阅读说明：本技术 用于线性调频调制无线电信号的发射机、接收机和方法 (Transmitter, receiver and method for chirp modulated radio signals ) 是由 O·B·A·塞勒 C·J·J·德沃塞勒 于 2021-05-24 设计创作，主要内容包括：线性调频调制无线电信号的发射机、接收机和方法。用于线性调频调制无线电信号的发射机,包括：线性调频生成器,被配置为生成一系列线性调频信号,其中每个线性调频携带被编码为循环移位的信息的元素,并且具有对依赖于线性调频的循环移位的纠错码进行编码的相位；进一步包括被配置为将一系列线性调频调制到无线电信号上的调制器和发射无线电信号的无线电发射机。用于线性调频调制无线电信号的接收机,包括时钟单元和被配置用于解调一系列接收的线性调频信号的解调器,该解调器具有去线性调频单元,该去线性调频单元被配置用于基于接收的线性调频的相位确定每个接收的线性调频相对于基本线性调频和纠错码的循环移位,该接收机具有同步校正单元。(A transmitter, receiver and method for chirp modulated radio signals. A transmitter for chirp modulated radio signals comprising: a chirp generator configured to generate a series of chirp signals, wherein each chirp carries elements of information encoded as a cyclic shift and has a phase encoding a cyclically shifted error correction code that is dependent on the chirp; further comprising a modulator configured to modulate a series of linear tones onto the radio signal and a radio transmitter to transmit the radio signal. Receiver for chirp modulated radio signals comprising a clock unit and a demodulator configured to demodulate a series of received chirp signals, the demodulator having a dechirping unit configured to determine a cyclic shift of each received chirp with respect to a basic chirp and an error correction code based on the phase of the received chirp, the receiver having a synchronization correction unit.)

1. A transmitter for chirp modulated radio signals, the transmitter comprising: a chirp generator configured to generate a series of chirp signals, wherein each chirp carries elements of information encoded as a cyclic shift and has a phase encoding a cyclically shifted error correction code that is dependent on the chirp; the transmitter further includes a modulator configured to modulate the series of chirps onto a radio signal and a radio transmitter to transmit the radio signal.

2. A transmitter as claimed in claim 2, wherein the phase of the chirp is the sum of a phase modulation term of the encoded error correction code and an alignment term that depends quadratically on the cyclic shift of each chirp, whereby the alignment term compensates for the phase shift introduced by the cyclic shift.

3. The transmitter of claim 3, wherein the chirp generator is configured to classify the chirps into a set of sub-classes based on a cyclic shift of each chirp, and to determine a phase modulation term for each chirp based on the sub-class of each chirp.

4. The transmitter of claim 2, wherein the chirp generator is configured to classify the chirps into a set of sub-classes based on a cyclic shift of each chirp, and to determine a phase change for each chirp based on the sub-class of each chirp, wherein the phase change of each chirp is added to the phase of a preceding chirp to determine the phase modulation of each chirp.

5. A receiver for chirp modulated radio signals comprising a clock unit and a demodulator configured to demodulate a series of received chirp signals, the demodulator having a dechirp unit configured to determine a cyclic shift of each received chirp with respect to a base chirp and an error correction code based on the phase of the received chirp, the receiver having a synchronization correction unit configured to detect and/or correct errors in the clock unit based on the error correction code.

6. The receiver of claim 5, wherein the step of determining the cyclic shift comprises computing an oscillating signal by multiplying each sample of the received chirp with a corresponding sample of the complex conjugate of the basic chirp, computing a Fourier transform of the oscillating signal, determining the peak position in the Fourier transform.

7. The receiver of claim 6, wherein the step of determining an error correction code comprises determining a complex phase of a Fourier transform at the peak location.

8. The receiver according to claim 5, wherein the synchronization correction unit is configured to correct a cyclic shift of the received chirp and/or to adjust a clock unit of the receiver based on an error correction code in the continuous chirp.

9. The receiver of claim 8, wherein the synchronization correction unit is arranged to classify received chirps into sub-classes based on their estimated cyclic shifts, and to determine synchronization based on phase differences between successive chirps.

10. A method of transmitting data as a chirp modulated radio signal, comprising:

generating a series of chirps each having cyclically shifted encoded data and a phase;

wherein the phase of each chirp encodes an error correction code that depends on the cyclic shift of the chirp;

transmitting a series of chirps in the radio signal.

