EP4248224A1

EP4248224A1 - Spectrum analyser, system and method for transmitting data from a spectrum analyser

Info

Publication number: EP4248224A1
Application number: EP21815411.0A
Authority: EP
Inventors: Thorsten Chmielus
Original assignee: Aaronia Ag
Current assignee: Aaronia Ag
Priority date: 2020-11-18
Filing date: 2021-11-15
Publication date: 2023-09-27
Also published as: WO2022106330A1; DE102020007046B3; US20240003946A1

Abstract

The invention relates to a spectrum analyser (10) having: a signal input means or receiver (11) for receiving a signal; an analogue-to-digital converter (12) which is designed to scan the received signal and to generate a data stream of IQ data; a digital processing circuit (13) for generating compressed data from the data stream of IQ data; and a data interface (19) for transmitting the compressed data from the spectrum analyser (10).

Description

SPECTRUM ANALYZER, SYSTEM AND METHOD FOR OUTPUTTING DATA FROM A SPECTRUM ANALYZER

TECHNICAL AREA

The invention relates to spectrum analyzers and electronic computing devices designed to communicate with a spectrum analyzer. The invention relates in particular to spectrum analyzers with a data interface for extracting data for storage and/or further processing by a computer or server.

BACKGROUND

Spectrum analyzers provide a large amount of data. The data typically includes amplitude and phase information and often exists as IQ data with an in-phase (I) and quadrature (Q) component.

WO 2017/121623 A1 describes a spectrum analyzer with an increased real-time bandwidth.

US 2008/0270440 A1 describes a data compression method for generating spectral traces.

It is desirable to be able to output data from the spectrum analyzer to an attached computer or server for storage and/or further processing. The continuous output of IQ data quickly reaches its limits due to the large amount of data generated by the spectrum analyzer and the typically available data rate of the connection between the spectrum analyzer and the computer. This limits the real-time bandwidth of the IQ data that can be continuously extracted from the spectrum analyzer. For example, using a USB 3.0 connection with a data rate of 5Gbit/s results in a 40MHz real-time bandwidth limitation of the outgoing data.

For many applications it is desirable to be able to extract data from the spectrum analyzer with larger real-time bandwidths.

BRIEF DESCRIPTION OF THE INVENTION

The object of the invention is to provide improved spectrum analyzers. The invention is based in particular on the object of providing a spectrum analyzer and a method which enables data with a higher real-time bandwidth to be output via a data interface.

According to the invention there is provided a spectrum analyzer and a method as defined in the independent claims. The dependent claims define preferred or advantageous embodiments.

A spectrum analyzer according to the invention comprises: a signal input or receiver for receiving a signal, an A/D converter which can be arranged to sample the received signal and generate a data stream of IQ data, a digital processing circuit for generating compressed data the data stream of IQ data and a data interface for extracting the compressed data from the spectrum analyzer. The digital processing circuitry may be configured to perform real-time compression of the IQ data to generate the compressed data.

The digital processing circuit can be set up to compare the IQ data or data derived therefrom in the spectral space (e.g. an amplitude or power spectrum) with at least one threshold value in order to generate the compressed data. Values below the threshold can be replaced with a constant value, such as the threshold or 0.

On top of that, by replacing sub-threshold values with the constant value, the data can be efficiently derived. For example, information about the frequency ranges in which the constant value is present can be derived. Compared to deriving the original IQ data or the amplitude or power spectrum, there is a significant data reduction, since values of the IQ data or the amplitude or power spectrum that are below the threshold value do not have to be output separately for each of the frequencies.

The threshold can be fixed or user-configurable.

The electronic processing circuit can be designed to determine the threshold value automatically, for example as a function of an average noise level.

The digital processing circuit can be set up to determine an amplitude or power spectrum of the IQ data, in particular by means of a fast Fourier transformation (FFT), and to generate at least part of the compressed data from the amplitude or power spectrum.

Part of the amplitude or power spectrum can be extracted as compressed data. Since the amplitude or power spectrum does not contain phase information (which is often irrelevant to the user), the real-time bandwidth of the extracted data can be increased.

The digital processing circuit can be set up for data reduction of a noise of the amplitude or power spectrum and/or the IQ data in order to generate the compressed data.

The data reduction of the noise can have an at least partial smoothing of the noise of the amplitude or power spectrum and/or the IQ data.

The data reduction can include smoothing of background noise. To do this, parts of the spectrum in which only background noise is present can be replaced by a constant value or the mean value of the amplitude or power spectrum and/or the IQ data in the corresponding frequency range. Instead of the noisy amplitude or power spectrum, only the start and stop frequency (or other information about the position of the frequency range, such as center frequency and width) and the average level in the corresponding frequency range must be specified for the corresponding frequency range in the compressed data about the Data interface are discharged.

The digital processing circuit can be set up for data reduction of at least one peak of the amplitude or power spectrum and/or the IQ data in order to generate the compressed data. The data reduction of the at least one peak can include a smoothing of the at least one peak.

The data reduction of the at least one peak can include an approximation of the peak using a predefined carrier type.

