DE102024115600A1 - Method for operating a lidar device - Google Patents

Abstract

The present invention relates to a method for operating a lidar device, wherein in the course of the method the lidar device (12) emits optical signals (18), receives optical signals (20) reflected back from objects in an environment, and distances between the lidar device (12) and the objects are determined from the transit times (22) of the received signals (20).
According to the invention, the optical signals (18) to be transmitted are generated by creating digital sequences (24) of a first period (T1), each containing a header with a transmitter ID and a timestamp. The digital sequences (24) are combined with a digital spreading sequence (26) to obtain combined digital sequences (42). An analog optical carrier signal is modulated using the combined digital sequences (42) to obtain the resulting optical signals (18), which are transmitted with the first period (T1). The received back-reflected optical signals (20) are demodulated and separated using the inverse spreading sequence to determine the transmitter ID and the timestamp. At least one channel estimation is performed using each received sequence.

Description

The present invention relates to a method for operating a lidar device according to the preamble of claim 1.

In a generic method for operating a lidar device, the lidar device emits optical signals and receives optical signals reflected back from objects in its environment. Distances between the lidar device and the objects are determined from the signal travel times.

This system can include a front end of the vehicle, in which the lidar device is installed, that scans a detection area defined by the lidar device and emits optical signals into this area. Laser beams reflected from objects in the vehicle's vicinity are also received by the vehicle's front end and evaluated by a processing unit.

Instead of measuring the transit time of individual pulses, a continuous, periodic signal with a training sequence can also be transmitted. The signal transit time can then be determined from the time difference between the transmitted and received signals. Additionally, classical digital channel estimation methods can be used to determine the channel characteristics with respect to amplitude, phase, and polarization, which in turn provide information about the environmental conditions (e.g., rain or fog, properties of the reflected object, e.g., fabric or metal).

Various problems can arise when determining distances using known methods for operating a lidar device.

If periodic optical signals with a sequence length corresponding to one period are emitted, then for propagation times greater than one sequence length, the propagation time of the signal (and consequently the distance to the object in the vicinity) can no longer be uniquely determined. This problem will be further addressed below using the following examples: 1 This will be explained in more detail later. Therefore, either very long sequences with long periods must be transmitted, which involves additional signal processing and leads to very slow distance estimates. Alternatively, an upper limit can be set for determining travel times and thus for distance measurement, which reduces the detection range of the lidar device.

Problems can also arise if another lidar system is simultaneously scanning the same detection area. Such an additional lidar device might be installed on the same vehicle or a different vehicle. Optical signals received from one of the first lidar systems cannot then be reliably separated. A "foreign" lidar signal can be received and detected, leading to an incorrect propagation delay calculation.

It is therefore the object of the present invention to provide a method by which these problems can be reliably solved.

This problem is solved according to the invention by a method according to the characterizing feature of claim 1.

In a method according to the invention, digital sequences with a first period are generated, each containing a header with a sender ID and a timestamp. Providing a sender ID offers the advantage that reflected signals from the "own" lidar device can be identified and separated from signals from other lidar devices. Providing a timestamp can also serve to distinguish one's own signals from those of other devices, since the lidar device knows when an optical signal was actually transmitted, allowing received optical signals with an unknown timestamp to be disregarded. Furthermore, providing a timestamp offers the advantage that even optical signals with a long propagation delay, particularly delays greater than the period of a transmitted optical signal, can be reliably evaluated. A timestamp thus allows determination of whether a propagation delay is present that is, for example, several times the period of a transmitted optical signal, since the transmission time is known.

However, simply sending a digitally modulated optical signal consisting of sequences with ID and timestamp is not particularly suitable for digital channel estimation or precise transit-time determinations. Therefore, according to the invention, the digital sequence is combined with a periodic signal as follows.

According to the invention, the digital sequences are combined with a digital spreading sequence to obtain combined digital sequences. These combined digital sequences correspond to a baseband signal. An analog optical carrier signal is then modulated using the combined digital sequences to... to obtain optical signals which are emitted with the first period.

The received reflected optical signals are demodulated and separated to determine the transmitter ID and timestamp.

