Packet Switching and Statistical Multiplexing: Improving Network Efficiency, Slides of Data Communication Systems and Computer Networks

Packet switching and statistical multiplexing as solutions to increase network efficiency. Packet switching involves fragmenting long messages into smaller packets and transmitting them individually, reducing the need for large buffers and improving link utilization. Statistical multiplexing allows for the dynamic allocation of time slots based on traffic demand, making it suitable for bursty data traffic. The document also covers connectionless and connection-oriented packet switching, their advantages and disadvantages.

Typology: Slides

2011/2012

Uploaded on 07/05/2012

raoo
raoo 🇵🇰

4.3

(22)

45 documents

1 / 11

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Communication and Digital
Switching
Docsity.com
pf3
pf4
pf5
pf8
pf9
pfa

Partial preview of the text

Download Packet Switching and Statistical Multiplexing: Improving Network Efficiency and more Slides Data Communication Systems and Computer Networks in PDF only on Docsity!

Communication and Digital

Switching

PACKET SWITCHING

  • If the message length is long, the intermediate nodes

are required to have large buffers for storage and this

increases the cost of the node. Also as the amount of

storage required increases the space occupied by these

nodes also increases.

  • A possible solution for the above problem is to

fragment the long messages into small size units,

known as packets. It is these packets that are

transmitted instead of the single long message. This

method is slightly different from Message switching

and is called Packet switching.

Packet switching

  • Another advantage is that no link is engaged for a long time

since the packets are of smaller size than the single

message. This permits better sharing of the links among

multiple users.

  • However the scheme just discussed has two major

drawbacks.

  • The packets of the same message traveling through different paths may arrive at the destination at different times due to different delays encountered in different paths. - The sequence number increases the overhead and require more network bandwidth
  • Some of the paths may not be very good and some packets may get lost. This worsen the quality. - They require retransmission which in turn requires more processing time and more bandwidth

STATISTICAL MULTIPLEXING (why ?)

• Every user has a different time slot in time

division multiplexing (TDM). In frequency division

multiplexing (FDM) every user gets a different

frequency band.

• The total number of different slot/bands is fixed

and thus the number of users also get fixed in

each of the schemes.

• These schemes are not efficient when traffic is

variable bit rate and/or heterogeneous like voice,

data and video because of their rigid structure.

STATISTICAL MULTIPLEXING

  • Statistical Time Division Multiplexing is also known as

Asynchronous TDM.

  • TDM with no regularity in slots.
  • The statistical multiplexer exploits the bursty nature of data

transmission by dynamically allocating time slots on demand.

  • Using this technique we never have to send an empty time

slot if there is no data to be transmitted.

  • Statistical multiplexing is best suited to bursty traffic (on time

is less than off time). Data traffic is very bursty in nature and

so we can use Statistical multiplexing in this case.

  • If statistical multiplexing is to be employed then packet

switching is the natural choice

Connectionless Packet Switching

• The kind of packet switching which is being

referred to is really connectionless packet

switching where no connection is pre-decided

for the packets.

• Because of this switching lacks the definition

of connection between the source and the

destination, this type of packet switching is

called connectionless (CL).

Connection Oriented Packet Switching

  • It is worth while to have a technique which is via media between the two extremes.
  • In circuit switching the entire path is dedicated to the single communication for the whole duration whereas in CL packet switching (CLPS) no path is even selected.
  • To take advantage of both the techniques it was suggested that let a path from source to destination be earmarked for a call but it should not be dedicated to the communicating users.
  • The path can be shared by other users.
  • All packets from a source to destination will travel the same chosen path. The packets would have to wait at the nodes till the forward link is available. Obviously all the packets will maintain their original sequence and the probability of packet loss would be less.
  • This switching is still packet switching but is connection oriented (CO).

CONCLUDING REMARKS

• Each of the different switching techniques has

some merits and some drawbacks. The choice

of a particular technique depends on the

application.

• A general conclusion can be that the

attraction of packet switching is in its ability to

utilize the lines better and to support

heterogeneous variable rate traffic. The circuit

switching gives the best reliability.