Line Coding

Sources

1. Lecture Slides
2. Class Lecture

What is Line Coding

• Also known as Pulse Modulation
• The conversion of binary data to digital signals for baseband transmission purposes.¹
• Each pulse unit is referred to as a symbol²
• Common pulse modulation techniques:
  • NRZ-L
  • NRZI
  • Bipolar-AMI
  • Pseudoternary
  • Manchester
  • Differential Manchester
• Categories
  • Unipolar, polar or bipolar
    • Unipolar: zero, positive, or negative
    • Polar: two polarities―positive and negative
    • Bipolar: two polarities and a zero
  • NRZ, RZ, Phase encoded or Multilevel binary
    • NRZ: two levels
    • RZ: three levels
    • Phase-encoded: uses phase inversion
    • Multilevel Binary

Line Coding Properties

Signal/Data Levels

• Signal levels refers to the number of possible values a signal can have.
• In contrast, data levels refer to the number of values for representing data³

Pulse/Symbol Rate

• Pulse rate refers to the number of symbols per second
• A pulse per bit interval exists for a number of line coding schemes. Nonetheless, it is possible to have no pulse at specific data segments.
  • For each data element, different line coding schemes will have a different number of signaling elements.
    • is ratio of the number of data elements () over a number of signaling elements ().
• Average signal rate
  • Multiply the case factor (, where ) to obtain the average signal rate

Spectrum

• Desired characteristics:
  • Lack of high frequency components—lowers the bandwidth requirement
  • Lack of a DC component—enables the usage of transformers and capacitors for coupling
  • Concentration of power at the bandwidth’s center
• Power spectral density (PSD)
  • Refers to the distribution of a signal’s power over frequency
    • • is the PSD
      • is the signal’s Fourier transform

Clocking

• Synchronizes the transmitter and receiver signals so that they can differentiate the symbols/bits from one another
  • Synchronize means that the phase and frequency are the same for the transmitter and receiver signal
  • The receiver must have a clock synchronized to the transmitter clock because the timing of the transmitter’s output waveform is synchronized to its clock
• Clocks can provide references to determine the bit intervals of the signal
• A self-clocking code is preferred because using a separate clock line/channel is expensive
  • Without self-clocking in the coding scheme, speed is limited due to the the chances of losing synchronization and misreading data

Error Detection

• Some line coding schemes can detect error
  • For instance, one that discerns errors through the occurrence of 2 successive positive/negative pulses
  • They are only effective when they can detect a significant amount of errors
• Error detection/correction blocks exists, hence it is not that necessary to have an error detection mechanism present in your coding scheme

Noise Immunity

• Different coding scheme posses different noise tolerance
• The stronger the signal is relative to the noise, the better. This can be determined using the signal to noise ratio (SNR)
  • The SNR can mitigate the negative effects of noise by specifying requirements to be classified as a high or low signal. (i.e. it specifies at what amplitude should the signal be to be considered as high/low)
  • Lower SNR indicates lower speed, as illustrated in the bandwidth formula

Cost and Complexity

• Some line coding schemes are more complex to implement, which also often means it costs more to carry out.

Common Pulse Modulation Techniques

NRZ

Non-Return-To-Zero Level (NRZL)

• Two different voltages for 0 and 1
• Voltage is constant in a bit interval
  • No return to zero voltage
• Usually, negative voltage for one value and positive for another (polar)
  • This is because using 1 and 0 (positive voltage and no voltage) will result in a positive nonzero average level, which is undesirable
    • A positive nonzero voltage corresponds to a dc signal, which is not preferred because they cannot pass in some communication circuits, does not contain information, and cost more power.
• Prevalent in digital logic systems

Differential NRZ Encoding (NRZ-I or NRZ-M)

• 1 is represented by a change/transition in level
• 0 is represented by no signal transition/change
• This tries to provide an error correcting property
• Used in magnetic recording
• NRZ-S is the inverse of this coding scheme, where 0 is represented by change and 1 is represented by staying the same (in terms of level).

Pros and Cons

• Pros
  • Simple implementation
  • Efficiently uses bandwidth
  • Used for magnetic recording
• Cons
  • DC component
  • Asynchronous
  • Not used much for signal transmission

Return-to-Zero (RZ)

Unipolar RZ

• 1 is represented by a half-width pulse
• 0 is represented by the absence of pulse
• Used for baseband transmission and magnetic recording

Bipolar RZ

• 1 and 0 is represented by a half-bit pulse of opposite polarities
• A pulse is present for every bit
• Used for baseband transmission and magnetic recording
• A self-clocking type of signal that can provide information about the signal’s frequency and phase
  • This is at the cost of higher bandwidth

Multi-level Encoding

Bipolar AMI (Alternate Mark Inversion)

• Mark means 1
• 1 is either a positive or negative pulse; it alternates between the polarities
• 0 is a no line signal
• It has error detection capabilities
• Compared to NRZ, it has no net dc component; nonetheless, it is not as efficient
• lower bandwidth than RZ
• maintains synchronization in long strings

Pseudoternary

• The same as Bipolar AMI but alternates with 0 instead of 1; therefore, 1 is a no line signal.

Footnotes

1. Check DIGDACM - Lesson 4 for more information about baseband transmission ↩

2. Check DIGDACM - Lesson 4 for more information about symbols ↩

3. For example, binary has two data levels because it uses 2 values for representing data ↩