International audience

In resource-constrained Internet of Things (IoT) environments like Low-power and Lossy Networks (LLNs), efficient communication protocols are essential. In this context, IP multicast protocols play a crucial role, facilitating the transmission of data packets from a single source to multiple recipients, thereby conserving bandwidth, power, and time for numerous LLN applications, such as over-the-air programming, information dissemination, and device configuration. Despite their usefulness, existing multicast solutions face several challenges, including scalability, energy efficiency, and reliability. To tackle such issues, this paper introduces Protocol Independent Multicast for LLNs (PIM-LLN). PIM-LLN employs a multicast distribution tree anchored at the border router, a multi-path data dissemination mechanism, and an efficient retransmission technique to route streams exclusively to regions with group members reducing energy consumption and bandwidth usage while improving response times and reliability. Through comprehensive simulations and public testbed experiments, we meticulously assess PIM-LLN’s performance, benchmarking it against state-of-the-art solutions under different scenarios. Our findings underscore the scalability, reliability, reduced latency, and efficient resource utilization of PIM-LLN in terms of memory, bandwidth, and energy. Notably, PIM-LLN, as compared to state-of-the-art solutions, achieves a similar level of reliability while reducing overhead by up to 50%.

Arbitrary point-to-point (P2P) routing is becoming a necessity in a multitude of Internet-of-Things (IoT) applications, including building automation, smart cities, and smart manufacturing. For this reason, new P2P routing protocols such as lightweight on-demand ad hoc distance-vector routing protocol next generation (LOADng) and ad hoc on-demand distance vector routing-based RPL protocol (AODV-RPL) are being standardized. Both protocols are inspired by AODV, and hence, they might suffer from broadcast storms caused by flooding route requests (RREQs) and related issues. Thus, while AODV-RPL takes advantage of RPL mechanisms, LOADng uses blind flooding to forward RREQs. However, both lack effective techniques for avoiding unnecessary RREQ transmissions when routes are found. In this article, we first deploy stopping Trickle timers to ensure scalable, efficient, and reliable dissemination of RREQs. Second, we devise and present new techniques to suppress unnecessary RREQs with optimizations for radio duty-cycled networks. Finally, the two mechanisms are combined in a third proposal for better efficiency. These mechanisms provide individually and collectively great enhancements to P2P routing protocols, such as AODV-RPL and LOADng, while staying backward compatible with their base specifications. The performance of the proposed mechanisms when applied to LOADng has been validated using both extensive time-accurate simulations and large-scale public testbeds. The obtained results have shown the effectiveness of the proposed mechanisms with around 50% less overhead and savings in energy consumption, along with a 20% gain in route discovery ratio at the expense of an increase in discovery delays.

This paper presents BoostSole; a smart insole based system for automatic human gait recognition.It consists of a smart instrumented insole connected to the cloud via the patient's smartphone using low-power wireless communication.First, the design of BoostSole is introduced with discussions of sensors choice, placement, calibration, and data communication.Next, an adaptive multi-boost classification algorithm is deployed to accurately identify different gait patterns.The algorithm is fast and lightweight and can be implemented in ordinary smartphones with a small footprint in terms of computational requirements, energy consumption, and communication usage.Raw and ondevice classified data can be securely uploaded to a distant cloud server for continuous monitoring and analysis.Indeed, they can be visualized and exploited by doctors to identify/correct walking habits and assess the risks of chronic pain associated with an abnormal walk.The system has been evaluated on a dataset containing three gait patterns, namely: shuffle walk; toe walking; and normal gait.Obtained results are promising with more than 97% classification accuracy accompanied by low response time and computational demands.

Human Activity Recognition (HAR) is an important area of research in ambient intelligence for various contexts such as ambient-assisted living.The existing HAR approaches are mostly based either on vision, mobile or wearable sensors.In this paper, we propose a hybrid approach for HAR by combining three types of sensing technologies, namely: smartphone accelerometer, RGB cameras and ambient sensors.Acceleration and video streams are analyzed using multiclass S upport Vector Machine (S VM) and Convolutional Neural Networks, respectively.S uch an analysis is improved with the ambient sensing data to assign semantics to human activities using description logic rules.For integration, we design and implement a Framework to address human activity recognition pipeline from the data collection phase until activity recognition and visualization.The various use cases and performance evaluations of the proposed approach show clearly its utility and efficiency in several everyday scenarios.

Finding the right resource at the right time and space is a key enabler for a wide adoption and spread of the Internet of Things (IoT).The Constrained Application Protocol (CoAP) and related standards are among the most prominent efforts working towards such a goal.Indeed, CoAP-related standards provide interesting mechanisms for resource discovery in both centralized and distributed architectures based on the CoAP's GET method.In this paper, we, first, highlight the limitations of GET-based discovery mechanisms.The paper, then proposes a new solution using the recently standardized FETCH method and develops its specifications, rules and semantics.The proposed solution is implemented in the recently released, secure and reliable OpenThread platform and compared with GET-based approaches in different home automation scenarios.Obtained results demonstrate the performance of FETCH-based discovery in achieving fine-grained, time-efficient and reliable discovery while preserving network resources.

We propose to augment the Trickle algorithm with contextual information freely and locally available in low-power wireless networking technologies. The aim is to equip Trickle with hints and heuristics allowing it to propagate updates faster using link indicators such as received signal strength and link quality indicators along with available neighbourhood and network state information. The proposed augmentations are carefully designed so to preserve Trickle strengths in terms of simplicity, reliability, scalability, and load balancing while minimising its latency. Extensive simulation evaluations, conducted under TinyOS, show that the resulting algorithm, dubbed Trickle++, propagates updates more than twice faster than Trickle. Obtained results also show that Trickle++ preserves Trickle performance regarding overhead, load balancing, and code footprint.

In Low-power and Lossy Networks (LLNs), multicast communications accommodate many interesting applications ranging from firmware installation and upgrades to resource discovery and utilization. However, performing multicast in LLNs has shown to be less efficient and lacks reliability when compared with unicast. This is especially the case under radio duty cycling mechanisms employed by LLNs in order to save energy and hence extend the network lifetime. In this paper, we present a technique to enhance the performance of multicast burst forwarding in duty-cycled LLNs. The proposed mechanism has been implemented in Contiki OS and evaluated in both testbed experiments and cycle-accurate simulations. Results show a considerable increase of multicast burst throughput (more than 8 times) accompanied with an important decrease in energy consumption (about three times) and transmission latency (at least 35%). Finally, the proposed mechanism was integrated with a multicast-based service discovery protocol and demonstrated noticeable improvements.

The emerging trend towards all-IP networks provides network-layer interoperability for the Internet of Things (IoT). Service oriented architecture is expected to provide the application-layer interoperability and hence achieve IoT objectives. The first component into realizing this vision is efficient discovery of things and their services. In this paper, we present a fully distributed opportunistic approach to optimize the discovery of constrained-node services. We also discuss the impact of radio duty cycling on the performance of such a multicast-based approach. Simulation and testbed results have shown the capacity of such an approach to cope with network dynamics and provide high discovery rates, fast response times and low energy consumption.