The predefined bearer types can have spectral masks of one or more communication standards.

The spectral masks can have spectral masks according to IEEE 802.11 or cellular communication standards such as LTE (4G) or LTE-A (5G), for example.

The digital processing circuit can be set up to output a unique identifier of the identified carrier type via the data interface.

The carrier type can be parameterizable. Typical parameters may include, for example, level, width (in frequency space), and/or sectional slopes in carrier type (as a function of frequency).

The digital processing circuitry may be configured to determine a plurality of parameters of the predefined carrier type and to output the plurality of parameters as part of the compressed data.

The plurality of parameters may include a carrier type signal level and/or a carrier type width, the signal level and/or the width being automatically determined such that with this parameterization the carrier type reflects the actual peak in the IQ data and/or the amplitude or power spectrum approximated.

The spectrum analyzer can be set up to output both compressed data derived from the IQ data and compressed data dependent on the amplitude or power spectrum via the data interface.

Real-time compression of the IQ data can be performed before dumping.

In the real-time compression of the IQ data, values in the I data and the Q data that are less than a threshold may be replaced with a constant value (which may be equal to the threshold or equal to 0, for example).

This allows the data relevant for a spectrum display (e.g. information about channel utilization, transmission duration, signal strength, spurious signals, interference signals, etc.) to be transmitted with a larger real-time bandwidth, while the transmission of the IQ data (i.e. the data with amplitude and phase information) can be limited to a smaller real-time bandwidth. For example, IQ data (i.e. data with amplitude information) can be selectively output only for one or more frequency ranges that are particularly relevant to the user.

The derived IQ data may be associated with a first frequency range and the compressed data generated from the amplitude or power spectrum without phase information may be associated with a second frequency range.

The second frequency range can be larger than the first frequency range.

The digital processing circuit can be set up to determine a carrier type for each of a plurality of frequency ranges in order to generate the compressed data, which has a peak (in the IQ data or in the amplitude or power spectrum) in the corresponding frequency range approximated and to derive a unique identifier for the carrier type. The digital processing circuit can be set up to determine a respective level of the carrier type in the frequency range for generating the compressed data for each of a plurality of frequency ranges and to output it via the data interface as part of the compressed data. Information about the frequency range (e.g. start and stop frequency or center frequency and width) and/or about the carrier type (e.g. unique identification of the carrier type by an identifier) can also be extracted as part of the compressed data.

The spectrum analyzer can also have a memory coupled to the digital processing circuit for storing a plurality of predefined carrier types and/or a plurality of predefined frequency ranges (channels).

The digital processing circuit can be set up to call up information about a number of predefined carrier types and/or a number of predefined frequency ranges (channels) via the data interface or an interface of the spectrum analyzer that is separate from the data interface.

Regardless of whether a database of predefined carrier types is kept non-volatile locally in the spectrum analyzer or is called up by the spectrum analyzer from an external device, the multiple predefined carrier types can be spectral masks of at least one communication standard, in particular a radio standard (e.g. IEEE 802.11 or cellular communication standards such as LTE ( 4G) or LTE-A (5G)). The several predefined frequency ranges channels can have at least one communication standard, in particular a radio standard (e.g. IEEE 802.11 or cellular communication standards such as LTE (4G) or LTE-A (5G)).

The digital processing circuit can be set up to generate and output compressed data for a plurality of time intervals.

The compressed data can be extracted in real time.

The digital processing circuit can be set up to use one of the compression techniques described here for each of the time intervals in order to generate the compressed data.

The digital processing circuit can be set up to determine a carrier type and a level of the carrier type in the frequency range for each of the time intervals for each of a plurality of frequency ranges and to output them via the data interface as part of the compressed data. Information about the frequency range (e.g. start and stop frequency or center frequency and width) and/or about the carrier type (e.g. unique identification of the carrier type by an identifier) can also be extracted as part of the compressed data for the corresponding time interval.

The digital processing circuit can be set up to determine a number of identical temporal repetitions of a signal and to output them as part of the compressed data.

The data interface can be or have at least one of the following interfaces: a USB interface, an Ethernet interface, a wireless interface, in particular a WLAN interface or a mobile radio interface.

The digital processing circuitry may include at least one Field Programmable Gate Array, FPGA. According to a further aspect of the invention, an electronic processing unit is specified which is configured to store and/or process compressed data output by the spectrum analyzer.

The electronic processing unit can have at least one integrated semiconductor circuit, in particular at least one processor, which is designed to generate a lossy representation of the IQ data and/or the amplitude or power spectrum in the frequency domain and/or frequency-time domain from the compressed data generate.

The electronic processing unit can be configured to access a database, which assigns corresponding signal forms to the carrier types, as a function of identifiers for carrier types contained in the compressed data. The database can be stored locally in the electronic processing unit in a non-volatile manner. Alternatively or additionally, the electronic processing unit can be designed to call up the database of signal forms from the spectrum analyzer.

The electronic processing unit can be configured to perform an extrapolation between a signal shape that is associated with a carrier type defined by the compressed data and a level defined by the compressed data (and optionally having a width defined by the compressed data, and surrounding frequency ranges.

