News:

Precision Simulator update 10.173 (24 February 2024) is now available.
Navburo update 13 (23 November 2022) is now available.
NG FMC and More is released.

Main Menu

Nuclear Subs' acoustic capability.

Started by Dominic Manzer, Fri, 19 Jun 2009 08:01

Dominic Manzer

A few days ago I stated that Nuclear Subs would help search for Flt 447's Flight Recorders (FR) and that they potentially could "hear" them 100 or more miles away. As it has been announced that a french sub has joined the search I'll explain the the basis for this statement. I have no personal experience with sub's or sub defenses so I'm not revealing any classified information, only extrapolating from publicly available information and capabilities in related fields and should be taken as speculation, details may differ from reality but the underlying concepts are sound.

First, the range for detecting the FR's assumed they are not buried in debris, are sitting exposed on flat ocean floor in still unstratified water. Unfortunately reports of the bottom topology in the area make it clear this search is in one of, if not the most, difficult location a search of this type has ever been conducted.

Many of the sonar techniques come directly from radio technology and require a substantial amount of background to understand. AM radio transmits a carrier signal that stays constant in frequency and is modulated (the method the program is added to the signal) by changing the amplitude (power) of the carrier signal. Thus the name Amplitude Modulation or AM. Modulation causes the transmitted power to be spread across a band of frequencies around that of the carrier (center) frequency. The width of the band of effected frequencies is a measure of the the amount of information that the signal can transmit. The terms bandwidth and broadband come directly from radio theory but today are mostly associated with computer internet connections. In very ruff terms, the bandwidth of a signal measured in cycles per second is the approximate number of bits per second a signal can carry.

Amplitude Modulation, AM, is not very effective at spreading out the carrier power, creates a small radio bandwidth producing pore sound quality. Frequency Modulation, FM, directly shifts the frequency of the carrier signal in response to the program material spreading the carrier power more effectively than AM. FM provides better quality of program transmission at a cost. FM requires exclusive usage of a wider band of frequencies from the limited radio frequency spectrum as well as a higher signal to noise ratio to work. The higher required signal to noise ratio of FM limits reception range. Also FM is more difficult to analyze for prepossess of design. Both the transmitter and receiver are more difficult to build, requiring more parts. Because FM moves the carrier frequency around it's center frequency it is possible for the radio to automatically make fine adjustments to optimize reception. Early (mid to late 1960's) analog FM receivers exhibited the lock in effect to the user. In geographic regions with few FM stations, as you tuned the knob to change stations you would hear low level noise until suddenly the station appeared, strong, at full quality, and would stay locked onto the station until the receiver was de-tuned a substantial amount.

More advanced modulation methods increase the difficulty of design and construction, increase the lock in effect, and increase bandwidth. GPS receivers have a substantial lock in effect, including what is called software gain. When an early GPS receiver is first powered, it has to find the spacecraft. As it searches it locates the spacecraft's then locates the ephemeris signal. When it has received the complete ephemeris data set it can calculate it's position. Essentially it synchronizes it's internal clock frequencies with those in the GPS satellites. Initial acquisition took 3 minutes in early receivers but when contact is momentarily lost (in trees or a tunnel) the internal clocks and ephemeris are still close enough to instantly reacquire the spacecraft's signals. Be patient, I'll get to sub's search for the pingers soon.

AM has one big advantage over other modulation methods, an AM signal is much easier then others to find if both are transmitted at similar frequencies. Nearly all of the AM carrier power is concentrated in a narrow band of frequencies. If you look at an AM signal with a radio frequency spectrum analyzer (similar to the spectrum analyzer in stereos) the signal looks like a very steep mountain range. The central peak (the carrier frequency) is much higher than the rest, Other peaks and valleys rapidly get smaller to the sides until the signal fades into the background noise. Advanced modulation and coding techniques reduce the hight of this central peak by spreading more of the transmitted power over a wider band of frequencies. This is good for bandwidth but very bad for detectability. (Coding, similar to compression, tailors the distribution of one's and zeros to optimize bandwidth use for the particular modulation method and medium (radio, telephone line, TV cable.))

The acoustic beacons' (pinger) attached to the FR's are designed to produce a narrow band signal at a fairly precise frequency to maximize detectability. I believe the pingers are designed (or allowed) to shift frequency wile transmitting to reduced false alarms. Many man made and natural things produce a constant tone. Fewer produce a sweeping tone that precisely repeats. The sweep makes the signal identifiable even if the pinger is off frequency from damage. Narrow bandwidth is important for another reason, the received noise power is less with smaller bandwidths. The smaller bandwidth signal you are looking for the narrower you can set the filters in a spectrum analyzer without fear of filtering out the signal. As the filter bandwidth is reduced, the background noise is reduced in proportion.  Using this technique and very large high gain antennas NASA's deep space network can locate and receive data from spacecraft literally at the edge of the solar system (well beyond the planet Pluto.) using a transmitter that is no more powerful than a cell phone. (The data rate in this case is reduced to a few bits per second (even a few seconds per bit) to allow enough integration time to positively identify one's and zero's.)

