| Summary: | The intercell interference in massive multiple-input-multiple-output uplink consists of both correlated and uncorrelated components, where the former is due to pilot contamination and the latter is usually neglected in most of the existing literature. In this paper, we show that the uncorrelated intercell interference is actually not negligible for reasonably many antennas at the base station. Hence, a novel and general uplink performance analysis framework is established to take the uncorrelated interference into account. Specifically, we investigate the tradeoff between pilot overhead and uplink signal-to-interference-plus-noise ratio (SINR) in four uplink transmission schemes. Without cell coordination, the conventional uplink scheme and the soft pilot reuse scheme are evaluated. Compared with the former one, the soft pilot reuse scheme has higher mitigation efficiency on correlated intercell interference. However, the uncorrelated intercell interference cannot be mitigated without cell coordination. With cell coordination, we first evaluate the scheme that the pilot information can be shared via backhaul. Hence, both the correlated and uncorrelated interference can be canceled effectively. We also propose and analyze a novel pilot extension scheme in the cell coordination scenario, which could address different preferences of both cell-edge and cell-center users. In all the above four schemes, the asymptotic expressions of uplink SINR for arbitrary user are first obtained. Based on them, we derive the distribution of uplink SINR by considering the randomness of interfering users' locations. It is shown that the analytical results match the numerical simulation tightly. Besides, both analytical and numerical results demonstrate the performance gain of the proposed scheme.
|
|---|
