Digital radio transmitter and data transmission method

Abstract

Digital radio transmitters and data transmission methods are provided. Chirp-based spread spectrum digital radio transmitters are configured to suppress or attenuate discontinuities in the chirp frequency profile by an interpolation or smoothing procedure. The interpolation significantly reduces out-of-band emissions and does not affect demodulation and decoding. In the case of LoRa communications, the transmitter of the invention has low constellation error and can communicate with conventional unmodified receivers. The procedure does not introduce unwanted phase shifts and can be used to control the phase of the signal, also inserting arbitrary phase shifts on consecutive segments of the chirp.

Description

Technical Field

[0001] This invention relates to a digital radio transmitter suitable for low-power wireless networks using spread spectrum chirp-modulated signals, and corresponding transmission and modulation methods. Background Technology

[0002] In recent years, wirelessly connected devices have become a focus of considerable attention and investment. Improved wireless communication technologies play a crucial role in the creation and development of the "Internet of Things" (IoT). Against this backdrop, various wireless communication protocols have been proposed and utilized. The LoRa™ communication system (known, among others, through patents EP 2449690 B1, EP 2763321 B1, EP2767847 B1, EP 3247046 B1, and EP 2449690 B1) uses chirped spread spectrum modulation to achieve long transmission ranges with low complexity and low power consumption.

[0003] In the context of this disclosure, for the sake of brevity, the term "LoRa" refers to a communication system based on the exchange of radio signals, which includes multiple frequency chirps, each chirp being defined within a finite time interval and a finite bandwidth. The chirps include a reference chirp (the instantaneous frequency in the reference chirp follows a given function from the beginning to the end of the time interval) and a modulation chirp (the modulation chirp is the result of cyclic shifting the reference chirp). The shift values ​​encode symbols taken from the modulation alphabet. This definition includes known LoRa™ products and standards, as well as other forms of chirp modulation sharing the same basis, including possible but not yet implemented variants of the generalized concept.

[0004] Signals containing a series of chirps at many frequencies exhibit abrupt frequency jumps at the boundaries between chirps and within chirps modulated by cyclic shift modulation (such as in LoRa), as the frequency wraps around between minimum and maximum permissible frequency values. Such frequency jumps generate out-of-band emissions that may exceed regulatory limits.

[0005] Traditionally, bandpass or lowpass filters are used in the time domain (i.e., for the signal itself) to suppress out-of-band emissions. However, this is not a satisfactory solution because it modifies both the phase and amplitude of the signal, removing the constant envelope property. Furthermore, filtering incurs a non-negligible computational cost. Summary of the Invention

[0006] This invention proposes a transmitter and communication method based on chirped spread spectrum modulation, which overcomes at least some of the shortcomings and limitations of the prior art.

[0007] According to the present invention, a digital radio transmitter is configured to: encode digital data into a series of frequency-converted chirps; modulate a carrier based on the series of frequency-converted chirps; and transmit the carrier as a radio signal. The digital radio transmitter is characterized by being configured to: locate one or more interpolation segments in the series of frequency-converted chirps; modify the series of frequency-converted chirps by replacing the instantaneous frequency values ​​of the series of frequency-converted chirps in each interpolation segment with modified values ​​obtained by interpolating the instantaneous frequency values, thereby modulating the carrier with the modified series of frequency-converted chirps.

[0008] Advantageously, the interpolation segment can encompass discontinuities in the instantaneous frequencies of a series of frequency-converted chirps, and through interpolation, these discontinuities are eliminated or at least replaced by one or more smaller steps or discontinuities. In this way, the transmitter reduces out-of-band emissions.

[0009] Interpolation can be applied to linear segments, piecewise linear functions, polynomial functions, exponential rational functions, cosine functions, or hyperbolic functions; however, this list is not exhaustive. Many mathematical functions can be used within the framework of this invention, and it is impossible to enumerate them all.

[0010] Preferably, the interpolation is confined to a finite number of compact interpolation segments, and the frequency profile outside these segments is not modified. The interpolation segments may have a common, predetermined width, or the width of the interpolation segments may be adaptively varied based on the height of the discontinuous steps they enclose, the slope of the chirp, or any other useful parameter.

[0011] Advantageously, interpolation does not introduce unwanted phase shifts. Each interpolation in a segment can produce zero net phase shift or introduce a desirable phase shift (positive or negative).

