DSP Blockset
  Go to block:
    Search    Help Desk 

Sample Vectors and Sample Matrices

Both vectors and matrices can be used to represent collections of data points sampled at the same time. This section explains how to create sample vectors and sample matrices in Simulink, and how to import sample-based data from the workspace.

Working with Sample Vectors

A length-N sample vector represents a snapshot of N independent signals at a particular sample instant. The sample vector bundles these N concurrent samples into a single unit.

As an illustration, consider the three discrete signals below.

If you simultaneously sample all three signals at t=3, you acquire the following values for each:

Multiplexing these three concurrent samples into a single vector yields the sample vector

Repeating this at t=16 yields the sample vector

Each sample vector is a snapshot of the three signals at a particular moment in time.

Note
Sample vectors are represented as row vectors.

Creating Sample Vectors in Simulink.    You create a sample vector when you acquire a single scalar sample from each of N distinct signals, and multiplex the N samples together into a vector.

For example, the output from the Simulink Mux block in the figure above is a sample vector containing a single sample from each of the three input signals:

All three samples correspond to the same instant in time.

Importing Sample Vectors into Simulink.    Often, the data that you acquire (or import) into Simulink from external sources is already in a sample-based format. For example, if you are modeling a communications system, you might import a few seconds of speech data from the MATLAB workspace. The figure below shows an N-by-3 workspace variable, voice, that contains three channels of voice data. When you import this data into a model using the Signal From Workspace block, the data is output row-wise, as a sequence of sample vectors.

Each sample vector in the output sequence is a snapshot of the three speech signals at a particular sample time. See "Working with Sources and Sinks" in Chapter 3 for additional information about importing data.

Working with Sample Matrices

Sample-based data can also be represented by a matrix. (See also "Understanding Matrices" in this chapter.) In this case, each of the elements of the M-by-N matrix represents an independent signal channel. Just like a sample vector, a sample matrix is a snapshot of multiple (M*N) signals at a single instant in time.

The figure below shows a sequence of three 6-by-6 matrices, each containing the instantaneous values of 36 independent signal channels. In this example, each matrix might be a collection of pixel brightness values; the value of each matrix element represents the corresponding pixel brightness at a particular instant.

Creating Sample Matrices in Simulink.    You create a sample matrix when you concurrently acquire a single scalar sample from each of N*M distinct signals, and multiplex the N*M samples together into a matrix.

For example, the output from the Reshape block in the figure above is a sample matrix containing a single sample from each of the four input signals,

All four samples in the matrix correspond to the same instant in time. (The Reshape block used in the model is simply an aid to tracking and verifying matrix sizes, and does not actually rearrange the values in the signal. See "Specifying Matrix Dimensions" in "Understanding Matrices" for additional information about the matrix format.)

You can also create a sample matrix by multiplexing together concurrent sample vectors, as shown below. In both cases, the resulting matrix represents a snapshot of concurrent scalar signals.

Importing Sample Matrices into Simulink.    Data that you acquire (or import) into Simulink from external sources may already be in sample-based format. For example, you might import a sequence of matrices representing a video segment. The figure below shows a 6-by-6-by-N workspace variable, video, that contains 36 (6 times 6) independent channels of data. When you import this data into a model using the Matrix From Workspace block, the data is output page-wise, as a sequence of sample matrices.

Each matrix in the output sequence contains a snapshot of the video picture at one sample instant. See "Working with Sources and Sinks" in Chapter 3 for additional information about importing data.


[ Previous | Help Desk | Next ]