What he's chosen to do is not easy and would be a pretty respectable project for a final year of a degree at a reasonable university.
There are plenty of cheap affordable radio modules available. The trouble is finding something with accurate enough timing to work out the position of a player on the field and no, you can't just go by the intensity of the signal because it will depend on the weather conditions and obstacles.
To give you an idea of the timing, assuming the field is 105m long and there's a receiver at either end. If the player is 5m away from one end of the field, calculate how long the signal will take to be received by each receiver.
Light travels at about 300*106m/s (c in the formula used below)
The time it takes for the signal to reach the nearest receiver:
t1 = d/c = 5/300*106 = 16.67-9s
The time it takes for the signal to reach the furthest receiver:
t2 = d/c = 100/300*106 = 333.3-9s
The difference between both receivers would be:
333.3-9 - 16.67-9 = 316.7-9s = 316.7ns
The problem is, if you connect the two receivers together via cable or they communicate wirelessly, the delay in transmission would be longer than the time measured so you'll need to subtract it.
You also only need three receivers (not transmitters) placed around the field to locate the position of the ball or a player, using a process known as trilateration.
One idea is to have transceivers (both transmitters and receivers) inside the objects on the field (ball and players) and two receivers and one transmitter around the field. Most of the time the transceivers are just listening, the transmitter sends out a code (the ball or a player number), as soon as this is received by the appropriate transceiver, it transmits a signal which is receiver at the edge of the field. If the delay between the transceiver receiving and transmitting is known and is accurate, then it should be possible to work out the position from the times when the signal is received.
The trouble with this whole system is timing: all delays and processing times must be consistent and accurate, any change will mess things up. What you want to do is emulate the GPS system used in satellite navigation on a smaller scale; something which is not easy to do.
Just another thought, you don't want more than one transmitter on simultaneously to save interference and avoid the need to too much processing.
See the attached diagram.
The base unit is the part which connects to the PC.
First you need to know how far XC1 and XC2 are away:
Send a signal to XC1, wait how long it takes for XC1 to send the signal back. Repeat for XC2
Now you know how far XC1 and XC2 are, assuming they're at 90 degrees from the base.
Send a signal to the object from the base, time how long it takes for it to come back to base.
Send a signal to XC1 telling it to send a signal to the object and time how long it takes to get back to XC1, the time code can be sent back to the base or the base could time how long it takes to receive the signal from XC1. Repeat for XC2.
Perform trilateration (see Wikipedia)
To calibrate you could place the object to the left of the field at the half way line, then at the right of the field at the half way line: you'd then know the distances so you can calculate the delay for the electronics inside of the object. For the most accurate results, you'll need to calibrate each object individually and store the results in a table.