Resumen
The Internet of Things (IoT) brings Internet connectivity to devices and everyday objects. This huge volume of connected devices has to be managed taking into account the severe energy, memory, processing, and communication constraints of IoT devices and networks. In this context, the OMA LightweightM2M (LWM2M) protocol is designed for remote management of constrained devices, and related service enablement, through a management server usually deployed in a distant cloud data center. Following the Edge Computing paradigm, we propose in this work the introduction of a LWM2M Proxy that is deployed at the network edge, in between IoT devices and management servers. On one hand, the LWM2M Proxy improves various LWM2M management procedures whereas, on the other hand, it enables the support of QoS-aware services provided by IoT devices by allowing the implementation of advanced policies to efficiently use network, computing, and storage (i.e., cache) resources at the edge, thus providing benefits in terms of reduced and more predictable end-to-end latency. We evaluate the proposed solution both by simulation and experimentally, showing that it can strongly improve the LWM2M performance and the QoS of the system.