Packet correlator for a radio transmission system with time-varying frequency drift

Description

[0001] Embodiments of the present invention relate to a data receiver and, more particularly, to a data receiver for receiving a signal subject to a time-varying frequency drift. Some embodiments relate to a packet correlator for a radio transmission system with time-varying frequency drift.

[0002] When transmitting a signal which has a data packet with at least two spaced-apart pilot sequences or which, as described in [2] or [4], has a plurality of partial data packets, each of which has a pilot sequence and which is distributed in time and frequency according to a hopping pattern, the temporally spaced-apart pilot sequences can be subjected to a time-varying frequency shift (frequency drift), for example caused by a relative change in movement between the data transmitter and data receiver or a time-varying mismatch between the frequency generators (e.g. oscillators) of the data transmitter and data receiver.

[0003] Conventionally, a data packet or partial data packets transmitted with such a signal can no longer be detected based on the pilot sequences, in particular if these are subject to a larger time-varying frequency drift, such as a frequency drift of more than 1 / 8 of the symbol rate used in Hertz.

[0004] US 2008 / 158050 A1 describes a GPS receiver with sampling time error and frequency offset compensation. GPS signals are resampled at multiple sampling rates for each of a plurality of predetermined frequency offsets to compensate for a sampling time error and a frequency offset of a crystal oscillator. The resampled GPS signals can be cross-correlated for each of the predetermined frequency offsets. In some embodiments, the sampling time error and frequency offset of the crystal oscillator can significantly exceed a time and frequency drift of reference clocks on GPS satellites.

[0005] The present invention is therefore based on the object of creating a concept which makes it possible to detect pilot sequences which are spaced apart in time and which are subject to a larger time-varying frequency drift.

[0006] This problem is solved by the independent patent claims.

[0007] Advantageous further developments can be found in the dependent patent claims.

[0008] Embodiments provide a data receiver, wherein the data receiver is configured to receive a signal from a data transmitter [e.g., stationary station, satellite, or relay], wherein the signal comprises at least two pilot sequences distributed [e.g., spaced apart from one another] in time [e.g., and optionally in frequency] according to a pilot pattern, wherein at least a second pilot sequence of the at least two pilot sequences has a time-varying frequency shift [e.g., frequency drift] [e.g., due to a Doppler effect] compared to a first pilot sequence of the at least two pilot sequences, wherein the data receiver comprises a detector having a correlator configured to detect the at least two pilot sequences based on a correlation pattern [e.g., corresponding to the pilot pattern or derived from the pilot pattern], wherein the correlation pattern is distributed [e.g.,to the time-varying frequency shift], e.g. to reduce the influence of the time-varying frequency shift [e.g. to compensate].

[0009] In embodiments, the data receiver [e.g., the detector of the data receiver or the correlator itself] may be configured to adjust the correlation pattern in frequency, e.g., to reduce [e.g., to compensate] an influence of the frequency shift.

[0010] In embodiments, the signal may comprise a data packet, wherein the data packet comprises the at least two pilot sequences.

[0011] In embodiments, the signal may comprise at least two partial data packets, each of the at least two partial data packets comprising a pilot sequence of the at least two pilot sequences.

[0012] In embodiments, the correlation pattern may describe a distribution of the at least two pilot sequences in time [and optionally in frequency].

[0013] In embodiments, at least one of the data transmitter and the data receiver can move relative to the other of the data transmitter and the data receiver, wherein the time-varying frequency shift results from a relative change in movement between the data transmitter and the data receiver.

[0014] In embodiments, the time-varying frequency shift may also additionally or alternatively result from a time-varying mismatch between oscillators of the data transmitter and the data receiver.

[0015] In embodiments, the data receiver may be configured to estimate the time-varying frequency shift of the at least one second pilot sequence relative to the first pilot sequence and to adjust the correlation pattern in frequency based on the estimated time-varying frequency shift.

[0016] In embodiments, the correlator may comprise at least two correlation units operating in parallel, wherein a first correlation unit of the at least two correlation units operating in parallel is configured to perform correlations of the at least two pilot sequences with reference sequences based on the correlation pattern in order to obtain a set of correlation results of the first correlation unit, wherein a second correlation unit of the at least two correlation units operating in parallel is configured to perform correlations of the at least two pilot sequences with reference sequences based on the correlation pattern in order to obtain a set of correlation results of the second correlation unit, wherein the data receiver is configured to adapt the correlation pattern used in the at least two correlation units operating in parallel differently in frequency, e.g.in order to reduce [e.g. compensate] influences of different frequency shifts, wherein the data receiver is configured to select a set of correlation results from the sets of correlation results of the at least two correlation units operating in parallel [e.g. select a set of correlation results from . the set of correlation results of the first correlation unit, and the set of correlation results of the second correlation unit,] based on which the detection of the at least two pilot sequences takes place, depending on values ​​of the respective set of correlation results.

