No abstract

No abstract

No abstract

Time-division duplex (TDD) is the most efficient technique for acquiring channel state information (CSI) in massive MIMO systems where the reciprocity between the uplink and downlink channels is utilised by the pilot signals to extract the channel parameters. In this paper, we consider the pilot contamination problem in TDD multicell multiuser massive MIMO systems and examine two different pilot signal allocation schemes for which we derive the lower bounds on the achievable rate on the uplink for the cases of maximum-ratio combining (MRC) and zero-forcing (ZF) detectors. To achieve further performance enhancements, we propose a new algorithm for pilot sequences allocation in which the multiplicity of the pilot sequences over the number of users in each cell is exploited. Our results show that when pilot contamination is severe, allocating more system resources for channel estimation results in a better system performance especially in limited mobility environments. Moreover, we show that when the signal to interference plus noise ratio (SINR) is low, MRC is superior to ZF, and vice versa. Finally, we demonstrate that our proposed allocation algorithm can significantly improve the spectral efficiency of the network compared to the conventional pilot allocation method.

Abstract-The pilot contamination problem is one of the major obstacles that limit the performance of time-division duplex (TDD) multicell massive multiple-input multiple-output (MIMO) systems. Pilot contamination results from the re-use of the same set of pilot sequences in the different cells of the system. In this paper, we propose an Adaptive Pilot Allocation (APA) algorithm for pilot contamination mitigation. The proposed algorithm divides the users in the system into two groups according to the inter-cell interference they cause to the other users in the system. To improve the system performance, the algorithm allocates orthogonal pilot sequences to the users that can cause high inter-cell interference. Simulation results show that the proposed algorithm outperforms the conventional (random) allocation method. Also it show that the proposed algorithm improves both the minimum and the mean achievable rates.