Increasing key randomness in physical layer key generation based on RSSI in LoRaWAN devices

Abstract

In several Wireless Sensor Networks applications, the sensitive nature of collected information makes the security of data transmission a major concern. Different cryptography techniques have been considered to provide reliable communication between nodes. However, the secure distribution of encryption keys is still a challenging task. In this work, we use physical layer key generation to simultaneously generate symmetric encryption keys at the end device and the gateway. Specifically, we present three contributions to the physical layer key generation process. Firstly, we propose a preprocessing method based on the difference between nonconsecutive Received Signal Strength Indicator (RSSI) samples. This preprocessing stage allows us to obtain random keys even in scenarios with low-mobility nodes. Secondly, we present a new quantization algorithm, named Lossy Cumulative Distribution Function (LCDF), to reduce the key bit mismatch. And finally, we propose a joint Reed–Solomon and Convolutional (RSC) key reconciliation mechanism that further reduces the bit mismatch. The results show that the preprocessing stage together with LCDF and the RSC reconciliation allow us to obtain long and reliable keys in terms of randomness with zero mismatched bits.