Many large subs are equipped with hull attached hydrophones as well as towed arrays of hydrophones. The towed arrays are known to be up to several miles long and can be deployed both upward and downward from the sub. This provides a collection area measured in square miles (Talk about a large antenna.) (A hydroplane array of 1 sq. mile operating at 5 kHz has the same theoretical performance as the deep space network's 100 meter dishes operating at 10 gHz.) By arranging delays of some hydrophones relative to others the array performance can mimmic a dish antenna pointed in any direction. At the operating frequency of the pingers, computers can easily do the calculations in real time to make the array perform like several high gain antenna pointed in different directions at once.

Of course the sub will be listening for the pingers any time they are within range and traveling slow enough to do so. Effective sonar speed is highly classified, so I don't have a clue, but is probably quite fast at depth for the hull mounted hydrophones. I would expect the sub to first perform vertical and horizontal surveys of the area to establish the acoustic topology for planning the search methodology. Salinity and temperate gradients along with currents create sound channels, reflective layers, acoustic holes and other phenomenon that can help or hinder the search. That done, given the mountainous terrane of the bottom, I would expect a preliminary search with the array configured to sweep a vertical continuos fan from one side to the other with maximum gain to the sides and a reasonable gain down ward. Simultaneously searching the bottom with active sonar for debris while listening for the pingers, pass over the widest valleys near the probable impact location at the deepest depth with favorable acoustics. After this initial search, configure the array for one sided and down operation to circle the outside of the search area. With luck, the pingers will be up high on a mountain and will be herd from a distance or will be in one of the initially searched valleys.

This initial search, taking only a day or two, is certainly completed by now. We know that one or more surface ships are now equipped to investigate any targets identified by the sub. Next, configure for best performance down, start mowing the lawn preferentially following the deepest parts of the valleys avoiding mountain peaks. This search "grid" should be scaled for completion in about 3/4 of the remaining expected pinger life. After the grid search, check any shadow areas or "acoustic holes" that could not be covered.

As their nickname implies, the pingers do not transmit continuously. The signal is short pulses separated by relatively long gaps. This is done primarily to extend battery life, but secondarily it may have been more compatible with 1950's active sonar that used similar signals and were thus optimized to receive them. Ordinarily, the quite times would be a mixed blessing. Humans, especially divers trying to locate the FR in a pile of debris can distinguish and locate a pulsing tone easer than a continuos one. But for long range detection, as this case certainly requires, pausing limits the amount of signal energy available to integrate, forcing the sub to slow, reducing area coverage. But modern radio theory may have mitigated the down side of pulsed operation and made it an advantage.

Now for the spooky (classified) stuff. In the late 1970's the concept of frequency agile radio was being publicly discussed. Frequency agility is simple, build radios that can rapidly change frequency in a non repeating suddo-random pattern so an enemy can't listen to you if he can't change frequencies fast enough or doesn't know what pattern of frequencies you are going to use. The problem is, the enemy can either buy, build or steal the same type of radio you are using. With a little ingenuity and a spectrum analyzer, he can tell when you are transmitting and will figure out your patterns.  To fix this, the military put money into the mathematical theory of frequency agility. Theory advanced rapidly and became known as spread spectrum.

Spread spectrum and advanced modulation & coding, both have the property of increasing bandwidth and making the transmission look more like noise. When beginning a dial up internet connection, modems at opposite ends of the line determine (negotiate) how much bandwidth is available producing the series of tones followed by white noise. Most of the speed increases above 28kb, is produced by coding during the white noise portion of the negotiation. Both dial-up and wide-band internet connections probably use modulation & coding at least partially developed by funding from US military.

Cell phones and other shared spectrum radios use spread spectrum techniques. Conventional radio can have only one broadcast in each frequency band at once. Spread spectrum allows many simultaneous users of a given band. The FM broadcast band can accommodate only about 35 broadcasters at once. Spread spectrum can accommodate hundreds of broadcasters at once and thousands of potential broadcasters. Spread spectrum has the nice property of degrading slowly with heavy use. Increasing numbers of users increases transmission noise levels (the noise herd on the phone.) With cell phones, if the network becomes overloaded simply increase the number of cell sites.  This decreases the number of users per cell, decreases the average distance to each user and improves the radio frequency signal to noise ratio.

For the military, it was found that as you spread the transmission over a wider spectrum the signal looks more like noise eventually disappearing into the background noise. Broadcasts no longer can be seen on a spectrum analyzer and can only be detected if you know the pattern. This is called low probability of intercept radio, it is possible to intercept it, but the other side can't have enough computers or radios to stumble onto and keep up with the changes.

Spread spectrum also maximizes the lock in effect. When you get the signal you've got it load and clear. Remember the signal can be found in what looks like noise to other technologies. When Seti (the Search for ExtraTerrestrial Intelligence) was proposing it's first generation of ultra wide-band, narrow channel receiver-spectrum analyzers, the US military had to be convinced that it would not be able to break the low probability of intercept radio's code. Seti is looking for the opposite of spread spectrum, narrow band transmissions designed to be found but at every possible frequency. Low probability of intercept transmissions will look like noise to Seti. The military was happy and let Seti continue.

