Your browser does not support JavaScript!
Technical Committee on Human Factors - Guidelines for real-time communication services
 
 
 
 
 

 Packet loss

 
 
 
 
 
Previous Lesson part Next Lesson part

Frequently asked questions

  1. Why is TCP-IP not suited for real-time person-to-person communication with constraints on low delay?
  2. What are the three main reasons for Packet loss?
  3. How can a user detect Packet loss?
  4. What is Burst packet loss?
  5. How does the receiving site know the correct order in which packets should be used?

1. Why is TCP-IP not suited for real-time person-to-person communication with constraints on low delay?

TCP-IP works as well as UDP-IP when there are no Packets lost. However, when a packet is lost, TCP-IP will retransmit the packet, and do so until correct reception has taken place. This implies a non-predictable delay.

2. What are the three main reasons for Packet loss?

  • Too late arrival of a packet,
  • None arrival at all,
  • Packet too much corrupted.

3. How can a user detect Packet loss?

  • For audio: Distortion, noise or silence,
  • For video: Quadratic blocking or frozen picture.

4. What is Burst packet loss?

If two or more sequential packets are lost, we call this burst packet loss.

5. How does the receiving site know the correct order in which packets should be used?

All packets have a sequence number. The communication software will sort the receiving packets if they arrive in the wrong order.