1
1
Digital Communication
Systems
Dr. Shurjeel Wyne
Lecture 3
Formatting
2
zThe first block in any Digital Communication
System
zPurpose:
zTo insure that the message (or source signal) is
compatible with digital signal processing
Formatting
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Guidelines and tips
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Formatting-Digital Communication Systems-Lecture Slides, Slides of Digital Communication Systems
Dr. Shurjeel Wyne delivered this lecture at COMSATS Institute of Information Technology, Attock for Digital Communication Systems course. In this he discussed: Formatting, Digital, System, Processing, Compatible, Textual, Data, Character, Coding, Encoding
Typology: Slides
2011/2012
1 / 14
This page cannot be seen from the preview
Don't miss anything!
On special offer
Related documents
Partial preview of the text
Download Formatting-Digital Communication Systems-Lecture Slides and more Slides Digital Communication Systems in PDF only on Docsity!
Digital CommunicationSystems
Dr. Shurjeel WyneLecture 3
Formatting
z The first block in any Digital CommunicationSystem
z Purpose: z To insure that the message (or source signal) iscompatible with digital signal processing
Formatting
z Formatting Analog signals z Sampling, Quantization, Encoding z Formatting Textual data (character coding) z Encoding
Formatting
What about computer-to-computer communication?
Formatting analog signals z To transform an analog waveform into a form that is compatible with digitalcommunication, the following steps aretaken: z z SamplingQuantization and encoding z Baseband transmission Format Sourcesignal Sample Quantize Encode Bit streammodulate^ Pulsewaveforms Pulse
z^ SamplingSampling signal, x is the processes of converting a continuous-time analog time intervals a(t),^ into a discrete-time signal by taking “samples” at discrete- x (^) a(t)
Samples T (^) s
- second are required compared with if the waveform is rapidly varying• component present in the signal If the signal is slowly varying, a fewer number of samples perThe optimum sampling rate depends on the maximum frequency
Aim analog signal intodiscrete signal sothat we can: convert an perform digitalprocessing andafterwardsreconstruct the original analogsignal
Sampling theorem z If the sampling is performed at a proper rate, no info is lostabout the original signal and it can be properly reconstructed
z Sampling theorem: components abovevalues sampled at uniform intervals of A bandlimited signal with no spectralHz can be uniquely determined by z z Alternatively, the sampling rateThe sampling rate, must satisfyis called Nyquist rate.
Analogsignal^ Samplingprocess Discrete time signal
Sampling Time domain Frequency domain
x (^) s ( t )= x δ( t )× x ( t )
x ( t ) | X ( f ) | | X δ( f ) | x (^) s ( t ) | X (^) s ( f ) |
x δ( t ) X (^) s ( f )= X δ( f )∗ X ( f )
z Anti-Aliasing Filter Aliasing can be prevented by first passing the analog signalthrough an anti-aliasing filter before sampling is performed The anti-aliasing filter is simply a LPF with cutoff frequencyless than or equal to half the sample rate, fs / Bandwidth of the sampled signal is forced to satisfy therequirement of the Sampling Theorem
Preventing Aliasing effect
z Other methods of preventing aliasing^ Preventing Aliasing effect z Over sampling (increasing f (^) s ) z High-order filters (increase slope of filter roll-off)
The highersampling ratefaliasing by (^) s’ eliminates separating thespectralreplicates Sharper cut-offfilters eliminatealiasing
Practical Sampling Rates z Speech
- Telephone quality speech has a bandwidth of 4 kHz(actually 300 to 3300Hz)Most digital telephone systems are sampled at 8000 z Audio: - samples/secThe highest frequency the human ear can hear is
- approximately 15kHzCD quality audio are sampled at rate of 44,000samples/sec z Video - The human eye requires samples at a rate of at least 20frames/sec to achieve smooth motion
Quantization
z Quantization is the process of representing a large(possibly infinite) set of values with a much smaller set.
