Implementing LWM2M in constrained IoT devices

LWM2M is an emerging Open Mobile Alliance standard that defines a fast deployable client-server specification to provide various machine to machine services. It provides both efficient device management as well as security workflow for Internet of Things applications, making it especially suitable for use in constrained networks. However, most of the ongoing research activities on this topic focus on the server domain of LWM2M. Enabling relevant LWM2M functionalities on the client side is not only critical and important but challenging as well since these end-nodes are invariably resource constrained. In this paper, we address those issues by proposing the client-side architecture for LWM2M and its complete implementation framework carried out over Contiki-based IoT nodes. We also present a lightweight IoT protocol stack that incorporates the proposed LWM2M client engine architecture and its interfaces. Our implementation is based on the recently released OMA LWM2M v1.0 specification, and supports OMA, IPSO as well as third party objects. We employ a real world application scenario to validate its usability and effectiveness. The results obtained indicate that the memory footprint overheads incurred due to the introduction of LWM2M into the client side IoT protocol stack are around 6-9%, thus making this implementation framework very appealing to even Class 1 constrained device types.