A system has a spectrum analyzer according to the invention and an electronic processing unit coupled to the data interface, in particular a computer or server, for processing and/or storing the compressed data.

A method for extracting data from a spectrum analyzer, particularly for extracting data continuously from a real-time spectrum analyzer, comprises the steps of: A/D converting acquired signals to generate a data stream of IQ data, generating compressed data from the data stream of IQ data and outputting the compressed data from the spectrum analyzer via a data interface.

The method can be performed with the spectrum analyzer or system according to an embodiment.

Other features of the method and the effects achieved thereby correspond to the features and effects described with reference to the spectrum analyzer.

The spectrum analyzer according to the invention and the method according to the invention allow data to be extracted with greater real-time bandwidth than conventional methods.

BRIEF DESCRIPTION OF THE FIGURES

Embodiments of the invention are described with reference to the figures.

Figure 1 is a block diagram of a spectrum analyzer.

Figure 2 is an exemplary amplitude spectrum.

Figure 3 illustrates a data-reduced representation of the amplitude spectrum of Figure 2 generated by compression.

Figure 4 shows example I data.

Figure 5 shows example Q data. Figure 6 illustrates a data-reduced representation of the I-data from Figure 4 generated by compression.

Figure 7 illustrates a compressed data reduced representation of the Q data of Figure 4.

Figure 8 illustrates a data-reduced representation of the amplitude spectrum of Figure 2 generated by compression.

Figure 9 illustrates a data-reduced representation of the amplitude spectrum of Figure 2 generated by compression.

Figure 10 is a block diagram of a spectrum analyzer.

Figure 11 is a schematic representation of a database of carrier types stored in or retrievable by the spectrum analyzer.

FIG. 12 is a schematic representation of a frequency-time space for explaining the operation of the spectrum analyzer.

Figure 13 is a block diagram of a spectrum analyzer.

Figure 14 illustrates a data reduced representation of the IQ data of Figure 4 and Figure 5 generated by compression.

Figure 15 is a block diagram of a system including a spectrum analyzer and an electronic processing unit according to one embodiment.

Figure 16 is a representation of an amplitude spectrum reconstructed from compressed data by an electronic computing unit.

DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS

Embodiments of the invention are described with reference to the figures, in which corresponding or similar units are denoted with corresponding or similar reference numbers. The features of different exemplary embodiments can be combined with one another unless this is expressly ruled out in the following description.

FIG. 1 is a block diagram of a spectrum analyzer 10 according to one embodiment. The spectrum analyzer 10 can be a real-time spectrum analyzer. The spectrum analyzer 10 can be set up to carry out frequency sweeps in order to determine time-dependent information about the frequency and phase position of analog signals received at a receiver 11 or an input interface. The spectrum analyzer 10 can be designed as a hand-held device or as a portable device.

The spectrum analyzer 10 has an A/D converter 12 . The A/D converter can have an A/D front end and is designed for sampling the analog signal.

The spectrum analyzer 10 has a digital processing circuit 13 . Digital processing circuitry 13 may include one or more integrated circuits. The digital processing circuit 13 may include a field programmable gate array (FPGA). The digital processing circuitry 13 may alternatively or additionally comprise one or more of a processor, a controller, an application specific integrated circuit (ASIC) or other semiconductor integrated circuits. The processing circuit 13 can have a circuit set up in terms of hardware or programming for carrying out a compression 15 . The compression generates compressed data from the IQ data, which are discharged from the spectrum analyzer 10 via a data interface 19 .

Processing circuitry 13 may include circuitry configured in hardware or programming to perform a Fast Fourier Transform (FFT) to generate the IQ data.

The term "compressed data" as used herein means data that has a reduced amount of data compared to the IQ data provided by the A/D converter 12 .

The data interface 19 can be a wired or wireless interface for data communication. During operation of the spectrum analyzer 10, the data interface 19 can be coupled to an electronic processing unit, for example a computer or server, in order to output the compressed data. The data interface 19 can be a USB interface, an Ethernet interface or a wireless interface, in particular a WLAN interface or a mobile radio interface.

The compressed data can be generated in different ways.

The processing circuitry 13 may be arranged to apply real-time compression to the IQ data (i.e. data containing amplitude and phase information). The compressed IQ data generated by real-time compression can be output via the data interface 19 as compressed data.

As will be explained in more detail with reference to Figures 4 to 7, the real-time compression of the IQ data can include smoothing of the IQ data at least in sections (e.g. by replacing the noise floor with a constant value). The smoothing, at least in sections, facilitates the compression. For example, instead of IQ data, which only represents background noise, only an indication of the start and stop frequency of the frequency range in which the IQ data was replaced by the constant value must be transmitted.

The phase information is often of little or no relevance for further processing and/or storage in a computer or server coupled to the spectrum analyzer 10 . For example, the power that is received in different frequency ranges (e.g. due to channel occupancy or secondary signals) is relevant for many users, but not the phase position of the respective signals. Information for what is known as a spectrum display (such as channel utilization, transmission duration, signal strength, secondary signals and/or interference signals) can be transmitted by transmission of the amplitude spectrum, ie a spectral display without phase information.

