Release 11 Enhancements (Release 11 New Features)

Release 11 New Features

Fixed-Point Blockset 2.0

Release 11 contains two versions of the Fixed-Point Blockset: Version 1.2, which was included with Release 10 (MATLAB 5.2), and Version 2.0. The 2.0 blockset is located in /toolbox/fixpoint and the 1.2 blockset is located in
/toolbox/fixpoint/obsolete.

The Fixed-Point Blockset 2.0 features are discussed below.

Fixed-Point Blocks

The Fixed-Point Blockset 2.0 includes a number of building blocks to assist you in designing and simulating dynamic systems using fixed-point arithmetic. The fixed-point blocks are grouped together as shown below.

Arithmetic Blocks

Block Name
Description

FixPt Constant
Generate a constant value

FixPt Gain
Multiply the input by a constant

FixPt Matrix Gain
Multiply the input by a constant matrix

FixPt Product
Multiply or divide inputs

FixPt Sum
Add or subtract inputs

Block Name	Description
FixPt Constant	Generate a constant value
FixPt Gain	Multiply the input by a constant
FixPt Matrix Gain	Multiply the input by a constant matrix
FixPt Product	Multiply or divide inputs
FixPt Sum	Add or subtract inputs

Conversion Blocks

Block Name
Description

FixPt Conversion
Convert from one Fixed-Point Blockset data type to another

FixPt Conversion Inherited
Convert input two to the data type of input one

FixPt Gateway In
Convert a Simulink data type to a Fixed-Point Blockset data type

FixPt Gateway Out
Convert a Fixed-Point Blockset data type to a Simulink data type

Block Name	Description
FixPt Conversion	Convert from one Fixed-Point Blockset data type to another
FixPt Conversion Inherited	Convert input two to the data type of input one
FixPt Gateway In	Convert a Simulink data type to a Fixed-Point Blockset data type
FixPt Gateway Out	Convert a Fixed-Point Blockset data type to a Simulink data type

Look-Up Table Blocks

Block Name
Description

FixPt Look-Up Table
Approximate a one-dimensional function using a selected look-up method

FixPt Look-Up Table (2D)
Approximate a two-dimensional function using a selected look-up method

Block Name	Description
FixPt Look-Up Table	Approximate a one-dimensional function using a selected look-up method
FixPt Look-Up Table (2D)	Approximate a two-dimensional function using a selected look-up method

Logical and Comparison Blocks

Block Name
Description

FixPt Logical Operator
Perform the specified logical operation on the inputs

FixPt Relational Operator
Perform the specified relational operation on the inputs

FixPt Relay
Switch output between two constants

FixPt Saturation
Bound the range of the input

FixPt Switch
Switch output between input one or input three based on the value of input two

Block Name	Description
FixPt Logical Operator	Perform the specified logical operation on the inputs
FixPt Relational Operator	Perform the specified relational operation on the inputs
FixPt Relay	Switch output between two constants
FixPt Saturation	Bound the range of the input
FixPt Switch	Switch output between input one or input three based on the value of input two

Discrete-Time Blocks

Block Name
Description

FixPt FIR
Implement a fixed-point finite impulse response (FIR) filter

FixPt Unit Delay
Delay a signal one sample period

FixPt Zero-Order Hold
Implement a zero-order hold of one sample period

Block Name	Description
FixPt FIR	Implement a fixed-point finite impulse response (FIR) filter
FixPt Unit Delay	Delay a signal one sample period
FixPt Zero-Order Hold	Implement a zero-order hold of one sample period

Filters and Systems

The Fixed-Point Blockset 2.0 provides several useful fixed-point filter and system realizations. These realizations are intended to be used as design templates so you can easily see how to build filters and systems suited to your particular needs. The filters and systems are described below.


Filter or System Name	Description
FixPt State-Space Realization	Implement a fixed-point realization of a state-space system
FixPt Integrator: Trapezoidal	Implement a fixed-point realization of an integrator based on trapezoidal numerical integration
FixPt Integrator: Backward	Implement a fixed-point realization of an integrator based on backward numerical integration
FixPt Integrator: Forward	Implement a fixed-point realization of an integrator based on forward numerical integration
FixPt Filtered Derivative	Implement a fixed-point realization of a filtered derivative
FixPt Derivative	Implement a fixed-point realization of a derivative
FixPt Lead or Lag Filter	Implement a fixed-point realization of a lead filter or lag filter

Data Types

The Fixed-Point Blockset 2.0 supports several fixed-point and floating-point data types, which are collectively referred to as the "Fixed-Point Blockset data types." The supported data types and related features are described below.

Fixed-Point Data Types

Integer, fractional, and generalized fixed-point data types are supported.
Unsigned and two's complement formats are supported.
The fixed-point word size can range from 1 to 128 bits.
The radix (binary) point is not required to be contiguous with the fixed-point word.

Floating-Point Data Types

IEEE-style singles and doubles are supported.
A nonstandard IEEE-style data type is supported. For this data type, the fraction (mantissa) can range from 1 to 52 bits and the exponent can range from 1 to 11 bits.