[0017] For example, the data receiver may be configured to receive the at least two pilot sequences based on one of the set of correlation results of the first correlation unit, and the set of correlation results of the second correlation unit, which have the highest values.

[0018] In embodiments, the correlator may be a multi-stage correlator having a first correlation stage and at least one second correlation stage following the first correlation stage [e.g., a second correlation stage and optionally a third correlation stage], which operates based on correlation results of the first correlation stage, wherein the detector is configured to detect the at least two pilot sequences in the received signal, wherein the data receiver is configured to frequency-adjust at least one correlation pattern used in at least one correlation stage from the at least one second correlation stage of the multi-stage correlator, e.g., to reduce [e.g., compensate] the influence of the frequency shift.

[0019] For example, the multi-stage correlator may comprise a first correlation stage and a second correlation stage, wherein the data receiver is configured to frequency-adjust a correlation pattern used in the second correlation stage of the multi-stage correlator, e.g., to reduce [e.g., compensate] the influence of the frequency shift.

[0020] For example, the multi-stage correlator may comprise a first correlation stage, a second correlation stage, and a third correlation stage, wherein the data receiver is configured to frequency-adjust a correlation pattern used in the second correlation stage of the multi-stage correlator and / or a correlation pattern used in the third correlation stage of the multi-stage correlator, e.g., to reduce [e.g., compensate] the influence of the frequency shift.

[0021] For example, the signal may comprise at least two partial data packets [e.g., a plurality of partial data packets] distributed in time and frequency according to a hopping pattern, wherein the at least two partial data packets each comprise a pilot sequence of the at least two pilot sequences [e.g., a plurality of pilot sequences], wherein the data receiver may be configured to detect the at least two partial data packets based on the respective pilot sequences in the received signal. The pilot pattern may, for example, be the same as the hopping pattern or, for example, a time-shifted version of the hopping pattern [e.g., if the hopping pattern defines absolute times of the partial data packets or the pilot sequences are located at different positions within the partial data packets than those defined by the hopping pattern [e.g., beginnings of the partial data packets vs. midambles]].

[0022] In embodiments, the at least two pilot sequences may be distributed in time and frequency according to a pilot pattern, wherein the multi-stage correlator is configured to detect the at least two pilot sequences in the received signal or a derived version thereof [e.g., a plurality of subband signals].

[0023] In embodiments, the received signal may comprise a plurality of subband signals, wherein the plurality of subband signals comprise different [e.g., partially overlapping] subbands of the signal [e.g., broadband signal].

[0024] For example, the data receiver may be configured to obtain a received signal comprising the plurality of subband signals based on the signal [e.g., wideband signal].

[0025] In embodiments, the plurality of subband signals may be used directly for the correlation performed by the multi-stage correlator.

[0026] In embodiments, the multi-stage correlator may be configured to perform a multi-stage correlation of at least a subset of the plurality of subband signals to detect the at least two pilot sequences in the subset of the plurality of subband signals.

[0027] In embodiments, the first correlation stage may be configured to correlate the received signal or a version derived therefrom with a plurality of pilot sequence sections corresponding to different sections of the at least two pilot sequences [e.g., the plurality of partial data packets] to obtain a plurality of section correlation results, wherein the first correlation stage is configured to combine the plurality of section correlation results to obtain a set of correlation results or a subset of correlation results as correlation results of the first correlation stage.

[0028] In embodiments, the at least two pilot sequences may be the same, e.g., map the same bit sequence.

[0029] In embodiments, the first correlation stage may be configured to normalize the plurality of section correlation results [e.g., by forming magnitude squares].

[0030] In embodiments, the first correlation stage may be configured to normalize the plurality of section correlation results depending on a determined [e.g., calculated] power [p[n]] of the received signal or the version derived therefrom [e.g., the filtered and / or stored version of the signal to be received].

[0031] For example, the first correlation stage may be configured to normalize the section correlation results by forming magnitude squares, dividing by the determined power, and calculating the roots of the quotients.

