When several aircraft are arriving at the same airport within a short period of time, the controller has to arrange for a proper sequence of the traffic, since they cannot not all land at the same time.
For that purpose he may use radar vectors to the final approach course to establish a mileage interval between aircraft and keep a safe spacing on final.
However, when using procedural control (see, things are a little bit different in the sense that the spacing between aircraft will considerable be increased..... )
The controller may have to deal with traffic reaching the IAF at the same time or almost and of course can't have them starting the approach procedure all together. In addition, according to the type of procedure and to local rules, it happens that no aircraft can leave the IAF and start an approach procedure until the preceeding aircraft has landed.
Then the controller will have to keep some of the arrivals into the holding pattern(s) and give regulation times. That is times they may leave the holding pattern again.
The first one will start the procedure directly and the other ones will stay into the holding pattern(s) until they are authorized to leave it. These regulation times are called EAT (Expected Approach Time), representing the time at which ATC expects to be able to give a clearance to an aircraft to leave the IAF and start the instrument approach procedure.
When given an EAT, the pilot is responsible for managing the holding pattern so that he/she overflies the IAF at the given time. He may shorten the published legs of the pattern (say 30 seconds outbound/ inbound legs instead of the published 1 minute for instance).
If the EAT has changed by 5 minutes or more, ATC must advise the pilots ASAP!
A Pilot who experience a radio failure, should leave the Holding as close as possibel to the last received EAT.
In certain countries, the fact of assigning a time to leave the holding to start a apporach is called TIMED APPROACHES procedure.
Say, the average time between IAF and runway is 7 mins for the types of aircraft involved, and no aircraft can leave the IAF until the preceeding has landed. Say the IAF altitude is 3000 ft.
Given the following aircraft : A expecting IAF at 10:12, B expecting it at 10:14 and C at 10:17.
According to the above, 2 of them will have to wait a little bit before proceeding beyond the IAF.
A is cleared as soon as arriving and leaves the IAF at, say, 10:13 and then should land at 10:20.
Keep B in the holding at 4000 ft and give EAT 10:20. When C arrives, keep it at 5000 ft and give EAT 10:27.
B will have to wait for 6 mins and C for 10 mins.
If the pilots comply with the given times, B will reach the IAF when A has landed and will be authorized for the approach.
Make sure to have B at 3000 ft over the IAF, then give 4000 ft to C as soon as 4000 ft are released by B.
7 mins later, it should be OK for C too.
This is why it is important the pilots respect the given times. If B reaches the IAF 2 minutes late, then C will have to be given a new EAT at 10:22. If other aircraft are to come behind, such delays may involve their holding to be quite long.
This is a simple case. Things may be much more complicated if the types of aircraft are really different (small turboprops and 4-engine commercial jets for instance). The timed approaches procedure / EAT assignments have to be adapted to each airfield. Some guidelines may be indicated on divisions websites.
Normally, the rule is first at IAF, first to land. However, the controller may have to make some tactical choices like authorizing the second arriving aircraft first if this can improve the approach sequence.