LoRa物理层同步及解调性能研究

针对现有的基于FFT的同步算法误差较大的问题，提出了一种优化的LoRa物理层前导码同步方法。首先，根据Chirp信号的时频转换特性，分别采用差分算法和插值算法对前导信号的归一化的小数频偏和小数时延进行估计，在此基础上采用基于FFT同步的联合估计算法对整数频偏和整数时延进行估计；之后，从理论上分析残余的同步误差对LoRa信号解调性能的影响。仿真结果表明，所提出的同步方案可以在相应信噪比极限下满足解调性能要求。

Research of synchronization and demodulation performance for LoRa physical layer

The existing synchronization algorithms based on FFT have large synchronization error. To solve this problem, this paper proposed an optimized synchronization method for LoRa preamble signal. Firstly, according to the time-frequency conversion characteristics of Chirp signal, it estimated the fractional frequency offset and fractional time delay of the preamble signal respectively by using differential algorithm and interpolation algorithm. Based on this, it derived the integer frequency offset and integer time delay by using a joint estimation algorithm based on FFT. Then, it theoretically analyzed the effect of the residual synchronization error on the demodulation performance of LoRa signal. The simulation results show that the proposed synchronization method can meet the demodulation performance requirement at the corresponding required signal-to-noise ratio(SNR).