[0032] In embodiments, the power for normalization can be determined over several subbands.

[0033] In embodiments, the first correlation stage may be configured to separately normalize the plurality of section correlation results, wherein the power is determined separately for each pilot sequence section or jointly for all pilot sequence sections.

[0034] In embodiments, the first correlation stage may comprise a plurality of queue buffers [e.g., ring buffers] configured to buffer the respective section correlation results, wherein the plurality of queue buffers have different storage lengths, wherein the storage lengths of the plurality of queue buffers are dependent on the respective pilot sequence sections.

[0035] In embodiments, the first correlation stage may be configured to correlate at least two subband signals of the plurality of subband signals [e.g., a plurality of subband signals of the plurality of subband signals or all subband signals of the plurality of subband signals] with the plurality of pilot sequence sections, respectively, to obtain a subset of correlation results [e.g., (normalized) correlation amplitudes or a one-dimensional array of (normalized) correlation amplitudes] for each subband signal of the at least two subband signals, wherein the first correlation stage is configured to provide, as correlation results of the first correlation stage, a set of correlation results comprising the subsets of correlation results.

[0036] For example, the set of correlation results may include the one-dimensional subsets of correlation results.

[0037] In embodiments, the set of correlation results of the first correlation stage may be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results [e.g., a sequence of] describes sampling times of the received signal [e.g., time direction], wherein a second dimension of the two-dimensional array of correlation results describes subbands of the received signal [e.g., frequency direction].

[0038] In embodiments, the first correlation stage may comprise an [e.g., multi-channel] output queue buffer [e.g., ring buffer] configured to buffer the set of correlation results of the first correlation stage.

[0039] In embodiments, the first correlation stage may be configured to perform a maximum calculation over correlation results of adjacent subband signals and to discard the smaller values.

[0040] In embodiments, the at least two pilot sequences may be a plurality of pilot sequences, wherein at least two groups of pilot sequences of the plurality of pilot sequences have the same relative group pilot pattern in groups [e.g., or wherein at least two groups of partial data packets of the plurality of partial data packets have the same relative group hopping pattern in groups], wherein a second correlation stage of the at least one second correlation stage of the multi-stage correlator is configured to select groups of correlation results from the set of correlation results of the first correlation stage based on a group correlation pattern derived from the group pilot pattern [e.g., or group hopping pattern] and to combine them in groups to obtain a set of correlation results of the second correlation stage.

[0041] For example, a second pilot sequence of the first group of pilot sequences may have the same time and frequency spacing from a first pilot sequence of the first group of pilot sequences as a fourth pilot sequence of the second group of pilot sequences has from a third pilot sequence of the second group of pilot sequences.

[0042] In embodiments, the second correlation stage may be configured to select the groups of correlation results from the set of correlation results of the first correlation stage in the time and / or frequency direction based on the group correlation pattern.

[0043] In embodiments, the set of correlation results of the first correlation stage may be a two-dimensional array of correlation results, wherein the group correlation pattern indicates time and frequency spacings of the correlation results of the two-dimensional array of correlation results of the first correlation stage that correspond to the relative time and frequency spacings of the group hopping pattern of the groups of pilot sequences.

[0044] In embodiments, the set of correlation results of the second correlation stage may be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results describes a [e.g., relative] temporal position of the group of pilot sequences [time direction], wherein a second dimension of the two-dimensional array of correlation results describes a [e.g., relative] frequency position of the group of pilot sequences [frequency direction].

[0045] In embodiments, at least one dimension [e.g., frequency direction] of the two-dimensional array of correlation results of the second correlation stage may be smaller than the respective at least one dimension of the two-dimensional array of correlation results of the first correlation stage.

[0046] In embodiments, the second correlation stage may comprise an [e.g., two-dimensional] output queue buffer [e.g., ring buffer] configured to temporarily store the set of correlation results of the second correlation stage.

[0047] In embodiments, the at least two groups of pilot sequences may form a sequence, wherein the at least two groups of pilot sequences have a relative group sequence pilot pattern [e.g., or wherein the at least two groups of partial data packets have a relative group sequence hopping pattern to each other], wherein a third correlation stage of the at least one second correlation stage of the multi-stage correlator is configured to select groups of correlation results from the set of correlation results of the second correlation stage based on a group sequence correlation pattern derived from the group sequence pilot pattern [e.g., or group sequence hopping pattern] and to combine them in groups to obtain a set of correlation results of the third correlation stage.

