Research On Key Technologies For Filter Bank Multi-Carrier Systems

The classical Orthogonal Frequency Division Multiplexing (OFDM) technology has been intensively studied and applied after several decades of development. In the meanwhile, OFDM has the disadvantages of low spectrum efficiency, weak ability to resist the carrier frequency offset and high out-of-band power leakage. In order to meet the needs of the next generation wireless communication technology, some new multicarrier modulation methods was put forward, the Filter Bank Multi-Carrier (FBMC) technology which has many nice features has been considered as a promising candidate and has gained widely attention of lots of scholars.Compared with the traditional OFDM system, the FBMC system using the prototype filter banks of good time-frequency focusing properties on each sub-carrier has better resistance to inter-symbol interference and sub-carrier interference, and the omission of cyclic prefix also brings improvement on power efficiency and reduces the out-of-band attenuation in the meanwhile. Although the FBMC system has many nice features, in pursuit of maximizing the spectrum efficiency, the orthogonality between sub-carriers can only be met in the real domain, which brings about the intrinsic imaginary interference. Because of the inherent interference, the FBMC system encounters some challenges in channel estimation and multi-antenna detection.In terms of FBMC channel estimation, this paper focuses on the channel estimation method of preamble-based pilot structure considering the characteristics of FBMC system. Firstly, the FBMC channel estimation principle and prototype filter interference weight coefficient matrix are studied. According to the analysis, it is noticed that the larger the amplitude of the demodulated equivalent pilot symbol is, the stronger the resistance to Gaussian noise will be. Therefore, we can make reasonable pilot arrangement to reinforce the imaginary interference so as to improve the channel estimation performance. After that, four derived methods that based on interference approximation are studied and the pilot structure of the IAM methods are improved so that it can be applied to the MIMO-FBMC systems. Finally, we propose a method based on the IAM pilot structure to do MIMO-FBMC channel estimation using iterative interference estimation.In terms of multi-antenna detection of the spatial multiplexing MIMO-FBMC system, the classical linear detectors are firstly introduced. Due to the intrinsic interference, which disables the minimum Euclidean distance rule, the optimal maximum likelihood detector cannot be directly applied to MIMO-FBMC systems. A linear detector is employed to do tentative interference estimation then the interference is estimated and removed from the received signals so that the maximum likelihood detector can be applied to the MIMO-FBMC systems. Finally, in order to further improve the performance of the detector and reduce the complexity, a K-Best algorithm based iterative detector is proposed, which utilizes iterative interference reconstruction and cancellation to exploit the undesirable intrinsic interference.