What is a Jammer?

Tucson company Raytheon just won a major contract to develop upgraded radar jammers to the US Navy for the F/A-18s. I thought this would be a good opportunity to explain radar jamming. My first notion of radar jamming came from the 1987 Star Wars spoof, Spaceballs. So, I felt the need to include a clip of the radar jamming scene below.

This scene inspired me to ask 25 years ago, “What is radar jamming?” 

Jamming a radar is a way to prevent antagonist from effectively detecting an object. Before I get into jamming, let me give a brief explanation of radars. Radars are able to detect objects distance from a radar receiver, the speed of the objects they are detecting, and in some cases the angle of the object relative to the radar’s current position. If you do not want the radar to know this information, you jam the radar.

What a jammer does is make the radar not be able to do all of those things. However, you know when you are being jammed. So, think about as if someone does not want you to see them and they poke you in the eye, you cannot see and know you have been poked in the eye. On a side note, when someone cannot see you and they do not know you are there that is stealth.

The way the most jammers work is they figure out what frequency the antagonist radar is at (think of a radar frequency as just like a radio frequency, it is basically the station you are transmitting your radar at.) Then they send out a stronger signal of noise (gibberish) at a higher magnitude or power than your radio to drown out your information from the radar. It is like when you are watching TV when someone is vacuuming in your house to cover up the sound of the vacuum you increase the volume of the TV. Think about jamming as turning up the volume on their signal to try to disrupt it except instead of sound radars use radio signals.



 

 

To counter jamming, the radar can change its frequency like constantly changing stations so the jammer cannot block the message. Like when you change stations when the static for one station gets too bad.

Radar jamming is one type of way to prevent a radar from figuring where you are and how fast you are going. Electronic attack (EA) is what jamming and other ways to prevent a radar from working are called. The list of things that can be done to prevent EA is called electronic counter-measures (ECM). I will talk about more of these in another blog.

You Have an AM/FM Radio, What's TM?

Tucson company TM Technologies Inc announced they have demonstrated a new way to transmit information in electromagnetic waves according to announcements in the Arizona Daily Star and TVTechnology.

For me, this is a local interest story since I live and work in Tucson; however, this technology has the potential to revolutionize communication. According to their press release they demonstrated a new way to modulate electromagnetic waves which they call transpositional modulation (TM). A TM wave is opposed to a frequency modulation or amplitude modulation which you may recognize as FM and AM on your car radio.

FM is a way transmit information by changing the frequency (the number of wavelengths/second that travel across a particular point) of the wave. AM transmits data by increasing the power of the transmission wave. There is a third way of transmitting data called phase modulation (PM) which changes the phase of wave, basically by shifting the phase from a sine to cosine. PM and FM are very similar because FM is the derivative of PM and many people consider PM and FM to be the same. 

Top: base signal
Middle: base signal with AM to carry signal
Bottom: base signal with FM to carry signal

When the receiver (like the radio in my truck) receives the signal it measures the changes in the wave and converts that data into a usable medium (like music in your car speakers).

What TM does is change the same of the wave by adding indentions to the wave. By doing this extra information can be stored in the waves. In the Daily Star, the chief scientist for the Medusa Scientific, the parent company of TM Technologies, Rick Gerdes said, “At the least, the technology allows at least double the throughput, but in some cases four or five times more, and in some extreme cases 30 times more data.” This method would be a fourth way to transmit data on a wave.

I am interested to see how this would work, since detecting minor changes to the sinusoidal signal would have to be detected by the receiver by separating it from the noise and the loss of information when the wave is quantized or converted from analog to digital. On top of separating the signal from the noise, the receiver needs to sample the data at a high enough rate to detect the indentations in the signal wave. In September, they plan on doing a transmission of a UHD signal, which is a fairly low frequency UHF signal which is between 400 – 800 MHz (for comparison WiFi transmits at 5GHz).

TVTechnology lists two patents pending for Mr. Gerdes,(Patent #5200715 –Waveform modulation and demodulation methods and apparatus and Patent #5327237– Transmitting data with video) that they believe the technology associated from the technology is derived from. If this totally is a killer app, I would expect most of the technology would be kept as a trade secret. 

Predicting Future Locations Paper by Salilek and Krumm

 “Far Out: Predicting Long-Term Human Mobility” by Adam Sadilek and John Krumm is an interesting paper that says you can predict long-term were some will be. There is a good non-technical summary by Camille Sweeney and Josh Gosfield of Fast Company.

