An entirely Indian listening device

Interview with S.K. Shivakumar, ISTRAC Director.

FOR deep-space missions, the Indian Space Research Organisation has established an impressive communications infrastructure called the Indian Deep Space Network (IDSN) at Byalalu, a village about 45 kilometres from Bangalore, as part of the ISRO Telemetry, Tracking and Command Network (ISTRAC) system. Comprising an indigenously built 32-metre-diameter antenna and a German 18-m antenna, the IDSN will be the hub of communications from the ground and the centre of activity for the entire duration of the mission.

In an interview given a week before the launch to Frontline, ISTRAC Director S.K. Shivakumar talked in detail about the new infrastructure and the nature of operations after the satellite separates from the launch vehicle. Excerpts from the interview:

What are the critical issues involved in telemetry, tracking and communications in general associated with deep-space missions?

When we say deep space, deep is with respect to the distance. When we talk of satellites in near-earth orbit, we mean about a 1,000 km altitude or more, or near-earth space of about 2,000- or 2,500-km range from the surface of the earth. But when we say deep-space mission, we mean lakhs of kilometres. For example, when we talk of the moon mission, it means that the distance is not less than the distance between the earth and the moon, which is about 400,000 km. Internationally, there is a way of categorising deep space and near earth, but a common way of defining them would be the moon distance and beyond.

In deep-space missions, as the space probe moves farther away from the earth, the strength of the electromagnetic signals from it becomes weaker and weaker. The real challenge is to catch those weak signals. Mathematically, from antenna theory and all that, we know that we have to put up larger and larger dishes. ISTRAC has so far been involved with smaller dimension dishes, about 10 m-11m diameter dishes. But now for a deep space mission, it jumps to something like 32 m. To make such an antenna to meet the Chandrayaan mission requirements, especially through the indigenous industry, was a big challenge. We looked at [systems] the world over and found that the nominally working deep-space antenna you get to see is 30 m, 32 m, 34 m, etc. We decided to make a 32-m antenna in Bangalore that would give us the strength to talk to our satellite from our own soil and also to collect the signals from Chandrayaan about 400,000 km away both in terms of satellite control capability that is, TTC [Telemetry, Tracking and Command] operations or satellite housekeeping and also the science data coming from the various on-board experiments.

But wisely, this DSN-32 has been done not only for the Chandrayaan mission but for all deep-space missions to come in the future. It puts us in the category of deep-space antennae found anywhere else in the world. That is the whole essence of building an IDSN facility. Starting with Chandrayaan, we are pretty sure that we can track any other object deeper than this. If we are doing a Mars mission, we do not have to worry at that point of time whether we have to build some more things. We have built a world standard facility that meets all the international standards. That means it can track any other [deep-space] object of any other space agency. So, simply stated, it is state-of-the-art interoperable and cross-support compatible facility that meets the Indian requirements with good margins and also the requirements of any other space agency.

For deep-space applications, when we say that we are capable of receiving signals of weaker strengths with this antenna, we should similarly be able to pump fairly strong signals to the satellite an appropriate uplink facility for commanding the spacecraft. Once the diameter of the dish increases, that [the pumping of fairly strong signals] is very easily done with higher-power amplifiers. About 2 kilowatt was our normal use. This time we have put up a 20 kW [klystron-based] high-power amplifier. That much power with a big dish is enough for the satellite to receive and execute the command functions. This is another world standard that has been met by the IDSN. Of course, this antenna will also be capable of doing what is called the two-way ranging required for determining the position of the spacecraft.

In addition, we have put up a reception facility for the science experiments. The science data received here can then be sent to different processing systems for producing the various data products. This, in short, is what it means to put up a deep-space antenna in terms of utilising it for the upcoming [Chandrayaan] mission. All this needed a lot of critical technologies to be developed and everything had to be done through Indian industry.

You chose 20 kW for your power requirements. Why do you need such high power?

We want to pump as much as possible to talk to the satellite. Certainly, it is on the higher side for a moon mission. But always you want to have the power with you. It is up to you to control and use it. If you have more power, in any bad situation spacecraft attitude loss and such when communication is difficult one needs extra power to command the spacecraft. Somehow you want to send command to the satellite and you feel the need for extra power. Twenty kilowatt is essentially for commanding the spacecraft at your requirement. So, that much we have built into the system. We can work with 2 kW also for Chandrayaan distances. But if you want to go farther, you need that kind of power. Most of the deep-space ground stations the world over use similar systems, 2 kW in the main and 20 kW redundant. With 2 kW you can do a lot of work. You keep 20 kW for any extraneous situations and contingency requirements. Of course, the first time you are doing such things, you want to live with a slightly higher level of comfort than you could have done with.

Why do you call it a network when you will be using only a single dish to do the tracking operations?

