The most ofICI avoidance algorithms divide UEs into

cell edge and cell centre users based on static thresholds

such as path loss, SINR and geometry factors. However, in

an actual communication environment, the UE distribution

as well as the shape of a cell can be changed. Therefore

selection of the cell edge bands is difficult and needs to be

optimized according to the time variation environment.

Reference [6] proposed an ICI avoidance algorithm that can

change cell edge bands dynamically. For each active UE and

every time gap (the update duration of CEU/CCU), the

threshold is updated based on SINR in order to keep the

ratio of number of CEUs and CCUs constant. Then, cell

weight is calculated to follow the instant changing of

network.