Spacebridge to Moscow Marks Joint U.S./Russian Venture in Telemedicine
Digital Audio Radio Broadcast Systems Tested
New-Technology Satellites Would Provide T1 and Higher Rate Service
Advanced Traveling-Wave-Tube Circuit Simulated and Designed
Cassini Mission Ka-Band TWTA Developed
Electrical Properties of Heterostructures for Advanced High-Speed Electronics Determined
Graphical Display Designed for Communications Satellite Test Bed
Real-Time Compression of Digital Video Achieved
Shared-Memory-per-Beam Architecture and Simulation Completed
Using a two-satellite system in a "double hop" configuration with NASA Lewis acting as the gateway facility, the Spacebridge began operation in September 1993 and completed the last session in June 1994. The Spacebridge used the Russian western Space Data Relay Network (SDRN) satellite positioned in geostationary orbit at 16deg. west longitude over the Atlantic Ocean for the Moscow-to-Cleveland portion of the link and the domestic commercial communications satellite GTE GStar II for the U.S. portion of the link. The gateway facility in Cleveland comprised two Earth terminals--a domestic Ku-band, 5-m station capable of transmitting and receiving broadcast-quality video, and a special SDRN Earth terminal supplied by the Russians for this project. The four U.S. medical centers took turns hosting different sessions as the U.S. uplink site. The remaining U.S. medical centers participated through terrestrial telephone lines in a teleconference setup while viewing the video transmissions originating in both the United States and Russia.
Two notable highlights of the 16-session demonstration project were the inaugural session originating from the U.S. Senate Hart Building in Washington, D.C., and the emergency trauma sessions held in October 1993. The inaugural session occurred on November 5, 1993, and included a demonstration of teleradiology and telepathology using the Loral-Siemens medical diagnostic imaging support system and an excellent demonstration of the benefits of rural medical consultation from a doctor present at the inaugural site to a small town in West Virginia. A special T1 line between Elkins, West Virginia, and the Senate Hart Building was set up for this purpose.
Another major achievement of the Spacebridge to Moscow was in disaster assistance. The early autumn of 1993 was a tumultuous and chaotic time in the former Soviet Union. During this unrest the storming of the Russian Parliament regrettably resulted in numerous casualties and trauma cases. The Spacebridge link was used on two separate, unscheduled occasions to hold special sessions where many of these trauma cases were addressed.
Bibliography
NASA Lewis is hosting the laboratory testing of seven proposed systems. There is no comparable program in the world to so thoroughly test all known systems. The tests are being conducted by the Electronic Industries Association's Consumer Electronics Group (CEG). Proponents delivered their test systems to Lewis in January 1994. Laboratory tests are scheduled to be completed in late December 1994. Then the systems will be field tested in the San Francisco, California, area.
The laboratory tests have been designed to comprehensively evaluate the technical performance of each system under a variety of impairment conditions. The impairments include noise, simulated mobile multipath, simulated airplane flutter, and cochannel, first-adjacent-channel, and second-adjacent-channel interference. For in-band systems the tests also measure analog-to-digital and digital-to-analog interference.
To test system performance, critical audio segments have been recorded that stress the audio coding used in the systems. These critical segments are played through each system under each impairment condition. Each system's audio output is recorded for various carrier-to-noise ratios at and around the point of failure and the threshold of audibility observed in the laboratory. Each audio sample is subjectively evaluated at the Communications Research Center (CRC) in Ottawa, Ontario, Canada, by an expert listener panel that rates the audio quality in a double-blind setting.
Although ideally suited for precise determination of noise and interference thresholds, laboratory tests cannot adequately measure performance in actual broadcast transmission through a nonlinear, dispersive radio channel. Thus, the CEG test and evaluation program will include extensive field tests planned for February 1995.
With the successful completion of these two tests the United States will be able to establish realistic standards for domestic digital audio radio broadcasting and be a key player in establishing it internationally.
Lewis contact: James E. Hollansworth, (216) 433-3458
Headquarters program office: OSAT
New-technology satellites appear to be very competitive for shared or switched services. In optimum (minimum service cost) systems very-small-aperture terminals (VSAT's) are used to access a major hub with very bursty data. The major differences are that the satellite plays the role of the hub (performing all message detection and routing) and that maximum user data rates are three to seven times greater than with conventional VSAT's. Because these systems favor bursty traffic, voice and real-time video are excluded as appropriate markets. However, these systems are suitable for multimedia communications, where video clips and sound bites are intermittent. Consequently, the prime application is viewed as multimedia communications within "VSAT-like" networks and is expected to be prevalent with the common use of Microsoft Windows and OS/2 operating systems. Because conventional VSAT technology cannot efficiently process such communications, the new-technology satellites are suggested as an alternative.
Within the accuracy of the study cost modeling, optimization techniques show that for private-line T1 and higher rate services large Earth terminals (4 to 5 m) are required to minimize user cost. The corresponding satellite technology includes multibeam antennas, onboard processing, and advanced bandwidth-efficient modulation techniques. Optimum satellite capacity exceeds 10 gigabits per second (Gbps) per spacecraft. Thus, the multibeam antenna serves a dual role of high spectrum reuse as well as closing a link on the wideband services.