11. The method of claim 10, comprising

Receiving a radio signal in a receiver, demodulating a series of received chirp signals;

determining a cyclic shift of each received chirp signal;

determining an error correction code based on a phase of the received chirp signal; and

detecting and/or correcting time errors or frequency errors based on the error correction code.

12. A method according to claim 11, comprising determining the phase offset introduced in the chirp signal, which comprises adding an error correction code to an alignment term which depends quadratically on the cyclic shift of each chirp signal, whereby the alignment term compensates for the phase shift introduced by the cyclic shift.

13. A method according to claim 12, comprising correcting the cyclic shift of the received chirp and/or adjusting the clock unit of the receiver based on the difference between error correction codes in successive chirps.

14. The method of claim 12, comprising classifying the chirps into a set of sub-classes based on a cyclic shift of each chirp, and determining a phase modulation term for each chirp based on the sub-class of each chirp.

15. The method of claim 12, comprising classifying the chirps into a set of sub-classes based on a cyclic shift of each chirp, and determining a phase change for each chirp based on the sub-class of each chirp, wherein the phase change of each chirp is added to the phase of a preceding chirp to determine the phase modulation of each chirp.

Technical Field

In an embodiment, the invention relates to a receiver and a transmitter for a chirp modulated spread spectrum radio signal.

Prior Art

A known problem in the field of radio communications is that data is exchanged wirelessly over greater distances despite interference, fading and attenuation in the radio channel. When wireless communication is used to connect portable devices, appliances and sensors, as is increasingly the case, there is an additional requirement to reduce power consumption. Chirp modulated signals as embodied by LoRa modulation have for example been successfully used in this context.

Chirp modulated signals are used in Semtech corporation's LoRa^TMIn RF technology, the LoRa is used in the following of the present disclosure^TMThe RF technology will be referred to as LoRa for short. As disclosed in, among other documents, documents EP 2763321 a1, EP 3264622 a1 and EP2449690 a1, LoRa is based on the transmission and reception of symbols obtained by cyclic shifting of a "basic" symbol, which is a frequency chirp with a determined slope and bandwidth.

Document US 8971379B 2 discloses a chirp spread spectrum communication system in which data is transmitted by symbols comprising a chirp characterized by a slope, a cyclic shift and a complex phase shift.

It is also known from US 6549562B 1 to generate chirps in various radio frequency circuits, US 6549562B 1 describes a method for generating a modulated chirp signal, and EP 0952713 a2 discloses a synchronization process based on a chirp signal.

US 6940893B 1 and US 6614853B 1, among others, disclose the generation and use of chirps by passing pulsed signals through a wave dispersion filter, and communication schemes based thereon.

A clear advantage of using chirp modulated radio signals in machine-to-machine communication is that they can also be demodulated and processed by devices with low power consumption and processing speed. However, there is a need to improve the error detection and correction capabilities of the receiver, as well as the ability to align the internal time reference of the receiver with the time reference at the transmitter side.

Although the LoRa communication system has the ability to detect and correct time or frequency synchronization errors between the transmitter and the receiver, this is limited to errors expressed in time that are less than one sampling period. Larger errors are not detected as such, but are perceived by the receiver as cyclic shifts, creating demodulation errors for an integer number of modulation values.

Although the LoRa data frame includes correction codes that allow for the detection and correction of such errors, these errors should not occur on every symbol. The starting symbol of each frame cannot be so corrected because the error at the start will propagate to all symbols. The same applies to the first symbol received after fading of the signal inside the frame.

It is an object of the present invention to provide a modulation method and a corresponding receiver which are able to fulfill these tasks with a simple architecture and which are therefore able to be produced at low cost and to be deployed in large numbers.

Disclosure of Invention

According to the present invention, these objects are achieved by the objects of the appended claims, and in particular by: a transmitter for chirp modulated radio signals, the transmitter comprising: a chirp generator configured to generate a series of chirp signals, wherein each chirp carries information elements encoded as a cyclic shift and has a phase encoding a cyclically shifted error correction code dependent on the chirp; the transmitter further comprises a modulator configured to modulate a series of linear tones onto the radio signal and a radio transmitter to transmit the radio signal; and a receiver for chirp modulated radio signals, comprising a clock unit and a demodulator configured to demodulate a series of received chirp signals, the demodulator having a dechirp unit configured to determine a cyclic shift of each received chirp relative to a base chirp and an error correction code based on a phase of the received chirp, the receiver having a synchronization correction unit configured to detect and/or correct errors in the clock unit based on the error correction code; and a corresponding method.

