The following three processes arrive simultaneously and are to be scheduled on a single CPU for the specified burst length, measured in abstract time units. Assume that no process does any I$/$O$,$ that the overhead of a context switch is one time unit, and that the idle process will always be run when there are no other processes scheduled.

Process ID CPU burst length

P$110$

P$220$

P$330$

$1)$Calculate the total turnaround time of the three processes when scheduled using Round Robin $($RR$)$ assuming a time quantum of two time units.

$2)$Calculate the total turnaround time of the three processes when scheduled using Shortest Job First $($SJF$).$

$3)$Unfortunately$,$ the CPU burst length of a process is not actually known when it arrives. Assume that it takes the SJF scheduler $10$ time units to estimate the CPU burst length of a process before its execution. Recalculate the total turnaround time using SJF$.$

$4)$Compare your results for RR $($i$)$ against those for SJF when the burst lengths are known $($ii$)$ and unknown $($iii$).$ When does the total turnaround time for SJF exceed that for RR$?$ Which parameter has the most affect on the relative performance of these two schedulers, and why?