Real-time Remote Sensing Imagery Streaming over Peer-to-Peer Network

Abstract

Motivated by the widespread use of real-time video streaming techniques over peer-to-peer networks, we propose a design of a peer-to-peer solution for real-time remote sensing imagery browsing. In this paper we firstly identify the characteristics of image browsing and mapping them to P2P-based streaming. Our design makes use of Tracker in locating remote sensing data resources. Further, a dynamic image tiles downloading queue technique which optimize image transmission is adopted. Considering that the conventional data piece selection algorithm of P2P systems has many limitations in supporting image streaming, our new approach gives each image tile a priority level before being transmitted. We use a parameter K to strike the balance between high priority tiles and low priority tiles. Experimental results show that the prototype system performances much better than the traditional C/S mode when an appropriate parameter K is chosen.