Look-Up Table Based Implementation of Ultra-Low Complexity Narrowband OFDM Transmitters

In cyclic-prefix orthogonal frequency-division multiplexing (CP-OFDM) based radio access, the coexistence of different systems without precise time-frequency synchronization is limited due to high out-of-band emissions. Therefore, filtered-OFDM (F-OFDM) type spectrum enhancement can play a key role to relax the synchronization requirements as it provides well-contained spectrum through subband based filtering. This allows higher degree of flexible and dynamic spectrum use with minimized interference. However, this approach increases computational complexity compared with CP-OFDM. This paper presents a low-complexity solution for narrow-band F-OFDM transmitters based on the use of a look-up table (LUT) to store the F-OFDM waveform. This approach can be applied equally well for filtered version of the discrete Fourier transform-spread-OFDM (DFT-s-OFDM) waveform with small number of subcarriers. DFT-s-OFDM is commonly used in the uplink of OFDM-based systems to mitigate high peak-to-average power ratio (PAPR) of OFDM. The scheme is particularly interesting for the transmitters of Internet-of-Things (IoT) or massive machine type communication (mMTC) devices operating at low data rate. The LUT approach requires only memory units and relatively low number of additions. Moreover, we propose a low-complexity solution to deal with the CP-length variations within the transmission frame. The required memory wordlengths are evaluated through simulations. Comparisons with time-domain filtering and fast-convolution-based filtering solutions are included as well.