Beruflich Dokumente
Kultur Dokumente
Presented By:
Adnan Mir
Email: adnanmir@ciit.net.pk Room No. 328 Department of Electrical Engineering Comsats Institute of Information Technology Abbottabad 22060 Pakistan
1.1
Syllabus: Part 1
1.
2.
3. 4. 5.
6.
1.2
SOURCE
SINK
Any practical communication system can be mapped on to this generalised framework. Objective of the communication system is to reproduce source signal at sink. Functions of individual elements and their signal processing requirements will now be considered.
1.3
1.2 GCS Elements and their Signal Processing (SP) Requirements 1.2.1 Source
SOURCE
INFORMATION (ESSENTIAL)
REDUNDANCY (NON-ESSENTIAL)
Many sources can be processed to remove some of the redundancy and hence allow the transmission channel to be used more efficiently. This is a function of the source encoder. Also, analogue sources, e.g. speech, may be digitized; this is a further function of the source encoder and involves sampling and quantization. As more is known of source characteristics, so a more efficient source encoder can be designed.
1.4
Data Word 101 010 110 111 100 011 000 001 Total
Probability 0.6 0.2 0.05 0.05 0.04 0.03 0.02 0.01 1.00
Average number digits after encoding average number of digits before encoding for compression to be successful. Signal processing involves statistical analysis of source data and code design.
1.5
SPEECH SPECTRUM
FREQUENCY 0 f1 f2 f3 f4 4kHz
Only information on high energy regions is digitized and transmitted, together with voiced/unvoiced decisions. Thus, effective sampling rate is reduced. Signal processing requires real-time digital spectrum analysis routines, digital filtering and analogue-to-digital conversion (ADC).
1.6
PIXEL
x
After digitization, typically employs 2-D digital spatial transform to describe areas of some pixel type efficiently; also frame-to-frame differential encoding, and variablelength compression coding (MPEG algorithm). Signal processing requires 2-D transforms, differential encoding, data compression coding and ADC.
1.7
INFORMATION
PARITY CHECKS
CODEWORD
Involves modulation to match the coded data to the type of channel being used (baseband or RF carrier). As more is known about the channel characteristics, so a more efficient channel encoder can be designed. Signal processing involves time and frequency domain analysis of waveforms and spectra, together with code design and ADC.
1.8
1.2.4 Channel
Typical effects: Gaussian and/or non-Gaussian noise; Multipath; Doppler Shifts; Delay; Fading. Signal processing to counter channel effects on the received signal may include noise suppression, equalisation, frequency offset correction, automatic gain control and diversity combining.
1.9