The idea is that people move in patterns, and you can predict where someone will be in the future based on where they are now. The authors recorded the movements of 703 subjects (307 people and 396 vehicles) from 7 to 1247 contiguous days with the average number being 45.9 days and a standard deviation of 117.8  days (I’m guessing the 1247 was one of the authors.) They had 33,268 days of location data.

They used Fourier analysis to find the periodicities in movements of the subjects and used singular value decomposition (SVD), a type of principle component analysis (PCA), to reduce the dimensionality of the data and to form predictive weights.

They broke the surface of the globe into triangular cells to make the locations and movements more finite, and broke up the day into finite blocks as well. The authors formed the data by breaking up the probabilities that a subject would be at 11 particular locations by 24 hour blocks and by days with a separate block for holidays.

Using the past data the authors, formed the predictive models to predict the locations of people up to 80 weeks in the future. The results were above 80% accurate and better than their baseline.

Normally, I do not like using PCA for dimensionality reduction because the top PCA (aka eigen) features may not be the features you need for your modeling goals. For the radar automatic target recognition work, my team and I picked features manually (using stuff like length and width of targets) to identify them because we knew the size of the vehicles and physical characteristics we were looking for, but with huge data sets where you may not know what features are best, PCA could be a better way to go. I have also used discrete cosine transform (DST) for really large 3D volumes because my PC did not have enough RAM to handle the matrix transforms of PCA.

I was skeptical of this paper when I first looked at it because I did not think people’s movement were that regular to be good for prediction, but with their accuracy people are more predictable then I thought. When you consider the accuracy values you also have to consider most people sleep 6 to 9 hours a night or 42 to 63 hours a week and work 8 hours a day or 40 hours a week and those schedules are fairly regular. So, the authors really need to account for the approximately 72 hours in a week. 

A system like this has both applications for marketing and demographics as well as security and defense applications. 

Basic Statistics terminology

•          Random Experiment:  experiment where outcome cannot be predicted with certainty.

As an example, you are getting dressed in the morning, you are in a rush and instead of picking socks from your drawer you randomly grab a pair of socks without looking. In this case, your picking of the socks is a random experiment.
 

•          Sample space: collection of all outcomes of the experiment (SS)

Your sample space (SS) is your drawer, it doesn’t contain all the socks in the world but has the collection of all the outcome of the experiment.

•          An Event: an outcome from one experiment (E)

An event would be the socks you ended up picking.

•          A Sample:  a collection of events from repeated trials/experiments (S).

If you randomly pick socks for a week those seven pairs of socks would be a sample.

•          Random variable:  function X(E) assigning one real value to each element from a sample space  ( features? )

A random variable for the socks could be a description of the socks like its style (athletic or business type socks) or the color.  

•         Probability:  p = P(E), a number assigned to event representing the fraction of experiments resulting with the event/outcome A ==> p(x) = P(x = X), the fraction of time a random variable x = X

In the sock example, you are planning on wearing black pants so you want black socks, so you would want to know the probability of getting a pair of black socks. Let’s say you have 20 pairs of socks and 11 of them are black so p(x=black) would be .55. 

•         Probability Density Function: f(x) = p(x=X) on SS.


My sock example falls apart when I start describing pdfs but these are normally represented as graphs showing the probability of something happening based on different outcomes. Think of a grades in class, so 100 students take a test 10 get As on the test, 20 get Bs, 45 get Cs, 15 get Ds and 10 get Fs. The pdf would have p(x=A) = 0.1, p(x=B) = 0.2, p(x=C) = 0.45, p(x=D) = 0.15, and p(x=F) = 0.1.

 

•         Probability Distribution Function: F(x) = p(x<=X) on SS.

A probability distribution function also called a cumulative density function (cdf) shows the probability of something happening cumulatively as the values increase so for the grades example if F is the lowest grade and A the highest, p(x <= F) = .1 (notice stays the same), p(x <= D) = .25 (or p(x=D) + p(x=F)), p(x <= C) = 0.7 (or p(x=C) + p(x=D) + p(x=F)), p(x <= B) =0.9 (or p(x=B)+p(x=C) + p(x=D) + p(x=F)), and p(x <= A) = 1.0 (or cumulative value of all the outcomes)

•         In summary, shown in the bottom figure, a sample space (SS) is all possible outcomes, a sample (S) is made up of experiments (E), the sample is used to model the SS with variables (X) to form a pdf (f(x))