Miyaji Laboratory

In recent years, with the proliferation of IoT, there is a growing demand for communication technologies that achieve low power consumption and stable long-range connectivity in environments where numerous devices transmit small amounts of data at low frequency. LPWA (Low Power Wide Area) technologies attract attention as communication methods that satisfy these requirements. However, LPWA communication is vulnerable to jamming attacks.
As a countermeasure, Emekcan et al. propose a channel hopping technique that dynamically switches communication channels. However, the concrete implementation of the channel update algorithm and its validation in real-world environments are not addressed. In addition, applying channel hopping to resource-constrained IoT devices poses challenges, including the need for lightweight channel update methods and synchronization schemes between transmitters and receivers.
To address these challenges, Ritesh et al. conduct practical evaluations of time synchronization schemes and channel hopping in a LoRaWAN environment. However, the effectiveness of their approach against jamming attacks is not investigated.
This study aims to evaluate the effectiveness of channel hopping against jamming attacks, focusing on a private LoRa environment. Specifically, considering deployment on IoT devices, we propose a basic channel update method based on the approach of Ritesh et al., in which the communication channel is dynamically selected using lightweight computational processing. Furthermore, we improve the basic channel update method and propose an adaptive approach that dynamically adjusts channel selection probabilities according to communication success or failure, thereby actively avoiding jammed channels.
Experimental evaluations using real devices demonstrate that both proposed methods improve communication success rates compared with scenarios without channel hopping.
In particular, the adaptive channel hopping method improves communication success rates by a factor of at least 1.05 under many conditions against jammers targeting fixed channels, compared with the basic channel hopping method.