Note: This page has been translated by MathWorks. Please click here

To view all translated materals including this page, select Japan from the country navigator on the bottom of this page.

To view all translated materals including this page, select Japan from the country navigator on the bottom of this page.

Communications System Toolbox™ provides sinks and display devices that facilitate analysis of communication system performance. You can implement devices using either System objects, blocks, or functions.

You can use blocks or functions to generate random data to simulate a signal source. In addition, you can use Simulink blocks such as the Random Number block as a data source. You can open the Random Data Sources sublibrary by double-clicking its icon (found in the Comm Sources library of the main Communications System Toolbox block library).

The `randsrc`

function generates random matrices
whose entries are chosen independently from an alphabet that you specify,
with a distribution that you specify. A special case generates
bipolar matrices.

For example, the command below generates a 5-by-4 matrix whose entries are independently chosen and uniformly distributed in the set {1,3,5}. (Your results might vary because these are random numbers.)

a = randsrc(5,4,[1,3,5]) a = 3 5 1 5 1 5 3 3 1 3 3 1 1 1 3 5 3 1 1 3

If you want 1 to be twice as likely to occur as either 3 or
5, use the command below to prescribe the skewed distribution. The
third input argument has two rows, one of which indicates the possible
values of `b`

and the other indicates the probability
of each value.

b = randsrc(5,4,[1,3,5; .5,.25,.25]) b = 3 3 5 1 1 1 1 1 1 5 1 1 1 3 1 3 3 1 3 1

In MATLAB, the `randi`

function generates
random integer matrices whose entries are in a range that you specify. A
special case generates random binary matrices.

For example, the command below generates a 5-by-4 matrix containing random integers between 2 and 10.

c = randi([2,10],5,4) c = 2 4 4 6 4 5 10 5 9 7 10 8 5 5 2 3 10 3 4 10

If your desired range is [0,10] instead of [2,10], you can use either of the commands below. They produce different numerical results, but use the same distribution.

d = randi([0,10],5,4); e = randi([0 10],5,4);

In Simulink^{®}, the Random Integer
Generator and Poisson Integer Generator blocks
both generate vectors containing random nonnegative integers. The
Random Integer Generator block uses a uniform distribution on a bounded
range that you specify in the block mask. The Poisson Integer Generator
block uses a Poisson distribution to determine its output. In particular,
the output can include any nonnegative integer.

In MATLAB^{®}, the `randerr`

function generates
matrices whose entries are either 0 or 1. However, its options are
different from those of `randi`

, because `randerr`

is
meant for testing error-control coding. For example, the command below
generates a 5-by-4 binary matrix, where each row contains exactly
one 1.

f = randerr(5,4) f = 0 0 1 0 0 0 1 0 0 1 0 0 1 0 0 0 0 0 1 0

You might use such a command to perturb a binary code that consists
of five four-bit codewords. Adding the random matrix `f`

to
your code matrix (modulo 2) introduces exactly one error into each
codeword.

On the other hand, to perturb each codeword by introducing one error with probability 0.4 and two errors with probability 0.6, use the command below instead.

```
% Each row has one '1' with probability 0.4, otherwise two '1's
g = randerr(5,4,[1,2; 0.4,0.6])
g =
0 1 1 0
0 1 0 0
0 0 1 1
1 0 1 0
0 1 1 0
```

The probability matrix that is the third argument of `randerr`

affects
only the *number* of 1s in each row, not their
placement.

As another application, you can generate an equiprobable binary
100-element column vector using any of the commands below. The three
commands produce different numerical outputs, but use the same *distribution*.
The third input arguments vary according to each function's particular
way of specifying its behavior.

binarymatrix1 = randsrc(100,1,[0 1]); % Possible values are 0,1. binarymatrix2 = randi([0 1],100,1); % Two possible values binarymatrix3 = randerr(100,1,[0 1;.5 .5]); % No 1s, or one 1

In Simulink, the Bernoulli Binary Generator block generates random bits and is suitable for representing sources. The block considers each element of the signal to be an independent Bernoulli random variable. Also, different elements need not be identically distributed.

