Consider an Internet video conferencing application. Assume that the compressed video streams are packetized, and that a packet is either lost (due to over long delay or mis-delivery or detected bit errors within a packet) or received intact. Assume that the effective packet loss rate is quite high, about 10%, and that there is no feedback channel available. The average end-to-end delay between the source and destination is about 50 msec. Further assume that the maximum end-to-end delay allowed between the source and destination is 150 msec.
What would you propose as mechanisms to control and conceal the effect of packet loss? Which error resilient coding tools will be effective and which tools will not be effective in this case?
Consider the previous scenario again, but assume that the network has a lower packet loss rate, say 1%. Would your solution be different?
Again, consider the same scenario, but assume this time that a feedback channel is available between the destination and the source and that it usually takes 50 msec to deliver a feedback message. What are the solutions you propose in this case? What would you send on the feedback channel?