[0048] In embodiments, the third correlation stage may be configured to select the groups of correlation results from the set of correlation results of the second correlation stage in the time and / or frequency direction based on the group sequence correlation pattern.

[0049] In embodiments, the set of correlation results of the second correlation stage may be a two-dimensional array of correlation results, wherein the group sequence correlation pattern indicates time and frequency spacings of the correlation results of the two-dimensional array of correlation results of the second correlation stage that correspond to the relative time and frequency spacings of the group sequence hopping pattern.

[0050] In embodiments, the set of correlation results of the third correlation stage may be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results describes a [e.g., relative] temporal position of the groups of pilot sequences [time direction], wherein a second dimension of the two-dimensional array of correlation results describes a relative frequency position of the groups of pilot sequences [frequency direction].

[0051] In embodiments, at least one dimension [e.g., frequency direction] of the two-dimensional array of correlation results of the third correlation stage may be smaller than the respective at least one dimension of the two-dimensional array of correlation results of the second correlation stage.

[0052] In embodiments, the third correlation stage may comprise an [e.g., multi-channel] output queue buffer [e.g., ring buffer] configured to temporarily store the set of correlation results of the third correlation stage.

[0053] In embodiments, the data receiver may be configured to pass the set of correlation results in a suitable form to a subsequent packet detection.

[0054] In embodiments, the data receiver may be configured to adapt the group correlation pattern in frequency, e.g., to reduce [e.g., to compensate] the influence of the frequency shift, and / or wherein the data receiver may be configured to adapt the group sequence correlation pattern, e.g., to reduce [e.g., to compensate] the influence of the frequency shift.

[0055] In embodiments, the group sequence correlation pattern used in the third correlation stage of the multi-stage correlator may be a first group sequence correlation pattern, wherein a fourth correlation stage of the at least one second correlation stage of the multi-stage correlator, which is parallel to the third

[0056] Correlation stage operates, is configured to select groups of correlation results from the set of correlation results of the second correlation stage based on a second group sequence correlation pattern derived from the group sequence pilot pattern [e.g. or group sequence hopping pattern] and combine them in groups to obtain a set of correlation results of the fourth correlation stage, wherein the data receiver is configured to adjust the first group sequence correlation pattern and the second group sequence correlation pattern differently in frequency to reduce different frequency shifts, wherein the data receiver is configured to adjust the at least two pilot sequences based on the set of correlation results of the third correlation level, and / or the set of correlation results of the fourth correlation level.

[0057] In embodiments, the data receiver may be configured to receive a set of correlation results from the set of correlation results of the third correlation level, and the set of correlation results of the fourth correlation level, based on which the detection [e.g. and subsequent decoding] of the at least two pilot sequences takes place, depending on values ​​of the respective set of correlation results.

[0058] For example, the data receiver may be configured to receive the at least two pilot sequences [e.g., or the at least two partial data packets] based on one of the set of correlation results of the third correlation level, and the set of correlation results of the fourth correlation level, which have the highest values ​​[e.g. and use for subsequent decoding].

[0059] In embodiments, the data receiver may comprise a decoder configured to decode data comprising the signal based on the at least two pilot sequences, wherein the data receiver is configured to receive the set of correlation results from the set of correlation results of the third correlation level, and the set of correlation results of the fourth correlation level, based on which the detection and decoding of the plurality of partial data packets takes place, depending on a previous decoding result.

[0060] In embodiments, the data receiver may comprise a decoder configured to decode data comprising the received signal, wherein the decoder is configured to account for or correct for the time-varying frequency shift when decoding the data.

[0061] Further embodiments provide a method. The method comprises a step of receiving a signal from a data transmitter [e.g., stationary station, satellite, or relay], wherein the signal has at least two pilot sequences distributed [e.g., spaced apart] in time [e.g., and optionally in frequency] according to a pilot pattern, wherein at least a second pilot sequence of the at least two pilot sequences has a time-varying frequency shift [e.g., frequency drift] [e.g., due to a Doppler effect] compared to a first pilot sequence of the at least two pilot sequences. Furthermore, the method comprises a step of detecting the at least two pilot sequences based on a correlation pattern [e.g., corresponding to the pilot pattern or derived from the pilot pattern], wherein the correlation pattern is frequency-adapted to the time-varying frequency shift, e.g.,to reduce the influence of the time-varying frequency shift [e.g. to compensate].

