Encryption of analog signals

[sandeep polasa]

Hi, i am doing a project on encryption and decryption.
For my project i need to encrypt and decrypt DTMF(Dual Tone Multi Frequency) signals.
These are analog signals in voice frequency range.
Can anybody suggest me how to do encryption of DTMF signals?

 

[zorro]

A common method used for voice is spectrum inversion: the signal is shifted in frequency (DSB modulation with a carrier of frequency near the upper end of the voice band) and then filtered. Decription is performed in the same way.
For DTMF, frequency translation can work as well: SSB with a low-frequency carrier.
Do you need an all-analog solution, or can use digital signal processing?
Regards

Z

 

[betwixt]

I have to ask.... why do you want to do this?

There are only 16 DTMF tones so it wouldn't take a genious to break the encryption, especially as spectral inversion produces the same output each time the same tone pair is decrypted. Encrypting the original numbers used to select the DTMF would be a far more secure system as it allows for rolling codes.

Brian.

 

[zorro]

There are only 16 DTMF tones so it wouldn't take a genious to break the encryption, especially as spectral inversion produces the same output each time the same tone pair is decrypted.

Of course, these are not secure methods. In this context, the term "encription" means (al least in my understanding) something that hiddens the information to standard DTMF decoders.

Encrypting the original numbers used to select the DTMF would be a far more secure system as it allows for rolling codes.

Of course, but that would not be "encrypt and decrypt DTMF(Dual Tone Multi Frequency) signals" as the original post asks, explaining that "These are analog signals in voice frequency range."

Z

 

[betwixt]

That's why I asked Zorro. I presumed the reason for obscuring the tones is so they can't be decoded but that begs the question 'why?'. The easiest analog method, assuming a standard DTMF transmitter and receiver are being used it to change the clock frequency on them both but if the intention is to hide something, the encryption needs to be more than a frequency shift or inversion. The simpest digital method is to swap the data bits around but all these methods, analog and digital give a 1:1 relationship of input data and output data so they are easy to interpret.

Brian.

 

[thylacine1975]

One generic approach to "securing" analog signals is by masking (obscuring) them with a bigger signal. If the receiver can strip (subtract away) the bigger signal, the original message can be recovered.
I've previously worked with chaotic masking systems which use a chaotic "random" signal generator as the masking signal, and frequently rely of self-synchronising chaotic receiver circuits to locally regenerate the masking signal for subtraction.

If you Google "chaotic masking" you'll find some good leads and there's practical examples using Chua's circuit and implementations of Lorenz' equations that can be easily replicated with a handful of opamps.