error:
Error using KDTreeSearcher/knnsearch (line 74)
Y must be a matrix with 8 columns.
Error in knnsearch (line 142)
idx = knnsearch(O,Y,'k',numNN, 'includeties', includeTies);
Error in OFDM_image (line 139)
rec_syms = knnsearch([real(symbol_book) imag(symbol_book)], [real(X_hat) imag(X_hat)]) - 1;
Accepted Answer
Walter Roberson
on 5 Dec 2015
[real(symbol_book) imag(symbol_book)] together makes a matrix with 8 columns, such as if there are 4 columns in symbol_book. But [real(X_hat) imag(X_hat)], the locations to search for, does not have 8 columns. The locations to search for (the second parameter) must have the same number of columns as the first parameter.
3 Comments
Walter Roberson
on 6 Dec 2015
You had a row vector of 4 elements. real() of that is something that is 1 x 4. imag() of it was also 1 x 4. []'ing those together resulted in something 1 x 8. Your set of points to be searched through was therefore a single point with 8 dimensions. But that is not what you wanted: you wanted 4 points with two dimensions. Using symbol_book(:) reshapes from 1 x 4 to be 4 x 1. Use [] to put two 4 x 1 together and you get 4 x 2 -- 4 points each with 2 dimensions.
More Answers (0)
See Also
Categories
Find more on QPSK in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!