At least one channel estimate is performed for each received sequence. This allows the information about the transmitter ID and the timestamp to be evaluated.

Thus, each bit with a duration of TB1 is combined with a digital, periodic spreading sequence, for example, by multiplication. The spreading sequence consists of a number of chips with a duration of TB2. The sum of all chips in a spreading sequence can correspond exactly to one bit duration of TB1. This multiplication can be calculated digitally at the transmitter. The resulting signal is then optically modulated and transmitted.

The periodic chip sequence can advantageously be designed such that it can be optimally detected at the receiver in order to synchronously recombine the received signal with the corresponding spread sequence, for example, by multiplying it. The chip sequence can also be optimized to suppress foreign signals during spread estimation when multiple signals are simultaneously acquired using different chip sequences. Furthermore, the chip sequence can be optimized for periodic channel estimation. Optimization with respect to two or three of these properties is also possible.

At the receiver, the optical signal is demodulated and digitally detected, followed by synchronization of the clocks for the bit and chip sequences. "Detection" here does not refer to the conversion of an optical signal into an electrical signal by a sensor. Rather, detection here means the unfolding or separation of the sequence and the spread sequence. A detailed propagation delay difference ΔT can be determined from the offset between the transmitted and received chip sequences. First, the received signal is unfolded with the corresponding chip sequence to determine the transmitted information from the ID and timestamp. The propagation delay can then be corrected using the timestamp by integer multiples of the bit duration TB1.

Furthermore, according to the invention, at least one channel estimation is performed using each received sequence. Channel estimation—as is known, for example, from wireless communication—determines the channel characteristics of a connection. Here, a known signal is transmitted, and changes in the received signal are analyzed with knowledge of the transmitted signal. In the case of a lidar device, this is the propagation channel of the optical signals from the lidar device to the object in the environment and, after reflection from the object, back to the lidar device. The header and the timestamp preferably correspond to the transmitted information from the "sender to the receiver." Since the transmitter and the receiver are integrated side-by-side in the lidar device, the transmitter information is already known at the receiver, which is utilized within the scope of the invention. The time offset of the transmitted signal and the received signal can thus be determined relative to a system time. Additionally, this allows interference and propagation delays in signal processing and modulation at the transmitter and receiver to be calibrated out.

Various channel properties can be determined in this way, such as the phase response and attenuation of a signal. Changes in the polarization of an optical signal can also be determined. The channel estimation therefore contains all channel information, including attenuation, phase (dispersion), polarization, and propagation time. However, a particularly important parameter for the present method is precisely the propagation time of the signal. Thus, as an alternative to using the timestamp, or preferably in addition to using the timestamp, the propagation time obtained from the channel estimation can be used to determine the distances between objects.

Since at least one channel estimate is performed for each received sequence, the information obtained through channel estimation is continuously updated. If the transmission of the channel estimation sequences occurs significantly faster than the time-varying changes in the channel or the spatial sampling, disturbances such as noise can be suppressed by averaging successive channel estimates.

Because the signal can preferably be transmitted and received continuously, there are no interruptions in channel estimation (distance determination) and spatial scanning.

Preferred embodiments of the inventive process are set out in the dependent claims.

Preferably, the sequences to be transmitted consist of a first sequence of digits. The first bit has a certain length. The spreading sequence consists of a second sequence of digital bits and has a second period.

Preferably, the period of the spreading sequence is equal to the bit length of the sequence. This means that each bit of the sequence is combined with the complete spreading sequence.

Preferably, a channel estimate is performed with each bit, i.e., with each received chip sequence. This significantly increases the number of channel estimates performed within a given time interval, since a channel estimate is now performed not just with each received sequence, but with each received bit of a received sequence. Depending on the number of bits used per sequence, the number of channel estimates is thus multiplied, which can be used, for example, to perform averaging over different channel estimates. This allows signal propagation times to be determined even more reliably. When multiple received sequences are averaged during high-bit-rate transmission, the signal-to-noise ratio is improved. Coherent reception (a necessary requirement for a lidar device) also inherently increases the signal-to-noise ratio compared to pure power detection.

In a further advantageous embodiment of a method according to the invention, the combination includes multiplying each sequence by the spreading sequence. A corresponding inverse spreading sequence can be used to detect a received signal. It can also be provided that a received sequence is multiplied by the original spreading sequence to perform a detection.

