Pilot scheduling based on water-filling algorithm in massive MIMO

Pilot contamination becomes a bottleneck of performance improvement in massive MIMO (multiple-input-multipleoutput) multiple-cell system, due to the signal interference coming from the users that reuse the same pilots in the adjacent cells. Therefore, it is necessary to explore effective approaches to reduce the inter-cell signals interference in order to increase the whole system capacity for massive MIMO. Few former schemes, which are used to mitigate pilot contamination, consider the specific characteristics of the different channels and the interfering degree of different pilots. In this paper, we take these characters into consideration and propose a low-complexity pilot scheduling scheme by applying water-filling algorithm to improve the cell sum-rate. We assign these pilots that are less interference to the users with favourable wireless channel states, and assign severely interfering pilots to the users with inferior channel states. From the perspective of the whole system capacity, pilot contamination can be largely reduced with the proposed scheme, and the system sum-rate can be significantly increased. The simulation results verify the proposed pilot scheduling scheme.