We’d like to draw parallels and compare and show to readers How 4G LTE is Different from GSM. The underlying multiplexing schemes in both 2G (GSM) and LTE appear similar: in both cases, multiple users are handled by ensuring that no two users send messages on the same frequency bucket at the same time (i.e., both technologies use a combination of frequency division and time division).
However, GSM is quite rigid in its resource allocation: a user is assigned one frequency bucket at the start of the session and he gets to send messages within that frequency bucket once in every eight time slots. Further, the allocation is frozen at the beginning of the session and remains fixed throughout the session.
LTE on the other hand is quite flexible: many frequency buckets can be assigned to a user, for any number of time slots. Further, the allocations on both frequencies and time slots can change continuously, in order to accommodate bursts in traffic or changing environmental conditions. This allows for a more efficient usage of spectrum.
The two technologies also differ in the way the frequency buckets are chosen. Due to real world imperfections, signals meant to travel on a particular frequency bucket usually also spread out in adjacent frequencies thus causing interference on adjacent buckets. It has to be ensured that signals on adjacent frequency buckets do not interfere with each other.
GSM uses a brute force approach wherein adjacent frequency buckets are separated by an unused 200kHz zone. LTE, on the other hand, manages a much tighter 15kHz separation between adjacent frequency buckets by elegantly making use of a property called ‘orthogonality’. By choosing the adjacent buckets at these ‘silent points’, interference is avoided and a denser packing of buckets is achieved.