A Proposal for AMS Ground Support Equipment
The text below is an early proposal for the AMS GSE. Many things has been
changed since writing this, and part of the text is at least missleading,
if not completely incorrect. However, the principal idea of having a
front-end(s) combined with an analysis computer(s) was more or less
correct. These pages will go through an update/modification cycle in the
near future.
Introduction
In order to clarify the development and operations phases of AMS experiment, the terminology how to call
the different phases should be defined. The following proposal is derived more or less directly from earlier
space projects we were involved in.
1) The first phase is 'subunit' level (called also box level)
2) Second phase is 'unit' level
3) Third phase is 'system' level
4) Fourth phase is 'flight operations'
The distributed nature of the AMS experiment designing and manufacturing leads to the conclusion that most
of the hardware for "subunit level" tests will be made by the associated laboratories which are responsible
of the AMS hardware development. The "subunit level" here means the phase of development when all the subunits
are separated, and when they do not yet communicate or interface to the AMS CPU or digital electronics (i.e.
DAQ or H/K & control computer). Testing of the subunits which handle the AMS interfaces to the Spacecraft,
however, can be done, at least partly, by using the "unit level" GSE.
When the subunits are being integrated to the AMS experiment, the GSE in use is the "unit level" GSE. The
responsibility of such GSE is to help identifying the malfunctions of each subunit with the aid of AMS H/K &
control computer's house keeping data. The GSE may also be used to calibrate the subunits, and to overall
monitor AMS's functions. On the other hand, such GSE is not likely the right tool for hardware level development
of AMS analog electronics or particle detectors.
Development phases
Now that you are completely confused of all the jargon above, I'll try to explain more what each phase is
going to look like. Note that there might be different ideas how things should be defined, but here's mine.
Subunit level GSE
During the "subunit level" testing and developing there is no AMS CPU available, and therefore the testing
of subunits must be done without the aid of the GSE. Obviously few hardware-only level testing equipments
are needed, but the AMS hardware builders should know best what kind of test equipment fits their needs best.
When developing the 1553 databus and associated electronics there might be possibility to use the GSE also
as a developing tool. This is because the GSE will contain a commercial interface unit to AMS 1553 data bus.
Testing of the fiber interface might need special tools to be developed. It is not yet known whether such
tools are manufactured by some AMS laboratories or bought from NASA or from a third party provider.
Unit level GSE
During the "unit level" testing the AMS subunits are being integrated together, or they are already integrated.
Hence AMS and its subunits are ready to communicate directly to the GSE system via the standard 1553 databus.
This assumes that the AMS CPU and 1553 interface subunit are the first subunits available when the integration
of AMS takes place.
The GSE system consist of several computers, all of which are responsible of well defined list of tasks. The
system will contain the "interface computer" to the AMS experiment, the "data analysis unit" and several
workstations which are used by the individual subunit builders to check out their subunit's functionality.
The "interface computer", which will later be called the "front end", will either be built by one of the AMS
laboratories, or perhaps some parts of it can be bought from the NASA. Also, it is not yet known whether during
the Shuttle flight there will be a "ISSA simulator" into which AMS is connected, or whether the AMS is connected
directly to the Space shuttle. Therefore there might also be a need to build an "ISSA simulator" which probably
will be a flight item.
The front end is responsible of receiving the telemetry from AMS and doing the translation of the data packets
to a format which is more convenient for the GSE "data analysis unit" (DAU) to handle. On the AMS side the front
end receives the 1553 data packets and handles the sending of the telecommands with appropriate protocol. The
front end also contains the power distribution unit (simulator) which is used to deliver the power to the AMS.
The protocols which will be used to transfer data between the front end and the DAU are UDP/IP and TCP/IP.
Depending on which operating system controls the front end, the DAU can also mount the front end's disks. This
is not necessary as the front end may be also a diskless workstation. Also, the front end can send the data
directly to DAU by connecting to a AMS specific TCP port on DAU.
The "data analysis unit" is a powerful UNIX workstation capable of providing remote file services and raw computation
power to a few client computers. The DAU also archives the data and builds some databases from the received science
and house keeping data. The analysis software running on DAU will be produced using some high level program generation
and development tool, like IDL. There will also be available at least C and FORTRAN compilers running on DAU.
System level GSE
The GSE system which will be used during system level is identical to the unit level GSE, except that there might be
need to modify or replace the front end. This is due to the fact that the interface to AMS must be done via the
satellite provider's ground segment. Other than the front end modifications it should not be necessary to alter the
GSE configuration.
Also, the hardware level development tools involved with the GSE, if any, will not be usable during system level.
GSE during the flight operations
The GSE during flight operations will basically be the same computer system as during system level testing. However,
more data processing power may be needed but it really depends on the complexity of the anysis software and the other
real-time tasks running on the GSE.
The data delivery system will be build around the main GSE system which resides at the Science Operations Center.
Such activities include processing of the AMS science data stream to a more suitable form, for example to build a
relational database which can be accessed from a Structured Query Language, and creating the archive media. The
structure of a system which suits best such tasks will be evaluated as soon as there is better understanding how
much processing power, disk cpacity etc. is needed.
GSE modules
So, the GSE will contain few modules which interface to each other via different buses and wires. To get idea
of what the system will look like here are some pictures. Basecally, there will allways be the front end and the DAU.
We will have two front ends as the interface to the data source will be changed when we step from unit level to
system level. Also, during pre-flight operations we might need to alter something as our experiment might be in
a "non-accessible" location, like in shuttles cargo bay.
Current status
Click
here
to see what's going on in GSE development. (15 May 1996, i.e. badly outdated)