[0062] Further embodiments provide a data receiver, wherein the data receiver is configured to receive a signal from a data transmitter [e.g., stationary station, satellite, or relay], wherein the signal comprises at least two pilot sequences distributed [e.g., spaced apart] in time [e.g., and optionally in frequency] according to a pilot pattern, wherein at least a second pilot sequence of the at least two pilot sequences has a time-varying frequency shift [e.g., frequency drift] [e.g., due to a Doppler effect] relative to a first pilot sequence of the at least two pilot sequences, wherein the data receiver comprises a detector having a correlator configured to detect the at least two pilot sequences based on a [e.g.,to detect correlation patterns selected from a set of correlation patterns [corresponding to or derived from the pilot pattern], wherein the correlation patterns of the set of correlation patterns reduce [e.g. compensate] different influences of time-varying frequency shifts.

[0063] In embodiments, the signal may comprise a data packet, wherein the data packet comprises the at least two pilot sequences.

[0064] In embodiments, the signal may comprise at least two partial data packets, each of the at least two partial data packets comprising a pilot sequence of the at least two pilot sequences.

[0065] In embodiments, the detector may be configured to estimate the time-varying frequency shift of the at least one second pilot sequence relative to the first pilot sequence, wherein the detector is configured to select a correlation pattern from the set of correlation patterns depending on the estimated frequency shift to obtain the selected correlation pattern.

[0066] In embodiments, the data receiver may comprise a decoder configured to decode data [e.g., a data packet or at least two sub-data packets] comprising the signal, wherein the decoder is configured to take into account or correct, when decoding the data, the time-varying frequency offset that is reduced [e.g., compensated] by the correlation pattern selected for detection.

[0067] In embodiments, the signal may comprise a plurality of sub-data packets distributed in time and frequency according to a hopping pattern, wherein the at least two pilot sequences are a plurality of pilot sequences, wherein the plurality of sub-data packets each comprise a pilot sequence of the plurality of pilot sequences, [e.g., wherein the pilot pattern is the same as the hopping pattern or is a time-shifted version of the hopping pattern], wherein the set of correlation patterns is derived from the hopping pattern.

[0068] In embodiments, the set of correlation patterns may be derived from at least one hopping pattern defined in ETSI TS 103 357.

[0069] Further embodiments provide a method. The method comprises a step of receiving a signal from a data transmitter [e.g., stationary station, satellite, or relay], wherein the signal has at least two pilot sequences distributed [e.g., spaced apart] in time [e.g., and optionally in frequency] according to a pilot pattern, wherein at least a second pilot sequence of the at least two pilot sequences has a time-varying frequency shift [e.g., frequency drift] [e.g., due to a Doppler effect] compared to a first pilot sequence of the at least two pilot sequences. Furthermore, the method comprises a step of detecting the at least two pilot sequences based on a correlation pattern selected from a set of correlation patterns, wherein the correlation patterns of the set of correlation patterns reduce different influences of time-varying frequency shifts [e.g.,compensate].

[0070] Further embodiments provide a computer program for carrying out a method according to one of the embodiments described herein when the computer program runs on a computer, microprocessor or SDR receiver (SDR = software defined radio, German: receiver with software-implemented signal processing).

