2020/03/13 by Hamed Hosseiny, Hosseiny, Hamed, Arman Farhang +3
Computer Science · Engineering · #Advanced Wireless Network Optimization #FOS: Electrical engineering #PAPR reduction in OFDM #Signal Processing (eess.SP) #Wireless Communication Networks Research #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2003.06435
openalex publication_date 2020/03/13 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
In this paper, we consider channel estimation problem in the uplink of filter\nbank multicarrier (FBMC) systems. We propose a pilot structure and a joint\nmultiuser channel estimation method for FBMC. Opposed to the available\nsolutions in the literature, our proposed technique does not rely on the\nflat-channel condition over each subcarrier band or any requirement for placing\nguard symbols between different users' pilots. Our proposed pilot structure\nreduces the training overhead by interleaving the users' pilots in time and\nfrequency. Thus, we can accommodate a larger number of training signals within\nthe same bandwidth and improve the spectral efficiency. Furthermore, this pilot\nstructure inherently leads to a reduced peak-to-average power ratio (PAPR)\ncompared with the solutions that use all the subcarriers for training. We\nanalytically derive the Cramer-Rao lower bound (CRLB) and mean square error\n(MSE) expressions for our proposed method. We show that these expressions are\nthe same. This confirms the optimality of our proposed method, which is\nnumerically evaluated through simulations. Relying on its improved spectral\nefficiency, our proposed method can serve a large number of users and relax\npilot contamination problem in FBMC-based massive MIMO systems. This is\ncorroborated through simulations in terms of sum-rate performance for both\nsingle cell and multicell scenarios.\n