A particularly preferred method according to the invention is one in which both channel estimation and timestamp are used to determine the propagation time of the received optical signal. Preferably, the timestamp is used to determine how many integer periods of a sequence the propagation time comprises. The channel estimation can be used to determine the additional time offset (+/-) less than one period of a sequence that the propagation time includes. In other words, the timestamp is used to roughly estimate the distance to an object at intervals of one period, in integer multiples of that period. The channel estimations can be performed at much shorter intervals, for example, with each bit of a received sequence, and thus much more frequently than the evaluation of the timestamp, which requires receiving a complete sequence. Since channel estimation is used to determine the time offset smaller than a period of a sequence that additionally includes the propagation delay, the propagation delay is recalculated with each bit of a received sequence in terms of how it has changed compared to the propagation delay determined from the last timestamp (as a rough estimate in multiples of the period).

A further preferred method according to the invention is one in which the spreading sequence includes a CAZAC sequence. Such CAZAC frequencies or signal shapes (Constant Amplitude Zero Auto Correlation) are known from communications engineering and are particularly optimized for performing channel estimation.

In a further preferred embodiment of the method according to the invention, different spreading sequences are used for different lidar devices, each spreading sequence potentially including an individual gold sequence. Such gold sequences are used in telecommunications and satellite navigation. While gold sequences are somewhat less suitable for channel estimation, performing channel estimation is still possible. The particular advantage of using gold sequences is the fact that such gold frequencies are uncorrelated, allowing signals from different lidar devices with different gold sequences to be reliably separated.

Overall, the spreading sequence can therefore be optimized with respect to its properties for channel estimation in digitally coherent lidar receivers or with respect to multi-user interference. Alternatively, the spreading or channel estimation sequence can also be systematically or statistically modified over time to avoid multi-user interference when multiple lidar devices are present, thus eliminating the need for systematic coordination of the sequences of all users.

The invention will be explained in more detail below using the figures as examples.

• 1 ein Fahrzeug mit einer Lidar-Vorrichtung;
• 2 und 3 vereinfacht dargestellte Beispiele für ausgesendete und empfangene optische Signale; und
• 4 die erfindungsgemäße Kombination eines digitalen Signals mit einer Spreizsequenz.

This shows:

• 1 a vehicle equipped with a lidar device;
• 2 and 3 Simplified examples of transmitted and received optical signals; and
• 4 the inventive combination of a digital signal with a spreading sequence.

1 Figure 1 shows a vehicle 10 equipped with a lidar device 12, which has a front end 14 and a processing unit 16. The front end 14 of the lidar device 12 has a transmitter unit (not shown) that emits optical signals 18. These optical signals 18 are reflected by an object (also not shown) in the vicinity of the vehicle 10 and return as received optical signals 20 to the front end 14 of the lidar device 12, for example, to a receiver unit of the lidar device 12, which converts them into electrical signals. The processing unit 16 evaluates the electrical signals.

Based on the 2 and 3 In a highly simplified representation, transmitted optical signals 18 and received optical signals 20 are shown superimposed in their temporal sequence. The simplified representation depicts signals 18 and 20 as modulated analog signals in the form of a repeating sawtooth wave. In reality, in a method according to the invention, a digital sequence is used as the transmitted optical signal 18 (and similarly as the received optical signal 20), as will be shown below. 4 will be explained in more detail.

The 2 Figure 1 shows an emitted optical signal 18 represented as signal strength s(t) with a period T1 over time t. A received optical signal 20 is also represented as signal strength r(t) over time t. The optical signal requires a certain travel time ΔT (reference symbol 22) to travel from the vehicle 10 to the reflecting object and back to the lidar device 12 of the vehicle 10. This travel time 22 is represented by the received optical signal 20. 2 represented as a displacement of the sawtooth pattern relative to the emitted optical signal 18. The propagation time 22, and thus the displacement by ΔT, is in the situation of 2 so small that it is, in particular, smaller than the period T1 of the emitted optical signal 18. The corresponding path length can be calculated via the travel time 22 and the speed of light. In the lidar device, in the situation of 2 It can simply be assumed that the first sawtooth of the received optical signal 20 corresponds to the first sawtooth of the emitted optical signal 18, so that the displacement ΔT reflects the entire travel time 22 and actually the correct distance to an object.

