An improved demodulation for LoRa through a threshold-based detection of error

This paper deals with an improved demodulation technique for LoRa signal. The principle consists of detecting the possible errors demodulation, by comparing highest peak periodogram dechirped received signal weighted a coefficient β other peaks periodogram. If error is detected, second stage processed at oversampling rate (OSR) higher than first stage. Otherwise, result kept. A thorough performance analysis suggested technique, based on order statistics, provided, in terms detection rate, and complexity. It shown that can be tuned to adjusts trade-off between Moreover, simulation results show it actually reach OSR while keeping complexity lowest OSR.