The panel visited some 15 satellite agencies or companies in japan. Table 2.25 depicts the wide diversity of OBP capabilities noted. As before, the reader is reminded that the panel may have not observed many skills of the visited organization, and therefore the chart may be incomplete. Regardless, Tabl e 2.25 shows a broad spectrum of demonstrated capabilities. The following is an enumeration of the primary OBP capabilities noted during the panel's visits in Japan.
ATR-O. ATR has made significant contributions in R& amp;D in OBP/OBS-related technologies:
ATR is a relatively young laboratory (founded in 1986). It has excellent facilities to support the research which has been started. The laboratory has initiat ed research in many areas essential to advanced telecommunications satellite development, and they have a highly qualified staff. Their work in ISL pointing, phased array antennas, and MMIC devices is showing good results and promises significant technol ogical advances.
CRL (MPT). CRL is a prime architect and developer for the COMETS spacecraft. Dr. Iida's group developed the mm wave personal communications systems (PCS) concept for COMETS in 1984. COMETS implements many OBS/OBP technologies:
On-Board-Processing Technology Matrix - Japan
As with COMETS, CRL is a prime architect and developer for the ETS- VI spacecraft. ETS-VI includes experiments for an S-band Intersatellite Communications link (SIC), and O-band (mm wave) communications experiment (OCE), a fixed and mobile communications experiment (FMC), and a Ka-band single access communications experi ment (KSA).
SIC uses a phased array multibeam antenna. It has 19 elements; all 19 are used on receive and 16 are used on transmit. The antenna develops one transmit beam and two receive beams. MELCO built the antenna.
CRL objectives in pursuing further OBP/OBS technologies include optical ISLs and very large space antennas for mobile communications, an "intelligent array" antenna for COMETS broadcast subscribers, and store-and-forward and ISL experiments. A new start "small sate llite" is to be launched (hopefully) in 1997. CRL has made a commitment to developing Ka-band technologies.
CRL is looking at adaptive data rate TDMA systems and developing what they call ADR/TDMA modems which will look at downlink BERs and use these to adjust transmission speeds in response to rain fades.
Fujitsu Limited, Kawasaki Research & Manufacturing Facility. Fujitsu is involved in building the "dexterous hand" and associated equipment for the ETS- VII on-board r obot. This robot is being designed to replace connectors, drive screws, etc. in space, and it will be used to assemble the space antennas planned by CRL/MPT.
Fujitsu's specific developments in computers and on-board processors include the data proces sor/on-board computer (DP/OBC) for the Solar-A satellite launched in 1991. Operational data from orbit show that the computer is functioning very well. An improved version of the DP/OBC was being prepared for the ASTRO-D satellite scheduled for launch i n February 1993.
Fujitsu has an on-going development program for OBPs (computers). The present target system is for the Solar-A and ASTRO-D spacecrafts. The Solar-A solar-flare observation spacecraft was launched in 1991 and included DP/OBC. The AS TRO-D astronomical observation spacecraft includes DP/OBC and for STT.
Data processors perform collection and formatting of telemetry data, control of subsystems by command, and buffering and data processing of observed data. The OBC has the followin g characteristics: 80C86 (4 MHz) CMOS GaAs, 256 kbytes of SRAM, 3 x 10(superscript 3) rad hardness, fault tolerant (protection against single event upset, SEU), and other space qualifications.
KDD R&D Laboratories. Major KDD R&D Labor atory (KDDL) developments include on-board linearizer; subchannel switched frequency division multiple access (FDMA) with surface acoustic wave (SAW) devices; 14/11 GHz TDMA; Sequence Scheduling Burst Time Plan (SSBTP) for satellite switched TDMA (SS-TDMA ); variable-rate, soft-decision Viterbi decoder; high-performance, multiple- beam earth station (multiple feeds for single reflector); HDTV modems at 120 and 140 Mbits/sec (may be used for B-ISDN) in the future; video codecs (seen in labs); digital voice communications for maritime use.