3 This illustrates a situation that can arise when longer runtimes 22 occur than in the situation of 2 or if another lidar device also transmits signals that could be received by the first lidar device that transmitted the optical signal 18. If the received optical signal 20 originates from the same lidar device, it is still not clear which sawtooth of the transmitted optical signal 18 corresponds to the first sawtooth of the received signal. The evaluation of a corresponding received optical signal 20 can therefore contain errors in determining the propagation time 22, which are several times the period T1. Naturally, a resulting ambiguous distance determination is very problematic. Even if the received optical signal 20 originates from a different lidar device, this can of course lead to problems if a distance to an object in the vicinity of a vehicle is to be determined.

A method according to the invention is described below with reference to the 4 An example will be explained in more detail. 4 This shows a digital sequence 24 with +1 bits 30 and -1 bits 32. Sequence 24 includes a header with a sender ID and a timestamp, which are not shown in detail. Naturally, the header and timestamp could contain more bits 30 and 32 than shown, or other bit sequences. All bits 30 and 32 have an identical bit duration TB1, but this is not necessarily the case.

The digital sequence 24 to be transmitted is combined with a spreading sequence 26 before transmission. In the representation of the 4 A (simplified) complete sequence 24 to be transmitted is shown together with two consecutive spreading sequences 26. Each spreading sequence 26 also consists of +1 bits 34 and -1 bits 36. The corresponding bits 30 and 32 of the spreading sequence 26 are also referred to as chips 28. In the depicted implementation of the method, each bit 30 and 32, or each chip 28, of the spreading sequence 26 has an identical chip duration TB2. However, this is not strictly necessary, so the chips 28 can also have different widths. It is advantageous, however, to design the spreading sequence 26 with a period T2, which, as shown, corresponds to the bit duration TB1 of the sequence 24 to be transmitted. Thus, each of the spreading sequences 26 continuously generated with period T1 can be combined with a bit 30 or 32 of the sequence 24 and modulate it. It should be taken into account that in the 2 Only two spreading sequences 26 are shown, which combine the two hatched bits 30 and 32 to generate the combined digital sequence 42, also hatched. Of course, further spreading sequences 26 are provided to combine the other bits of sequence 24 as well.

Further schematically indicated in the 4 a combination method is used here. In the case of the 4 The combination takes place such that the transmitted sequence 24 is multiplied by the spreading sequence 26 (x-character). Accordingly, the multiplication of a +1 bit 30 of sequence 24 with the spreading sequence 26 results in an identical representation of the spreading sequence 26 in the transmitted optical signal 18. Conversely, the modulation of a -1 bit 32 of sequence 24 with the spreading sequence 26 leads to an inverted spreading sequence 26 in the combined transmitted sequence 42. Further details or sequences of the transmitted optical signal 18 are described in the 4 Not shown for the sake of simplicity.

The combined (here multiplied) sequence 42 represents a digital baseband signal and is used in a modulator not shown in detail to modulate an analog signal of a laser (carrier signal) in order to produce the emitted optical signal 18 of the 1 to obtain.

According to the 1 After reflection from an object in the vicinity of the lidar device 12, the reflected optical signal 20 is received and evaluated by the lidar device 12. Naturally, the modulated transmitted optical signal 18 must first be demodulated and then detected after reflection, in this embodiment by multiplying it again by the spreading sequence 26. A signal with bits 30 and 32, period T1, and the other form of the sequence 24 is then available for further processing.

Since sequence 24 contains a header with a transmitter ID and a timestamp, the processing unit 16 of the lidar device 12 can easily determine when the signal was transmitted and calculate corresponding propagation times 22. The transmitter ID is used to verify that a received optical signal 20 actually originates from the same lidar device 12. Otherwise, it may be intended that a received optical signal 20 is not used for a distance measurement at all (which would be inaccurate). Using the timestamp, the processing unit 16 can also determine a propagation time 22 at the end of each period T1. Since the timestamp specifies exactly when the optical signal 18 was transmitted, the propagation times 22 are calculated based on the 3 Problems with long runtimes were not explained in section 22. In the image of the 3 Put simply, it is always clear which sawtooth of the received optical signal 20 corresponds to which sawtooth of the emitted optical signal 18.

