I am having a problem implementing an encryption system on a speech signal. The input is a speech sequence that is converted to a digital signal where the encryption is to be performed. When the encryption is performed the signal is converted back to an analogue signal. Unfortunately nearly all the literature on the internet concerns digital coding with a digital transmission line, where I have an analogue transmission line - any ideas for a robust system where the encryption can be implemented using a key to encrypt and decrypt the signal? The level of security is not critical, but ~10.000 different keys should be available (A simple algorithm with low security is preferred instead of complex algorithm with high level of security). Regards Ole Thomassen
Ideas for a scrambling system...
Started by ●July 23, 2003
Reply by ●July 23, 20032003-07-23
Thomassen wrote:> > I am having a problem implementing an encryption system on a speech > signal. The input is a speech sequence that is converted to a digital > signal where the encryption is to be performed. When the encryption is > performed the signal is converted back to an analogue signal. > Unfortunately nearly all the literature on the internet concerns > digital coding with a digital transmission line, where I have an > analogue transmission line - any ideas for a robust system where the > encryption can be implemented using a key to encrypt and decrypt the > signal? > > The level of security is not critical, but ~10.000 different keys > should be available (A simple algorithm with low security is preferred > instead of complex algorithm with high level of security). > > Regards > Ole ThomassenWhat is a digital transmission line? I've never seen one. A digital signals can be sent on an analog line with a modem at each end. Why won't that work for you? Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●July 23, 20032003-07-23
"Jerry Avins" <jya@ieee.org> wrote in message news:3F1E8E3C.B6B24551@ieee.org...> Thomassen wrote: > > > > I am having a problem implementing an encryption system on a speech > > signal. The input is a speech sequence that is converted to a digital > > signal where the encryption is to be performed. When the encryption is > > performed the signal is converted back to an analogue signal. > > Unfortunately nearly all the literature on the internet concerns > > digital coding with a digital transmission line, where I have an > > analogue transmission line - any ideas for a robust system where the > > encryption can be implemented using a key to encrypt and decrypt the > > signal?(snip)> What is a digital transmission line? I've never seen one. A digital > signals can be sent on an analog line with a modem at each end. Why > won't that work for you?I think that is what he is asking about. But consider that the analog side of the modem is probably going through a channel with the same bandwidth as the digitized signal. I am assuming a telephone scrambler system. I believe that the modern day system is to compress the digital signal so that a more common modem can get the signal through. A 64K b/s digitized voice signal should compress to less than 28K, then put through the encryption system, and an ordinary 28.8K modem. -- glen
Reply by ●July 23, 20032003-07-23
Jerry Avins <jya@ieee.org> wrote in message news:<3F1E8E3C.B6B24551@ieee.org>...> Thomassen wrote: > > > > I am having a problem implementing an encryption system on a speech > > signal. The input is a speech sequence that is converted to a digital > > signal where the encryption is to be performed. When the encryption is > > performed the signal is converted back to an analogue signal. > > Unfortunately nearly all the literature on the internet concerns > > digital coding with a digital transmission line, where I have an > > analogue transmission line - any ideas for a robust system where the > > encryption can be implemented using a key to encrypt and decrypt the > > signal? > > > > The level of security is not critical, but ~10.000 different keys > > should be available (A simple algorithm with low security is preferred > > instead of complex algorithm with high level of security). > > > > Regards > > Ole Thomassen > > What is a digital transmission line? I've never seen one. A digital > signals can be sent on an analog line with a modem at each end. Why > won't that work for you? > > JerryI haven't seen an analogue line neither, though what I meant by an analogue line was that the output stream of the system is to be send over a wireless connection (FM) and due to the construction of the transmission and receiver chain it is not possible to send a digital "0" or "1", as more or less can be done on a cable connection. Everything that passes through the encryption block is to be modulated before transmitted and reverse in the receiver. I hope this clarifies the mystery... Regards Ole Thomassen
Reply by ●July 23, 20032003-07-23
Thomassen wrote:> > Jerry Avins <jya@ieee.org> wrote in message news:<3F1E8E3C.B6B24551@ieee.org>... > > Thomassen wrote: > > > > > > I am having a problem implementing an encryption system on a speech > > > signal. The input is a speech sequence that is converted to a digital > > > signal where the encryption is to be performed. When the encryption is > > > performed the signal is converted back to an analogue signal. > > > Unfortunately nearly all the literature on the internet concerns > > > digital coding with a digital transmission line, where I have an > > > analogue transmission line - any ideas for a robust system where the > > > encryption can be implemented using a key to encrypt and decrypt the > > > signal? > > > > > > The level of security is not critical, but ~10.000 different keys > > > should be available (A simple algorithm with low security is preferred > > > instead of complex algorithm with high level of security). > > > > > > Regards > > > Ole Thomassen > > > > What is a digital transmission line? I've never seen one. A digital > > signals can be sent on an analog line with a modem at each end. Why > > won't that work for you? > > > > Jerry > > I haven't seen an analogue line neither, though what I meant by an > analogue line was that the output stream of the system is to be send > over a wireless connection (FM) and due to the construction of the > transmission and receiver chain it is not possible to send a digital > "0" or "1", as more or less can be done on a cable connection. > Everything that passes through the encryption block is to be modulated > before transmitted and reverse in the receiver. I hope this clarifies > the mystery... > > Regards > Ole ThomassenThanks. As Glen wrote, compress�>encrypt�>modulate/transmit�>receive/demodulate�>decrypt�>decompress. You may not need compression if the modem speed is high enough and the bandwidth can be made low enough. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●July 23, 20032003-07-23