On the contrary, for shared services the optimization process indicates that the Earth terminals should be small (0.5 to 2 m) for T1 service. The corresponding satellite technology includes very large (5 to 7 m at Ka band) multibeam antennas for closing the link with the very small terminals. Spectrum reuse is an obvious additional advantage, but these satellites will probably be power limited so that any achieved spectrum reuse would be incidental. As an alternative to the large spacecraft antennas, it is suggested that the optimization process be altered so as to constrain the minimum Earth terminal size. The resultant user costs triple, but the corresponding spacecraft would be far less risky.
Lewis contact: Grady H. Stevens, (216) 433-3463
Headquarters program office: OSAT
The models are being used to design a new type of ring-plane TWT circuit. The copper circuit consists of a series of beam-circling rings supported by slotted planes. Two vanes stretch the length of the circuit to improve the frequency bandwidth. Simulations show that the new circuit has considerably higher RF power and greater power efficiency than a conventional coupled-cavity TWT circuit without a significant sacrifice in frequency bandwidth.
Bibliography
Designated the Hughes 955H, the TWT has demonstrated an overall saturated efficiency of over 40%. This goal has been reached by including a unique NASA Lewis-supplied dynamic velocity taper (DVT) helix and advanced multistage depressed collector (MDC) designs, along with MDC electrode surface treatment to suppress secondary electron emission, and innovations in mechanical and thermal design and fabrication introduced by Hughes. Specifically, the DVT in the TWT's output section is characterized by a continuous nonlinear reduction in helix pitch from a synchronous value near the output end of the circuit. This results in better synchronization between the circuit wave and the electron bunches in the electron beam than can be realized with a constant-pitch helix. The MDC design procedure defines optimum electrode surface configurations by predicting electron trajectories, taking into account secondarily emitted electrons. Further, the oxygen-free, high-conductivity copper MDC electrode surfaces were treated at NASA Lewis with an ion-bombardment process to produce a robust, highly textured surface with secondary electron emission characteristics sharply lower than those of untreated copper.
Two flight model TWT's have been fabricated that meet or exceed the Cassini mission requirements, along with engineering models of the TWT and electronic power conditioner. Space qualification testing, including thermal-vacuum, vibration, and pyroshock, is in progress. The mass of the flight-packaged TWT is 750 g and the mass of the EPC is 2.67 kg, for a total TWTA weight of just over 3.4 kg. NASA Lewis has managed the contract with Hughes for model development. JPL has specified system requirements.
This TWTA development represents a significant advance in achieving high efficiency for low-power amplifiers at this frequency level. It also demonstrates the value of vacuum electron devices in low-power microwave applications--frequently conceded to solid-state devices in system planning. The production of flight hardware as part of a research effort has reduced overall cost and shortened delivery schedules.
Lewis contact: Arthur N. Curren, (216) 433-3519
(or fax (216) 433-8705)
Headquarters program office: OSAT
In the past decade new types of devices have been developed for high-speed electronics, for both digital and analog applications. These devices are based mainly on III-V compound semiconductor heterostructures, but new materials such as silicon-germanium/silicon are now being introduced. The most common device, a modulation doped field-effect transistor (MODFET), includes an active channel layer and several auxiliary layers (i.e., for contacts, barrier, buffer, etc). Carrier mobility is higher in the active channel layer than in any of the other layers. Although predicting proper device operation requires knowing the concentration and mobility of the carriers in the active channel only, the normal Hall effect technique can give only an average of these parameters in all the layers--which can be very different from the active channel values. Therefore, the Hall effect technique cannot be used as a good, quantitative predictor of material quality.
In the work done at NASA Lewis the room-temperature Hall effect technique and analysis were completely revised, without requiring hardware modifications, to obtain the desired active channel parameters. In the new technique the Hall effect measurements are repeated at many values of the magnetic field. The experimental results are the longitudinal and transverse resistances as functions of the magnetic field. As we are interested only in the active channel parameters, we analyze the data in terms of only two layers: the active channel and an average of all other layers, for a total of four parameters. A least-squares fit of all experimental points is used to obtain the value of these four parameters. Special care is needed in this least-squares fit, as the absolute value of the experimental results includes both very large and very small numbers. The results are checked for accuracy by using a much more complex technique that can directly measure the carrier concentration (but not the mobility) in the active channel, called the Shubnikov-de Haas technique. Our results for the carrier concentrations estimated by the two methods agree to better than 15%. Representative results obtained by the new technique for two samples are given in table I. In the table n and m represent carrier concentration and mobility in units of 10[12 ] per square centimeter and square centimeters per volt-second, respectively; the index 1 represents the active channel; the index 2 is associated with the average combined parallel layers; and the index H represents the regular Hall effect results. It is obvious that the Hall effect data are not useful for the active channel description.
