What's the difference between the bandwidth of the link and the "effective bandwidth"?
kayvonf
Great question.
Here, the bandwidth of the channel is B bits per second.
However, if I send a message that is n bits in size, it takes T0 units of time for the first bit to arrive, and then n/B units of time for all the bits to arrive. Therefore the total time to send the message is (T0 + n/B). Assume that after the first message has arrived, the channel is used to send another message as shown here. Therefore from the perspective of the sender 'n' bits get sent every (T0 + n/B) units of time. That's the effective bandwidth.
As the message size n gets increasingly large, the start-up latency would become increasingly insignificant, and the effective bandwidth would trend toward the channel bandwidth B.
Another way to realize the full channel bandwidth is to overlap the start-up latency of one message which communication of data from another. More on this on the next two slides.
What's the difference between the bandwidth of the link and the "effective bandwidth"?
Great question.
Here, the bandwidth of the channel is B bits per second.
However, if I send a message that is
nbits in size, it takes T0 units of time for the first bit to arrive, and then n/B units of time for all the bits to arrive. Therefore the total time to send the message is (T0 + n/B). Assume that after the first message has arrived, the channel is used to send another message as shown here. Therefore from the perspective of the sender 'n' bits get sent every (T0 + n/B) units of time. That's the effective bandwidth.As the message size n gets increasingly large, the start-up latency would become increasingly insignificant, and the effective bandwidth would trend toward the channel bandwidth B.
Another way to realize the full channel bandwidth is to overlap the start-up latency of one message which communication of data from another. More on this on the next two slides.