The processing circuit 13 can be set up to generate the compressed data from a spectral representation, in particular an amplitude spectrum or a power spectrum, and to export it via the data interface 19 . The amplitude spectrum can be processed by the processing circuit 13 according to, for example, %( ) = be calculated. A part of the amplitude spectrum or the entire amplitude spectrum can be output via the data interface 19 . Alternatively, a power spectrum can be calculated by the processing circuit 13 according to P( )=/( ) ² +<2(/) ² , for example will. A part of the power spectrum or the entire power spectrum can be output via the data interface 19 . Only the term amplitude spectrum is used below. It should be understood that the techniques described below can also be applied when the power spectrum (or other spectral representation of the analog signal that quantifies the amplitude or power) is determined and further compressed.

The calculation and output of part of the amplitude spectrum can be repeated (e.g. after completion of a frequency sweep) in order to derive the information relevant to a spectrum display as a function of time.

The processing circuit 13 can be set up to subject the amplitude spectrum and/or the complex-valued spectral representation, which is represented by the I and Q data, to one or more further processing steps for data reduction before the compressed data is output. Some possible processing steps for further data reduction are illustrated with reference to Figure 2 to Ligur 14.

The processing circuit 13 can be set up to smooth the amplitude spectrum. The smoothing can be done at least in sections. The smoothing can be done depending on whether there is only noise in the frequency domain or a peak of the amplitude spectrum (i.e. a carrier). The smoothing can take place as a function of one or more predefined frequency ranges (for example frequency ranges assigned to a plurality of channels of a data communication technology, e.g. channels according to IEEE 802.11, LTE (4G) or LTE-A (5G)).

The processing circuit 13 can be set up to identify and smooth a background noise in the spectrum. The identification of the noise floor may involve a threshold comparison with a threshold 39 . Values below the threshold can be replaced with a constant value. The constant value can be equal to 0 or equal to the threshold. The smoothing of the noise floor can thus correspond to a "clipping" of signal values lying below the threshold value. The constant value with which the noise floor is replaced can be determined as a function of the noise floor, for example by averaging the noise floor in the corresponding frequency range. The smoothing can include an averaging of the noise floor. The averaging can take place over a continuous frequency range in which only background noise is present.

The spectral representation does not have to be transmitted in the compressed data for a frequency range in which there is only background noise. For example, the processing circuitry 13 may generate the compressed data such that, instead of the noise floor, the compressed data includes an indication of the frequency range (e.g. an indication of start and stop frequency or some other indication such as center frequency and width) and (if the noise level is of interest) a Have an indication of the smoothed noise level of the noise floor. The smoothed noise level can be determined by averaging the noise floor in the corresponding frequency range.

Figure 2 and Figure 3 illustrate this compression technique. FIG. 2 shows an exemplary amplitude spectrum 30 with a plurality of peaks 31, 32. In the frequency ranges from fi to fz, T to fj and T to fs, the amplitude spectrum only has background noise but no carrier.

The compressed data can be generated and extracted in such a way that the corresponding frequency ranges fi to f2, T to fj and T to fs, in which only noise floor is present, and a level of the background noise in the respective frequency ranges (which can be determined, for example, by averaging the amplitude spectrum in the corresponding frequency range).

The compressed data can be generated and extracted in such a way that they contain at least the amplitude information for the frequency ranges in which peaks 31, 32 are present (or more generally in relevant frequency ranges which can be determined, for example, by the channels of a data communication standard). For example, in the frequency ranges from £2 to £3 and from £4 to ft, the full amplitude spectrum can be included in the compressed data. By smoothing the noise floor, significant compression is achieved, which allows data to be extracted with a higher real-time bandwidth.

Compression, which is a smoothing of the noise floor with subsequent extraction of only the cut-off frequencies of the frequency range in which only the noise floor is present, and optional extraction of the constant value that replaces the noise floor in the corresponding frequency range, can be applied not only to an amplitude or power spectrum, but can also be applied to the IQ data. This is explained with reference to FIG. 4 to FIG.

FIG. 4 and FIG. 5 show, by way of example, I data 40 and Q data 50 in the frequency domain. The I data 40 has one or more peaks 41,42. The Q data 50 has one or more further peaks 51 .

The evaluation circuit 13 can subject the I data 40 and Q data in the time or frequency space to a threshold value comparison with a threshold value 49 (which can be chosen to be the same or different for the I and Q data). The evaluation circuit 13 can smooth out background noise by setting all data falling below the threshold value 49 to a constant value. For example, as described above, the constant value can be equal to the threshold value, equal to 0, or equal to an average value of the noise floor.

For the frequency ranges from fi to £2, £3 to £4 and ft to £5, in which there is only background noise in the I data, the smoothed I data I _c is transmitted in compressed form in the frequency space (Figure 6) in that only an indication of the frequency ranges (e.g. by specifying the limit frequencies fi to £2, £3 to £4 and ft to £5) and optionally an instance of the constant value is output. The peaks 41, 42 can be output without additional compression or, as described in more detail below, also further compressed. A similar technique can also be used in time.

