A Theoretical and Experimental Evaluation on the Performance of LoRa Technology
Menno J. Faber, Klaas Minne Van Der Zwaag, Willian Garcia Viega dos Santos, Hélder R. O. Rocha, Marcelo E. V. Segatto, Jair A. L. Silva
Abstract
A theoretical study and an experimental evaluation on the performance of the long-range (LoRa) technology are presented in this paper. After an instructive mathematical review of important concepts, we assess the LoRa performance in additive white Gaussian noise (AWGN) channels through a numerical model, to elucidate its communication capabilities in negative signal-to-noise ratio (SNR) conditions and, as consequence, in the long distances promised by such low power and narrowband technology. Because of the significance of forward error correction codes in the performance of modern systems, we propose an analytical bit-error-rate (BER) expression that considers the influence of LoRa code rate (CR) parameter. The agreement between the theoretical and the numerical results of a Hamming coded system validated the closed-form BER, evaluated in terms of SNR in AWGN channels. Moreover, an experimental setup with off-the-shelf LoRa wide area network (LoRaWAN) equipment was prepared to demonstrate the feasibility of this promising technology, as well the impact of the CR parameter on the packet error rate (PER), in a link with a single end-device. As expected, communication in urban and open areas with distances up to ≈ 3 and ≈ 10 km, respectively, were achieved, according to measurements of the average received signal indicators around -100 dBm for an average SNR ≈ 5 dB. The usefulness of the practical analysis was confirmed by the agreement between the theoretical and the measured noise floors, and by the verification of the substantial impact of the parameter CR on the system PER.