Uplink Narrowband IoT Data Rate Improvement: Dense Modulation Formats or Non-Orthogonal Signal Waveforms?

Narrowband Internet of Things (NB-IoT) is widely used in low power wide area network (LPWAN) applications due to its long distance coverage and low power consumption. According to the 3GPP NB-IoT standard, the maximum modulation format that NB-IoT can support is QPSK, which limits the data rate sensitive IoT applications. To overcome this limitation either higher order modulation formats such as 8PSK or advanced non-orthogonal signal waveforms can be used. In this work, we propose an enhanced NB-IoT framework, eNB-IoT, which applies a non-orthogonal spectrally efficient frequency division multiplexing (SEFDM) signal waveform beyond the typical orthogonal frequency division multiplexing (OFDM). Since the uplink signal recovery is at base station (BS), in this work we use an efficient and sophisticated minimum Euclidean norm search detector, termed sphere decoding (SD) detector. Simulation results indicate that to achieve the same data rate improvement (i.e. 50% improvement) and BER performance, eNB-IoT, employing the non-orthogonal signal waveform, requires 3 dB less transmission power than typical NB-IoT using the dense modulation approach at BER=10
^-5
. The saved power can either extend the battery life of IoT devices or extend the signal transmission distance. This work also proposes an overlapped SD (OSD) detector to simplify the BS signal processing. Its parallel architecture, employing multiple small size SD kernels, can speed up the signal detection. Results show that by using the OSD detector, eNB-IoT still outperforms NB-IoT but with one order of magnitude complexity reduction.