For the frequency ranges from fi to £2 and £3 to £5, in which there is only background noise in the Q data, the smoothed Q data Q _c is transmitted in the frequency space (FIG. 7) in compressed form by only specifying the frequency ranges (eg by specifying the limit frequencies fi to £2 and £3 to £5) and optionally an instance of the constant value is derived. The peak 51 can be output without additional compression or, as described in more detail below, also further compressed. A similar technique can also be used in time.

In a further refinement, the signal components or carriers which correspond to the peaks 31, 32, 41, 42, 51 in the spectrum can also be simplified in terms of their complexity in order to reduce the amount of data. For example, a smoothing operation can be applied to peaks 31, 32, 41, 42, 51. The noisy peaks 31, 32, 41, 42, 51 can be included in the compressed data as a smoothed and thus data-reduced carrier. Due to the smoothing, the peaks 31, 32, 41, 42, 51, for example, with a frequency resolution Af in the frequency ranges from £2 to £3 and from £4 to fi, which is coarser than the frequency resolution of the IQ data.

Alternatively or additionally, the data-reduced extraction of the peaks 31, 32, 41, 42, 51 in the compressed data can take place with higher compression in that signals/carrier types and their signal form are assigned to predefined carrier types that are stored in a database locally in the memory of the spectrum analyzer 10 stored or can be called up by the spectrum analyzer 10 from a separate device. The processing circuit 13 can be set up to include only the carrier type and its signal strength and position in the spectrum in the compressed data and to output it from the spectrum analyzer. This can also be applied to the noise floor.

For example, a spectrum with a single carrier 31, 32, 41, 42, 51 in the spectrum and with the noise floor surrounding the carrier can then be transmitted as follows: noise with level "xi" at start frequency "fi" and stop frequency "£2", Type 'A' signal with level 'X2' at start frequency '£2' and stop frequency '£3', noise with level 'X3' at start frequency '£3' and stop frequency '£4' etc. For each of several frequency ranges Thus the processing circuit 13 generates the compressed data in such a way that it contains, for example, a unique identifier for one of several predefined carrier types (e.g. different signal forms in the frequency domain) or noise, a signal level and an indication of the frequency range (e.g. an indication of the start and stop frequency or a Center frequency and width). The specification of the frequency range can be simplified or omitted if different channels are predefined and/or carriers are clearly assigned to respective frequency ranges.

Spectrum analyzer 10 can be configured to output a list of database based carrier types with frequency and level in the compressed data. This is illustrated by way of example in FIG. The spectrum 30 of Figure 2 can be processed in such a way that, instead of the peaks 31, 32, an identifier for a carrier type (e.g. for a peak in the form of a Gaussian curve, a Lorenz curve, etc.), the signal level (which indicates the height of the amplitude of the indicates peaks) and the frequency range (indicating the width of the peak) in the compressed data. The peaks 31, 32 are thus effectively replaced by data-reduced peaks 37, 38, which are determined on the basis of carrier types defined in a database.

The various compression techniques for data reduction can be combined. A specific carrier can thus be transmitted as IQ data (in particular including phase information, optionally including noise), the background noise can be smoothed and all other carriers can be transmitted with reference to a database of carrier types. With reference to Figure 2 and Figure 9, the processing circuit 13 can be set up in such a way that the peak 31 as amplitude or power data (ie without phase information) or as IQ data (including phase information not shown in Figure 9), the noise floor in the frequency ranges of fi to fi, fi to fi and fi to fi are smoothed and the peak 32 is transmitted by reference to a carrier type in a database (and thus as a data reduced peak 38). FIG. 10 is a block diagram of a spectrum analyzer 10. The spectrum analyzer has a memory 18 in which predefined carrier types and/or predefined frequency ranges (for example carrier frequencies of a data communication standard) are stored. The processing circuit 13 is set up to match carriers (ie peaks in the spectral representation) with the carrier types in the memory and/or to retrieve information about those frequency ranges in which background noise can be replaced by smoothed, constant noise levels. This information can be used to generate the compressed data for ejection.

The carrier types and/or predefined frequency ranges stored in the memory 18 can be user-configurable, for example depending on the intended use of the spectrum analyzer.

The carrier types stored in memory 18 can have spectral masks of at least one communication standard, in particular a radio standard (e.g. IEEE 802.11, LTE (4G) or LTE-A (5G)).

The predefined frequency ranges stored in the memory 18 channels can have at least one communication standard, in particular a radio standard (e.g. IEEE 802.11, LTE (4G) or LTE-A (5G)).

FIG. 11 shows a database 60 of carrier types by way of example. Various signal forms 61, 62 are stored in the database 60 with a unique identifier assigned to them. The waveforms can be parameterizable waveforms. For example, gradients and/or widths of sections of a signal form 62 can be variable parameters. While only two waveforms are exemplified, database 60 may store more waveforms (e.g., more than ten, more than twenty, etc.) waveforms.

