| 摘要 |
<p>1393082 Secret transmission GENERAL ELECTRIC CO Ltd 13 April 1972 [15 April 1971] 9559/71 Headings H4R and H4L In a secret transmission system a first pseudorandom digital pulse sequence is linearly added to an analogue information signal whereafter a second pseudo-random digital pulse sequence is added to the combination signal and the final resultant is transmitted over a line or radio link. It will be helpful if reference is first made to Fig. 1, which relates to the prior art. In the transmitter of Fig. 1, an input voice frequency signal at microphone 10 is differentiated at 12 in order to increase the number of zero-crossings, particularly of high frequency components, and these are detected by a high gain saturating amplifier or Schmitt trigger 14 to provide an output binary signal. The latter is supplied as one input of a module two adder 16 whose other input is obtained from a pseudo-random pulse generator 18 having a bit rate of 2,000 bits per second. The output of differentiator 12 is also supplied to an envelope amplitude detector consisting of a rectifier 20 and low pass filter 22 to provide a unidirectional signal indicative of the voice signal amplitude. The unidirectional signal is then linearly added to or, more preferably, amplitude modulated on the pulse output from adder 16 to provide a composite signal for application, possibly via a low pass filter for reducing its bandwidth, to a conventional radio transmitter 26. At a receiver (Fig. 2, not shown), the amplitude information is retrieved from the incoming signal so as to amplitude modulate the resultant of adding a synchronized pseudo-random pulse sequence to the incoming pulse signal in another module two adder (which thus removes the masking pulse sequence). Whereafter the signal is integrated and applied via a low pass filter to a receiving transducer. In the present invention the amplitude information is itself subjected to masking by another pseudo random pulse sequence. In the receiver, Fig. 4, of this arrangement, the incoming signal is converted to a binary sequence by limiting amplifier 16 and then added to the (frequency) masking pulse sequence provided by generator 64 in a module two adder 62. Amplitude information is recovered by module two adding the decoded output of adder 62 to the amplitude masking pulse sequence from generator 68 so that after correcting 70, 72 for the inherent non-linearity of this method of coding an amplitude controlling signal can be combined with the output of adder 42 whereby after integration 44 and filtering 46 the required audio output is obtained. It may be noted that in the absence of a speech representative signal in the input, the output of adder 66 is all zero, but as the speech level increases the pseudorandom sequence becomes suppressed and the output of adder 66 approaches a 50% "one" level state. In the transmitter the first (amplitude) pulse sequence is added to the voice signal after it has been differentiated but before it has been digitized but as it is at a very low level it is effectively suppressed from the voice signal when the latter has a high amplitude. The pseudo-random generators may be shift registers whose inputs are fed from the outputs of exclusive OR gates connected to the last and one other stage of each register. Synchronization is effected by comparing an input pulse sequence, in the absence of speech, with the locally generated pulse sequence produced by a generator having an unconnected input, until coincidence is maintained for a predetermined period whereupon the input connection is restored. The pulse sequences may have three or more levels and they may be added or subtracted to the input signal.</p> |