[0012] In the context of this disclosure, "interpolation" refers to a mathematical procedure to determine the value between two points having known values. In the context of a time series of sampled values, interpolation can result in the determination of one or more sampled values ​​between two known samples. The (continuous or discrete) time of the known values ​​can be considered as the left and right boundaries of the "interpolation segment".

[0013] Possible variations relate to methods for transmitting digital data and corresponding transmitters. The method includes: encoding data into a series of frequency-converted chirps; modulating a carrier wave with the series of frequency-converted chirps; transmitting the carrier wave as a radio signal; receiving the radio signal in a receiver; reconstructing the digital data based on the received radio signal; and, particularly in the transmitter, including the step of smoothing the frequency-converted chirps by applying a low-pass filter to the instantaneous frequency or phase value of the signal; and modulating the carrier wave with a modified series of frequency-converted chirps.

[0014] Advantageously, the low-pass filter can be a first-order digital IIR (iterative) digital filter, but other solutions are also possible, such as higher-order filters or FIR low-pass filters. The low-pass filter can be applied to the entire chirp profile or only to a selected segment. Attached Figure Description

[0015] Exemplary embodiments of the present invention are disclosed in the specification and illustrated with reference to the accompanying drawings, wherein: Figure 1 The structure of a radio modem according to one aspect of the invention is shown in an illustrative and simplified manner.

[0016] Figure 2a The instantaneous frequencies of the reference chirp and the modulated chirp according to one aspect of the invention are plotted. The phase of the same signal is in Figure 2b The Chinese side indicated that, and Figure 2c The real and imaginary parts of the reference chirp and the modulation chirp are plotted in the time domain and in the baseband representation.

[0017] Figure 3 The graph shows the instantaneous frequency of the chirped signal with step discontinuity and the interpolated continuous version of the chirped signal with step discontinuity.

[0018] Figure 4 Different interpolation processes are shown, which adapt the interpolation width to the step height.

[0019] Figure 5a and 5b This illustrates the effect of interpolation on the signal phase. Figure 5a The signal phase remains unchanged and Figure 5b The signal phase is ahead.

[0020] Figure 6 An example of partial interpolation is shown.

[0021] Figure 7 This demonstrates that interpolation in the frequency domain can be used to introduce a phase shift in any segment of the chirped profile.

[0022] Figure 8 The comparison test results between the interpolated and uninterpolated signals are shown. The spectrum illustrates the improvement in out-of-band emission brought about by the present invention. The figures given in the legend correspond to the EVM of the modified LoRa signal.

[0023] In the accompanying drawings, key elements are identified by reference markers that are repeated in the text. The same reference marker can be used to identify different elements that are identical, similar, logically related, or technically equivalent. When there are many identical, similar, related, or equivalent elements in the accompanying drawings, some reference markers may be omitted to avoid cluttering the drawings. Detailed Implementation

[0024] The patents mentioned in the "Background Art" section above (such as EP 2449690 B1, EP 2763321 B1, and EP 3247046 B1) disclose the basic principles of chirped modulation. A brief review of these basic principles is provided here.

[0025] Figure 1 The radio transceiver schematically shown is a possible embodiment of the present invention. The transceiver includes a baseband section 200 and an radio frequency (RF) section 100. The transceiver includes a baseband modulator 150 that generates a baseband complex signal based on digital data 152 at its input. This baseband complex signal is then converted to a desired transmission frequency by the RF section 100, amplified by a power amplifier 120, and transmitted via an antenna through an RF switch 102.

[0026] Once a signal is received at the other end of the radio link, the signal is... Figure 1 The receiving section of the transceiver performs processing. This receiving section includes a low-noise amplifier 160, which is then connected to a down-conversion stage 170. The down-conversion stage 170 generates a baseband signal (the baseband signal is also a complex signal, for example, represented by two components I and Q, which include a series of chirps). The baseband signal is then processed by a baseband processor 180, which performs the opposite function of the modulator 150 and provides a reconstructed digital signal 182.

[0027] As discussed in EP2449690, the signal to be processed comprises a series of chirps, the frequency of which varies at predetermined time intervals from an initial instantaneous value. Change to final instantaneous value For the sake of simplicity, it will be assumed that all chirps have the same duration T, although this is not an absolute requirement of the invention.

[0028] Chirps in a baseband signal can be described by the time profile f(t) of its instantaneous frequency, or by a function ϕ(t) whose signal phase is defined as a function of time. Importantly, the processor 180 is arranged to process and identify chirps with a variety of different profiles, each profile corresponding to a symbol in a predetermined modulation alphabet.

