Inspect the network obtained in problem 4.1. You will likely notice that it consists of m columns of n/2 processors each, where m is a function of n obtained from your analysis. It is required to...

Inspect the network obtained in problem 4.1. You will likely notice that it consists of m columns of n/2 processors each, where m is a function of n obtained from your analysis. It is required to exploit this regular structure to obtain a sorting network consisting of a single column of n/2 processors that sorts a sequence of length n in 0(m) time. The idea is to keep the processors busy all the time as follows. The input sequence is fed to the processors and an output is obtained equal to that obtained from the first column of the bitonic sorting network. This output is permuted appropriately and fed back to the processors to obtain the output of the second column. This continues for m iterations, until the sequence is fully sorted. Such a scheme is illustrated in Fig. 4.8 for n = 8.