Actually, two antennae are there now. Besides DSN-32, an 18-m antenna is also there. The ISTRAC network itself comprises about 20 antennae. Since we had crossed our 11-m barrier, and DSN-18 can also be used for the Chandrayaan mission, at least two of them can form a network to start with and it wont stop at that, I am sure. There will be requirements once we go further in deep-space missions in the future; there will be requirements for us to build.

Will you be linked to other deep-space network centres of the world?

It is an Indian effort and it is meant for India as envisaged, but we can get connected to or can support international deep-space missions. As I said, our antenna can be used for tracking the space object of somebody else and it can be used for sending commands or receiving data or whatever is required from our stations. [For this mission] it wont be connected to anything else other than other ISTRAC network centres. But one ground station in Hawaii will be tracking the first three to four days of the mission and providing TTC support [on contract] till the satellite gets to 100,000 km during the orbit-raising phase [data from which will be transmitted via its network control centre at Pennsylvania, United States]. [Basically, since satellite visibility at Byalalu will not be available all the time at lower ranges of the satellite trajectory, a station at the diametrically opposite side of the earth has been chosen to track during such hours of the day. This information was provided by L. Srinivasan, the head of operations at the Byalalu DSN site.]

All the data will be sent to the spacecraft control centre [of ISTRAC at Peenya], and the science data will be sent from this facility to the Space Science Data Centre [or SSDC, located next to the antennae at Byalalu]. So this is how connectivity has been established as far as our use is concerned. But if somebody wants to use [our] antenna per se, we can certainly do that. Then, special connectivity will have to be established, maybe for a short duration, but that can be done.

Will the science data you receive be only from the Indian experiments or from all the 11 on-board experiments?

Data from all experiments will come in an integrated manner as a single transmission from the satellite. This antenna will receive that total signal, and we do what we call dechannelisation. We segregate the signal into individual experiment data streams, and these are given to the computers for processing. And an individual scientist will get his experiments data. For convenience of transmission and all that, enough techniques are available to combine all the data into one single data stream complete with compression, coding etc. and then send it. That is being exploited on board and a similar thing is done on ground also. Whatever encoding you have done will be decoded here.

But will you be sending the raw data to the individual groups?

We can send raw data or processed data depending upon who wants what. Clear plans have already been drawn up for that.

In terms of the amount of data that you will be receiving, what will the bandwidth requirements be? Could you give a comparison with what you handle in low-earth orbit (LEO) missions?

Of course, in deep space everything is [at] a premium, naturally. Actually, IRS [Indian Remote Sensing] satellites, which are in 700-900 km orbit, produce much more data than what Chandrayaan will produce. [For] the imagery that you collect with 1 m and 5 m resolutions, the data are quite voluminous. But, to my knowledge, we are [already] in the higher level of data transmission from Chandrayaan. We will be transmitting the data at 8.4 Mbps, whereas many people do it at much lower rates. Just for comparison, IRS satellites transmit images at 100 Mbps data rate.

But the comparison is not quite appropriate because here we have to travel that much more distance. So you cannot afford to do in deep space everything that you do near-earth. Naturally, one does a technical assessment of what we are capable [of] and how much we can bring down. As far as Chandrayaan itself is concerned, we are certainly at the higher end. Since the incoming data are at 8.4 Mbps, we have certainly done well in terms of organising ourselves for transmitting these data. Just to give an idea, the data we receive from Chandrayaan at our SSDC will be redistributed the way in which principal investigators of experiments want. We have put up really high-speed dedicated links, which will be a mix of both terrestrial and satellite-based, and for the first time we have established such a large data network.

For Indian payloads, we have a Payload Operation Centre at the Space Applications Centre [SAC] in Ahmedabad, which has built both the Terrain Mapping Camera (TMC) and HyperSpectral Imaging [HySI] camera, to which we have a 16 Mbps link from Bangalore. For ISRO Satellite Centre [ISAC], which was responsible for the Lunar Laser Ranging Instrument [LLRI] and the High Energy X-ray [HEX] spectrometer, and is also a co-investigator in Chandrayaan-1 X-ray Spectrometer [C1XS], we have put up a 2 Mbps link because it is closer and the data volume is less. For the Vikram Sarabhai Space Centre [VSSC], Thiruvanathapuram, we have given a 512 kbps link and investigators from the U.S., who have produced the Moon Mineralogy Mapper [M3] and Miniaturised Synthetic Aperture Radar [MiniSAR], we have put up a 6 Mbps link between Maryland, U.S., and Bangalore. In addition to that, since some people did not want dedicated links because they wanted [their data] to be in the public domain, we have put up a high-speed Internet link of 16 Mbps.

These are all, I would say, first in our domain. ISTRAC has never handled so many high-speed links, but these are all meant for science data distribution required by different scientists. In addition to that, of course, we have got normal mission operation links, which will be at 256 kbps between different ISTRAC stations. To get high-speed data from Pennsylvania to Bangalore, a 3 Mbps link has also been set up because of the volume of data received from there.

Are there any issues with regard to calibration that you needed to sort out before you started your operations? How did you actually calibrate the system?