[0029] According to an important feature of the invention, the received or transmitted signal may include a reference chirp (hereinafter also referred to as an unmodulated chirp) having a specific and predefined frequency profile, or one of a set of possible modulated chirs obtained from the reference chirp by cyclically time-shifting the reference frequency profile. Figure 2a and 2b The illustration demonstrates the chirping initiation moment. Until the end of the chirping The possible frequency and phase profiles between the reference chirp 30 and a modulation chirp 32, and Figure 2c The corresponding baseband signal in the time domain is shown. The horizontal range corresponds to, for example, the symbol, and although the graphs are plotted as continuous, in practice they actually represent a finite number of discrete samples. As for the vertical range, it is normalized to the expected bandwidth, or normalized to the corresponding phase or amplitude span.

[0030] The modulated chirp 32 exhibits a step discontinuity 39, where the frequency abruptly changes from the maximum possible value BW / 2 to the minimum possible value -BW / 2. Although not easily visible in the graph, the unmodulated chirp 30 exhibits a similar discontinuity at the boundaries of the graph.

[0031] exist Figure 2b The phase is represented as an unbounded variable, but in practice, it may actually span multiple cycles.

[0032] In the example depicted, the frequency of the reference chirp starts from the initial value. linearly increase to the final value Here, BW represents bandwidth extension, but downsweep chirps or other chirp profiles are also possible. Therefore, information is encoded in the form of chirps that have one of several possible cyclic shifts relative to a predetermined reference chirp, each cyclic shift corresponding to a possible modulation symbol. In other words, processor 180 needs to process signals comprising multiple frequency chirps (which are cyclically time-shifted copies of the reference chirp profile) and extract the message encoded in the time-shifted sequence.

[0033] The signal may also include a conjugate chirp, which is the complex conjugate of the reference unmodulated chirp. Such a chirp can be considered a downsweep chirp with a frequency ranging from... Descending to .

[0034] The operation of evaluating the time shift of the received chirp relative to a local time reference, hereinafter referred to as "dechirping," can be advantageously performed via a despreading step involving sample-by-sample multiplication of the received chirp with the complex conjugate of a locally generated reference chirp. This produces an oscillating digital signal whose dominant frequency can be expressed as proportional to the cyclic shift of the received chirp. Demodulation may subsequently involve a Fourier transform of the despread signal. The location of the Fourier component with the maximum amplitude is a measure of the cyclic shift and the modulation value. In mathematical terms, This represents the k-th received symbol, and the corresponding modulation value is given by... Given, among which express Conjugate with reference chirp The Fourier transform of the product between them. However, other methods are also possible, such as demodulating the signal and extracting the cyclic shift of each symbol.

[0035] Figure 3 This is a hypothetical instantaneous frequency curve of a chirped signal. The frequency curve 34, represented by the dashed line, has continuous portions and discontinuities 39. This curve only shows a short time period of the signal. The signal can be a modulated LoRa signal, but the invention is not limited thereto.

[0036] As mentioned, discontinuity 39 translates into undesirable out-of-band emissions, and the present invention proposes a method to eliminate or reduce discontinuities by smoothing or trimming the frequency profile, thereby mitigating this drawback.

[0037] Perhaps the chirp profile 34 can be smoothed in the frequency domain or the phase domain using linear filters (e.g., IIR or FIR low-pass filters). However, these linear filters introduce constellation errors—biases—of the modulation values ​​relative to the ideal position (which may manifest as excessively high EVM (error vector magnitude)) and phase shifts. Linear filters applied to instantaneous frequency or phase are linear operations on the phase that are linked to the frequency via a linear transformation (integration); however, this is not applied to the signal itself because the relationship between phase and signal is non-linear. Accordingly, linear filters applied to instantaneous frequency or instantaneous phase are only effective to a certain extent and it is difficult to find an acceptable trade-off between out-of-band (OOB) emission and error vector magnitude (EVM).

[0038] According to an aspect of the invention, the original signal is replaced with a modified signal whose instantaneous frequency (represented by continuous segment 37) is a smoothed version of the instantaneous frequency of the original signal, the smoothed instantaneous frequency being continuous or at least exhibiting less drastic discontinuities with smaller frequency jumps.

[0039] In the example, interpolation is performed by a predetermined half-width Δ in interpolation segments 35 that span discontinuities 39. These segments are highlighted with a gray background in the figure. Outside of interpolation segments 35, the instantaneous frequency of the signal remains constant.