The dependent claims relate to advantageous but not essential elements of the invention, including phase offsets introduced in the chirps, wherein an error correction code is added to an alignment term which depends quadratically on the cyclic shift of each chirp, whereby the alignment term compensates for the phase shift introduced by the cyclic shift, and a method of determining the cyclic shift in a received chirp, the method comprising calculating an oscillating digital signal by multiplying the received chirp by the complex conjugate of the basic chirp, determining the dominant frequency of the oscillating signal by means of a fourier transform, and deriving the cyclic shift from the dominant frequency, sorting the chirps into a set of subclasses in the transmitter, and determining their phase modulation or phase change on the basis of the subclasses. Also, the receiver of the present invention may be arranged to detect and correct errors based on phase differences between successive chirps, and to classify the received chirps into subclasses based on their modulation values.

Drawings

Exemplary embodiments of the invention are disclosed in the specification and illustrated by the accompanying drawings, in which:

fig. 1 shows in a schematically simplified manner the structure of a radio modem according to an aspect of the present invention.

Fig. 2a plots the instantaneous frequency of a fundamental chirp and a modulation chirp in accordance with an aspect of the present invention. The phase of the same signal is represented in fig. 2b, and fig. 2c plots the real and imaginary parts of the complex components of the basic and modulation chirps in the time domain as well as in the baseband representation.

Fig. 3 schematically shows the structure of a data frame exchanged between two devices in a frame of the invention.

Fig. 4 is a graph showing the relationship between the modulation value and the chirp phase in the embodiment of the present invention.

Examples of embodiments of the invention

Several aspects of the chirp modulation technique employed in the present invention are described in european patent EP2449690, which is hereby incorporated by reference and will be briefly recalled herein. The radio transceiver schematically represented in fig. 1 is a possible embodiment of the invention. The transceiver includes a baseband section 200 and a radio frequency section 100. It comprises a baseband modulator 150, the baseband modulator 150 generating a baseband complex signal based on digital data 152 at its input. It is then converted to a desired transmission frequency by the RF section 100, amplified by the power amplifier 120, and transmitted by the antenna.

Once the signal is received at the other end of the radio link, it is processed by the receiving part of the transceiver of fig. 1, which comprises a low noise amplifier 160 followed by a down-conversion stage 170, which down-conversion stage 170 generates a baseband signal comprising a series of chirps (which are also, for example, composed of two components)I、QThe complex signal represented) and then processed by a baseband processor 180, the function of the baseband processor 180 being the inverse of that of the modulator 150 and providing a reconstructed digital signal 182.

As discussed in EP2449690, the signal to be processed comprises a series of chirps whose frequencies are from an initial instantaneous value along a predetermined time intervalf ₀Change to the final instantaneous frequencyf ₁. For simplicity of description, it will be assumed that all chirps have the same durationTAlthough this is not an absolute requirement of the present invention.

Chirp in baseband signals may be determined by the time distribution of their instantaneous frequenciesOr may be described by a functionTo describe the functionDefining the phase of a signal as a function of time. Importantly, the processor 180 is arranged to process and identify chirps having a plurality of different profiles, each profile corresponding to a symbol of a predetermined modulation alphabet.

According to an important feature of the invention, the reception signal Rx may comprise a basic chirp (in the following also called unmodulated chirp) having a specific and predefined frequency distribution, or one among a set of possible modulation chirps obtained from the basic chirp by cyclically time shifting the basic frequency distribution. Fig. 2a and 2b illustrate by way of example the basic chirp and a modulation chirp at the moment when the chirp startsAnd the moment when the chirp endsWith possible frequency and phase distributions, while fig. 2c shows the corresponding baseband signal in the time domain. The horizontal scale corresponds to, for example, one symbol, and although the graphs are drawn as continuous, in particular implementations they actually represent a limited number of discrete samples. As for the vertical scale, they are normalized to the expected bandwidth or corresponding phase span. The phase is represented in fig. 2b as if it were an unbounded variable, but in a specific implementation it may actually span several revolutions (revolution).

In the depicted example, the frequency of the fundamental chirp is from an initial valueLinearly increasing to a final valueWhere BW denotes bandwidth extension, but down-chirp or other chip distributions are also possible. Thus, the information is encoded in the form of a chirp, which chirpHaving one of a plurality of possible cyclic shifts relative to a predetermined base chirp, each cyclic shift corresponding to a possible modulation symbol, or in other words, the processor 180 needs to process a signal comprising a plurality of frequency chirps that are cyclically time shifted copies of a base chirp distribution and extract the messages encoded in the time shifted sequence.

