The invention relates to sensor systems and methods. More particularly but not exclusively it relates to a rangefinder system and method. Preferably, the invention relates to a laser rangefinding system and method.

The main limitation with current laser rangefinding techniques is the need for a constant repetition rate to maintain a temporal correlation between the emitted pulse of the laser and a detected photon reflected from a target. In its simplest form, a single pulse is emitted and the elapsed time to detect the reflection determines the range. The measurement might be repeated in rapid succession with a burst of pulses to improve accuracy by averaging and to reduce the effect of scintillation. The repetition rate of the burst must be longer than a maximum value, f<c/2R, where R is the actual range, to obtain the absolute range. If f exceeds this value then the absolute range can be determined only by counting the number of pulses that have been emitted before detection of the return. The range is then given by R=mc/2f+5R, where m is an integer, f the pulse repetition frequency and 5R is an offset determined by the temporal correlation. However systems are known using an absolute rangefinding method using random, pseudorandom or non-repeatable pulse patterns to obtain the absolute range directly. In an earlier version of the technique the range is determined by calculating the auto (cross) correlation between the outgoing pulse stream and the returned (detected) pulse stream. This means that a method must be used to generate a known pulse pattern either by preprogrammed pattern generation or by direct detection of the transmitted pattern. These approaches require an optical modulator or pulse pattern driver. In this current version, a pattern waveform generator and optical modulator are not required.

An existing system is disclosed in GB 2306825A - Laser ranging using time correlated single photon counting , WALLACE ANDREW M; MASSA JOHN S; BULLER GERALD S; WALKER ANDREW C, Heriot-watt University; published 1997-05-07.

The invention aims to overcome the problems with existing systems.

According to the invention there is provided a range finder in which radiation pulses are derived from a pulsed radiation source which produces a pulse train at a constant repetition frequency.

According to the invention there is further provided a method of range finding in which the range to an object is determined by measuring the degree of temporal correlation between an emitted pulse and the detection of a single photon in a returned (reflected) pulse. The invention will now be described with reference to a fixed wavelength laser source as this is the most practical as the laser emission is directional and the wavelength is stable with temperature. However, it will be appreciated that any suitable radiation source may be used.

In this present invention, use is made of the physical nature of all low-power, weak, light beams. The number of photons per second (i.e. the photon flux) fluctuates about a mean value according to Poisson statistics. The Poissonian probability distribution gives the number of photons n expected in a time interval 5t for a mean arrival rate (average power) of μ. For n=0 the photon stream is discontinuous and gaps appear in the stream, thus discrete pulses of light in the beam can be identified. These light pulses are random in their length, their separation in multiples of 5t and in the number of photons in the pulse (energy or intensity). The light beam may therefore be thought of as a pattern of random pulses. This natural characteristic of light can be used directly as a laser rangefinder. The optical power required to make use of this natural pulsing depends on the recovery time (or deadtime) 5t _d of the photon counting system (comprising data acquisition electronics and a single photon detector) and the mean power of the beam. In fact, 5t _d is the time interval that is used to characterise the beam as a random pulse stream. For typical values of 5t _d , the average power is <lnW. This is so low that it is virtually impossible to detect and differentiate it from natural solar background (noise). This is because the pulsed beam is effectively Poissonian noise.

This presents several advantages:

1. no optical modulator is required. The beam is naturally "pulsed".

2. the beam is virtually impossible to detect and therefore covert and eyesafe. Covert operation provides the capability to range on a target without detection (i.e. without triggering a laser warning receiver).

3. the technique can be used for covert optical communications.

Clearly the returned beam must be detected in order to calculate the auto(cross) correlation and determine the range. The transmitted beam must also be measured because it is unknown due the random fluctuations of low light (Poissonian) light. Mathematical analysis of this process has been conducted and design rules formulated for ensuring optimum performance for the range precision and operational distances of interest.

Although the principle applies to any light source, a fixed wavelength laser source is most practical because the laser emission is directional and the wavelength is stable with temperature. This allows very narrow pass band filters to be used to reduce the out-of-band solar background and increase the signal to noise ratio.

The average power used is determined by the analysis below:

Calculation of Bit Sequence Generated by Low Power Poissonian Noise. Consider a photon bit sequence where each bit is a laser pulse of duration 5t. It is assumed that there is no solar background and no detector noise. The probability distribution of bits (pulses) containing n photons is given by;

Since the bit sequence is generated by noise it is unknown and must be measured. The detection of a photon destroys it, therefore bits containing n=0 and 1 photons are effectively logic 0 bits ¹ . The bit sequence has to be detected a second time on its return, therefore only bits (pulses) containing n>l photons are useable and, once detected, represent logic 1 in a digital bit stream.

