Cooperation mechanisms for pervasive mobile health applications

Abstract

The emergence of mobile health and Web-based services along with new ubiquitous mechanisms are providing new methods to deliver services that overcomes geographical and temporal barriers, delivering information regardless place and time. With the proliferation of mobile devices, online markets have been growing with many health applications, allowing users to have access to health records, treatment plans, alerts and health goals establishment (e.g., weight loss). This recent ubiquitous paradigm is much possible due to Web Services capabilities and new data-interchange notations, as well as the evolution of mobile software development kits. However, m-health architectures that depend on wireless networks have several constraints such as mobile devices battery, processor and memory resources, as well as issues regarding network connectivity and communication delays. Cooperation mechanisms have proven to be a promising solution to approach these constraints. In a typical cooperation strategy, information transmitted through wireless channels is usually relayed through a relay node and/or a relay station (i.e., a base station), using a packet forwarding cooperation model. While some cooperation approaches aim multiple constraints, such as bandwidth performance, wireless frequency management or localization improvement, other approaches aim one specific network or limitation inherent to mobile devices, such as battery power or processing power issues. In this work a novel cooperation strategy for m-health services following service oriented architectures is proposed in order to approach two common drawbacks in mobile health systems: the Internet connectivity and infrastructure dependencies. A reputation-based model is used, where a Web Service is responsible for nodes reputation management, as well as for the access control. At the client-side (i.e., in the mobile device) four software modules are used in order to manage and control the ubiquitous cooperation process. The ultimate goal is to provide an alternative for remote access, where mobile devices without Internet connectivity could retrieve remotely stored health data through cooperation. Packet forwarding should occur through short and low energy consuming communications, specifically through Bluetooth interface. This results in a free of charge alternative to cellular data network connections and independent of WiFi access points. Although the referred mechanisms aim any mobile health application, this work was carried with SAPO – Portugal Telecom and for test purposes a specific mobile health application, namely SapoFit, was used. Cooperation mechanisms were created and integrated in SapoFit, and a cooperative Web Service was built. A performance evaluation in a real scenario with different mobile devices is performed and presented in this work. The request and response message delays are measured, while varying the number of uncooperative nodes, and verifying the required time for each individual communication process. Furthermore, the memory footprint of the mobile cooperation mechanisms is revealed.