Lecture 2 encoding

Lecture 2

Communication basics.(2 hours)

01/09/2012 1

ContentsFurther readings: Behrouz, Chapters 3, 4, 5, 7 (section 7.1), 10

• Modulation and encoding Techniques– Digital-to-Digital– Digital-to-Analog ( for wireless)– Analog-to-Digital

01/01/04 2

Analog and digital signal representations

01/01/04 3

Digital-to-Digital - coding techniques (1)

• To create a good clock transition• To simplify receiving clock recovery.• To limit interference effect (high energy transition, eg: -1

---> +1, or +1 ---> -1)• To cancel DC components• To achieve high transmission• Code were developed – 8B/6T Tenary coding scheme used in 100Mbps

Ethernet on 100Base-T4, category 3. Every 8-bit of information (256 possibilities) is converted to 6 bits of codes (+1, 0, -1) (36 = 729 code words to chose from)

01/01/04 4

Digital-to-Digital - Non-return to zero

• NRZ-L used for terminals and other devices. • If the digital input is level 0, produce an output of positive

voltage (level +v), for the duration of the bit. • If the digital input is level 1, produce an output of negative

voltage (level -v), for the duration of the bit. • NRZI (also called differential encoding)used for magnetic

recording. • If the next level digital input is 0, stay on the same level (no

transition). • If the next level is 1, make a change to the opposite ( +/-) of the

the current level• Simple/ low frequency response, but has dc component,

lack of synchronization (long zeros), no good for transmission.

01/01/04 5

01/01/04 6

Digital input, base

NRZ-L

NRZI

Digital-to-Digital - Multilevel Binary• Solves the NRZ problems: No dc components (alternating 1

and 0 pulses), Low bandwidth compare to NRZ. But require more power to differentiate between 3 - levels, and the bit rate > NRZ. Used for ISDN for low data rate transmission.

• Alternate Mark Inversion (AMI): • If the level of the digital input is 0, the output line is zero. • If the level of the digital input is 1, the output signal alternates

between levels (+v) and (-v). There is no loss of synchronization for strings of 1’s.

• Pseudo-Tenary: It is the opposite of AMI. • If the level of the digital input is 1, the output line is zero. • If the level of the digital input is 0, the output signal alternates

between levels (+v) and (-v). There is no loss of synchronization for strings of 0’s.

01/01/04 7

01/01/04 8

Digital input, base

AMI

Pseudo-Tenary

• MLT-3 encoding (Multi-Level Transmit) is a line code that uses three voltage levels : -1, 0, +1, and 0. It moves to the next state to transmit a 1 bit, and stays in the same state to transmit a 0 bit.

• MLT-3 was first introduced by Cisco as a coding scheme for FDDI and CDDI. It was also used in the Ethernet 100BASE-TX.

01/01/04 9

Digital-to-Digital - Bi-Phase• No dc component. Less interference. Easy to detect error, many

transitions. However, the modulation rate is twice the bit rate. The signal is self-clocked.

• Used in LAN CSMA/CD (802.3), Token ring (802.5) with STP, up to 10Mbps. Not used in WAN require high signaling rate.

• Manchester Encoding: • In the middle of any bit for the input digital signal, there is a transition –

• For 0, the transition is from level high to low. (or vise versa, keep consistency)• For 1, the transition is from level low to high. (or vise versa, keep consistency)• The changes at the beginning of the bits match the changes of the next middle.

• Differential Manchester Encoding: If the next level in the input digital signal is 0 there is a transition – the output signal goes to the opposite level. If the next level in the input signal is 1 there is no transition at the beginning of a bit. There is always a transition – to the opposite level- in the middle of the bit for self-clocking.)

01/01/04 10

01/01/04 11

Digital input, base

Manchester

DifferentialManchester

Digital-to-Digital - coding techniques (2)• For instance– 00000000 ----> 0 +1 0 -1 +1 0– 00000001 ----> 0 -1 +1 0 -1 0 – 00000010 -----> 0 -1 0 1 -1 0

• 4B/5B code for full duplex 100Base-Tx category 5 and 100Base-Fx Ethernet. Every 4 bits are converted into a 5 bits binary code. The remaining 16 codes are used for frames boundaries

• 8B/10B code (Fiber Channel Code) developed by IBM and licensed to 1Gbps Ethernet, for full duplex. The rules for choosing a code word amount 1024, no code word should have more than 4 identical bits in a row, no code may have more than 6 (0’s) or 6 (1’s) .