TABLE I.-CARRIER CONCENTRATIONS AND MOBILITIES IN MODFET'S Sample n1 n2 nH u1 u2 uH
A 1.09 3.22 2.80 7570 1850 5080
B .557 4.13 2.92 8080 1810 4060
Lewis contact: Dr. Samuel A. Alterovitz, (216) 433-3517
Headquarters program office: OSAT
The CDD utilizes modern client-server technology and runs in the popular DOS/Windows environment. Windows Dynamic Data Exchange is used to transfer data between the client and server modules. This graphical display was developed by using an off-the-shelf graphical user interface development package and object-oriented programming techniques.
The CDD consists of a main display window containing buttons representing the test bed components. The test bed is designed to accommodate three Earth terminals, each of which includes three data sources and three data destinations. Clicking on a button in the main display window calls up a secondary window that displays more detailed information about specific components. This information includes data type, data rate, data destination, bit error rates, block diagrams, status, signal power levels, and the signal-to-noise ratio.
The CDD provides the engineer with an easy-to-use, real-time, graphical display of data collected from tests varying from component characterization to networking experiments.
Lewis contact: Elaine S. Daugherty, (216) 433-3456
Headquarters program office: OSAT
The algorithm is based on differential pulse-code modulation (DPCM), a predictive compression technique where the anticipated value (prediction) of an incoming pixel is subtracted from the actual value. This difference is then assigned to a level from a limited set of quantization groups. For this implementation all predictions of an incoming pixel are done on an intrafield basis to eliminate motion degradation and minimize the complexity of the processing circuits.
To improve on DPCM, the compression algorithm additionally uses a nonuniform quantizer, a nonadaptive predictor, and a multilevel Huffman coder. The nonuniform quantizer improves resolution of the reconstructed video, the nonadaptive predictor increases edge preservation, and the multilevel Huffman coder provides additional data rate reduction.
Several issues must be resolved to transmit Huffman-coded digital video over a constant-rate serial channel. Two such issues are rate conversion (from variable to a fixed rate) and recovery from communication errors. An intelligent data rate buffer was implemented to perform the rate conversion while also efficiently storing the compressed data and guarding against memory underflow and overflow. An error concealment circuit was implemented to allow the decoder to mask and gracefully recover from bit errors.
A real-time system has been developed to implement the enhanced DPCM video compression algorithm. The hardware reduces the digital video information from 114 Mbps to 26 Mbps (.1.8 bits per pixel). Quality of the reconstructed video is excellent with no motion degradation.
Lewis contact: Thomas P. Bizon, (216) 433-8121
Headquarters program office: OSAT
This ISP design is based on a shared-memory-per-beam approach (ref. 2). The architecture has three major subsystems: the input module, the control module, and the shared-memory module. In addition, a test bench has been added to simplify packet generation.
The input module enables the correct destination dwell first-in-first-out memory (FIFO), depending on point-to-point or multicast (point to multipoint) connection. Although true multicast is allowed at a beam level, the switch performs only broadcast (all dwells) at a dwell level (ref. 1).
The control module synchronizes the system by generating subpacket, subframe, and frame clocks for both the uplink and downlink. It also monitors and controls the dwell FIFO's, the address control memories, the address-pool FIFO, and the shared random-access memory.
The shared-memory module stores the uplink data and transmits them back to the ground. It also performs temporal switching by writing sequentially and reading randomly the address control memories. A multicast signal is created to prevent multicast packets' addresses from being written back prematurely into the address-pool FIFO.
The ISP test bench consists of a subpacket generator module. This module creates subpackets that are easily traced, by filling their fields with numbers representing the frame, subframe, subpacket, source, and destination. In this way we can easily verify proper operation of the architecture. When an error is detected in the received data, these self-identifiable subpackets can lead us to the potential problem. In total, 69 downlink subpackets of the available 80 are used in this example, with dwell usage ranging from 60 to 100%.
During the process of simulating the architecture, we made several modifications to the original concept. The bus width was reconfigured to be 32 bits to comply with existing digital electronic speeds and memory access times. For simulation purposes only, the uplink frame was reduced to 256 subpackets of data per subframe (instead of 8192) as it seemed impractical to simulate and verify so many subpackets. This resulted in the proportional scaling back of the downlink frame to 80 subpackets per subframe. This scaled-back system configuration greatly reduced the amount of central processing unit time needed to simulate and debug the architecture.
The shared-memory-per-beam architecture has been simulated successfully by using VHDL. This approach allows the designer to test for multiple configurations and eases the debugging process by increasing the system's flexibility. This successful simulation greatly diminishes the probability of errors when testing and debugging hardware. If errors occur, the code is easily modified and all the programmable logic reprogrammed. This flexibility results in a faster and more efficient implementation of the design.
References
Last updated 1995
Responsible NASA Official:
Gynelle.C.Steele@nasa.gov
216-433-8258
Point of contact for NASA Glenn's Research & Technology reports:
Cynthia.L.Dreibelbis@nasa.gov
216-433-2912
SGT, Inc.
Web page curator:
Nancy.L.Obryan@nasa.gov
216-433-5793
Wyle Information Systems, LLC