Since the transmitted optical signal 18 also includes the chip duration TB2 of the chips 28 of the spreading sequence 26, it has a significantly higher bandwidth (~1/TB2) than the originally unmodulated sequence 24 with its much longer bit duration TB1. Likewise, the period T1 of the bit sequence 24 is considerably longer than the period T2 of the chip sequence 26. The frequencies are inversely proportional to this. This higher frequency of the spreading sequence 26 can now be used to perform a channel estimation upon reception of each spreading sequence 26. Thus, a channel estimation can be performed for each bit 30, 32 of the sequence 24 after modulation with the spreading sequence 26, and due to the short time intervals of the channel estimations, averaging over several channel estimations is possible to obtain a more reliable result.

In particular, such a channel estimate, and optionally an averaging of several channel estimates, can be used to derive the time offset +/- ΔT smaller than a period T1 of a sequence that the runtime 22 includes in addition to several integer periods T1 of the sequence. In other words, the timestamp is used specifically to determine how many integer periods T1 of a sequence the runtime 22 includes, and this for each period T1. Within a period T1, multiple channel estimates can be performed to determine updated time offsets ΔT in between, thus enabling faster and more accurate determination of distance changes.

Reference symbol list

10

vehicle

12

Lidar device

14

Front-End

16

Processing unit

18

emitted optical signals

20

received optical signals

22

Duration

T1

Period sequence

24

digital sequence (header + timestamp)

26

Spreading sequence

28

Chips

T2

Period Spreading sequence

30

+1 bit

32

-1 bit

34

+1 Chip Spread Sequence

36

-1 Chip Spread Sequence

TB2

Chip duration

TB1

Bit duration

42

combined sequence

Claims (10)

A method for operating a lidar device (12), wherein, in the course of the method: the lidar device (12) emits optical signals (18), receives optical signals (20) reflected from objects in an environment, and distances between the lidar device (12) and the objects are determined from the transit times (22) of the received signals (20), characterized in that the optical signals (18) to be emitted are generated by: generating digital sequences (24) of a first period (T1), each of which has a header with a transmitter ID and a timestamp, combining the digital sequences (24) with a digital spreading sequence (26) to obtain combined digital sequences (42), modulating an optical carrier signal by means of the combined digital sequences (42) to obtain the optical signals (18) which are emitted with the first period (T1), the received reflected optical Signals (20) are demodulated and separated to determine the transmitter ID and timestamp, and at least one channel estimation is performed using each received sequence. Procedure according to Claim 1 , characterized in that the digital sequences (24) consist of a first sequence of digital bits (30, 32) and have a first bit length (TB1), and the spreading sequence (26) consists of a second sequence of digital bits (34, 36) and has a second period (T2). Procedure according to Claim 2 , characterized in that the period (T2) of the spreading sequence corresponds to the bit length (TB1) of the sequence (24). Method according to one of the preceding claims, characterized in that a channel estimation is performed with each received spreading sequence. Method according to one of the preceding claims, characterized in that the combination includes a multiplication of each digital sequence (24) with the spreading sequence (26). Method according to one of the preceding claims, characterized in that the channel estimation and the timestamp are used to determine the transit time (22) of the received, demodulated optical signal (20). Procedure according to Claim 6 , characterized in that the timestamp is used to determine how many integer period durations (T1) of a sequence (24) the runtime (22) contains. Procedure according to Claim 7 , characterized in that the channel estimation is used to determine what time offset (+/- ΔT) smaller than the period (T1) of a sequence (24) the runtime additionally includes. Method according to one of the preceding claims, characterized in that the spreading sequence (26) includes a CAZAC sequence. Method according to one of the preceding claims, characterized in that different spreading sequences (26) are used for different lidar devices (12), each spreading sequence (26) includes an individual gold sequence, and a lidar device (12) only demodulates such received sequences that were modulated with the gold sequence assigned to it.