Fig. 2c is a graph of the real and imaginary parts I and Q of the baseband signal in the time domain corresponding to a basic chirp.

As will become more apparent below, the signal may also include a conjugate chirp that is the complex conjugate of the substantially unmodulated chirp. One can view these as down-chirp, where the frequency is selected fromFall down to。

The operation of evaluating the time shift of the received chirp with respect to the local time reference may be referred to hereinafter as "despreading" and may advantageously be carried out by a despreading step involving multiplying the received chirp sample by sample with the complex conjugate of the locally generated basic chirp. This results in an oscillating digital signal whose main frequency can be shown to be proportional to the cyclic shift of the received chirp. Demodulation may then involve a fourier transform of the despread signal. The position of the fourier maximum is a measure of the cyclic shift and the modulation value. In mathematical terms, useMarking the k-th received symbol, whereinkIs a symbol index, andjis a sample index, the corresponding modulation value isIs given inSignAnd the conjugate of the fundamental chirpThe fourier transform of the product between. However, other ways of demodulating the signal and extracting the cyclic shift of each symbol are possible.

Preferably, the signals transmitted and received by the present invention are organized in frames comprising a suitably encoded preamble and data section. The preamble and data sections include a series of modulated and/or unmodulated chirps that allow the receiver to time align its time reference with the time reference of the transmitter, retrieve information elements, perform actions or execute commands. In the frames of the present invention, several structures of data frames are possible depending on, among other things, the channel conditions, the data or commands transmitted. Fig. 3 schematically represents a frame structure that may be employed in various aspects of the present invention.

In the example presented, the frame has a preamble followed by a data header 415 and a data payload 416. The preamble starts with a detected sequence 411 of basic (i.e. unmodulated or in case the cyclic shift is equal to zero) symbols. The detection sequence 411 is used in the receiver to detect the start of the signal and preferably performs a first synchronization of its time reference with the time reference in the transmitter. By demodulating the detection sequence, the receiver can determine the shift amount and adapt the frequency and phase of its clock to the frequency and phase of the transmitter, allowing subsequent data to be decoded.

The end of the detection sequence is marked by one or more (preferably two) frame synchronization symbols 412, which frame synchronization symbols 412 are chirps modulated with a predetermined value, e.g. the first chirp has a modulation value of 4 and the second chirp has the opposite modulation N-4. These modulation symbols are used to achieve frame synchronization as disclosed in EP 2763321 a1 and EP 3264622 a 1.

The frequency synchronization symbol 413 consists of one or more (preferably two) chirps that are complex conjugates of the substantially unmodulated chirps so that they have a slope that is opposite to all other symbols. Preferably, these symbols are followed by silence 420 to allow alignment of the receiver, as unmodulated substantially chirped fine synchronization symbols 414 for estimating and correcting residual timing drift.

According to an important aspect of the invention, the information is transmitted by symbols, each symbol being a chirp of determined length and slope, and the information is encoded in a cyclic shift of the symbol, as in the standard LoRa modulation, and each symbol is synthesized with a given phase offset for transmitting an error correction code to the receiver, in addition to this modulation by cyclic shift.

In the original LoRa modulation, the complex phase of each symbol is defined modulo the undetermined offset. In many implementations, such as those in which symbols are synthesized by a voltage controlled oscillator, the phase may never show a discontinuity, and each symbol has an initial phase implicitly determined by its cyclic shift, such that the phase is continuous at symbol boundaries, as shown in fig. 2 b. However, the requirement can be relaxed and each symbol can be synthesized with an arbitrary phase offset. In the following, the expression "phase of the chirp" may refer to the phase offset of the entire chirp with respect to the phase required for inter-symbol continuity.

According to an aspect of the invention, after the dechirping operation, the phase is estimated from the phase at the fourier transform peak. If it is notIndicating the Fourier transform andindicating the modulation value, the phase of each symbol can be determined byAnd (4) determining. In the frequency spectrum, although other ways of estimating the phase are possible and useful for the present inventionThe algorithm is convenient already available and relatively unaffected by noise.

