Avaliação do desempenho do Lora com diversidade espacial em canais com desvanecimento

Abstract

Long Range (LoRa) technology stands out in Internet of Things (IoT) applications by enabling long-range communication with low energy consumption; however, its performance can be degraded by noise and, mainly, by fading in multipath environments. This work investigates LoRa link-level performance under different channel conditions, with emphasis on Rayleigh fading and the use of spatial diversity at the receiver. To this end, a Python-based simulator was developed to generate and demodulate LoRa baseband signals using Chirp Spread Spectrum (CSS) modulation, model Additive White Gaussian Noise (AWGN) and Rayleigh channels under a block-fading assumption, and evaluate Single Input Single Output (SISO) and Single Input Multiple Output (SIMO) configurations with Maximal-Ratio Combining (MRC). Performance was quantified through Bit Error Rate (BER) versus Signal-to-Noise Ratio (SNR) curves, varying parameters such as Spread Factor (SF) and the number of receive antennas. In the AWGN channel, the results confirm the expected gain with MRC, consistent with the theoretical diversity trend. In the Rayleigh channel, the observed behavior agrees with the literature, with more pronounced gains due to the mitigation of deep fades, while preserving the trends and relative separations between curves despite small numerical differences.