Design and Implementation of High Bit Rate Satellite Image Data Ingest and Processing System

Abstract

Remote sensing satellites orbit around the Earth around 600-900 kilometers in sun synchronous polar orbit. Each satellite transmits the image data in X-band and the data reception is carried out through a parabolic antenna along with the associated systems. The entire data acquisition system is complex and expensive as it has to meet the requirements of the fast orbiting remote sensing satellites. During the visibility of the satellite, the data is acquired at the acquisition systems, frame synchronized band separated and displayed in the monitor in real-time. The authors of this paper used the concept of archiving the RAW data directly to redundant array of inexpensive disks (RAID) in real-time and also to provide on the fly sub-sampled video image display using generic PC Based Linux Servers available in the industry. Handling of high speed image data from satellite involves de-randomization, decryption and decompression in real time. Innovative multiplexing techniques are used for digital correlation to increase the operating speed. Effective system design strategies rely on appropriate matching of systems hardware, software and environment. This needs the proper evaluation of the system environment. Schedulers for a realtime system are concerned with maximizing resource utilization subject to meeting system timing constraints. Priority-driven presumptive schedulers used in realtime systems require predictable cache performance. The design involved shared memory concepts, cache memory handling, image data handling and transfer to RAID while each system handles the data of around 500 GB per day. The paper deals with a generalized design of such system using a common platform as a part of Indian Remote Sensing Satellite Program and supplied to many users around the world. These systems are working error free all over