A pretty classic way for analog scrambling is to invert the audio spectrum around a pseudo random jumping frequency. For voice encryption, the jumps may occur with a rate of about 20 times/sec. The transmitter and receiver should be synchronized. That is easily achieved using sub band or upper band signaling. The other classic way is scrambling in the time domain. The signal is split into time windows of about 50ms long, which are transmitted in a pseudo random order. That also requires synchronization, and introduces an interleaving delay. Perhaps, the simplest yet more or less reliable way of analog scrambling is to split the spectrum into several bands, then shuffle and invert those bands. You do not need any synchronization. All of theese methods can be broken by a sophisticated analysis. However they are good enough against an amateur eavesdropper. Vladimir Vassilevsky, Ph.D. DSP and Mixed Signal Design Consultant http://www.abvolt.com Thomassen wrote:> > I am having a problem implementing an encryption system on a speech > signal. The input is a speech sequence that is converted to a digital > signal where the encryption is to be performed. When the encryption is > performed the signal is converted back to an analogue signal. > Unfortunately nearly all the literature on the internet concerns > digital coding with a digital transmission line, where I have an > analogue transmission line - any ideas for a robust system where the > encryption can be implemented using a key to encrypt and decrypt the > signal? > > The level of security is not critical, but ~10.000 different keys > should be available (A simple algorithm with low security is preferred > instead of complex algorithm with high level of security). > > Regards > Ole Thomassen
Reply by ●July 23, 20032003-07-23
"Vladimir Vassilevsky" <vlv@abvolt.com> wrote in message news:3F1F074D.32489481@abvolt.com...> > A pretty classic way for analog scrambling is to invert the audio > spectrum around a pseudo random jumping frequency. For voice encryption, > the jumps may occur with a rate of about 20 times/sec. The transmitter > and receiver should be synchronized. That is easily achieved using sub > band or upper band signaling.(snip of other analog methods) Yes, but in those days the digital circuitry available today didn't exist. Though such analog circuits may have gotten slightly cheaper, the digital circuits are now a LOT cheaper. As far as the digital/analog line question, yes, electrical cables are analog (a voltage on the cable). For short distances, one might claim no modulation, but usually it is modulated. There was a discussion claiming that ethernet was Manchester coded (and so digital), so I argued that, no, it was phase modulated with 0 and 180 degree phase allowed. Oh, what difference a few words can make. -- glen
Reply by ●July 24, 20032003-07-24
glen herrmannsfeldt wrote:> > "Vladimir Vassilevsky" <vlv@abvolt.com> wrote in message > news:3F1F074D.32489481@abvolt.com... > > > > A pretty classic way for analog scrambling is to invert the audio > > spectrum around a pseudo random jumping frequency. For voice encryption, > > the jumps may occur with a rate of about 20 times/sec. The transmitter > > and receiver should be synchronized. That is easily achieved using sub > > band or upper band signaling. > > (snip of other analog methods) > > Yes, but in those days the digital circuitry available today didn't exist. > > Though such analog circuits may have gotten slightly cheaper, the digital > circuits are now a LOT cheaper.So where is your problem then? 1. Convert the analog to digital. 2. Encrypt digital by whatever cipher you like. 3. Transmit encrypted data by any modem you like. All steps are pretty classic and straightforward. Numerous literature is awailable. Vladimir Vassilevsky, Ph.D. DSP and Mixed Signal Design Consultant http://www.abvolt.com
Reply by ●July 24, 20032003-07-24
Vladimir Vassilevsky <vlv@abvolt.com> wrote in message news:<3F1F4C05.A5D2E9AE@abvolt.com>...> glen herrmannsfeldt wrote: > > > > "Vladimir Vassilevsky" <vlv@abvolt.com> wrote in message > > news:3F1F074D.32489481@abvolt.com... > > > > > > A pretty classic way for analog scrambling is to invert the audio > > > spectrum around a pseudo random jumping frequency. For voice encryption, > > > the jumps may occur with a rate of about 20 times/sec. The transmitter > > > and receiver should be synchronized. That is easily achieved using sub > > > band or upper band signaling. > > > > (snip of other analog methods) > > > > Yes, but in those days the digital circuitry available today didn't exist. > > > > Though such analog circuits may have gotten slightly cheaper, the digital > > circuits are now a LOT cheaper. > > > So where is your problem then? > > 1. Convert the analog to digital. > 2. Encrypt digital by whatever cipher you like. > 3. Transmit encrypted data by any modem you like. > > All steps are pretty classic and straightforward. > Numerous literature is awailable. > > Vladimir Vassilevsky, Ph.D. > > DSP and Mixed Signal Design Consultant > > http://www.abvolt.comMy problem is that I have not found a proper method for the encryption. I have found some literature concerning digitally encryption that can be used when a line is available, but most of these methods cannot be used in my system. Furthermore most of these functions are without various user defined encryption keys. Regards Ole Thomassen
Reply by ●July 24, 20032003-07-24
Thomassen wrote:>...> > My problem is that I have not found a proper method for the > encryption. I have found some literature concerning digitally > encryption that can be used when a line is available, but most of > these methods cannot be used in my system. Furthermore most of these > functions are without various user defined encryption keys. >It might help if you described why the various encryption methods you did find don't meet your needs. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
