LoRa Preamble Detection With Optimized Thresholds

Long Range (LoRa) is one of the widely adopted techniques for Internet of Things (IoT). Preamble detection is a key initial task for LoRa systems. Meanwhile, the so-called threshold-based preamble detection is a common technique for compatibility with LoRa. However, the existing methods on the threshold-based LoRa preamble detection suffer from low performance because the detection thresholds are heuristically chosen. To tackle this issue, in this letter, we aim to optimize those thresholds by maximizing the preamble detection probability while satisfying a constraint on false alarm rate. For this purpose, coherent and noncoherent procedures for the preamble detection are presented in a universal manner, followed by conducting their performance analysis. Simulation results demonstrate the superiority and effectiveness of the proposed scheme.