The Q4HEALTH project experiments are motivated to study video performance in a scenario where immersive wearable live video assists first responders in public safety applications such as an ambulance paramedic attending to patient. During the ‘golden hour’ after the incident, pre-hospital actions taken are crucial and potentially lifesaving. “Eyes on’ live video from the first responder to the hospital or clinical oversight centre can help make more rapid decision. Medical policy makers are considering how to harness advances in ICT to enhance patient care and outcome. Responding to eHealth market demands, RedZinc, an SME, has developed an interactive wearable video platform for ambulances and paramedics. The platform enables emergency doctors located in the hospital to remotely see a pre-hospital patient with acute medical conditions and to provide additional treatment capability for that patient. During the ‘golden hour’ rapid diagnosis and earlier treatment has the potential to enhance patient outcome (e.g. more rapid delivery of clot busting drugs).
The PerformNetworks testbed is being upgrade during the project in order to support advanced 5G prototypes supporting QoS slicing, low latency communications and multicast communications. The latency reduction prototypes are fully compatible with standard LTE networks, aim to provide a unique playground for testing commercial platforms in future 5G environments, and also where to evaluate fog computing services. The slicing is provided by the use of the Rx interface by external applications. Multicast communications are being explored using SDN technologies.
The main difference with the classical 4G architecture is the way we enable in the OAI system virtualization of the of the physical resource blocks (VRB) and allow for allocation of VRBs per different tenant (in our case the normal users and the VELOX users). This is achieved by adding an abstraction layer, which acts as an interface between the shared Physical Resource Block (PRB) and the tenant -specific Resource mapper (SRM). The resource mapping exposes information to each SRM regarding buffer status in both Uplink and Downlink for each logical channels and users. In the current status we implemented a downlink UE scheduler for the agent-side that supports the dynamic introduction of new tenant to the RAN and the on-demand modification of the scheduling policy per operator. This is a major innovation towards building programmable ran elements and to the best of our knowledge is the first open-source offering.
All the activities carried out in the project can help to assess how these experimental platforms can enhance readiness for market applications.
