Study, 016-Coding

■ Basics

- Why coding: for digital processing

- Coding = transforming symbols from one type to another type

- Encoding = coding into a particular format

- Decoding = reverse of encoding

- Reversible coding: original and the decoded data are identical. e.g. lossless data compression

- Irreversible coding: e.g. GIF

■ Glossary

OOK: on/off key = a special type of ASK

ASK: amplitude shift key = digital AM modulation (signal 0 or 1)

FSK: frequency shift key

■ Line coding

* Definition:

- A code chosen for use within a communications system for transmitting a digital signal down a line.

- Line = physical channel = transmission medium or data storage medium

- Differential signaling on a physical transmission line

* Basics

- Line coding consists of representing the digital signal to be transported, by a waveform that is appropriate for the specific properties of the physical channel (and of the receiving equipment).

- The pattern of voltage, current or photons used to represent the digital data on a transmission link is called line encoding.

- The common types of line encoding are unipolar, polar, bipolar, and Manchester encoding.

- DC component or no DC component

- Digital baseband modulation, digital baseband transmission

* Applications:

- Line-coded signal (the "baseband signal") undergoes further pulse shaping (to reduce its frequency bandwidth) and then modulated (to shift its frequency) to create an "RF signal" that can be sent through free space.

- Line-coded signal can be used to turn on and off a light source in free-space optical communication, most commonly used in an infrared remote control.

- Line-coded signal can be printed on paper to create a bar code.

- Line-coded signal can be converted to magnetized spots on a hard drive or tape drive.

- Line-coded signal can be converted to pits on an optical disc.

* Binary line codes

Signal	Comments
NRZ–L	Non return to zero level. This is the standard positive logic signal format used in digital circuits. 1 forces a high level 0 forces a low level
NRZ–M	Non return to zero mark 1 forces a transition 0 does nothing (keeps sending the previous level)
NRZ–S	Non return to zero space 1 does nothing (keeps sending the previous level) 0 forces a transition
RZ	Return to zero 1 goes high for half the bit period and returns to low 0 stays low for the entire period
Biphase–L	Manchester. Two consecutive bits of the same type force a transition at the beginning of a bit period. 1 forces a negative transition in the middle of the bit 0 forces a positive transition in the middle of the bit
Biphase–M	Variant of Differential Manchester. There is always a transition halfway between the conditioned transitions. 1 forces a transition 0 keeps level constant
Biphase–S	Differential Manchester used in Token Ring. There is always a transition halfway between the conditioned transitions. 1 keeps level constant 0 forces a transition
Bipolar	The positive and negative pulses alternate. 1 forces a positive or negative pulse for half the bit period 0 keeps a zero level during bit period

* Code selection criteria

Minimize transmission hardware

Facilitate synchronization

Ease error detection and correction

Minimize spectral content

Eliminate a dc component

* Common line codes

AMI

Modified AMI codes: B8ZS, B6ZS, B3ZS, HDB3

2B1Q

4B5B

4B3T

6b/8b encoding

Hamming code

8b/10b encoding

64b/66b encoding

128b/130b encoding

Coded mark inversion (CMI)

Conditioned diphase

Eight-to-fourteen modulation (EFM): used in Compact Discs

EFMPlus: used in DVDs

RZ (return-to-zero)

NRZ (non-return-to-zero)

NRZI (non-return-to-zero, inverted)

Manchester code and its variants (differential Manchester, biphase mark code)

Pulse-position modulation: generalization of Manchester code

Miller encoding (= Delay encoding = modified frequency modulation) and its variant (modified Miller encoding)

MLT-3 encoding

Hybrid ternary codes

Surround by complement (SBC)

TC-PAM

* Optical line codes

Carrier-suppressed return-to-zero

Althernate-phase return-to-zero

TS-FO (three of six, fiber optical)

■ Channel coding

■ Source coding

■ Bit synchronization

■ Self-synchronization code

■ Clock recovery

- In receiving serial data transmission (hard drive, Ethernet), message sent w/o clock signal

- The receiver generates a clock from an approximate frequency reference, and then phase-aligns the clock to the transitions in the data stream with a phase-locked loop (PLL). This is one method of performing a process commonly known as clock and data recovery (CDR).

- Other methods include the use of a delay-locked loop and oversampling of the data stream

- For reliable clock recovery at the receiver, one usually imposes a maximum run length constraint on the generated channel sequence, i.e., the maximum number of consecutive ones or zeros is bounded to a reasonable number.

- A clock period is recovered by observing transitions in the received sequence, so that a maximum run length guarantees such clock recovery, while sequences without such a constraint could seriously hamper the detection quality.

■ Machester code (= phase encoding)

- 0/1, 1/0 transition occurs at the center of bit time.

- 동기화 용이, 오류 감소

[Ref]

OOK, ASK, FSK comparison: anthes, ash

Machester encoding: atmel-9164

Line coding: locicero, ee179, wpi, iit