In operation, the processing circuit 13 can determine whether a part (particularly a peak) in the amplitude or power spectrum 30, in the I-data 40 and/or the Q-data 50 can be approximated by one of the waveforms 61, 62 of the database. During operation, the processing circuit 13 can determine which of the signal forms 61, 62 of the database 60 optimally approximates a part (in particular a peak) in the amplitude or power spectrum 30, in the I-data 40 and/or the Q-data 50. For this purpose, a difference between the detected spectrum 30, 40, 50 and each of the several signal forms can be determined and evaluated with a similarity metric (e.g. a root mean square) in order to determine which of the signal forms 61, 62 in the database 60 has a part (in particular a peak) in the amplitude or power spectrum 30, in the I data 40 and/or the Q data 50 is optimally approximated.

In the case of a parameterizable signal form 62, one or more parameters, such as gradients in sections of the signal form and/or widths of sections of the signal form, can also be determined, so that the signal form with its parameterization forms a part (in particular a peak) in the amplitude or power spectrum 30, optimally approximated in the I data 40 and/or the Q data 50.

If a database 60 with several carrier types is used, then only a unique identifier for one of several predefined carrier types (e.g. different signal forms in the frequency domain) or noise, a signal level, an indication of the frequency range (e.g. an indication of the start and stop frequency or a mean frequency and width) and optionally, in the case of a parameterizable signal form, one or more additional parameters of the signal form are output via the data interface.

Alternatively or additionally, the processing circuit 13 can be set up to call up information about carrier types and/or predefined frequency ranges from a memory separate from the spectrum analyzer 10 . The retrieval can take place via the data interface 19 or a different data interface 19 .

The processing circuitry 13 can then use this information to generate the compressed data, for example by any of the methods described above.

Alternatively or additionally, the spectrum analyzer 10 can be set up to output information about carrier types and/or predefined frequency ranges in the database via the data interface 19 to the electronic processing unit, which receives the compressed data. In this way, the electronic processing unit in the spectrum analyzer 10 can call up non-volatile information about signal forms 61, 62 and their respective identifiers and use them to process the compressed data.

The determination and deriving of carrier types at different frequencies (e.g. at the channels of a communication standard) can be repeated time-sequentially. This can be done, for example, after a frequency sweep has been carried out by the spectrum analyzer 10 .

FIG. 12 shows an example of a frequency period. At a first time 61, spectrum analyzer 10 may collect a first set of IQ data. For this purpose, the spectrum analyzer 10 can carry out a frequency sweep. From the IQ data collected at the first time 61, the processing circuitry 13 can determine that an amplitude or power spectrum (or an FFT of the IQ data) at a first frequency 71 can be represented by a first carrier type, schematically represented as a shaded area 81 is shown. From the IQ data collected at the first time 61, the processing circuitry 13 can determine that the amplitude or power spectrum (or an FFT of the IQ data) at a second frequency 72 can be represented by a second carrier type, indicated schematically as a shaded area 82 is shown. Identifiers for these carrier types and optionally other parameters (level, cut-off frequencies/peak width) can be extracted from the compressed data. The carrier types can each be automatically recognized by database comparison.

At a second time 62, spectrum analyzer 10 may collect a second set of IQ data. For this purpose, the spectrum analyzer 10 can carry out a further frequency sweep. From the IQ data collected at the second time 62, the processing circuitry 13 can determine that an amplitude or power spectrum (or an FFT of the IQ data) at the first frequency 71 can be represented by a third carrier type, indicated schematically as a shaded area 83 is shown. From the IQ data collected at the second time 62, the processing circuitry 13 can determine that the amplitude or power spectrum (or an FFT of the IQ data) at the second frequency 72 can be represented by a fourth type of carrier, shown schematically as shaded area 84. FIG. Identifiers for these carrier types and optionally other parameters (level, cut-off frequencies/peak width) can be extracted from the compressed data. The carrier types can each be automatically recognized by database comparison.

At a third time 63, spectrum analyzer 10 may acquire a third set of IQ data. For this purpose, the spectrum analyzer 10 can carry out a further frequency sweep. From the IQ data collected at the third time 63, the processing circuitry 13 can determine that an amplitude or power spectrum (or an FFT of the IQ data) at the first frequency 71 can be represented by a fifth carrier type, indicated schematically as a shaded area 85 is shown. From the IQ data collected at the third time 63, the processing circuitry 13 can determine that the amplitude or power spectrum (or an FFT of the IQ data) at the second frequency 72 can again be represented by the second carrier type, schematically indicated as hatched Area 86 is shown. Identifiers for these carrier types and optionally other parameters (level, cut-off frequencies/peak width) can be extracted from the compressed data. The carrier types can each be automatically recognized by database comparison.

For some of the occupied channels 71, 72, data with phase information can be output selectively. For example, data containing phase information (e.g. compressed IQ data) can optionally be extracted for one of a plurality of channels 71, 72, while only amplitude or power information without phase information is used for another of a plurality of channels 71, 72 in order to generate compressed data.

