In RL3, we will receive the deluge of data from RL1 and RL2. This needs to be transported, and tagged so to be easily selected and filtered in an EHR secure distributed storage. Methods for thoroughly controlling, logging and auditing the information transfer will also be addressed in RL3. RL3 will research algorithms to make this information efficiently accessible so that in RL4 we can process the acquired material to generate actionable knowledge. The transport and gathering of data in RL3 will need to provide directive/instructions to the personal devices from RL1. These will include communication optimizations and time synchronization for the devices.

Within this very broad scope we will address the following research objectives and challenges:

  • Existing data sanitization techniques have undesirable utility/security trade-offs. We will re-search on practical mid points between these extremes, and on the interplay with verifiable secure computation which remains, as of yet, unstudied;
  • Traditional SQL databases are at the core of health data management but lack the security to be deployed in a cloud environment. We will extend an existing SQL database with Multi-Party Computation techniques that can be safely deployed on the cloud;
  • Using the user’s smartphone as the personal access authorization broker entity on the net-work;
  • A set of personal attribute authorities, an identity broker and RPs/SPs that request data owner’s permission to acquire temporary/conditional access to a specific user’s identity at-tributes.
  • Formally define transparency, security and privacy and their relations for an access control model;
  • Define new formal methods and automated tools to identify potential vulnerability points in HCI and automatically analyse systems’ socio-technical security;
  • Creation of a federated health related repository of data based on web services on top of Dual Model Architecture Standards (storage of symbolic, image, video, vital signs seamlessly from different devices/processes/sources; supports clinical, personal health data, genetic, nutritional and lifestyle data);
  • Add a level of confidence regarding the quality of the data, develop a quality measure for each data source (reputation);
  • Devise efficient methods and algorithms to deal with relational data;
  • Advance our tabling mechanisms to handle great volumes of data;
  • Advance our Map-Reduce approach for ILP;
  • Advance the parallelism capabilities of our YAP system;
  • Improve the timeliness of the sensed data in both sampling times, acting on clock synchronization and sensing scheduling, and communication times, acting on synchronization of neighbouring nodes and transmission scheduling at the access network level;
  • Devise aggregation mechanisms in the network that maintain timing accuracy;
  • Trim and optimise a telco-grade, standard data collection infrastructure for IoT to be used in limited resource devices for eHealth applications;
  • Specifically target the use of smartphones as gateways for body area networks, quantifying the battery consumption vs delay trade-off;
  • Develop a module for the smartphone gateway that uses the previous results to decide on sensor data transmission scheduling;
  • Input/adaptation/tagging of the gathered data into the EHR of the user.


This research line will be responsible for devising the infrastructure needed to manage health data, divided in 5 work packages. The gathering of all the data (WP5) in a repository holding the information of the individual (EHR, WP3) is crucial for enabling the scalable querying of it (WP4). The transport of this data should follow authorization and access control (WP2), and be maintained in a secure, private and scalable storage facility (WP1):

  • WP1: Secure outsourced data storage and operation
  • WP2: Identity management and authorization infrastructure
  • WP3: Big EHR
  • WP4: Efficient Data Access
  • WP5: Data gathering and selection


More information HERE.