Athanasios Antoniou

Energy efficient protocols for sensing multiple events in Smart Dust networks

Wireless sensor networks are comprised of a vast number of ultra-small, autonomous computing and communication devices, with restricted energy, that cooperate to accomplish a large sensing task. In this work: a) we propose extended versions of two data propagation protocols for such networks: the sleep-awake probabilistic forwarding protocol (SW-PFR) and the hierarchical threshold sensitive energy efficient network protocol (H-TEEN). These non-trivial extensions improve the performance of the original protocols, by introducing sleep-awake periods in the PFR protocol to save energy, and introducing a hierarchy of clustering in the TEEN protocol to better cope with large networks; b) we implemented the two protocols and performed an extensive simulation comparison of various important measures of their performance with a focus on energy consumption; c) we investigate in detail the relative advantages and disadvantages of each protocol; and d) we discuss a possible hybrid combination of the two protocols towards optimizing certain goals.

A peer-to-peer environment for monitoring multiple wireless sensor networks

In this work we present the basic concepts in the architecture of a peer-to-peer environment for monitoring multiple wireless sensor networks, called ShareSense. ShareSense, which is currently under development, uses JXTA as a peer-to-peer substrate. We demonstrate its basic functionalities using a simple application scenario, which utilizes multiple disparate wireless sensor networks. This application scenario involves monitoring of such networks using a separate management environment and a custom application GUI, as well as using Google Earth as an additional user interface.

Advanced observation and telemetry heart system utilizing wearable ECG device and a Cloud platform

Short lived chest pain episodes of post PCI patients represent the most common clinical scenario treated in the Accidents and Emergency Room. Continuous ECG monitoring could substantially diminish such hospital admissions and related ambulance calls. Delivering community based, easy-to-handle, easy to wear, real time electrocardiography systems is still a quest, despite the existence of electronic electrocardiography systems for several decades. The PATRIOT system serves this challenge via a 12-channel, easy to wear, easy to carry, mobile linked, miniaturized automatic ECG device and a Cloud platform. The system may deliver high quality electrocardiograms of a patient to medical personnel either on the spot or remotely both in a synchronous or asynchronous mode, enhancing autonomy, mobility, quality of life and safety of recently treated coronary artery disease patients.

Resource and service virtualisation in M2M and IoT platforms

Decoupling hardware, software and service provisioning through principles of virtualisation has enabled the exponential uptake of todays internet. The currently emerging internet of things IoT, be it through its more academic embodiment of wireless sensor networks WSNs or more industrial embodiment of machine-to-machine networks, is, however, still fragmented due to dispersed business models, isolated administrative control, heterogeneity of devices and platforms, proprietary communications protocols, different economic interest of stakeholders and a lack of well-crafted solutions for data, information and knowledge exchange. This work presents a novel platform supporting different aspects of WSN virtualisation to cope with requirements dictated by diverse IoT application domains. It introduces beyond state of the art technological advancements following the recent standardisation efforts in this research area. The implementation of the platform and its validation through a representative real-life application scenario exploring different virtualisation facets are also reported.

Using future internet infrastructure and smartphones for mobility trace acquisition and social interactions monitoring

Recent activity in the field of Internet-of-Things experimentation has focused on the federation of discrete testbeds, thus placing less effort in the integration of other related technologies, such as smartphones; also, while it is gradually moving to more application-oriented paths, such as urban settings, it has not dealt in large with applications having social networking features. We argue here that current IoT infrastructure, testbeds and related software technologies should be used in such a context, capturing real-world human mobility and social networking interactions, for use in evaluating and fine-tuning realistic mobility models and designing human-centric applications. We discuss a system for producing traces for a new generation of human-centric applications, utilizing technologies such as Bluetooth and focusing on human interactions. We describe the architecture for this system and the respective implementation details presenting two distinct deployments; one in an office environment and another in an exhibition/conference event with 103 active participants combined, thus covering two popular scenarios for human centric applications. Our system provides online, almost real-time, feedback and statistics and its implementation allows for rapid and robust deployment, utilizing mainstream technologies and components.

Monitoring physical space using mobile phones for inferring social and contextual interactions

In this work, we explore context-aware application scenarios that become possible utilizing semantically-rich information derived from real-world mobility and presence traces. Traces produced by people carrying personal mobile devices, capturing social and contextual interactions, serve as enables for Future Internet applications. We discuss the fundamental concepts, technical issues and related research challenges. We propose a reference architecture for setting up a system that collects such traces in a Smart City environment. We present the algorithms used to process the traces and infer interactions and interests for the observed populations. We conduct two 3-day trial deployments: one in an office environment and the other in the context of a Smart Conference application. We discuss our findings regarding the systems capability to track interactions and the overall efficacy of the application.

On the Deployment of Healthcare Applications over Fog Computing Infrastructure

Fog computing is considered as the most promising enhancement of the traditional cloud computing paradigm in order to handle potential issues introduced by the emerging Interned of Things (IoT) framework at the network edge. The heterogeneous nature, the extensive distribution and the hefty number of deployed IoT nodes will disrupt existing functional models, creating confusion. However, IoT will facilitate the rise of new applications, with automated healthcare monitoring platforms being amongst them. This paper presents the pillars of design for such applications, along with the evaluation of a working prototype that collects ECG traces from a tailor-made device and utilizes the patients smartphone as a Fog gateway for securely sharing them to other authorized entities. This prototype will allow patients to share information to their physicians, monitor their health status independently and notify the authorities rapidly in emergency situations. Historical data will also be available for further analysis, towards identifying patterns that may improve medical diagnoses in the foreseeable future.

Human Mobility Trace Acquisition and Social Interactions Monitoring for Business Intelligence Using Smartphones

Human mobility monitoring and respective traces are important for understanding human behavior, respective patterns and associated context. Such data can be potentially used in business intelligence-oriented systems, for providing added value commercial services or insight to internal enterprise procedures. At the same time, smart phones are rapidly becoming an indispensable tool for our everyday life, while their advanced networking and computing capabilities are increasingly being used as enablers for new applications. We discuss here a system using a stable computing and networking infrastructure along with smart phone applications, based on commodity technologies, meant to be deployed rapidly and provide analytics almost in real-time for such aspects. We also discuss a related scenario in order to provide insight as to where our system could be used. We briefly present the deployment of our system in two settings, an office building and a research exhibition event, along with our experiences. Our findings show that it is feasible and efficient to deploy and operate our system relatively easy, producing meaningful data.

Design and Evaluation of a Person-Centric Heart Monitoring System over Fog Computing Infrastructure

Heart disease and stroke are becoming the leading cause of death worldwide. Electrocardiography monitoring devices (ECG) are the only tool that helps physicians diagnose cardiac abnormalities. Although the design of ECGs has followed closely the electronics miniaturization evolution over the years, existing wearable ECG have limited accuracy and rely on external resources to analyze the signal and evaluate heart activity. In this paper, we work towards empowering the wearable device with processing capabilities to locally analyze the signal and identify abnormal behavior. The ability to differentiate between normal and abnormal heart activity significantly reduces (a) the need to store the signals, (b) the data transmitted to the cloud and (c) the overall power consumption. Based on this concept, the HEART platform is presented that combines wearable embedded devices, mobile edge devices, and cloud services to provide on-the-spot, reliable, accurate and instant monitoring of the heart. The performance of the system is evaluated concerning the accuracy of detecting abnormal events and the power consumption of the wearable device. Results indicate that a very high percentage of success can be achieved in terms of event detection ratio and the device being operative up to a several days without the need for a recharge.