Another possibility for further data reduction concerns the time domain. Thus, some of the previously mentioned techniques can also be applied to the time domain. Here there is another possibility for reducing the amount of data, which scales linearly with the size of the time buffer, but at the expense of a corresponding time delay. For example, instead of transmitting N > 2 time-sequentially repeated spectra of an identical signal, it could also be summarized as follows: noise with level "xi" at start frequency "fi" and stop frequency "f ₂ ", signal of type "A" with level "x ₂ ” at start frequency “f ₂ ” and stop frequency “f ₃ ”, noise with level “x ₃ at start frequency “f ₃ ” and stop frequency “f ₄ ”, etc. For a time-repeating signal, instead of identical transmission of this data - on Cost of transmission latency - an indication N of the number of retries in which compressed data is output.

13 is a block diagram of a spectrum analyzer 10. The processing circuit 13 of the spectrum analyzer 13 has a hardware or programming circuit for performing compression of the amplitude spectrum 16 and a hardware or programming circuit for selecting IQ data 17 to be transmitted.

IQ data to be transmitted (ie data that is output including phase information) can be selected, for example, depending on predefined frequency ranges and/or depending on signal levels of the carrier types that are stored in memory 18 . The IQ data to be transmitted can optionally be smoothed, for example by applying a smoothing operation to the I and Q data before transmission. The compression of the amplitude spectrum in those frequency ranges for which no IQ data is transmitted may include smoothing and/or by reference to carrier types stored in memory 18, as described above.

FIG. 14 illustrates this by way of example for the IQ data in the frequency domain which are shown in FIG. For a frequency range from f2 to f1. which may correspond to a channel of a wired or wireless communication link, data including phase information may be extracted. For this purpose, for example, a signal form that is defined as a signal mask for the corresponding standard can be read out from the database 60 and a unique identifier for the signal form can be derived. As a result, the peaks 41, 51 are effectively approximated by a signal form 62 from the database. For all frequency ranges outside of one or more channels (e.g. for frequencies lower than fz and frequencies higher than T) only amplitude information can be transmitted. This can be compressed as explained with reference to FIGS.

In each of the described embodiments, the processing circuitry 13 may be an FPGA or may include an FPGA.

In each of the described embodiments, the data interface 19 can be selected from the group consisting of a USB interface, an Ethernet interface, a wireless interface, in particular a WLAN interface or a mobile radio interface.

FIG. 15 is a block diagram of a system 10 which has a spectrum analyzer 10 and an electronic arithmetic unit 95 according to the invention. Spectrum analyzer 10 may include a user interface 91 through which information about channel occupancy, waveforms, detected signals in time or frequency domain, or other information may be output.

The electronic processing unit 95 can be or can be coupled to the signal analyzer 10 via a unidirectional or bidirectional data connection 98 . The electronic processing unit 95 is designed to process and/or store the compressed data output by the signal analyzer 10 .

The electronic computing unit 95 can be configured to reconstruct at least one amplitude or power spectrum from the compressed data. Due to the compression of the data for transmission, the reconstructed amplitude or power spectrum is typically lossy.

The electronic processing unit can have at least one integrated semiconductor circuit 96, in particular at least one processor, which is designed to generate a lossy representation of the amplitude or power spectrum in the frequency domain and/or frequency-time domain from the compressed data.

The electronic processing unit 10 can be configured to access a database 60, which assigns corresponding signal forms to the carrier types, as a function of identifiers for carrier types contained in the compressed data. The database 60 can be stored locally in the electronic processing unit 95 in a non-volatile manner. Alternatively or additionally, the electronic processing unit 95 can be designed to call up the database 60 of signal shapes from the spectrum analyzer 10 .

The electronic computing unit 95 can be configured to reconstruct the lossy representation of the amplitude or power spectrum, an extrapolation between a waveform 62 associated with a carrier type defined by the compressed data and having a level (and optionally a width defined by the compressed data) defined by the compressed data, and surrounding frequency ranges.

FIG. 16 illustrates such processing by the electronic arithmetic unit 95. For the purpose of illustration, it is assumed that an indicator for the signal form 61 of the database 60 with an associated level in the frequency range f2 to T was contained in the compressed data. For the adjacent frequency ranges from fi to fz and from T, indicators were included in the compressed data that only show noise floor.

For the (lossy) reconstruction of the original amplitude spectrum, the electronic arithmetic unit 95 can reconstruct the carrier 101 in the frequency range f2 to T from the indicator for the signal form 61 and the associated level. At the rising and falling edges, an extrapolation 102, 103 towards the adjacent constant value can be undertaken, which represents a (weak) background noise in the original amplitude spectrum.

The electronic processing unit 95 can output and/or further process the reconstructed information via a user interface 97 .

The spectrum analyzer according to the invention and the method according to the invention make it possible, due to the compression of the data before it is discharged from the spectrum analyzer, to output amplitude information (optionally at least also phase information for certain frequency ranges) with a higher real-time bandwidth.

The spectrum analyzer and the method according to the invention can be used to output data to a computer or server for further processing and/or storage, without being limited thereto.

