Fast Convolution Based Complexity Reduction in Universal Filtered Multi-Carrier Systems

Abstract

The Universal Filtered Multi-Carrier (UFMC) is a derivative of the Orthogonal Frequency Division Multiplexing (OFDM) scheme. It is considered as a candidate waveform for 5G. The considerably high computational complexity restricts the usage of UFMC based systems. This complexity is caused by the filtration process in the system. Since the flexible and efficient use of non-continuous unused spectrum for different network deployment strategies is important in the next generation communication, the selection of UFMC is inevitable because of its improved performance over OFDM and the Filter Bank Multi-Carrier (FBMC) systems. Hence, it is essential to reduce the complexity of a UFMC system. In this paper, we propose a Fast Convolution (FC) based UFMC system. The technique makes use of smaller FFTs and overlapped data frames. The proposed FC-UFMC system achieves a computation minimisation higher than 95% for systems having 64 or more fully utilised sub-channels compared to similar UFMC systems. The Peak to Average Power Ratio (PAPR) for the FC-UFMC system has a 2.27 dB improvement over the UFMC system. The BER performance of the FC-UFMC system in AWGN and ITU-R channels models are superior or similar to the corresponding UFMC system. The paper also presents the relationship of BER dependence on Signal to Noise Ratio (SNR) and Side Lobe Attenuation (SLA) of the sub-band filter.