[0040] exist Figure 3 In this context, interpolation is an exponential function, which can be written as: in Where Δ represents the half-width of the interpolation segment, and α is a coefficient controlling the exponential growth / decay. Represents the instantaneous frequency at the inflection point, and It is the half-high frequency jump.

[0041] The inventors have discovered that in many important use cases, the width Δ of the interpolation segment can be selected so that the degradation of the EVM does not exceed acceptable limits. Once the value of Δ is selected, the coefficient α in the exponential function can be determined, thereby optimizing OOB emissions.

[0042] The inventors also discovered that this interpolation reduces out-of-bounds emissions, and that many interpolation functions are suitable for achieving this goal. For example, interpolation can be linear, piecewise linear, cubic, polynomial, exponential, hyperbolic tangent, cosine, and so on. The possibilities are too numerous to list. The interpolated signal can be a weighted average of the initial signal and the mathematical interpolation result, thus providing a smooth transition between the portion replicating the original signal and the interpolated portion. For example, Figure 6 The chirp with frequency step 39 is shown. In the interpolation segment 35, the frequency step 39 is interpolated by a linear function 37a, an exponential function 37b defined by the above formula, a cosine function 37d, and a hyperbolic tangent function 37c. The hyperbolic tangent function 37c is selected such that at the boundary of the interpolation segment 35, it does not completely reach the level of the original chirp, so that it shows two small frequency jumps 38 instead of the original large jump 39.

[0043] Advantageously, a series of chirps interpolated in the transmitter produce a radio signal with a cleaner spectrum, exhibiting a lower OOB (out-of-band) transmission than the original signal, which can still be received and understood without altering the demodulation and decoding processes. It is compatible with conventional receivers. In the case of LoRa, the signal can be processed in the receiver using the above-disclosed dechirping method without modification. This disclosure will show that the error introduced by interpolation can be measured or simulated and expressed as an acceptable EVM ratio.

[0044] Typically, strict continuity of the interpolated instantaneous frequency is not required, and as... Figure 6 In the final example, interpolation can be partial, in the sense that one or more small frequency jumps are preserved at the boundaries of the interpolation segment and within it. This is a way to reduce OOB emissions without degrading constellation errors.

[0045] Advantageously, the width Δ of the interpolation interval can be adapted to the characteristics of the signal. An example is shown in... Figure 4 As shown, interpolation segment 35 has different half-widths Δ0 and Δ1. The latter is smaller than the former because it corresponds to a lower frequency hop. The width of the interpolation segment can also be adjusted based on the chirp slope. A lower slope corresponds to a longer symbol time, which means the EVM will be more robust, and a longer interpolation segment will be acceptable. Furthermore, a lower chirp slope (in LoRa, this translates to a higher spread factor parameter value and a lower data rate) provides better noise immunity.

[0046] Another advantage of this invention is that it does not introduce unwanted phase shifts. The interpolation function shown above exhibits central symmetry, and therefore, the net effect of interpolation on the signal phase is zero. This can be achieved... Figure 5a As can be seen, the area between the original signal and the interpolated signal is highlighted. Regions where the interpolation is lower than the initial signal introduce a phase delay, while regions where the interpolation is higher than the initial signal result in phase lead. Since the two highlighted regions are identical, the net area is zero, and therefore the phase shift is zero.

[0047] Importantly, this provides a way to arbitrarily control the phase without introducing discontinuities or artifacts into the signal. Figure 5b An example is shown where the interpolation function is intentionally asymmetric. The area of ​​the interpolated chirp below the original chirp is smaller than the area of ​​the interpolated chirp above the original chirp. This results in a net phase lead, the value of which can be arbitrarily chosen by the shape of the interpolation function.

[0048] This fine-grained phase control can be developed by interpolating the chirp even within a signal segment with a continuous, non-jumping instantaneous frequency. The phase shift can be obtained by placing an interpolation segment 35 at any point (with or without discontinuities) in the chirp sequence 34 and adding continuous "bumps" 36, the integral of which equals the desired phase shift, such as... Figure 7 As shown. This process can be viewed as introducing an artificial or conceptual discontinuity, for example, by changing the frequency values ​​of one or more samples and subsequently interpolating the frequency profile to restore at least partial continuity. As illustrated, this achieves the desired phase shift with negligible impact on the EVM and OOB spectrum.

[0049] Figure 8The figures illustrate comparative test results between different processing methods for a LoRa packet with a spread factor SF=7 and a nominal bandwidth BW=125kHz. These curves show the spectra obtained by applying different filters and interpolations in the frequency domain.

