Multiplexing Techniques

1
Multiplexing Techniques
2
Multiplexing Techniques

• There are 3 basic types
• Frequency-Division Multiplexing (FDM).
• Time-Division Multiplexing (TDM).
• Statistical Time-Division Multiplexing (STDM).

3
FDM

• High bandwidth medium when compared to signals to
  be transmitted.
• Widely used (e.g., TV, radio).
• Various signals carried simultaneously where each
  one modulated onto different carrier frequency,
  or channel.
• Channels separated by guard bands (unused) to
  prevent interference.

4
TDM

• TDM or synchronous TDM.
• High data rate medium when compared to signals to
  be transmitted.

5
TDM

• Time divided into time slots.
• Frame consists of cycle of time slots.
• In each frame, 1 or more slots assigned to a data
  source.

U1
U2
...
UN
1
2
...
1
2
N
...
N
Time
frame
6
Statistical TDM

• Dynamically allocates time slots on demand.
• Multiplexer scans input lines collecting data
  until frame is filled.
• Demultiplexer receives frame and distributes data
  accordingly.

7
STDM

• Data rate on muxed line lt sum of data rates from
  all input lines.
• Can support more devices than TDM using same
  link.
• Problem peak periods.
• Solution multiplexers have some buffering
  capacity to hold excess data.
• Tradeoff data rate and buffer size (response
  time).

8
Finally.Piggybacking

• When both endpoints transmit, each keeps 2
  windows, transmitter and receiver windows.
• Each send data and need to send ACKs.
• When sending data, transmitter can piggyback
  the acknowledgment information.
• When no data, send just the ACK.

9
ATM Switches
10
Switching

• We have already seen traditional crossbar
  switches (and space division switches that are
  made up of lots of small crossbar switches).
• We have also looked at time division switches
  (used in digital systems).
• Now we will look briefly at the ATM cell switch.

11
ATM cell Switch

• The ATM switch has a number of inputs (typically
  1024) and outputs (1024 again) that connecting
  incoming and outgoing lines.

12
ATM Switching Cycle

• The cells received on each incoming line are
  buffered.
• During each cycle, the last complete cell is
  taken from each input lines buffer.
• The header of each cell is examined by the
  internal switching fabric.
• The switching fabric determines the output line
  on which the cell should be transmitted.
• The cells are then transmitted on the appropriate
  output lines.

13
Duration of a Switching Cycle

• Cells arrive at the ATM switch at about 155 Mbps.
  This means there are 360,000 cells arriving on
  each line every second.
• This means that an ATM Switching cycle must be
  started every 2.7?sec.
• Fortunately, the ATM switch can be pipelined
  (that is to say that later stages of previous
  switching cycles can be executed concurrently
  with the current switching cycle).

14
High Speed Switching Cycle

• Things are worse for ATM switches that need to
  deal with 622 Mbps lines. The switching cycles
  for these must be initiated every 700nsec
  (computer memory typically has an access time of
  60ns).
• It is only because ATM uses small fixed sized
  packets that such switching speeds are practical.

15
Goals of an ATM switch

• All ATM switches have the following goals
• Switch all cells with as low a discard rate as
  possible
• Never reorder the cells on a virtual circuit
• In an emergency, a cell can be dropped but this
  should happen only rarely.
• A loss of 1 or 2 cells per hour is just about
  acceptable.

16
Queuing

• The second goal can cause some difficulty within
  an ATM switch.
• What happens when two cells want to use the same
  output line?
• The only way to overcome this conflict is to
  queue cells.
• A cell that wants to use the same output as a
  cell from another input is held back in the input
  buffer until the next cycle.

17
The Knockout Switch

• An alternative to queuing cells in the input
  buffer is to provide queuing in output buffers as
  in the knockout switch.

18
The Knockout Switch

• Incoming cells are broadcast in the same cycle
  that they arrive in.
• When each cell arrives, the hardware inspects it
  header information and enables the appropriate
  crosspoint.
• The cell travels along the bus until it gets the
  the enabled crosspoint.
• The cell is then directed towards the correct
  output line.

19
Advantages of Knockout Switch

• Several packets can be sent the the same output
  line (where they will be queued until the output
  is available for transmission).
• Typically there is room for only n cells in the
  queue. If the queue is full, then additional
  cells are discarded (this is, hopefully, unlikely
  to happen).
• Multicasting is also possible with the knockout
  switch (it is just a matter of enabling multiple
  crosspoints).

20
Practical Work
Consider transmitting the two binary bytes
11001010 00010101. We could send the bits as a
continuous stream of 1s and 0s. Here, each bit is
sent at fixed time intervals. This is
synchronous transmission. What would happen if we
ran out of data?
One way to overcome the problem is to send
individual bytes rather than a continuous stream
of bits. The start of each byte is indicated with
a start bit and ended with a stop bit.
21
Finally.Disadvantages

• The problem with the knockout switch is that it
  is basically a crossbar switch and the number of
  crosspoints increases quadratically with the
  number of lines.
• We solved that problem with the space division
  switch that breaks a large switch into lots of
  smaller switches.
• A variant of this multistage switch is called the
  Batcher-banyan switch.