First there are the standard test and evaluation procedures that we have in ISTRAC. Then, we tracked some of these LEO satellites with the big antenna; we tracked Cartosat, IRS-P4/Oceansat and others. But, of course, this does not satisfy anybody because you have to track something nearer to the moon. Very recently, we have started tracking Selene, the Japanese lunar orbiting satellite [launched in September 2007]. We were able to establish downlink with the spacecraft, thanks to JAXA [the Japan Aerospace Exploration Agency] cooperation. We were also able to bring uplink fairly quickly.

We have been able to track the satellite continuously with this antenna. That has given us ample confidence to say: Yes. Once Chandrayaan goes near the moon, we will be there to track it. To that extent our comfort level is quite high because if you have tracked a similar object that is closer to the moon and you have been able to establish links with good margins and all that, we dont have to speak much about our ability to do [the same] with Chandrayaan. In addition, we are planning to track another deep-space probe, Rosetta. We have got good hold on that in terms of information. This was another opportunity that was created thanks to the European Space Agency [which launched it in 2004].

That is one part of it. We have also tracked radio stars like Cygnus and Cassiopeia, as well as the sun and the moon and obtained signals [from them].

Do you think that any loss of transmission could occur for some reason say, owing to interference of any kind or something like that?

Well, such things are part of life. But when we chose this site [at Byalalu] we made sure that we were far away from the citys radio interference. On orbit yes, we have to be prepared for that. Of course we are already bandwidth limited; we are living within the band allocated to us. But a glitch or two is not uncommon in any such work. It does not last for too long. We should be able to recover in such cases. But, by and large, we dont anticipate it as a major issue. But once you are in the radio business you have to live with these interferences.

What will happen when the satellite orbit is on the far side of the moon?

L. Srinivasan: The [pole-to-pole] orbit around the moon will be continually changing, and communication will not be there when it goes on the far side. The orbit period is about two hours, out of which about 40-50 minutes will be on the far side on the average. But you store data on on-board solid-state recorders. When the complete orbit becomes visible, which will happen once every 14 days, it will be face on. You command and retrieve that data then.

What about solar events and such?

No, that is not to be expected. Solar flares, etc. do not cause any disruption. But scintillations can cause. But they last only for a short duration.

During the calibration exercises, at least you have had no problems yet.

No. So far, so good.

What are the critical technologies that had to be developed to establish this set-up?

The realisation of the entire antenna system itself was a big challenge because we were doing it for the first time. ISTRAC was responsible for building this. We called DSN-32 a project. We were the project leader. We chose ECIL [Electronics Corporation of India Ltd.], Hyderabad, as the prime contractor. They had the responsibility of building this in collaboration with various industries in the country. In turn, we worked with ECIL very closely. The primary responsibility for the reflector and the mount of the antenna was given to them.

Along with that, we chose BARC [Bhabha Atomic Research Centre] for the antenna control servo system, the major subsystem. The third is the RF [radio frequency] design, [which] was entrusted to ISAC, Bangalore. And ISTRAC and ISAC together developed the feed system the reflector, the sub-reflector and all that goes with the feed. These three are the heart of the whole system and these four agencies constituted the core team for executing the project.

But that is not all, because many subsystems had to be realised. So we went around scouting different industries in the country. We could identify sources with good capability within the country M/s L&T, Godrej & Boyce, SLN Technologies in Bangalore, HAL [Hindustan Aeronautics Limited] and many others. I think we had interface with about 40 industries to do this work. Of course, there would be other sub-subcontractors as well.

But you built DSN-18 with German assistance.

Because once we started this project, we did not have any ground infrastructure. And the launch was declared for March 2007. When you are building a 32-m antenna in the country for the first time, you are not sure if you can meet the schedule. We were fairly confident that we could do the work, but meeting the schedule is an important thing in projects. We cannot be saying that since the ground station is not ready, Chandrayaan mission is postponed. So it was [a] well-thought-out decision to get something in hand. That is why we did not go in for [procuring] a 32-m antenna.

We said that we would only get what would suffice for the Chandrayaan mission, build the ground station and be ready. The entire specifications were drawn up by ISTRAC, and the tender was floated and the German Vertex was chosen. We installed this antenna in October 2006. The reflector and the front-end electronics were done by them and the back-end electronics was done by ISTRAC. And so was the computer control system [done by us]. To that extent, by early 2007 we had checked out the entire system. Since then we have been using it for testing.

When will we actually start getting data signals from the experiments? When will we start that operation?

That will start once we enter the lunar orbit. Once we reach the mission orbit, the 100 km pole-to-pole circular orbit, then we will start what we call commissioning of payloads one by one, systematically. First you ensure that your commanding procedure is right and make sure that you switch on the experiment and that its health is all right, and then you start doing science. Each one has to be done against an operational plan. [First, we will] let scientists have a quick look at that to see if the experiment is okay. This should happen by the second week of November.