Construct noise generator blocks in Simulink to simulate communication links.

You can construct random noise generators to simulate channel noise by using the MATLAB Function block with random number generating functions. Construct different types of channel noise by using the following combinations.

Distribution | Block | Function |
---|---|---|

Gaussian | MATLAB Function | `randn` |

Rayleigh | MATLAB Function | `randn` |

Rician | MATLAB Function | `randn` |

Uniform on a bounded interval | MATLAB Function | `rand` |

See Random Noise Generators for an example of how Rayleigh and Rician distributed noise is created.

In MATLAB, the `wgn`

function generates
random matrices using a white Gaussian noise distribution. You specify
the power of the noise in either dBW (decibels relative to a watt),
dBm, or linear units. You can generate either real or complex noise.

For example, the command below generates a column vector of length 50 containing real white Gaussian noise whose power is 2 dBW. The function assumes that the load impedance is 1 ohm.

y1 = wgn(50,1,2);

To generate complex white Gaussian noise whose power is 2 watts, across a load of 60 ohms, use either of the commands below.

y2 = wgn(50,1,2,60,'complex','linear'); y3 = wgn(50,1,2,60,'linear','complex');

To send a signal through an additive white Gaussian noise channel,
use the `awgn`

function. See AWGN Channel for more
information.

You can use blocks in the Sequence Generators sublibrary of the Communications Sources library to generate sequences for spreading or synchronization in a communication system. You can open the Sequence Generators sublibrary by double-clicking its icon in the main Communications System Toolbox™ block library.

Blocks in the Sequence Generators sublibrary generate

The following table lists the blocks that generate pseudorandom or pseudonoise (PN) sequences. The applications of these sequences range from multiple-access spread spectrum communication systems to ranging, synchronization, and data scrambling.

Sequence | Block |
---|---|

Gold sequences | Gold Sequence Generator |

Kasami sequences | Kasami Sequence Generator |

PN sequences | PN Sequence Generator |

All three blocks use shift registers to generate pseudorandom sequences. The following is a schematic diagram of a typical shift register.

All *r* registers in the generator update
their values at each time step according to the value of the incoming
arrow to the shift register. The adders perform addition modulo 2.
The shift register can be described by a binary polynomial in *z*,
g_{r}*z*^{r} +
g_{r-1}*z*^{r-1} +
... + g_{0}. The coefficient g_{i} is
1 if there is a connection from the ith shift register to the adder,
and 0 otherwise.

The Kasami Sequence Generator block and the PN Sequence Generator
block use this polynomial description for their **Generator
polynomial** parameter, while the Gold Sequence Generator
block uses it for the **Preferred polynomial [1]** and **Preferred
polynomial [2]** parameters.

The lower half of the preceding diagram shows how the output sequence can be shifted by a positive integer d, by delaying the output for d units of time. This is accomplished by a single connection along the dth arrow in the lower half of the diagram.

The Barker Code Generator block generates Barker codes to perform synchronization. Barker codes are subsets of PN sequences. They are short codes, with a length at most 13, which are low-correlation sidelobes. A correlation sidelobe is the correlation of a codeword with a time-shifted version of itself.

Orthogonal codes are used for spreading to benefit from their perfect correlation properties. When used in multi-user spread spectrum systems, where the receiver is perfectly synchronized with the transmitter, the despreading operation is ideal.

Code | Block |
---|---|

Hadamard codes | Hadamard Code Generator |

OVSF codes | OVSF Code Generator |

Walsh codes | Walsh Code Generator |

The Comm Sinks block library contains scopes for viewing three types of signal plots:

The following table lists the blocks and the plots they generate.

Block Name | Plots |
---|---|

Eye Diagram | Eye diagram of a signal |

Constellation Diagram | Constellation diagram and signal trajectory of a signal |

An eye diagram is a simple and convenient tool for studying the effects of intersymbol interference and other channel impairments in digital transmission. When this software product constructs an eye diagram, it plots the received signal against time on a fixed-interval axis. At the end of the fixed interval, it wraps around to the beginning of the time axis. As a result, the diagram consists of many overlapping curves. One way to use an eye diagram is to look for the place where the eye is most widely opened, and use that point as the decision point when demapping a demodulated signal to recover a digital message.