Claims

PATENT CLAIMS

1. Spectrum analyzer, in particular real-time spectrum analyzer, comprising: a signal input or receiver (11) for receiving a signal, an A/D converter (12) which is set up to sample the received signal and a data stream of IQ data generate, a digital processing circuit (13) for generating compressed data from the data stream of IQ data and a data interface (19) for extracting the compressed data from the spectrum analyzer (10).

2. Spectrum analyzer according to claim 1, wherein the digital processing circuit (13) is set up for at least partial smoothing of the IQ data or spectral data derived therefrom, in particular an amplitude or power spectrum, in order to generate at least part of the compressed data, with optionally the at least partial smoothing has a threshold value comparison.

3. Spectrum analyzer according to claim 1 or claim 2, wherein the digital processing circuit (13) is set up for determining an amplitude or power spectrum (30) of the IQ data and for generating at least part of the compressed data from the amplitude or power spectrum (30 ).

4. Spectrum analyzer according to claim 3, wherein the digital processing circuit (13) is set up for data reduction of a noise of the amplitude or power spectrum (30) in order to generate the compressed data, the data reduction of the noise at least partially smoothing the noise of the amplitude or range of services.

5. Spectrum analyzer according to claim 3 or claim 4, wherein the digital processing circuit is set up for data reduction of at least one peak (31, 32) of the amplitude or power spectrum in order to generate the compressed data, optionally the data reduction of the at least one peak (31, 32) has smoothing.

6. Spectrum analyzer according to claim 5, wherein the data reduction of the at least one peak (31, 32) comprises an approximation of the peak by a predefined carrier type (61, 62), an identifier for the carrier type being derived in the compressed data, the digital Processing circuit (13) is optionally set up to determine multiple parameters of a parameterization of the predefined carrier type and to derive the multiple parameters as part of the compressed data, the predefined carrier type optionally having a spectral mask of at least one communication standard, in particular a radio standard.

7. Spectrum analyzer according to one of claims 3 to 6, wherein the spectrum analyzer (10) is set up to derive both compressed data generated from the IQ data and compressed data generated from the amplitude or power spectrum via the data interface (19), with optional the extracted IQ data are assigned to a first frequency range and the compressed data generated from the amplitude or power spectrum are assigned to a second frequency range, which is different from the first frequency range.

8. Spectrum analyzer according to one of the preceding claims, wherein the digital processing circuit (13) is set up to generate the compressed data for each of a plurality of frequency ranges in each case to determine a level of a carrier type in the frequency range and via the data interface (19) as part of the compressed derive data.

9. Spectrum analyzer according to claim 8, further having a memory (18) coupled to the digital processing circuit (13) for storing a plurality of predefined carrier types and/or a plurality of predefined frequency ranges, wherein optionally the multiple predefined carrier types have spectral masks of at least one communication standard, in particular a radio standard , and/or optionally wherein the plurality of predefined frequency ranges channels have at least one communication standard, in particular a radio standard.

10. Spectrum analyzer according to claim 8, wherein the digital processing circuit (13) is set up to retrieve information about a plurality of carrier types and/or a plurality of frequency ranges via the data interface or an interface of the spectrum analyzer that is separate from the data interface, with the plurality of predefined carrier types optionally containing at least one spectral mask Having communication standards, in particular a radio standard, and/or wherein optionally the plurality of predefined frequency ranges channels have at least one communication standard, in particular a radio standard.

A spectrum analyzer as claimed in any preceding claim, wherein the digital processing circuit (13) is arranged to generate and output compressed data for a plurality of times respectively, optionally wherein the digital processing circuit (13) is arranged to for at least one of the times for each to determine a carrier type, a position of the frequency range and a level of the carrier type in the frequency range from a plurality of frequency ranges and to output them via the data interface as part of the compressed data, with the digital processing circuit (13) also optionally being set up to record a number of time repetitions of a signal determine and output as part of the compressed data.

12. Spectrum analyzer according to one of the preceding claims, wherein the data interface (19) has at least one of the following interfaces: a USB interface, an Ethernet 18

Interface, a wireless interface, in particular a WLAN interface or mobile radio interface.

13. Spectrum analyzer according to one of the preceding claims, wherein the digital processing circuit (13) comprises at least one field programmable gate array, FPGA.

14. System, comprising the spectrum analyzer according to any one of the preceding claims and with the data interface (19) coupled or coupled electronic processing unit for processing and / or storing the compressed data, wherein the electronic processing unit is optionally configured to convert the compressed data into a to reconstruct a lossy spectral representation of the IQ data and/or an amplitude or power spectrum determined from the IQ data, with the reconstructing optionally including access to a database of predefined carrier types (61, 62) and an extrapolation (102, 103) between at least one of the predefined carrier types (61, 62) and a constant signal in adjacent frequency ranges.

15. A method for deriving data from a spectrum analyzer (10), in particular for continuously deriving data from a real-time spectrum analyzer, comprising the steps:

A/D converting acquired signals to generate a data stream of IQ data, generating compressed data from the data stream of IQ data and

Extracting the compressed data from the spectrum analyzer (10) via a data interface (19) of the spectrum analyzer (10).