[0050] Table 1 indicates Figure 8 The reference numbers, filtering or interpolation algorithms, and EVM used. Interpolation 37a and 37b use... Figure 6 An interpolation function is used for identification based on the same reference. Exponential interpolation has parameters Δ = 1.5 samples (1 sample = 1 / BW) and α = 2.5. The method of this invention produces an EVM comparable to or better than that produced by a low-pass filter, with superior reduction in OOB emissions, and introduces no unwanted phase shift compared to other forms of smoothing.

[0051] Explanation of reference numerals in the attached figures 30 baseline chirps 32 Modulated chirp 34 Original Frequency Profile 35 interpolation segments 36 Interpolation with continuous convexity 37. Smooth interpolation for discontinuities 37a Linear interpolation 37b Exponential Interpolation 37c Hyperbolic Tangent Interpolation 37d cosine interpolation 38. Residual discontinuity 39 Frequency jump 41. Spectrum of the original signal 42. Spectrum after LPF smoothing in the frequency domain 43. Spectrum after LPF smoothing in the frequency domain 44. Spectrum after linear interpolation in the frequency domain 45. The spectrum after exponential interpolation in the frequency domain 100 RF section 102 RF switch 110 Frequency Conversion 120 power amplifier 129 Oscillator, Time Base 150 modulator 152 Digital signal to be transmitted 154 Buffer 160 LNA 170 downconverter stage 180 Processor, Demodulator 182 Reconstructed digital signal 190 Controlled Oscillator 200 baseband section

Claims

1. A digital data communication method, the method comprising: - In the transmitter, digital data is encoded into a series of frequency-converted chirps. - The carrier wave is modulated based on a series of frequency conversion chirps. - Transmitting a carrier wave as a radio signal - Receive radio signals in the receiver. - Reconstructing digital data based on received radio signals, characterized by: In the transmitter, - Locate one or more interpolation segments within a series of variable frequency chirps. - A series of variable frequency chirps are modified by replacing the instantaneous frequency values ​​of a series of variable frequency chirps in each interpolation segment with the modified instantaneous frequency values. - The carrier wave is modulated by a modified series of frequency-converting chirps.

2. The method according to claim 1, wherein, The modified value is obtained by interpolating the instantaneous frequency values ​​in the interpolation segment, or by interpolating the leftmost and rightmost values ​​of the instantaneous frequency in each interpolation segment.

3. The method according to claim 1, wherein, The interpolation segment encloses the discontinuities in the instantaneous frequencies of a series of variable frequency chirps.

4. The method of claim 3, wherein the interpolation eliminates the discontinuity or reduces the height of the discontinuity.

5. The method according to claim 1, wherein, Interpolation is one of the following: linear interpolation, exponential interpolation, interpolation through a hyperbolic function, or interpolation through a cosine function.

6. The method according to claim 1, wherein, The interpolation segments have a common, predetermined width.

7. The method of claim 1, further comprising the step of adapting the width of one or more interpolation segments based on discontinuity height and / or based on chirp slope.

8. The method according to claim 1, wherein, Interpolation introduces a desired phase shift, which can be zero.

9. A digital radio transmitter configured to: encode digital data into a series of frequency-converted chirps; modulate a carrier wave based on the series of frequency-converted chirps; and transmit the carrier wave as a radio signal; characterized in that... It is configured to: locate one or more interpolation segments in a series of variable frequency chirps, and modify a series of variable frequency chirps by replacing the instantaneous frequency values ​​of a series of variable frequency chirps in each interpolation segment with modified instantaneous frequency values.

10. The transmitter according to claim 9, wherein, The modified value is obtained by interpolating the instantaneous frequency values ​​in the interpolation segment, or by interpolating the leftmost and rightmost values ​​of the instantaneous frequency in each interpolation segment.

11. The digital radio transmitter according to claim 10, wherein, The interpolation segment surrounds discontinuities in the instantaneous frequency profile of a series of variable frequency chirps, and the interpolation eliminates or reduces the height of the discontinuities.

12. The digital radio transmitter according to claim 9, wherein, Interpolation is one of the following: linear interpolation, exponential interpolation, interpolation through hyperbolic functions, or interpolation through cosine functions.

13. The digital radio transmitter according to claim 9, wherein, The interpolation segments have a common, predetermined width.

14. The digital radio transmitter of claim 9, further comprising the step of adapting the width of one or more interpolation segments based on discontinuity height and / or chirp slope.

15. The digital radio transmitter according to claim 9, wherein, Interpolation introduces a desired phase shift, which can be zero.

Images