The Eye Diagram block produces eye diagrams. This block processes discrete-time signals and periodically draws a line to indicate a decision, according to a mask parameter.

Examples appear in View a Sinusoid and View a Modulated Signal.

A constellation diagram of a signal plots the signal's value at its decision points. In the best case, the decision points should be at times when the eye of the signal's eye diagram is the most widely open.

The Constellation Diagram block produces a constellation diagram from discrete-time signals. An example appears in View a Sinusoid.

A signal trajectory is a continuous plot of a signal over time. A signal trajectory differs from a scatter plot in that the latter displays points on the signal trajectory at discrete intervals of time.

The Constellation Diagram block
produces signal trajectories. The Constellation Diagram block
produces signal trajectories when the `ShowTrajectory`

property
is set to true. A signal trajectory connects all points of the input
signal, irrespective of the specified decimation factor (```
Samples
per symbol
```

)

The following model produces a constellation diagram and an eye diagram from a complex sinusoidal signal. Because the decision time interval is almost, but not exactly, an integer multiple of the period of the sinusoid, the eye diagram exhibits drift over time. More specifically, successive traces in the eye diagram and successive points in the scatter diagram are near each other but do not overlap.

To open the model,
enter `doc_eyediagram`

at the MATLAB command
line. To build the model, gather and configure these blocks:

Sine Wave, in the Sources library of the DSP System Toolbox™ (

*not*the Sine Wave block in the Simulink Sources library)Set

**Frequency**to`.502`

.Set

**Output complexity**to`Complex`

.Set

**Sample time**to`1/16`

.

Constellation Diagram, in the Comm Sinks library

On the

**Constellation Properties**panel, set**Samples per symbol**to`16`

.

Eye Diagram, in the Comm Sinks library

On the

**Plotting Properties**panel, set**Samples per symbol**to`16`

.On the

**Figure Properties**panel, set**Scope position**to`figposition([42.5 55 35 35]);`

.

Connect the blocks as shown in the preceding figure. From the
model window's **Simulation** menu, choose **Model
Configuration parameters**. In the **Configuration
Parameters** dialog box, set **Stop time** to `250`

.
Running the model produces the following scatter diagram plot.

The points of the scatter plot lie on a circle of radius 1.
Note that the points fade as time passes. This is because the box
next to **Color fading** is checked under **Rendering
Properties**, which causes the scope to render points more
dimly the more time that passes after they are plotted. If you clear
this box, you see a full circle of points.

The Constellation Diagram block displays a circular trajectory.

In the eye diagram, the upper set of traces represents the real part of the signal and the lower set of traces represents the imaginary part of the signal.

This multipart example creates an eye diagram, scatter plot, and signal trajector plot for a modulated signal. It examines the plots one by one in these sections:

The following model modulates a random signal using QPSK, filters the signal with a raised cosine filter, and creates an eye diagram from the filtered signal.

To open the model,
enter `doc_signaldisplays`

at the MATLAB command
line. To build the model, gather and configure the following blocks:

Random Integer Generator, in the Random Data Sources sublibrary of the Comm Sources library

Set

**M-ary number**to`4`

.Set

**Sample time**to`0.01`

.

QPSK Modulator Baseband, in PM in the Digital Baseband sublibrary of the Modulation library of Communications System Toolbox, with default parameters

AWGN Channel, in the Channels library of Communications System Toolbox, with the following changes to the default parameter settings:

Set

**Mode**to`Signal-to-noise ratio (SNR)`

.Set

**SNR (dB)**to`15`

.

Raised Cosine Transmit Filter, in the Comm Filters library

Set

**Filter shape**to`Normal`

.Set

**Rolloff factor**to`0.5`

.Set

**Filter span in symbols**to`6`

.Set

**Output samples per symbol**to`8`

.Set

**Input processing**to`Elements as channels (sample based)`

.

Eye Diagram, in the Comm Sinks library

Set

**Samples per symbol**to`8`

.Set

**Symbols per trace**to`3`

. This specifies the number of symbols that are displayed in each trace of the eye diagram. A*trace*is any one of the individual lines in the eye diagram.Set