Primary areas of investigation at KDDL are (in descending order of importance): mobile satellite communications, VSAT, television transmission, information security, and SAW device applications.
KDDL has developed , or is developing, SAWs for SS-TDMA. This system was discussed relative to full OBP. An SCPC system using SAWs (vice wide-bandwidth TDMA) enables very small and inexpensive earth stations. INTELSAT SS-TDMA networks use an acquisition and synchronizati on unit (ASU) for transponder hopping -- each transmitter may hop up to four transponders with up to eight multiple access destinations per user; it uses a microwave switch matrix (MSM) to switch six beams.
KDDL has made significant advances in algori thm and software development for SSBTP to support SS-TDMA. This capability is essential for optimal implementation and, therefore, maximum efficiency of SS-TDMA. The software package supports on- board, satellite-switched MSM, development of acquisition and synchronization (SS- TDMA), design of network conference control, and development of BTP generation software for automatic burst transmission planning.
The functions of SSBTP software include assignment of network control, traffic loading of tr ansponders (e.g., six to eight transponders with thousands of bearer channels), switching sequence generation, and burst generation. The software output includes graphics which are easy to grasp, such as beams vs. time of day.
The first commercial us e of SS-TDMA was on INTELSAT VI. KDDL is also studying on-board precise time and its effect on SS-TDMA implementation. Difference timing on-board is being studied currently.
Some think that the jump to OBS will be made with full, autonomous OBP cont rol. KDDL's Deputy Director Ito thinks SS-TDMA is an interim step to OBS/OBP and therefore OBS/OBP may be available in five or six years. ESA's interest in OBP has been noted by KDDL. According to Dr. Ito, "OBS/OBP is difficult to realize but it i s the ultimate goal -- transparency is needed." Transparency was defined as not requiring fixed transmission rate nor fixed modulation type. (Digital transmission for terrestrial networks are well developed -- not so in space.) KDDL's view appears to be that SS-TDMA provides transparency and that if a multicarrier system is being used then OBP is the best approach.
Matsushita. No significant OBP/OBS activity.
MPT. MPT supports and funds COMETS, ETS-VI and ETS-VII (Note: ETS -VII is not an experimental satellite (primary focus isrobotics)). CRL, a subordinate element of MPT, is the prime agency for development of COMETS.
MITI. MITI is involved in the design and implementation of the ETS-VII satellite in terms o f the on-board robotics devices to experiment with the repair and retrofit of the satellite structure, subsystems and components.
MITI and MPT share supervision of certain "key technology centers" such as ATR, SCR, and the Space Communicatio ns Laboratory (SCL).
MELCO. MELCO built an S-band, phased array antenna for the ETS-VI ISL. The antenna system has 19 antennas. Each is driven by a BFN which is driven by a single phase shift controller. The system has a single-beam, 2.1 GH z mode and a two-beam, 2.3 GHz mode. Beams can be scanned at ñ10(superscript O) per minute.
MELCO manufactures space qualified frequency multiplexers, modems, and phase shift controllers.
Aside from prime contractor and spacecraft integrator res ponsibilities, MELCO produces antennas, solar panels, amplifiers, receivers, oscillators, and ion thrusters.
NEC. NEC is the contractor for the communications payload for COMETS which is to be launched in 1997. COMETS will have significant OB P/OBS technology on board.
NEC has extensive capabilities in development and production of on-board RF/IF switching. It has developed a 16 x 6 MMIC-based decoder and controller for OBS which has a throughput of 30 megabits per channel (developed for INTELSAT VI); a 4 x 4 switch processor with rate adaption (development of the 20 x 20 switch processor for INTELSAT IX expected); and an 8 x 8 MSM for 4 GHz (for ETS-VI and N-STAR).
