Performance Impact of User Grouping on Fractional Power Loading in the OFDMA Downlink

In the OFDMA-based downlink of wireless communication networks, in order to counteract the strong influence of multi-cell interference on the cell edge users, the so-called fractional power loading strategy was proposed. The key idea is to divide the whole frequency band into two parts, one part has lower transmission power, the other has higher transmission power and is orthogonal in the neighboring cells. With such power allocation, a natural thinking is to schedule the cell edge users onto the high power band, and to schedule the cell center users onto the other. Such strategy leads to the partitioning of users into the cell center/edge user groups. This paper proposes a way to do dynamic user grouping, and compares the cell performance for elastic traffic, with and without the user grouping. Extensive system level simulations with multi-channel proportional fair scheduler are conducted to reveal the effect of user grouping on the scheduler performance and QoS fulfillment.