The technique used to receive low probability radio will work at acoustic frequencies, if you know what pattern to look for. Every submarine, surface ship and even aircraft have their own individual acoustic fingerprint. Once that fingerprint (pattern) is fully known, the vessel can be identified and tracked at much longer range than without.   Hopefully, the French had previously cauterized (fingerprinted) the FR pingers. If not, another user of compatible equipment may have the finger print and supplied it to them. If not, they can quickly test one or two pingers on-sight and get a reasonable fingerprint to use. But they will not know how different Flt 447's are from the samples. Not knowing the variability, risks looking for too wide a range of variation and suffering reduced sensitivity or looking for a two narrow a range and miss them all together. There is a very good chance the French have a good idea of what to look for as navies record everything in the ocean so they can tell what they can ignore.

Phil Bunch

As best I could follow your summary, most of the things you describe apply to radio frequency signals.  I believe that the pingers for underwater "black box" detection are purely acoustic - almost all radio signals cannot transmit through water for any significant distance.  The only underwater radio signals I can recall are the extremely low frequency signals the US submarines use to receive very limited data transmissions while underwater.  From distant memory, they have huge antenna farms designed for sending these signals.

Please excuse my error(s) if I have misunderstood your note or the nature of the acoustic pingers they are searching for.

I also read that the US Navy has sent a team of specialists with some specialized equipment to search for the pingers, in cooperation with this search effort.
Best wishes,

Phil Bunch

Dominic Manzer

The radio techniques apply directly to acoustics. The only difference is that a hydophone is used too convert the sound waves to electrical signals on a wire instead of an antenna converting electro magnetic waves to electrical signals on a wire.

Dominic Manzer

I tried to make the difficult subject of signal processing understandable. I used a lot of familiar examples to link the techniques to everyday experience. The result is that the description tends to jump around. But signal processing is not linear subject. To fix the short communings of my description would have taken about a month, and thus not be timely, I sorry for that.

If anyone would like clarification of any point, or something that may or may not be related, please ask.

Knowing what is taking place in the background, I'm continuously amazed at how far things have advanced in the last 30 years. When I was in collage, only simple digital signal processing could be done requiring a large mane frame computer. Digital signal processing was barely mentioned in undergrad courses. Today, the graphic equalizer, that we take for granted, in music applications is sophisticated digital processing running in real time on a computer (MP3 player) that fits in a shirt pocket.

Holger Wende

#4
Hi Dominic,

Thanks for the effort to explain all the details, I appreciate that.

I just wonder why these blackbox transmitters are not implemented like a transponder, i.e. they reply only to interrogation signals. This might save a lot of lifetime. I read or heard somewhere that blackboxes send only for approx. 30 days, al least under water. Additionally locating a transponder is faster/easier, because as soon as you have received the transponder signal you know the distance, too...

A few years back I remember that the US DOD claimed, they might locate any submarine in the Atlantic ocean with their installed high-tech hydrophone network. Well, at least it does not seem work for the blackbox signals, which beep "loud", I guess. But this might have been as well just a deterrence trick or just annother rumor spread by newspapers/magazines :?

Regards, Holger

Dominic Manzer

Quote from: Holger WendeI just wonder why these blackbox transmitters are not implemented like a transponder,

A few years back I remember that the US DOD claimed, they might locate any submarine in the Atlantic ocean with their installed high-tech hydrophone network.
Regards, Holger

The pingers are as simple a device as possible to make them as reliable and as crash worthy as possible. I'd like to see a diagram, but guessing from the little I've herd the transducer is little more that a mechanical buzzer - when power is applied it moves away from a contact breaking the circuit, then falling back under spring tension. I have no guess how the timer is made, but it will be vary simple, and use little power. I believe the battery is water activated, saves an ON switch and no chance of discharging prematurely.

A transponder will necessarily be more complicated, and will need to be listening
for signal to respond to. Probably will not use less power, but a good thought. Also the round trip time will have allot of uncertainty, taking between 12 and 16 seconds. A long time to wait, this limits the number of pings to one every 4 seconds even with overlapping queries. The speed of sound limits how fast things can happen, this is like a radar transponder on the Moon, round trip time 6 seconds, Mars, about 20 min.

The claim for locating any sub refers to the SOSA network. SOSA is a line of hydrophones laid across the natural chock points in the Northern North Atlantic, Scotland - Iceland - Greenland. (and probably a few islands) Pretty much the transatlantic cable route. They pick them up crossing SOSA then can track them using a few carefully placed hydrophones. But I don't know how far south the coverage goes.

Remember the Boxes are probably in a deep canyon that blocks the sound from spreading out horizontally and can only be herd from above. If the half width of the sound cone is a generous 45 degrees the size of the spot on the surface is only 6 to 10 miles in diameter. A small bulls eye in a very big ocean.

When I saw the crash location, I knew it would be a major stroke of luck to have any sonar system from any country near by. The equator is not of strategic interest to anybody, except Brazil and they might not patrol it. It took several days for the French sub to get there, probably sailed from the French coast. How long does it take at 40 Knots? 10 times? as long a as plane flight, 60 - 80 hours, 3 - 4 days?