01/01/04 12

Digital-To-Analog - modulation• Uses public telephone line, transmitting voice frequency

in the range 300 to 3400 HZ. Utilize modems.• Amplitude-shift keying (ASK) modulation• Frequency-shift Keying (FSK) modulation– Binary FSK (BFSK)– Multiple FSK (MFSK)

• Phase-shift Keying (PSK) modulation– Binary PSK– Differential PSK (DPSK)– Multiple Level PSK

• Quadratic AM and PSK: combination of ASK and ASK

01/01/04 13

ASK• Used on voice-grade lines < 1200bps.• Used in some optical fiber.• Used in Infrared wireless LAN, intensity

modulation.• Inefficient.• ASK ( Amplitude Shift Keying) works as :– Binary 1 is transmitted by a sine wave of amplitude

A1 cos(2 f t).– Binary 0 is transmitted by a sine wave of amplitude

A2 cos(2 f t), Most implementations uses A2 as 0.

01/01/04 14

01/01/04 15

ASK

Digital

ASK

A

-A

FSK• Sender / Receiver may use different frequencies to allow

full-duplex transmission on the same channel, Modems 1070, 1270, 2025, 2225 Hz.

• Less susceptible to errors than ASK.• Can be used for higher frequencies (3-30MHz), radio

transmission and LANs.• Can support Multiple levels MFSK.• BFSK: ( Binary Frequency Shift Keying)– Binary 1 is transmitted as A cos(2 f1 t).

– Binary 0 is transmitted as A cos(2 f2 t).

01/01/04 16

01/01/04 17

FSK

Digital

BFSK

MFSK• MFSK: – Can support Multiple levels MFSK.– Where– fi = fc +(2i-1-M)fd

– fc is the carrier frequency

– fd is the difference frequency between two consecutive ones– M the number of different signals M = 2L, L is the number of

bits in a symbol.– The total bandwidth is wd = 2(M-1)fd

– The bit rate is 1/T and the signal element is held for a period of Ts = LT

01/01/04 18

MitfAtS ii 1)2cos()(

Example MFSK

• M=4 that each different signal represent two bits (n=2).• With fc = 250KHz and fd = 25KHz we have the following frequencies:

• f1 = 175KHz, f2 = 225KHz, f3 = 275KHz, f4 = 325KHz

• The bandwidth is Wd = 2*(4-1)*25 = 150KHz, the bit rate is 50Kbps.

01/01/04 19

T

Ts

175

225

275

325

11000010011111001101

Wd

PSK• BPSK: ( Binary Phase Shift keying)– for the binary 1, produce the sine wave A cos(2 f t).– for the binary 0, produce a sine wave A cos(2 f t + 180) =

- A cos(2 f t).• DPSK: ( Differential PSK)– Avoid using accurate local oscillator at the receiver.– 0 -> send a signal similar to the previous one.– 1->send a signal with phase shift to the previous one.

• MPSK:– QPSK Quadratic PSK uses phase shift of multiples of /4,

for more details see reading list.

01/01/04 20

01/01/04 21

PSK

Digital

A

BPSK

DPSK

Constellation Diagram (1)

01/01/04 22

2700

900

001800

2700

900

001800

QPSK: same amplitude4 different phases, to represent 4 different symbols, needing 2bits.

QAM-16: 4 different amplitudesand 4 different phases. Representing 16 different symbols, each symbolneed 4 bits. For a baud of 2400, the transmission rate will be 2400*4 =9600 bps

Analog-to-Digital• Voice sent with digital transmission.• CODEC - Digitization. • Pulse Code Modulation PCM: signal is sampled

into discrete levels and coded in binary them transmitted.

• Delta modulation DM: improve on the complexity of PCM but modulating the differences in the signal.

01/01/04 23

Digitization of analog waveforms

01/01/04 24

Sampling frequency >= 2* bandwidth of the signal.

Quiz1. What is the difference between Digital to Analog

and digital to digital modulations?2. What is a DC component?3. Where do we use analog to digital modulation?4. Give an example for 4B/5B code.5. What information does the constellation

diagram give?6. How does MFSK improve the bit rate over FSK?

01/01/04 25