An efficient object discovery and selection protocol in 3D streaming-based systems over thin mobile devices

3D streaming over thin mobile devices is considered challenging due to the mobile devices´ limited capabilities such as limited energy lifetime, processing power, storage capacity, and graphics´ hardware and accelerator capabilities, that make it very difficult for mobile devices to render and process large and complex 3D scenes. To address this issue, 3D streaming techniques have been proposed that aim at reducing the resolution of 3D objects to make it easy to deliver and render. However, streaming all of the 3D objects in a given virtual environment (VE) is not considered efficient. In this paper, we propose a novel object selection technique for 3D streaming based systems over thin mobile devices, which we refer to as OCTET. Our protocol, which is based on multi-level area of interest (AOI), allows for the selection of the 3D objects required for the user´s virtual scene, taking into account the user´s interests, the location of the 3D objects in the multi-level AOI, and the mobile device´s available resources.