Baseband Switching: NEC is now developing, with SCR participation, the breadboard model for a 3,000 x 3,000 64 kbit/sec switch for compatibility with ISDN, but not with ISDN protocols. The separate access control channel will use the slotted Aloha protocol. The controller will be on board; demand processing will be on board. The control will be implemented via SCPC on the uplink and TDM on the downlink. NEC and SCR have developed a 64 x 1000 switch using LSI which they intend to expand to the 3,000 x 3,000 switch also using LSI. NEC is also working on 3,000 x 64 kb it/sec on-board switching which will have a throughput data rate of 192 Mbits/sec.
In reference to the need for large memory on board to support on-board baseband switching, the NEC implementation does not use large memory size. SCPC and TDM obviate the need for large memory. (NEC does have a 64 kbit SRAM which is space qualified.) NEC segments the ground and space system to reduce the size of OBP equipment. With regard to performance gains from OBS/OBP, NEC believes that the use of SCPC with TDM provides very large gains (and thus eliminates or delays the need to go to full OBS/OBP). NEC is investigating SAW devices for SS/FDMA filtering but not for on-board digital switching.
NEC is not developing general purpose computers (stored p rogram) using microprocessors to support OBS. Stored-program computers are not being investigated. They agree with the premise that, further on, computers are the way to provide flexibility. NEC sees the satellite as a channel, not a node, and does not equate COMETS-like OBS to network layer protocol implementation. They believe these "problems" should be handled on the ground.
NEC is working in two areas of OBS protocols, LANS and WANs; LANS for the space station network using IEEE 802. 4 and token bus protocols; and IEEE 1553b protocols for OBS units. NEC's implementation of IEEE-1553b protocols for OBS units is not standard.
During the laboratory tour, a NEC on-board computer display was noted. Its parameters were 0.25 MIPS, 64 kbytes RAM, 32 kbytes ROM, 7 W power consumption, mass 2.5 kg, dimensions (in mm) 220W x 200L x 100H and an operating temperature range of -20 degrees centigrade to + 50 degrees centigrade. The unit was to be used for attitude control on MUSES-A.
NEC is implementing bit-by-bit processing for forward error correction for COMETS and also the SCR project. Viterbi decoding will be used.
In the future, NEC believes it will need to relate ISLs with uplink and downlinks with OBP for both GEO to GEO, an d GEO to LEO. NEC says processing is not being done for ISL to downlink. Each has been investigated separately. NEC has not investigated world-wide mesh networks via satellite. NEC sees the need for large memory and buffering for such networking. (NE C has developed a space qualified, 64 kbit SRAM.)
NHK (Japan Broadcasting Corporation). NHK has developed state-of-the-art 12 Ghz channel multiplexers.
NTT Yokosuka R&D Center. Two objectives (missions) have been defined forN TT Yokosuka R&D Center: (1) research the impact of multi-beam satellite systems, and (2) expansion of service areas for maritime mobile and land mobile services.
Some key accomplishments of NTT are development of MMIC for satellite OBP (1987), la unching of the CS-3 satellite (1988), and development of communications devices for ETS-VI (1989). ETS-VI is a precursor of N-STAR where many of the developments will be used in operational service. All technology developments are of very high quality a nd are state-of-the-art.
ETS-VI employs a 16 x 2 IF switch with RF/IF MMIC. Its estimated size is 12" x 16" x 3." ETS-VI uses an 8-bit digital CPU for switch control. The CPU unit was slightly smaller than the IF switch matrix.
T he purpose of NTT's MMIC program is downsizing. The program is evolving through five stages: single-layer IC, double-sided IC, MMIC, Uniplanar MMIC, and multi-layer MMIC. They are now achieving 80% to 90% yield in manufacturing. Switching power levels of MMICs are about 1 W to several W. ETS-IV RF/IF converters are conventional, microstrip-based MMIC technology. MMICs for frequency conversion result in mass reductions of up to 5/6 (end mass = 1/6) with average mass reduction of the units of 3/4 (end mass = 1/4). NTT has MMICs operating from UHF to 30 GHz.
The MMIC work was described in the laboratory, which appeared very well equipped. NTT has developed CPW MMICs. A 26 GHz communications system on a set of chips was shown (published at MTT symposia); also a chip for a frequency synthesizer at Ku-band. NTT makes the LSI chips. MMIC receivers are made by both MELCO and NTT.
