Lecture 12: Number representation and Quantization effects

Instructor: Dr. Gleb V. Tcheslavski Contact: [email protected] Office Hours: Room 2030 Class web site: http://ee.lamar.edu/gleb/ds p/index.htm p/index htm

ELEN 5346/4304

DSP and Filter Design

Fall 2008

Representation of numbers
Up to this point, we were considering implementations of discrete-time systems without any considerations of finite-word-length effects that are inherent in any digital realization, whether in hardware or software. realization software Let us consider first two different representations of numbers.

1. Fixed-point representation.
A real number X is represented as:

X = ( b A ,..., b1 , b0 , b1 ,...bB )r =

i = A

br
i

, 0 bi (r 1)

(12.2.1)

Where bi represents the digit, r is the radix or base, A is the number of integer digits, and B is the number of fractional digits. For example:

(1223.45)10 = 1102 + 2 101 + 3 100 + 4 101 + 5 102 (101.01)2 = 1 22 + 0 21 + 1 20 + 0 21 + 1 22

Representation of numbers
We will focus our attention on the binary representation as most important for DSP. In this case r = 2 and the digits {bi} are called binary digits or bits. They take the values {0 1} The binary digit b-A is called the most significant bit {0, 1}. (MSB) of the number, and the binary digit bB is called the least significant bit (LSB) of the number. The binary point between the digits b0 and b1 does not exist explicitly and the logics assumes location of this point. By using an n-bit integer format (A = n-1, B = 0), we can represent unsigned integer numbers from 0 to 2n-1. More frequently, the fractional format (A = 0, B = n-1) is used with a binary point between b0 and b1 that can represent numbers from 0 to 1-2-n. Any integer or mixed number can be represented in a fraction format by factoring out the term r A. There are three formats to represent negative numbers. The format for the positive numbers is the same: the MSB is set to zero

X = 0.b1b2 ...bB = bi 2i , X 0
i =1
ELEN 5346/4304

(12.3.1)

DSP and Filter Design

Fall 2008

Representation of numbers
The negative numbers can be represented by: 1) Sign-Magnitude format: MSB is set to 1 to represent -

X SM = 1 b1b2 ...bB 1.
2) Ones-Complement Format:

for f X 0

(12.4.1)

X 1C = 1.b1b2 ...bB

for X 0

(12.4.2)

Where bi = 1 bi is the complement of bi (i.e., we replace ones by zeros and zeros by ones for all bits). B

X 1C = 1 20 + (1 bi ) 21 = 2 X 2 B
i= i 1

(12.4.3)

3) Twos-Complement Format:

X 2C = 1.b1b2 ...bB 0 0...0 1 for X < 0

(12.4.4)

Where is modulo-2 addition. For example, -3/8 is obtained by complementing 0011 (3/8) to obtain 1100 and then adding 0001, which yields 1101 to represent -3/8 in the twos-complement format.

Representation of numbers
The basic operations of addition and multiplication depend on the format used. Most fixed-point digital signal processors use twos-complement arithmetic, therefore, the range for (B + 1) bit number ranges f th f th f b from -1 t 1-2-B. 1 to 1 2 B In general, the multiplication of two fixed-point numbers each of b bits in length results in a product of 2b bits of length. The product is either truncated or rounded back to b bits resulting either in truncation or rounding errors. A fixed-point representation allows to cover a range of numbers, say, xmax xmin with a fixed resolution:

xmax xmin m 1

(12.5.1)

where m = 2b is the number of levels and b is the number of bits.

ELEN 5346/4304

DSP and Filter Design

Fall 2008

Representation of numbers
2. Floating-point representation.
Covers a larger dynamic range by representing the number X as

X = 2E M
where M is a mantissa the fractional part of the number: 0.5 M 1, E (exponent) is either negative or positive number. Both mantissa and exponent require additional sign bits for representing negative numbers. For example:

(12.6.1)

X 1 = 5 M 1 = 0.101000; E1 = 011; X 2 = 3 8 M 2 = 0.111000; E2 = 101 ;

Representation of numbers
Overflow occurs in the multiplication of two floating-point numbers when the sum of the exponents exceeds the dynamic range of the fixed-point representation of the th exponent. t The floating-point representation allows us to cover a larger dynamic range than the fixed-point representation by varying the resolution across the range. The distance between two successive floating-point numbers increases as the numbers increase in size. Also, the floating-point representation provides finer resolution for small numbers but coarser resolution for large numbers.

ELEN 5346/4304

DSP and Filter Design

Fall 2008

Quantization
1. Fixed-point: truncation
To truncate a fixed-point number from (+1) bits to (b+1) bits, we just discard the least significant (-b) bits. The truncation error is denoted by

t = Q( X ) X

(12.8.1)

Here Q(X) is the truncated version of the number X. For a positive X, the error is equal t zero if all bit b i di h l to ll bits being discharged are zeros and i l d d is largest if all di h t ll discharged d bits are ones.

(2 b 2 ) t 0

(12.8.2)

Quantization
For a negative X, the truncation error will be different for three different formats: 1) Sign-Magnitude:

0 t 2 b 2
2) Ones-complement:

(12.9.1)

0 t 2 b 2
3) Twos-complement:

(12.9.2)

( 2 b 2 ) t 0

(12.9.3)

ELEN 5346/4304

DSP and Filter Design

Fall 2008

10

Quantization
2. Fixed-point: rounding
In case of rounding, the number is quantized to the nearest quantization level. The rounding error does not depend on the format used to represent negative numbers:

1 b ( 2 2 ) < r 1 ( 2 b 2 ) 2 2

(12.10.1)

In practice, >> b, therefore, 2- 0 in all expressions considered.

Quantization
3. Floating-point
Considering a floating-point representation floating point

Q ( X ) = 2E Q ( M )
X = 2E M

(12.11.1)

of a number

(12.11.2)

Quantization is carried out on the mantissa only in case of floating-point numbers. Therefore, it is more reasonable to consider the relative error.

Q( X ) X Q(M ) M = X M

(12.11.3)

In practice, a rounding quantizer can be modeled as follows:

Q ( X ) = 2 B round ( X 2+ B )
ELEN 5346/4304

(12.11.4)

DSP and Filter Design

Fall 2008