8 bits/pixel 4 bits/pixel
Tradeoff between qualityand compression Effect of quantization
Sourcesignal (^) Sample Quantize Format Encode^ Bit stream
Quantization z Mapping samples of a continuous waveform to a finite set of amplitudes. In
Out Average quantization noise powerSignal peak power^ Quantizedvalues quantization noise powerSignal power to average VV^ Lq (^) pp (^) pp^ = 2V= Number of quantization levels= step size, in volts, between quantization levels= Lq (^) p = peak-to-peak voltage range of analog signal
z^ Quantization error Quantizing error:of a quantizer The difference between input and output z z Granular or linear errors of quantizer Saturation errors happen for inputs outside dynamic range of happen for inputs within dynamic range quantizer z z Saturation errors are larger than linear errorsSaturation errors can be avoided by proper tuning of AGC
e ( t )= x ˆ( t )− x ( t )
x ( t ) ˆ x ( t ) xe ˆ(( tt )) −= x ( t )
AGC y^ Qauntizer^ = q ( x^ ) x Process of quantization noise x ( t ) x ˆ( t ) e ( t )
Model of quantization noise
z Uniform Quantization: distributed (case, the step size z The quantization noise is the same for all signal magnitudes. equally-spaced q is the same (uniform) over the When the quantization levels are uniformly) over the quantizer’s dynamic range. dynamic range. In this z Non-uniform Quantization: over the quantizer’s dynamic range.^ z^ (SNR)^ q^ is worse for low-level signals than for high-level signals. Un-equal size of quantization interval q z (^) signals.Fine quantization for weak signals, coarse quantization for strong
Uniform and non-uniform quantization
Non-uniform quantization for speech signals z In speech, weak signals are more frequent than strong ones.
z z Using equal step sizes (uniform quantizer) gives lowsignals and highAdjusting the step size of the quantizer by taking into account the for strong signals. for weak speech statistics improves the average SNR for the givendistribution of input signal levels.
1.00. Probability that abscissavalue is exceededNormalized magnitude of speech signal1.0 2.0 (^) 3. ⎜⎝⎛ NS^ ⎟⎠⎞ q ⎜⎝⎛ NS^ ⎟⎠⎞^ q
Input 25
Output
x ( t ) y ( t ) x
y = C ( x ) Compress Qauntize
A-Law Companding Standard (Europe, China, Russia, Asia, Africa)
where x and y represent the input and output voltagesA is a constant number determined by experiment A = 87.6 commonly usedA = 1 corresponds to uniform quantization
Types of Companding…
μ -Law Companding Standard (North & South America, and Japan) where z z x and y represent the input and output voltagesμ is a constant number determined by experiment z z In the U.S., telephone lines use companding withμ 0 corresponds to uniform quantization μ = 255
Types of Companding
- 0^0 2 - 1. 5. 5 012 0 0 2 3 0 0 2 4 0 0 2 5 0 0 2 6 0 0 2 7 0 0 2 8 0 0 2 9 0 0 3 0 0 0
- 0^0 2 - 1. 5. 5 012 0 0 2 3 0 0 2 4 0 0 2 5 0 0 2 6 0 0 2 7 0 0 2 8 0 0 2 9 0 0 3 0 0 0
- 0^0 - 1. 5. 5^010 1 0 0 0 2 0 0 0 3 0 0 0 4 0 0 0 5 0 0 0 6 0 0 0 7 0 0 0 8 0 0 0 9 0 0 0 1 0 0 0
- 0^0 - 1. 5. 5^010 1 0 0 0 2 0 0 0 3 0 0 0 4 0 0 0 5 0 0 0 6 0 0 0 7 0 0 0 8 0 0 0 9 0 0 0 1 0 0 0
x [ n ]=speech^ Example:^ μ -law Companding y Companded Signal[ n ]=C( x [ n ]) SegmentSegment of of y (^) [ xn [] n ] Companded Signal
Close View of the Signal