Inevitably, healthcare goes mobile. Recently developed mobile healthcare (i.e. m-health) services allow healthcare professionals to monitor mobile patient's vital signs and provide feedback to this patient anywhere at any time. Due to the nature of current supporting mobile service platforms, m-health services are delivered with a best-effort, i.e., there are no guarantees on the delivered quality of service (QoS). In this paper, we argue that the use of context information in an m-health service platform improves the delivered QoS. We give a first attempt to merge context information with a QoS-aware mobile service platform in the m-health services domain. We illustrate this with an epilepsy tele-monitoring scenario.
Inevitably healthcare goes mobile – providing m-health applications to users anywhere-anytime, and relying their delivery on the best-effort Quality of Service (QoS) of the underlying wireless networks. We examine a technical and business viability of QoSinformation system (QoSIS), which, based on Mobile Web 2.0 paradigm, predicts the QoS provided by networks available in a given m-health user location-time thus enabling this user an informed network choice.
Our daily affective states are a key element in the analyse is of our quality of life (QoL). Yet, to date a reliable and accurate solution for an affect analysis in our daily environments does not exist. This paper draws a research path towards a development of a Body Area Network (emoBAN) that enables long-term monitoring and analysis of affective states, as well as provision of appropriate feedback via adaptive interfaces. emoBAN can be realized with the use of unobtrusive psycho physiological measurement sensors (e.g., heart rate, respiration), supported by mobile-device based self-reports to investigate a range of affects we exhibit in our natural daily life environments. The acquired data can be automatically mined for patterns and correlated with different contexts, such that QoL could be quantified and areas of personal improvement pointed out. With the use of emoBAN, we can gain a deeper understanding of our personal experiences and resulting behaviors. The system has the potential to be exploited in many application domains and contribute to defining new research questions in disciplines ranging from behavioral to applied sciences.
This paper applies digital PLL (Phase-locked loop) approach to the body movement classification problem. PLLs have been used efficiently in telecommunication to retrieve modulated signals from the background noise. Acceleration sensor signal processing often require the same kind of noise tolerance and the relevant movement patterns are frequently periodic, e.g. walking. The paper presents the PLL-based algorithm developed for step counting on a mobile phone and evaluates it against a commercially available wearable step counter.
Mobile (smart)phones prevail in our daily life activities, and in our research we aim for it to provide pervasive services for wellness. Therefore, we assess the phone's feasibility to unobtrusively, continuously and in real-time track its user's physical activity and the resulting energy expenditure (EE). Activity Level Estimator (ALE) is an Android OS application developed for that purpose.We have assessed the accuracy of ALE against the BodyMedia SenseWear (SW) device and the gold standard for EE estimation, i.e., an indirect calorimetry (IC) method. ALE has mean accuracy of 86% (vs. SW) to 93% (vs. IC) for walking, and in 24h it underestimates EE by 23% ALE is currently used for a long-term behavioral trends study with the University of Geneva students and faculty.
In this paper, based on the Mobile Data Challenge data obtained from the Lausanne data collection campaign, our research aims first to derive network connectivity (e.g., WLAN, 3G) and its Quality of Service (QoS) and mobility patterns of the mobile users, as this connectivity and QoS relate to the user's application activity. Second, we aim to understand how these patterns relate to the overall Quality of Experience (QoE) of the user. Concerning the mobility patterns, we define indoor and outdoor activity for each mobile user. Moreover, we attempt to define semantic places using time filters and GIS techniques, which could also be correlated to the application activity of the users. By correlating the above with the application activity of the users, as well as the hour and weekday patterns, certain inferences can be extracted, concerning the users' spatial and temporal behaviour. These inferences could be used further in developing methods for assurance of the mobile users' QoE.
Access is a new algebraic specification formalism, which focuses on the fine description of the ``true'' concurrency and on a high degree of expressivity. Systems are specified by a set of local states, whose value changes under the occurrence of events. Both events and data structure are specified by abstract data types. Static properties, i.e. global constraints over events and data structure, are described by first order formulae, while dynamic axioms, explaining the behavior of events, are given by causality rules. Concurrency can be described in different ways, it can be interleaving or true concurrency. Finally, expressivity is given by fine descriptions of both static and dynamic properties. Access has demonstrate to be a natural generalization of a great variety of Petri Nets (as for example -Algebraic, -Coloured, -With arc extensions Petri Nets). It is also able to capture concepts from other formalisms as Gamma language or CO-OPN. This report presents a complete description of Access syntax and semantics, together with an example of Access specification based on Petri Nets with Arc Extensions. This report also explains how specifications written in other languages as Petri Nets, Gamma, CO-OPN, are written in Access.
We argue that principles from the design of dependable software, especially separation of concerns and the use of formality, can be applied beneficially in the construction of self-managing systems. We illustrate this approach by presenting an experimental architecture for dynamic and resilient computer-based systems which utilises component metadata to govern reconfigurations in accordance with formally stated policies. Initial experiments with the architecture are described. We argue that the architecture describes a self-organising system and, further, provides a basis for self-managing systems.
The objective of this paper are twofold. On the one hand, it sims to show the advantages of Co-ordinated Atomic actions (CA actions) as a design concept for dependable distributed system development, and on the other hand, it explains how the formal language CO-OPN/2 can be used to express the semantics of CA action design. A fault-tolerant distributed application is developed according to a simple development life cycle: informal requirements, specification, design, implementation. The design phase is built according to the CA action concept. The CO-OPN/2 language is used to formally express the design phase. The implementation is made in Java based on a library of generic classes implementing the CA action concept. The paper is to seve as a basis for a more general approach aimed at defining CA action semantics.
