Thursday, August 25, 2016

What is the impact of the Scorpene leak?

It appears that the Australian has gained access to the technical manual of  the Scorpene submarine variant being sold to India [1,2,3]. The leak appears to be in the DCNS, through a retired navy official who attempted to secure a Malaysian order through the unauthorized release of these IN Scorpene related documents. As you may be aware the Malay Scorpene deal has come under quite a cloud as several murders have been linked to it. It is not implausible that the data was not intended to be actually shown to the Malays but that somehow given the general shit show that pervades anything in KL these days, the chain of custody around the Scorpene data was not maintained properly.

Based on the crude descriptions available in the lay press, it appears the data leaked contains technical portions of the "manual" of the submarine as written by the manufacturer. These documents lay out the envelope of vehicle performance in which an operator might define a set of missions. It would be foolish to take these numbers at face value, as most of them will either be too broad or too specific to plan specific military counter-measures. Anyone receiving this information cannot be certain that anything highly specific is accurate, so all the typical naval intelligence research type activity will have to be carried out anyway.

Of the information released, some sub-systems can be upgraded or revamped to reduce the damage of the leak. However one part of this sub cannot be changed, and that is its "accoustic signature". It appears some of the typical frequency ranges of the acoustic signature of the Scopene's supplied to India has leaked out in the those documents.

As a submarine moves under water it will make sounds from its propeller. The nature of this sound varies with the speed at which the submarine travels. The physical origins of this sound are a combination of cavitation effects and mechanical noises from the propeller shaft. As the exact shape of each propeller and shaft is different for each submarine (even in the same general class) the sound of each sub is distinct and over various speeds it can form a unique pattern.

The propeller noise is extremely low frequency (100mHz -1Hz) and can be easily captured by even shore based sonar arrays. There are many different sources of such sounds (other ships, geophysical events, whales etc...) and you can detect a submarine *if-and-only-if* you can discriminate the sound of its propeller from background noises in the sea. Once discriminated from the background, using relatively standard DSP techniques, you can isolate the fingerprint of each sub. From a data mining perspective - when you are listening for a submarine in the waters near you, it really helps to refine the search to specific frequency regions where the sound may be found.  Once you have clearly identified a potential submarine signature, you can use more DSP tools to improve filter efficiency and catch the same signature in a variety of other data sets. In this way you can rapidly identify the submarine's patrol pattern and find the submarine whenever you want to.

It is extremely difficult to alter this aspect of the submarine's performance. Making mechanical changes to the submarine propeller design or the submarine shaft is prohibitively expensive.

In the more general sense, it is much easier to refine DSP techniques than to change submarine designs. So in the arms race between DSP techniques and submarine mechanical design - the DSP side always wins. And that is why submarine design and performance details are kept so secret. 

If enough Scorpene subs were in operating in the world, someone somewhere would be able to record the sound from a single submarine. And then once that was done, you could use such information to guide the search for other submarines of the same-ish design elsewhere. This process would have taken several decades.

Now thanks to this leak - the decades that the process usually takes will now reduce to a few years. This will put adversary navies at a position of advantage vis-a-vis the Indian navy. Not only is the Indian Navy compromised by this to some extent, the other navies that operate this design are also affected adversely. (That is why even the Australian won't publish the exact numbers).

I think the governments of all the nations involved, the French, Indian, Australian, Malay etc... will collaborate to investigate and plug the leak, but the specifics of the Indian subs are out there already and it best to proceed under the assumption that Pakistan and China already have this information.


Post a Comment

<< Home