Distributed collaborative visualization on mobile devices using interactive video streaming techniques

Abstract

Remote visualization using mobile devices has been a challenge for distributed systems for a long time. Large datasets, usually distributed on different servers require high network bandwidth and significant computational power for effective, real time rendering. The problem is getting more complex when data are visualized in collaborative environment, where every user can interactively participate in rendering session. In this paper we present a distributed system we have developed for the interactive visualization of remote datasets on variety of mobile devices such as laptops, tablets and cell phones. In our system mobile users can join sessions, where they can collaborate over remote data in real time. Every user can watch presentation or can become presenter. If needed, users can individually manipulate the data without affecting rest of participants. During these sessions all the data are generated on dedicated rendering servers, compressed on-the-fly by the encoding machines using video codec and progressively sent to participants as video streams. Every video stream is dynamically adapted to individual capabilities of users' devices and their network bandwidth. Our system works in a distributed environment, where every machine serve different functionality, like data storage, frames rendering or video compression. Successive parts of processed data are streamed between different servers in real time to achieve highly interactive visualization with minor latency. Based on this model we took off most of the computational power from client's application so it can be run on almost any kind of modern mobile device. We were also able to achieve very high video quality and frame rates. System can work with 2D, 3D and even animated 3D data, all of them being processed remotely in real time. At the end of this paper we present some preliminary results of performance test we have obtained using sample multidimensional datasets. © 2012 Springer-Verlag.