Content/Context-Aware Multiple Camera Selection and Video Adaptation for the Support of m-Health Services

In this paper we focus on the problem of delivering multiple health-related real-time video streams from an emer- gency scenario to a remote hospital by exploiting the uplink of an LTE wireless access network, in order to support efficient m-health tele-consultation services. In this context, the transmission of health-related information is a chal- lenging task, due to the variability and the limitations of the mobile radio link, the different qualities of the visual representations of the cameras and the heterogeneous end-to-end quality requirements of the contents to be delivered. We propose a solution based on: (i) a context-aware camera selection algorithm, which selects among the cameras deployed in the emergency scenario one or more video sources taking into account specific ranking criteria mainly related to the quality of the visual representation of the object of interest; (ii) a content-aware technique for the trans- mission of multiple scalable videos that jointly considers video aggregation and adaptation at the application layer of the transmitting equipment and takes into account the different quality requirements of diagnostic and ambient videos. Numerical results show that the proposed strategy permits to achieve a good end-to-end quality for both the diagnostic and the ambient videos even in the presence of rate limitations and fluctuations in the wireless link, due to the channel variations and the traffic load inside the LTE cell. When the wireless link capacity decreases, the proposed strategy appropriately discards the videos coming from the cameras providing the lowest visual quality, according to the camera ranking results, and, at the same time, adapts the rate of the transmitted videos to provide the requested quality with priority to diagnostic content.