How Interference Diversity in GSM?


Interference is dependent on time, frequency and mobile location. Without frequency hopping, some cell planning margins must be incorporated so that sufficient service quality can still be provided in an interfered situation. By changing frequency on every TDMA frame, a mobile only experiences interference on a particular frequency once in a number of hops. Similarly, interference on a particular frequency will be spread across many mobiles (i.e. averages out with other mobiles).

This is called interference averaging and results in interference diversity. With interference diversity, the perceived radio environment will be more even. As a result of frequency hopping, cell planning margins can be reduced which makes it possible to implement a tighter frequency plan.

Interference diversity is independent of the mobile speed, but is dependent on the mode of hopping, cyclic or random, and the type of frequency hopping used, baseband and synthesizer hopping. The greatest improvement is obtained when the interferer and the interfered connections use uncorrelated hopping sequences.

The lower the correlation, the higher the hopping gain. If both server and interferer use the same set of frequencies and also cyclic hopping, it is possible that some mobiles will hop “in phase” with each other. The effect is a total correlation, as if no hopping is done, and the resulting improvement will be very small. On the other hand, random frequency hopping would still show hopping gain, in a fully loaded system, because of its uncorrelated hopping. The number of hopping frequencies also affects interference diversity gain.

If interference can be spread over a larger bandwidth (i.e. more frequencies), the interference collisions will be fewer, which results in higher hopping gain.