How to signal divide into windows of lengths [1024] .

I have a signal x which is 43939 , which I want to divide into windows of lengths [1024] where each window overlaps with 50% of the previous window.using 1024 as an example I want the whole signal to be divided to windows of length 1024 where each window overlaps with the previous by 50% like 1:1024 then 512:1536 and store each window how can I implement that using a for loop or if statement then how to I use discrete wavelet transform (DWT) for the signal decomposion to 10 levels

Answer 1

[Dave B]

How about assembling up a matrix marking the indices of x you want. Of course you'll have to do something so that it's divisible by 512 (like pad the end).

x=rand(1,43939);
% pad at the end?
x=padarray(x,[0 512-mod(numel(x),512)],nan,'post');

start=(1:512:numel(x)-512)';
ind=start+(0:1023);

% just a view of what this looks like:
ind(1:3,1:5)

ans = 3×5

1 2 3 4 5 513 514 515 516 517 1025 1026 1027 1028 1029

ind(1:3,end-5:end)

ans = 3×6

1019 1020 1021 1022 1023 1024 1531 1532 1533 1534 1535 1536 2043 2044 2045 2046 2047 2048

y = x(ind);

For the wavelet bit, it sounds like wavedec, but that's about all I know.

[Rasheed Majeed]

Thank you very much

Answer 2

[Star Strider]

Use the buffer function. It will do everything desired.

signal = 1:100;
winlen = 10;                                            % WindowLength
overlap = fix(winlen/2);                                % 50% Overlap
out = buffer(signal, winlen, overlap)

out = 10×20

0 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 0 2 7 12 17 22 27 32 37 42 47 52 57 62 67 72 77 82 87 92 0 3 8 13 18 23 28 33 38 43 48 53 58 63 68 73 78 83 88 93 0 4 9 14 19 24 29 34 39 44 49 54 59 64 69 74 79 84 89 94 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 2 7 12 17 22 27 32 37 42 47 52 57 62 67 72 77 82 87 92 97 3 8 13 18 23 28 33 38 43 48 53 58 63 68 73 78 83 88 93 98 4 9 14 19 24 29 34 39 44 49 54 59 64 69 74 79 84 89 94 99 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

The wavelet decomposition then depends in part on the sort of wavelet that will give the best results for the intended decomposition. I have not used wavelets frequently enought to consider myself skilled in their use, so I do not feel comfortable commenting on that part.

.