3F4 Data Transmission Introduction

1
3F4 Data TransmissionIntroduction

• Dr. I. J. Wassell

2
Pre-requisites

• Familiarity with IB courses
• Signal and Data Analysis (Paper 7)
• Linear Systems and Control (Paper 6)
• Communications (Paper 6)

3
Booklist

• Couch, L. W, Digital and Analog Communication
  Systems, Prentice Hall (5th Edition). Covers all
  except DFE.
• Shanmugam, K. S. Digital and Analog Communication
  Systems, Wiley. All except DFE.
• Proakis, J. G, Digital Communications, McGraw
  Hill.
• Wicker, S. B., Error Control Systems for Digital
  Communication and Storage, Prentice Hall, 1995.

4
Applications

• Data transmission over copper cables and optical
  fibres, e.g.,
• computer local area networks (LANs)
• integrated services digital network (ISDN)
  connections, e.g., Basic rate (2BD) and primary
  rate (30B2D) channels. B64kBit/s, D16kBit/s.
• 30 channel pulse code modulation (PCM), i.e., 30
  telephony channels, 2.048MBit/s.
• Asynchronous transfer mode (ATM) in wide area
  networks (WANs), e.g., 25MBit/s, 155MBit/s

5
Applications

• Data storage,
• magnetic disk drives
• magnetic tape drives
• optical disk drives

6
Topics Covered

• Communications System Model
• Pulse Amplitude Modulation (PAM) for baseband
  data transmission
• Intersymbol Interference (ISI), noise and bit
  error rates (BER)
• Pulse Shaping for bandwidth control and
  elimination of ISI

7
Topics Covered

• Line coding schemes
• Optimum Transmit and Receive Filtering
• Equalisation to compensate for undesirable
  channel characteristics
• Error control coding (ECC)

8
Baseband Transmission

• The transmitted signal is limited to a range from
  -B Hz to B Hz.

X(f)
X(f)
0
0
f
B
-B
fc
-fc
f
Baseband
Bandpass

• Example baseband channels include
• copper cable, magnetic disk, CD-(ROM)

9
Comms System Model

• Transmission of digital data in communications
  channels
• True digital data, eg, comms in a computer
  network
• Analogue information which has been converted to
  a digital format, eg anti-alias LPF followed by
  A/D conversion

10
Transmission Model
11
Components of the Model

• Assume input source is in the form of symbols, eg
  bytes from a PC, or 16-bit audio samples
• Source encoding- Transforms digital symbols into
  a stream of binary digits (BITS), eg PCM
• Error Control Coding- Adding extra bits
  (redundancy) to allow error checking and
  correction.

12
Components of the Model

• Line Coding- Coding of the bit stream to make its
  spectrum suitable for the channel response. Also
  to ensure the presence of frequency components to
  permit bit timing extraction at the receiver.
• Transmit Filtering- Generation of analogue pulses
  for transmission by the channel.

13
Components of the Model

• Channel- Will affect the shape of the received
  pulses. Noise is also present at the receiver
  input, eg thermal noise, electrical interference
  etc.
• We will concentrate on the components to the
  right of the vertical dashed line.

14
Channel Response

• Assumptions
• Linear time-invariant (LTI) frequency response,
  ie, the channel frequency response Hc(w) is
  fixed, known and linear.
• Additive Gaussian noise- The channel noise has a
  Gaussian amplitude distribution (pdf) is often
  assumed to be uncorrelated (ie white, flat power
  spectral density) and additive.

15
Channel Response

• Thus the received signal may be expressed as,
• Y(w) Hc(w) X(w)N(w)
• Where,
• Hc (w) is the channel frequency response
• X(w) is the transmitted signal spectrum
• N(w) is the noise spectrum
• In practice,
• Channel response may be non-linear, time-varying
  or unknown
• Noise may be non-Gaussian, particularly
  interference.