Defining a "standard" phase shiftConveniently, the "standard" phase shiftIn this way dependent on the modulation valuemSuch that each demodulation peak shows the same phase after dechirping and fourier transformation. The standard phase shift can be considered as an alignment phase term. In case all symbols are provided with the same slope and the same duration, it depends in a quadratic way on the modulation value (cyclic shift) and is in fact the same as the phase shift implicitly introduced by the requirement of inter-symbol phase continuity.

According to an aspect of the invention, the phase offset of each symbol comprises an alignment term, and a phase modulation term added thereto, which encodes an error correction code that the receiver can use to detect and/or correct lack of synchronization (misalignment in time or frequency) between its clock and the transmitter clock. As mentioned in the introduction, in standard LoRa modulation, the time misalignment in the symbol caused by more than one sample is indistinguishable from the modulation-induced cyclic shift. The invention introduces redundant information in the symbols, which allows this drawback to be overcome.

The phase shift may be demodulated at the receiver in various ways including by looking at the phase of the demodulation peak after dechirping and fourier transformation. Since the symbols include an alignment term, the receiver can compare the phase of the corresponding modulation peaksThe difference between the phase shifts in the two symbols is determined.

By introducing a phase shift dependent on the cyclic shift in various ways, the desired detection and correction can be obtained, so that when the receiver demodulates the symbol, the cyclic shift and the phase shift can be compared and the synchronization error can be distinguished from the normal modulation.

In a possible embodiment of the invention, the possible modulation values are divided into several sub-classes, for example, according to the remainder obtained by the modulation value (cyclic shift) modulo a given divider, and cannot be detected or corrected. The receiver may then determine the synchronization error from the difference between the modulation peak phases of two symbols that do not have the same modulation value.

For example, modulation values can be divided into three sub-classes according to the remainder of the modulation value modulo three: c0, C1, and C2. Each sub-class is associated with a phase value.

Table 1.

With this arrangement, the phase difference between two symbols is determined by their subclasses. FIG. 4 is a graph showing modulation values according to this variation of the inventionAnd chirp phase. The phase of each chirp being a phase modulation termAnd the quadratic alignment term introduced aboveAnd, the phase modulation termAnd the quadratic alignment term introduced aboveDepending on the subclasses as shown in table 1.

For example, a synchronization error corresponding to one sample time will shift the received modulation value and change the sub-class to which the symbol belongs, but will have no effect on their relative phase difference. Thus, synchronization errors can be detected and corrected from observations of two or more symbols belonging to different sub-classes. The receiver is arranged to sort the received symbols into respective sub-classes, determine their phase differences and compare the phase differences with those of table 1. This comparison provides a synchronized state that contains an error between-1 and +1, as shown in table 2.

Phase difference

TABLE 2

The cases [ C0, C1], [ C0, C2], [ C1, C2] are not shown because they are opposite to the cases shown.

To determine the point of view with a specific example, if two consecutive symbols having modulation values of 87 and 46 are received with a phase difference of approximately +120 °, that is, the phase of the symbol having the cyclic shift 46 is 120 ° higher than the phase of the symbol having the cyclic shift 87. The receiver can determine that they belong to subclasses C1 and C0 and that the synchronization error is-1. Thus, the correct modulation values are 47 and 88.

A similar pattern can be devised to increase the correction range of synchronization errors. For example, the division of the symbols into five sub-classes according to the remainder of the modulation value modulo 5 allows the detection of synchronization errors of ± 2, ± 1 or 0.

Table 3.

With this arrangement, the receiver can detect and correct errors ranging from-2 to +2 modulation positions.

Phase difference

Table 4.

In a variant of the invention, the phase is differentially encoded: the modulation symbols are again divided into sub-classes, and each sub-class is associated with a phase change relative to the phase of the previous symbol. This mode allows for unrestricted detection of synchronization errors from the observation of two symbols and may be more robust than the previous example in some use cases, for example if the error is in the first received symbol or when the cyclic shift repeats itself in a series of symbols.

As in the previous example, the modulation symbols may be subdivided into three sub-classes according to the remainder of the modulation value modulo three. Each sub-class is associated with a phase change value relative to the phase of the preceding symbol.

Table 5.

In this mode, the phase difference between two consecutive symbols is uniquely determined by the subclass of the second symbol, as shown in the following table

Table 6.

The concept can be extended to five sub-classes to detect errors in the-2, …, +2 range

Table 7.

In general, the differential coding variant allows the detection of errors between-n and + n by dividing the symbols into n sub-classes, each associated with a phase changeWherein。