Performance analysis of SCHC compression for IPv6 in a real-world LoRaWAN deployment

Abstract

The Internet protocol (IP) is pivotal for the Internet of Things (IoT) paradigm, since it glues together the application-layer and the lower-level protocols of the stack, abstracting away the differences in the several access technologies. In this way, applications can access a uniform network interface, ignoring details about the actual underlying technology. The scalability requirements of IoT obliges to adopt IPv6, which brings a large overhead and payloads hardly compatible with limited resources of 'smart things'. The Static Context Header Compression (SCHC) protocol defines a compression/fragmentation scheme that exploits statically provisioned rules for squeezing the IPv6 (and UDP) header targeting Low Power Wide Area (LPWA) communication solutions. In particular, LoRaWAN specifications have been recently updated to officially include SCHC. The advantage is that the smart thing can be easily linked with any other Internet-based application, enabling a full end-to-end IP link. Despite several analyses about SCHC exist based on simulations, very few (if any) exist about real-world deployments. This work aims at filling in this gap, exploiting the SCHC solution from Acklio and defining a set of metrics that can be easily evaluated. Results for a Class C LoRaWAN node show that end-to-end latency for the uplink and downlink directions are in the order of 0.5 s and 2.5 s, respectively for both a local and remote end user transferring 16 B payload.