Why do we need to aggregate frames into a multiframe in GSM networks?

Question

I'm struggling to understand the need for frame aggregation into the multiframe and other higher level frames.

What exactly do we gain with the aggregation? Is this for synchronisation purposes across all the channels? Is this simplifying slot addressing?

Answer 1

There are two kinds of multiframes in GSM. They are:

• Traffic multiframe: 26 frames taking 120 ms. In a traffic multiframe, 24 frames are used for traffic. These are numbered 0 to 11 and 13 to 24. One of the remaining bursts is then used to accommodate the SACCH, the remaining frame remains free. The SACCH and free frame alternate between position 12 and 25.
• Control multiframe: 51 frames taking 235.4 ms. This multiframe is subdivided into logical channels which are time-scheduled. The control channels are scheduled during the control multiframe, with different frequencies as needed. These include the broadcast channel (BCH), sync burst and so on.

So multiframes help with scheduling and synchronization.

To make the timing of traffic multiframes and control multiframes unfirom, we have the superframe. It is 51 traffic multiframes long, i.e., it is 26 control multiframes long. Notice how a multiframe is 51*26 frames long, so it is an integer multiple of both the control multiframes and traffic multiframes, so it helps to align traffic and control multiframes in one superframe.

Last but not least, the biggest is the hyperframe, which is 2048 superframes long. That is over 3 hours. A few functions are designed to use this periodicity of the hyperframe. For example, the encryption counter is unique only within a hyperframe and then repeats in the next hyperframe. It would be best to never repeat the counter, but as a practical tradeoff, GSM uses the hyperframe for that periodicity.