The average photon flux (power) that generates a sequence for a given ratio of Is to Os (i.e. the sparsity) may be calculated. It is assumed that the quantum efficiency of the detector q is unity. From (1) the probability that a bit contains n=0, 1, 2, etc photons is: ) μδ

P(0,a - = e

1!

Because the bit sequence has to measured a second time on its return. P(n > l,St) = \-έ ^~μδ - {μά) ^β~μδ (j. _e . the total probability that a bit contains n>l photons).

The sparsity a is the ratio of logic 1 bits to logic 0 bits and is the fraction of bits (pulses) containing n>l photons in a sequence m-^ ⁰⁰ . For example the sequence [01260] has sparsity 2/5=40% for m=5. Therefore, for m->°°: p(n>\,a)

(o, a) + P(I, a)+p(n,a) ^{~ a nd}

l = p(, a) + P(I, a) + p(n >\,a).

In the limit that μδΐ->0, then:

(μά) ² =α. (2) Since e ^"x =l-x to within ~2% for χ=μδΐ=0.2, hence a~4%. Therefore for a sparsity of ~4% and 8t =10ns, which is an average power of ~4pW for λ=1 micron. For 6t=lns the average power must be less than ~40pW.

We now consider the quantum efficiency explicitly where 0<q<l. The quantum efficiency is the probability that a photon will be detected by a detector and excludes the coupling and transmission efficiencies of the transceiver. We define the detection of a pulse as the detection of a single photon in the pulse (i.e. there is a finite deadtime and the detector does not recover quickly enough to detect a second photon in the same pulse). For a pulse of n photons, the probability of detecting the pulse P _d is: P _d =q + (l-q)q + (l-q) ² q + (l-q) ³ q + + (ΐ-? 7 = l~(l-q)

Table 1 shows a breakdown of the number of pulses detected for a sequence of length m.

Table 1

Using (1), the number of detected pulses N _d containing n>l photons that a detected is:

-μδ

(4) n=2 ^η ·

Using:

⁰⁰ x

?* = l + x + y—

n=2

N _d = πιέ- ^μδ ^ ^μδ - 1 - μά - β ^{ι ~ ^ς)μ + 1 + (l - ς)μ&}

Note that as qf- 0, N _d 0 as expected, and as

N _d → me ^' "* ^ ^μδ - (l + μά)}.

For μδΐ -> 0 then N _d -> 0. Calculation of Bit Errors

The total number of bit errors in a detected sequence is the sum of:

1. the number of logic 0 bits of photon number n<2 recorded as logic 1 due to the solar background. The mean photon number (average optical power) is μ=Γμι+ο, where r is the round trip loss due to diffraction, scattering and absorption; μι is the mean transmitted power and b is the mean solar background power in-band to the detector.

2. the number of bits with photon number n>l (i.e. logic 1) which are not recorded in the transmitted sequence due to q<l.

3. the number of bits with photon number n>l (i.e. logic 1) which are not recorded in the received sequence due to q<l.

Accuracy

The accuracy 6R _P that can be achieved is set by 5t, the effective deadtime of the detector and data collect equipment since this parameter determines the timing resolution and 6R _P =5tc/2. For the case of continuous time-stamping (or time-tagging), photons are "stamped" with their time-of-arrival and stored in a memory buffer for signal processing. The time stamping accuracy of modern data acquisition hardware is <100ps, and considerably less than the deadline of single photon avalanche detectors (~10ns) or micro-channel plate, photomultiplier tubes (~lns). A range accuracy of lm requires a deadtime of better than 6ns.

If the bit pattern is considered to be a continuous stream of time intervals of duration, 5t, then the auto (cross) correlation has a minimum width of 5t. This minimum value can only be achieved with sparse waveforms which are long enough to provide auto (cross) correlation gain above the ambient noise level.

In this way laser pulses are derived from a continuous laser which produces a random pulse train due to the Poissonian nature of low intensity light. The range to an object is determined by measuring the cross-correlation between the emitted pulse train and the detection of the returned (reflected) pulse train. The accuracy of the range is determined by the time stamping resolution.

It will be appreciated that any thermal light source with a sufficiently narrow linewidth may be used. Advantageously, this means eyesafe systems can be contemplated. It will further be appreciated that this system has commercial applications in the following applications: Eyesafe laser rangefinding, terrain mapping, 3D lidar, high altitude mapping, satellite-based mapping of the earth's surface, as well as combined covert optical communications for, special forces and operations.