GSM TIMING ADVANCE

GSM TIMING ADVANCE

The Problem

When bursts of information are transmitted on a carrier frequency, they obviously have to travel a certain distance from the mobile to the BTS. This distance is not known to the mobile or to the BTS, and so the distance could be anywhere from a few metres to the maximum 35km.

As the speed of a radio wave travelling through the ether is constant (or near enough as to make no difference for these purposes), then you can see that the time taken for the burst to travel from the mobile to the BTS will vary greatly depending upon the distance involved.

The GSM system must take account of this, because it is absolutely critical that all bursts arrive at the receiver within their allocated timeslot. If they fail to do so, then some of the burst will not be correctly demodulated and decoded by the receiver in its allocated timeslot, and the rest of the burst will cause interference for the call in the next timeslot; both of course are highly undesirable effects.

The Solution

GSM ensures that the first burst sent by the mobile in a call is much shorter than the normal type of burst; this type of burst is known as the Access Burst. It is sized so that even if the mobile is at the maximum possible distance from the BTS (35 km), the burst will always arrive completely inside a timeslot. In effect the guard period of the burst has been dramatically increased.

But how do we know for sure that even a short burst will arrive at the BTS within a timeslot? Well, first remember that the mobile is not a particularly intelligent instrument; it does only what it is instructed to do by the BTS for most of the time. Therefore when in idle mode (i.e. before the call began), the mobile was listening to transmissions on the Broadcast Control Channel (BCCH) to gain various information about the cell. It received this information in bursts within timeslots in much the same format, as it now wants to transmit its bursts in the opposite direction. So effectively the mobile’s time base is given to it by the BTS, and so the mobile knows when the beginning and end of the timeslots should be.

However the BCCH information was subject to a delay in reaching the mobile due to the distance involved, so the mobile’s time base is in fact “running a bit late” relative to the Timing at the BTS’s. When the mobile transmits its first burst to the BTS, therefore, it is late before it even leaves the mobile, and is then subject to the same delay (assuming the mobile is not moving) again on the way to the BTS. So therefore the Access Burst is actually short enough to suffer twice the delay due to distance and still arrive completely inside a timeslot. Once received, the “lateness” of the burst is measured by the BTS and used to calculate the required timing advance, which is then told to the mobile on the Access Granting Channel (AGCH). All subsequent bursts can now be full size as they are transmitted early by the mobile and so will arrive on time at the BTS.

If the mobile moves relative to the BTS during the call, the timing advance being used may need to change; this is detected by the BTS because the full size burst will start to arrive either early or late, and so the BTS can again tell the mobile the required value of timing advance only this time using the Slow Associated Control Channel (SACCH).

The maximum speed at which a mobile can move relative to the BTS is 250 km/h; if the propagation path lengthens or shortens quicker than this, the BTS’s timing advance processing may not be fast enough to keep up with it and the call may fail due to bursts not being correctly decoded.