- 2 fundamental ways to share a communication channel:
- Sharing "space": divide up the frequency band of the communication channel
- Sharing "time": divide up the time of usage of the channel
- The signals used
to send messages
between computers are
electro-magnetic waves
- Electro-magnetic waves
can have
different frequences
- Sharing a channel by dividing the wave lengths:
- Electr-magnetic waves
of different frequences
(that are far enough apart)
do
not interfere
with one another
Example:
- Two radio stations can transmit siumultaneously on 2 different frequencies (that are far enough apart) without interfering with each other.
- Electr-magnetic waves
of different frequences
(that are far enough apart)
do
not interfere
with one another
- Frequency division multiplexing:
- The Frequency
Division Multiplexing (FDM)
(sharing) technique
divides the range of frequencies
of a communication channel
among the users
- Each user
uses a
subrange of frequencies
to transmit its messages
- Users must use non-overlapping ranges of transmission frequencies
- The Frequency
Division Multiplexing (FDM)
(sharing) technique
divides the range of frequencies
of a communication channel
among the users
- Note:
- This form of sharing
(frequency division multiplexing)
is not suitable for
computer communication
- Computers
must
communicate with each other
In order communicate, computer must use the same frequency (otherwise, they cannot hear other's transmissions)
- This form of sharing
(frequency division multiplexing)
is not suitable for
computer communication
- How to share a channel in the time dimension:
- Users take turn to transmit on the communication channel
- 2 fundamental ways
to share in the
time dimension
- Synchronous TDM :
- Users take turn transmitting in a pre-assigned order
- Asynchronous TDM :
- There is NO pre-assigned order on the transmission order of the users.
Asynchronous TDM is obviously more complicated
- Synchronous TDM :
- Synchronous Time Division Multiplexing:
- Time
is divided into
slots
- A fixed number of slots are grouped into a period
Example:
- Time
is divided into
slots
- A user
(e.g., a TCP connection) is assigned
(one of more) slots
inside a period.
The user is assigned the same slot position in each period :
- Transmission:
- When user has some data to
transmit, it waits for
its assigned slot
When the slot arrives, it can transmit the data
- When a user does not have any data to send, the assigned slot will contain no data (wasted bandwidth)
- When user has some data to
transmit, it waits for
its assigned slot
- Advantage of STDM:
- Simple to implement
- Provide very good
absolute (i.e., not statistical)
guarantees of service
(Because the transmission slots are reserved for a given user)
- Simple to implement
- Disadvantage of STDM:
- Wasteful
(When a user has no data to transmit, its reserved slots are wasted)
- Wasteful
- Asynchronous Time Division Multiplexing:
- Time
is divided into
slots
- The slots are
not organized
into
periods.....
- A slot is
given to a user
when it has data to transmit
- Users are given slots in a round-robin manner
Example:
Notes:
- Packets
from each flow
is stored in a
separate queue
- Router transmits one packet from each queue in a round robin manner
Idle flow:
- When a flow
does not have any packets queued ,
its turn
of transmission will be forfeited
(no transmission slots will be wasted as long as some flow is backlogged).
- Time
is divided into
slots
- Advantage of ATDM:
- Provide better bandwidth utilization
- Disadvantage of ATDM:
- More difficult to implement
than STDM
- More difficult to
provide
absolute guarantees of service
to flows
(Because we don't make reservations)
- More difficult to implement
than STDM