Non-Orthogonal Waveform Scheduling for Next Generation Narrowband IoT

Narrowband Internet of Things (NB-IoT), firstly introduced in 3GPP release 13, brings benefits such as low power and wide coverage. Repetitive transmission is the mechanism in uplink channels in NB-IoT for coverage extension. This is done at the expense of extra power consumption and time latency. In this work, we propose a power efficient framework termed enhanced NB-IoT (eNB-IoT), which adopts a non-orthogonal waveform that can bring additional power budget without using the repetitive transmission mechanism. The waveform halves the required bandwidth, for a given bit rate, whilst maintaining the same BER performance. We propose to divide the frequency band into a set of sub-bands and then apply adaptive waveform scheduling to the modulated data on the optimal frequency band with lower loss and therefore higher signal-to-noise ratio (SNR). Advantageously, the narrower bandwidth of fast-orthogonal frequency division multiplexing (Fast-OFDM) waveform can flexibly map the signal on the frequency bands associated with high SNR and avoid transmission in the higher loss frequency band. Hence, compared to the single-carrier-frequency division multiple access (SC-FDMA) signal, which is the default uplink signal format of NB-IoT, and which occupies all the signal bandwidth in bands of high and low SNR, simulation results show that ~84% power can be saved via using the Fast-OFDM signal in a defined non-line-of-sight channel. Experimental results show a ~11 dB error vector magnitude (EVM) improvement compared to the SC-FDMA signal. The saved power can extend signal coverage where repetitive transmission is limited.