Knowing that some typical checksums are only 16 bits long, it appears that a corrupted data frame has a 1 in 2¹⁶ chance of having the same checksum as its original data frame. Discuss this idea with respect to the design of checksum algorithms.

Provide 3 distinct reasons why a stream of data is broken into frames by the Data-Link layer?

Under what circumstances would the use of Data-Link Layer frame pipelining be unnecessary (a waste of either time or effort)?