GB2063614A - Detecting echoes - Google Patents

Abstract

The output of an echo signal receiving amplifier 10 is compared in a comparator 12 with a threshold value, the result of comparison being latched at regular intervals and entered by way of a latch 14 into a shift register 16. The contents of the shift register are transferred by a latch 18 to a full adder 20 for application to a clocked memory 22. <IMAGE>

Description

SPECIFICATION Signal receiver The present invention relates generally to a signal receiver employing electronic techniques for signal enhancement and retrieval. The invention is particularly concerned with the detection of weak echo pulses resulting from refietion of repeatedly transmitted signals.

The receiver of the present invention is intended to detect and retrieve repetitive electronic signals the ampiitude of which is smaller than that of the prevailing noise level.

According to the present invention, there is provided a method of detecting echoes of a repeatedly transmitted signal which comprises dividing the time following the transmission of each pulse into a predetermined number of windows, sampling the received signal during each window to produce a single binary bit indicative of whether the signal level lies above or below a predetermined value and accummulating the binary information in respect of each window in a respective store.

The invention is predicated upon the fact that the noise component of the received signal has equal probability of lying above and below zero. An echo signal caused by reflection of the transmitted signal from a boundary to be detected will occur always at the same time after transmission of a pulse and will therefore displace the mean level of the noise signal within the respective window.

Assuming, for example, a binary 1 indicates a positive value of r'eceived signal while a binary zero indicates a negative value. If the noise distribution is symmetrical about zero, then the register associated with each respective window will contain a number which will on average be one half of the number of complete cycles of transmitted pulses, since the binary bit corresponding to that time window will be 1 for only 50% of samples. Thus, for N samples or cycles the numbers stored would be expected to be N/2. The probable random deviation or noise increases as 4. If a signal exists which is coherent with the reference pulse and appears repetitively in one of the time windows, it will cause a shift in the probability distribution giving rise to a probable deviation in the stored number which is proportional to N the number of samples.Thus, the ratio signal to noise in the stored number increases in proportion to fland the signal is enhanced by repeated sampling.

The response curve of the system to signals of different amplitudes depends on the precise noise distribution. For a normal, that is to say Gaussian, distribution, the response will approximate to the error function n - =+(1 +erf(Vsig)) N where erf (Vsig)=

Since the samples are integrated digitally, the degree of enhancement is in principle unlimited. In practice, this will depend on the time available for sampling and the stability of the noise distribution.

A feature of the system is that no analogue information is retained from the actual signal received. This is justified so that the majority of analogue information consists of noise. This removes the need for multi-bit analogue-to-digital conversion at the input and therefore speeds up the process.

The invention will now be described further, by way of example, with reference to the accompanying drawing, which is a schematic and circuit diagram of a receiver in accordance with the invention.

In the drawing, there is shown a clock circuit 11 which produces a reference pulse at regular intervals. The reference pulses are synchronised with the transmission of a signal which may itself be either electromagnetic or acoustic. The clock circuit 11 also produces timing pulses at an output 1 3 which control the timing of the other circuitry to be described below.

The received signal is applied to a linear radio frequency amplifier 10, the output of which is fed to a fast comparator 12, for example a Plessey SP1651. The latter comparator may be latched for a period of 10 ns at intervals of 20 ns to sample the amplifier output.

The samples may be grouped in series of, say, 64 sequential 20 ns time windows. This series of binary bits is sent in sequence to a four-bit parallel access shift register 16, for example, model 74S195. The shift register samples the comparator output during the time that it is latched. In this way, a series of four-bits are arranged in order and these may then be written at 80 ns intervals into a second four-bit latch, for example, an 82S91, designated 18 in the drawling. After each set of four bits has been entered into the latch 18, the next four bits are read by the shift register 1 6. Each of the four latched bits is added by an 8-bit full adder 20, which may for example be made of 27483's, and entered into an eight bit register of a 64 X 8bit memory 22 made for example of 8 x N3101A registers. The latter memory receives the reference pulse from the clock circuit 11.

The contents of each register may be examined at intervals and either passed to a digital-toanalogue converter for direct display or to further storage for subsequent processing.

A feature of the described arrangement is that the system may be interleaved in time, to provide a closer spacing of samples of the incoming signal.

Claims (2)

1. A method of detecting echoes of a repeatedly transmitted signal which comprises dividing the time following the transmission of each pulse into a predetermined number of windows, sampling the received signal during each window to produce a single binary bit indicative of whether the signal level lies above or below a predetermined value and accumulating the binary information in respect of each window in a respective store.

2. A method of detecting echoes substantially as hereinbefore described, with reference to and as illustrated in the accompanying drawings.