Scribd
Upload
14 views2 pages

Binary Fixed Point

Binary Fixed-Point Representation is a method for representing fractional numbers using a fixed number of bits for integer and fractional parts, commonly used in digital signal processing and embedded systems. It features a scaling factor, defined range and precision, and employs two's complement for signed representation. While it offers efficient hardware implementation and consistent precision, it has a limited dynamic range and potential scaling errors.

Uploaded by

pjpatel
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
14 views2 pages

Binary Fixed Point

Binary Fixed-Point Representation is a method for representing fractional numbers using a fixed number of bits for integer and fractional parts, commonly used in digital signal processing and embedded systems. It features a scaling factor, defined range and precision, and employs two's complement for signed representation. While it offers efficient hardware implementation and consistent precision, it has a limited dynamic range and potential scaling errors.

Uploaded by

pjpatel
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Binary Fixed-Point Representation

Binary Fixed-Point Representation is a method for representing fractional numbers in a binary


system. In this approach, a fixed number of bits is assigned to the integer and fractional parts of
the number. It is commonly used in digital signal processing, embedded systems, and
applications where floating-point operations are too computationally expensive.

Key Features of Binary Fixed-Point Representation:

1. Structure:
o A binary number is divided into two parts:
 Integer part: Represents the whole number portion.
 Fractional part: Represents the fractional portion.
2. Scaling Factor:
o The value is scaled by a constant factor (e.g., 2−n2^{-n}2−n) depending on how
many bits are allocated for the fractional part.
o Example:
 If there are 3 fractional bits, the scaling factor is 2−3=0.1252^{-3} =
0.1252−3=0.125.
 The binary number 101.011101.011101.011 represents 5+0.375=5.3755 +
0.375 = 5.3755+0.375=5.375.
3. Range and Precision:
o Range: Depends on the number of bits allocated for the integer part.
 For NNN total bits and FFF fractional bits:
Range=[−2N−F−1,2N−F−1−2−F](for signed representation)\text{Range}
= [-2^{N-F-1}, 2^{N-F-1}-2^{-F}] \quad \text{(for signed
representation)}Range=[−2N−F−1,2N−F−1−2−F](for signed representatio
n)
o Precision: Determined by the smallest step size, 2−F2^{-F}2−F, where FFF is the
number of fractional bits.
4. Signed Representation:
o Often uses two's complement to represent negative numbers.
o Example for 4 bits, 2 fractional bits:
 010101010101 = 1.251.251.25
 101110111011 = −0.75-0.75−0.75
5. Operations:
o Addition and Subtraction: Similar to integer arithmetic but ensure that the
fixed-point alignment is consistent.
o Multiplication: Results need to be adjusted to maintain the fixed-point scaling
(e.g., shift by fractional bits).
o Division: More complex and may require normalization of the result.
6. Advantages:
oEfficient in terms of hardware implementation (requires less processing power
than floating-point).
o Consistent precision.
7. Disadvantages:
o Limited dynamic range compared to floating-point.
o Scaling errors can occur if the system isn’t well-designed.
8. Applications:
o Digital signal processing (DSP).
o Control systems.
o Embedded systems where memory and computational power are limited.

Example:

Suppose we want to represent the decimal number 6.3756.3756.375 using 8 bits with 4 fractional
bits:

1. Convert integer part (666):


o 666 in binary = 011001100110.
2. Convert fractional part (0.3750.3750.375):
o Multiply by 24=162^4 = 1624=16: 0.375×16=60.375 \times 16 = 60.375×16=6.
o 666 in binary = 011001100110.
3. Combine:
o 0110.01100110.01100110.0110.
4. Store as fixed-point (integer equivalent):
o 011001100110011001100110.

To interpret it:

 Divide by 24=162^4 = 1624=16: 102÷16=6.375102 \div 16 = 6.375102÷16=6.375.

You might also like