**Traces displayed**to`3`

.Set

**New traces per display**to`1`

. This specifies the number of new traces that appear each time the diagram is refreshed. The number of traces that remain in the diagram from one refresh to the next is**Traces displayed**minus**New traces per display**.On the

**Rendering Properties**panel, set**Markers**to`+`

to indicate the points plotted at each sample. The default value of**Markers**is empty, which indicates no marker.On the

**Figure Properties**panel, set**Eye diagram to display**to`In-phase only`

.

When you run the model, the Eye Diagram displays the following diagram. Your exact image varies depending on when you pause or stop the simulation.

Three traces are displayed. Traces 2 and 3 are faded because **Color
fading** under **Rendering Properties** is
selected. This causes traces to be displayed less brightly the older
they are. In this picture, Trace 1 is the most recent and Trace 3
is the oldest. Because **New traces per display** is
set to `1`

, only Trace 1 is appearing for the first
time. Traces 2 and 3 also appear in the previous display.

Because **Symbols per trace** is set to `3`

,
each trace contains three symbols, and because **Samples per
trace** is set to `8`

, each symbol contains
eight samples. Note that trace 1 contains 24 points, which is the
product of **Symbols per trace** and **Samples
per symbol**. However, traces 2 and 3 contain 25 points each.
The last point in trace 2, at the right border of the scope, represents
the same sample as the first point in trace 1, at the left border
of the scope. Similarly, the last point in trace 3 represents the
same sample as the first point in trace 2. These duplicate points
indicate where the traces would meet if they were displayed side by
side, as illustrated in the following picture.

You can view a more realistic eye diagram by changing the value
of **Traces displayed** to `40`

and
clearing the **Markers **field.

When the **Offset** parameter is set to `0`

,
the plotting starts at the center of the first symbol, so that the
open part of the eye diagram is in the middle of the plot for most
points.

The following model creates a scatter plot of the same signal considered in Eye Diagram of a Modulated Signal.

To build the model, follow the instructions in Eye Diagram of a Modulated Signal but replace the Eye Diagram block with the following block:

Constellation Diagram, in the Comms Sinks library

Set

**Samples per symbol**to`2`

.Set

**Offset**to`0`

. This specifies the number of samples to skip before plotting the first point.Set

**Symbols to display**to`40`

.

When you run the simulation, the Constellation Diagram block displays the following plot.

The plot displays 30 points. Because **Color fading** under **Rendering
Properties** is selected, points are displayed less brightly
the older they are.

The following model creates a signal trajectory plot of the same signal considered in Eye Diagram of a Modulated Signal.

To build the model, follow the instructions in Eye Diagram of a Modulated Signal but replace the Eye Diagram block with the following block:

Constellation Diagram , in the Comms Sinks library

Set

**Samples per symbol**to`8`

.Set

**Symbols displayed**to`40`

. This specifies the number of symbols displayed in the signal trajectory. The total number of points displayed is the product of**Samples per symbol**and**Symbols displayed**.Set

**New symbols per display**to`10`

. This specifies the number of new symbols that appear each time the diagram is refreshed. The number of symbols that remain in the diagram from one refresh to the next is**Symbols displayed**minus**New symbols per display**.

When you run the model, the Constellation Diagram displays a trajectory like the one below.

The plot displays 40 symbols. Because **Color fading** under **Rendering
Properties** is selected, symbols are displayed less brightly
the older they are.

See Constellation Diagram of a Modulated Signal to compare the preceding signal trajectory to the scatter plot of the same signal. The Constellation Diagram block connects the points displayed by the Constellation Diagram block to display the signal trajectory.

If you increase **Symbols displayed** to `100`

,
the model produces a signal trajectory like the one below. The total
number of points displayed at any instant is 800, which is the product
of the parameters **Samples per symbol** and **Symbols
displayed**.

Was this topic helpful?