• ACCELERATING REAL-TIME SPACE DATA PACKET PROCESSING

      Dowling, Jason; Welling, John; Aerosys, Loral; Nanzetta, Kathy; Bennett, Toby; Shi, Jeff; NASA (International Foundation for Telemetering, 1995-11)
      NASA’s use of high bandwidth packetized Consultative Committee for Space Data Systems (CCSDS) telemetry in future missions presents a great challenge to ground data system developers. These missions, including the Earth Observing System (EOS), call for high data rate interfaces and small packet sizes. Because each packet requires a similar amount of protocol processing, high data rates and small packet sizes dramatically increase the real-time workload on ground packet processing systems. NASA’s Goddard Space Flight Center has been developing packet processing subsystems for more than twelve years. Implementations of these subsystems have ranged from mini-computers to single-card VLSI multiprocessor subsystems. The latter subsystem, known as the VLSI Packet Processor, was first deployed in 1991 for use in support of the Solar Anomalous & Magnetospheric Particle Explorer (SAMPEX) mission. An upgraded version of this VMEBus card, first deployed for Space Station flight hardware verification, has demonstrated sustained throughput of up to 50 Megabits per second and 15,000 packets per second. Future space missions including EOS will require significantly higher data and packet rate performance. A new approach to packet processing is under development that will not only increase performance levels by at least a factor of six but also reduce subsystem replication costs by a factor of five. This paper will discuss the development of a next generation packet processing subsystem and the architectural changes necessary to achieve a thirty-fold improvement in the performance/price of real-time packet processing.
    • Shrinking the Cost of Telemetry Frame Synchronization

      Ghuman, Parminder; Bennett, Toby; Solomon, Jeff; NASA, Goddard Space Flight Center; RMS Technologies Inc. (International Foundation for Telemetering, 1995-11)
      To support initiatives for cheaper, faster, better ground telemetry systems, the Data Systems Technology Division (DSTD) at NASA Goddard Space Flight Center is developing a new Very Large Scale Integration (VLSI) Application Specific Integrated Circuit (ASIC) targeted to dramatically lower the cost of telemetry frame synchronization. This single VLSI device, known as the Parallel Integrated Frame Synchronizer (PIFS) chip, integrates most of the functionality contained in high density 9U VME card frame synchronizer subsystems currently in use. In 1987, a first generation 20 Mbps VMEBus frame synchronizer based on 2.0 micron CMOS VLSI technology was developed by Data Systems Technology Division. In 1990, this subsystem architecture was recast using 0.8 micron ECL & GaAs VLSI to achieve 300 Mbps performance. The PIFS chip, based on 0.7 micron CMOS technology, will provide a superset of the current VMEBus subsystem functions at rates up to 500 Mbps at approximately one-tenth current replication costs. Functions performed by this third generation device include true and inverted 64 bit marker correlation with programmable error tolerances, programmable frame length and marker patterns, programmable search-check-lock-flywheel acquisition strategy, slip detection, and CRC error detection. Acquired frames can optionally be annotated with quality trailer and time stamp. A comprehensive set of cumulative accounting registers are provided on-chip for data quality monitoring. Prototypes of the PIFS chip are expected in October 1995. This paper will describe the architecture and implementation of this new low-cost high functionality device.