The label "Fixed-Point Blockset data types" indicates that data types supported by this blockset are unique to it, and not directly compatible with Simulink. This means that a double generated by Simulink cannot be passed directly into a Fixed-Point Blockset block, and a double generated by the Fixed-Point Blockset cannot be passed directly into a Simulink block. Instead, the FixPt Gateway In and FixPt Gateway Out blocks must be used as interfaces between the Fixed-Point Blockset and Simulink.

Scaling

The Fixed-Point Blockset 2.0 supports two general scaling modes: radix point-only scaling and slope/bias scaling. Additionally, some blocks support scaling modes that maximize the precision for constant vectors or matrices. These scaling modes are described below.

General Scaling Modes

Fixed-point numbers can be scaled in these ways:

Radix Point-Only

This is "powers-of-two" scaling since it only involves moving the radix point. Radix point-only scaling does not require the radix point to be contiguous with the data word. The advantage of this scaling mode is the number of processor arithmetic operations are minimized.

Slope/Bias

With this scaling mode, you can provide a slope and a bias. The advantage of slope/bias scaling is that it typically provides more efficient use of a finite number of bits.

Constant Scaling for Best Precision

In addition to the general scaling modes described above, the Fixed-Point Blockset provides you with block-specific scaling modes for constant vectors and constant matrices. These scaling modes are based on radix point-only scaling and are designed to maximize precision.

Constant Vector Scaling

With this mode, you have the option of scaling a constant vector such that the precision is maximized for each element, or a common radix point can be found based on the maximum precision for the largest value of the vector.

Constant Matrix Scaling

With this mode, you have the option of scaling a constant matrix such that the precision is maximized for each element, or a common radix point can be found based on the maximum precision for the largest value of each row, each column, or the whole matrix.

The advantage of finding a common radix point is increased simulation speed, while the disadvantage is reduced precision.

Automatic Scaling Tool

A script is provided that automatically changes the scaling for each block that has generalized fixed-point output and does not have its scaling locked. The script uses the maximum and minimum values logged during the last simulation run. The scaling is changed such that the simulation range is covered and the precision is maximized.

As an alternative to (and extension of) the automatic scaling script, an automatic scaling GUI is provided. This interface allows you to easily control the parameters associated with automatic scaling and display the simulation results for a given model. With the automatic scaling GUI, you can:

Turn on or turn off logging for all blocks
Override the output data type with doubles for all blocks
Invoke the automatic scaling script
Run the simulation
Display the scaling results for each block that had its scaling changed

Locking the Output Scaling

If the output data type is a generalized fixed-point number, then you have the option of locking its scaling. When locked, the automatic scaling tool will not change the output scaling. Otherwise, the automatic scaling tool is free to adjust the scaling.

Rounding

Fixed-point numbers can be rounded in these ways:

Toward Zero

This mode rounds toward zero and is equivalent to MATLAB's fix command.

Toward Nearest

This mode rounds toward the nearest representable number, with the exact midpoint rounded toward positive infinity. Rounding toward nearest is equivalent to the MATLAB round command.

Toward Ceiling

This mode rounds toward positive infinity and is equivalent to MATLAB's ceil command.

Toward Floor

This mode rounds toward negative infinity and is equivalent to MATLAB's floor command.

Overflow Handling

Operations on fixed-point numbers that produce an overflow condition can be dealt with in these ways:

Saturate

Overflows are set to either the maximum or minimum value represented by the word.

Wrap

Overflows can be set to any value represented by the word.

Overriding with Doubles

The fixed-point data type can be overridden with doubles either globally or for individual blocks. This feature is useful when debugging a simulation.

Specialized Storage Capabilities

The maximum and minimum values encountered during a simulation can be logged to the MATLAB workspace. These values can then be accessed by the automatic scaling tool.

Standardization with Simulink

The fixed-point blocks are feature-compatible with standard Simulink blocks. Standardization with Simulink provides these expanded capabilities:

Vectorization of inputs and outputs
Variable number of input ports on appropriate blocks such as FixPt Sum
More powerful blocks that combine and expand the features of the basic blocks. For example, scalar addition and subtraction are combined in the vectorized FixPt Sum block.

Enhanced Model Construction

The Fixed-Point Blockset 2.0 makes it easier to construct models of systems due to these main blockset features:

Expanded features of old blocks such as vectorization
Inclusion of two new blocks -- FixPt Matrix Gain and FixPt FIR

Updating Obsolete Fixed-Point Blocks

Obsolete fixed-point blocks from previous Fixed-Point Blockset releases can be updated to current fixed-point blocks using the fpupdate command.

Code Generation

With the Real-Time Workshop, you can generate C code for execution on a fixed-point embedded processor. The generated code uses only integer types and automatically includes all operations, such as shifts, needed to account for differences in fixed-point locations. The code is structured so that key operations can be readily replaced by optimized target-specific libraries that you supply. You can also use the Target Language Compiler to customize the generated code.

All fixed-point blocks support code generation, but not every simulation feature is supported. For example, 13-bit numbers can be used for simulation but not for code generation.

Demos

The Fixed-Point Blockset 2.0 provides several demos that illustrate the main product features. You can access these demos via the fixed-point library's Demos block.

Online Help

Each Fixed-Point Blockset block, system, and filter has online HTML-based help. The help is accessed through the dialog box Help button.

[ Previous | Next ]