Uniplanar MMICs will be available later. They will be much thinner and lighter and their IF power requirement wil l be much lower. They will contain a higher level of integration and will be smaller in size. The smaller physical chip size is expected to result in a higher manufacturing yield and cost reduction. An NTT representative remarked, with caveats, that NT T hopes to have a multi-layer MMIC by 1999.
NTT built and integrated the following OBS/OBP related equipment on ETS-VI: 20/30 GHz antennas (four beams, 3.5/2.5 m reflectors) including feeds and frequency selective plates, the S-band feeds for 3.5 m r eflector, a C-band feed for a 2.5 m reflector, the overall antenna system design, and the 20/30 GHz S-band and C-band transponders. The 30 GHz receiver has two MMIC packages (down-converter and local oscillator) with several NTT-developed chips (in CPW). The 20 GHz 10 W TWTAs are built by NEC. The S-band power amplifier is a 100 W matrix amplifier (five beams, eight SSPAs). The 20/30 GHz transponder has an 8 12 IF matrix switch with MMIC switches. The S-band antenna generates five 2 degree HPBW beams and is a precursor of N-STAR which will implement the same antenna beams. Pointing accuracy is 0.015 degree rms using a beacon monopulse receiver at 30 GHz with the beacon station located in Hokkaido. The monopulse receiver drives an XY table which hol ds the subreflector of each 2.5/3.5 m dish. All this has been tested extensively at NTT Yokosuka. For instance, the 20/30 SS-TDMA matrix switch was tested through the PSTN through a telephone switch located at Yokosuka. Most hardware was/is a joint eff ort by NTT, NEC and MELCO.
NTT has developed microprocessors that are radiation hardened against dosages of 10(superscript 6) rads. MELCO is the manufacturer. They have an 8-bit microprocessor with SEU probability of 10(superscript -9) manufactured by OKI. NTT developed a 16 QAM digital microprocessor receiver/transmitter module for 4, 5, or 6 GHz.
NTT is involved in the development of high speed networks, specifically B-ISDN. Asynchronous Transfer Mode (ATM)-based networks at 156 Mbits/sec a re envisioned for dedicated service vice LAN or WAN applications. NTT has ten times the number of people working on ISDN compared with those working on satellite communications. Interim B-ISDN will provide LAN with ATM at high speed. Currently, NTT has about 100,000 ISDN subscribers, with service having been initiated in 1988. This narrowband ISDN can also be switched through satellites (CS 3 or JSCAT). NTT has implemented satellite-based ISDN for limited, remote areas; only six satellite earth termi nals have been deployed. (It is not clear how many ISDN users there are per terminal.) The system uses JSCAT and was deployed to provide service in special areas. Costs exceed revenues. ISDN is transparent to the user.
NTT Yokosuka predicts that f or trunk transit, 0.4% of terrestrial traffic will be carried via satellite after the terrestrial fiber network is established.
SCR. The primary focus of SCR has been deployable large scale antenna reflectors for mobile and broadcast applicati ons. The prime effort in the past years has been in defining a 16 ton geosynchronous space platform for mobile communications in the L-band which features two 30 m deployable antennas. The system calls for a 100 x 100 channel on-board baseband processor . SCR is involved in developing on- board, bit stream switching hardware.
Toshiba. Toshiba is the prime contractor for the data relay and tracking system, ground support system, satellite and orbit control for the ETS-VI satellite.
Tosh iba is developing space-based, phased array radar antennas. They are developing a 22 GHz multibeam Cassegrain reflector antenna for COMETS which is a six-beam design to cover the Japanese Islands. Toshiba has developed precision antenna pointing mechani sms (two orthogonal axes with 0.015 degree pointing accuracy). These will first see flight on ETS-VI. The software for the antenna positioning control electronics was developed in C or Fortran, then reduced to assembly code for the flight model.