[0071] Embodiments of the present invention are described in more detail with reference to the accompanying figures. They show: Fig. 1 is a schematic block diagram of a system with a data transmitter and a data receiver; Fig. 2 is a diagram showing the occupancy of the transmission channel during the transmission of a plurality of partial data packets according to a time-frequency hopping pattern and an exemplary structure of a partial packet consisting of data and synchronization symbols; Fig. 3 is a schematic block diagram of a system with a data transmitter and a data receiver; Fig. 4 is a diagram showing the occupancy of a transmission channel during the transmission of a plurality of partial data packets, wherein three groups of partial data packets of the plurality of partial data packets have the same relative group hopping pattern; Fig. 5a is a schematic block diagram of a data receiver with a multi-stage correlator, according to an embodiment;5b shows a schematic block diagram of a data receiver with a multi-stage correlator, according to another embodiment; Fig. 6 shows a schematic block diagram of a packet detector (multi-stage correlator) of a data receiver, according to an embodiment; Fig. 7 shows a diagram illustrating the occupancy of the transmission channel during the transmission of data packets using four different transmission methods; Fig. 8 shows a schematic view of the structure of a partial data packet, according to an embodiment; Fig. 9 shows a diagram illustrating the temporal and frequency arrangement of the partial data packets of a data packet in the assigned frequency band with bandwidth B; Fig. 10 shows a schematic block diagram of a first correlation stage of the multi-stage correlator of the data receiver, according to an embodiment;11 is a schematic block diagram of a second correlation stage of the multi-stage correlator of the data receiver, according to an embodiment; Fig. 12 is a schematic block diagram of a multi-stage correlator of a data receiver, wherein the multi-stage correlator has three correlation stages, according to an embodiment; Fig. 13 is a schematic view of an exemplary division of a preamble, according to an embodiment; Fig. 14 is a schematic block diagram of a section of the circuit shown in . Fig. 12shown first correlation stage as well as the filter bank (e.g. matched filter bank) and buffer (e.g. ring buffer) connected upstream of the first correlation stage, according to an embodiment; Fig. 15 a schematic view of the preamble section correlation of the subband signals temporarily stored in the ring buffer, including the combination of the section correlation results, carried out by the first correlation stage, according to an embodiment; Fig. 16 a schematic block diagram of a section of the in Fig. 12shown first correlation stage, according to an embodiment; Fig. 17 a schematic block diagram of the second correlation stage, according to an embodiment; Fig. 18 a schematic view of the group correlation performed by the second correlation stage based on the first correlation results buffered in the output queue buffer of the first correlation stage, according to an embodiment; Fig. 19 a schematic block diagram of a section of the in Fig. 12shown third correlation stage, according to an embodiment; Fig. 20 a schematic view of a reduction in the number of channels of the matched filter bank by an f / 8 matched filter bank and an f / 4 matched filter bank, according to an embodiment; Fig. 21 a schematic view of a section correlation with the preambles rotated with corresponding mixing frequencies or their sections as reference symbols; Fig. 22 a schematic view of a maximum formation between the first correlation stage and the second correlation stage of the data receiver, according to an embodiment; Fig. 23 a schematic block diagram of a first correlation stage of the data receiver, according to a further embodiment; Fig. 24 a schematic block diagram of a radio transmission system with a data transmitter and a data receiver, according to an embodiment of the present invention;25 shows a schematic block diagram of a correlator according to an embodiment of the present invention; Fig. 26 shows a schematic block diagram of a correlator according to a further embodiment of the present invention; Fig. 27 shows diagrams, starting from the zenith at time T = 0, a distance of a LEO satellite in km plotted against time in s, a relative speed of the LEO satellite in m / s plotted against time, a Doppler shift of a signal from the LEO satellite in Hz plotted against time, and a delta-Doppler shift of a signal from the LEO satellite in Hz / s plotted against time, each for a carrier frequency fc of 1 GHz; Fig. 28a shows a diagram of a Doppler shift in kHz plotted against an elevation angle of the LEO satellite; Fig. 28b shows a diagram of a delta-Doppler shift in kHz plotted against an elevation angle of the LEO satellite; Fig.Fig. 29 shows a diagram showing a conventional correlation pattern; Fig. 30 shows a diagram based on that shown in . Fig. 29shown correlation pattern, a modified correlation pattern, according to an embodiment of the present invention; Fig. 31 a schematic block diagram of a detector with a multi-stage correlator, according to an embodiment of the present invention; Fig. 32 a schematic block diagram of a detector with a multi-stage correlator having at least two third correlation stages (group sequence correlations) that use correlation patterns that are differently adapted in frequency, according to an embodiment of the present invention; Fig. 33 a schematic block diagram of a detector with a multi-stage correlator having five third correlation stages (group sequence correlations) that use correlation patterns that are differently adapted in frequency to compensate for different time-varying frequency shifts, according to an embodiment of the present invention;34 is a schematic block diagram of a data receiver with a detector having a multi-stage correlator, which has three third correlation stages (group sequence correlations), the correlation results of which are used in three packet detections to detect the pilot sequences, wherein only the detection results of one of the packet detections are provided to a decoder by means of a selector, according to an embodiment of the present invention; Fig. 35 is a schematic block diagram of a data receiver with a detector, a delta-Doppler compensator, and a decoder, according to an embodiment of the present invention; Fig. 36 is a schematic block diagram of a data receiver with a detector, a Doppler compensator, and a decoder, according to an embodiment of the present invention; Fig.Fig. 37 is a flowchart of a method for receiving a signal according to one embodiment of the present invention; and Fig. 38 is a flowchart of a method for receiving a signal according to another embodiment of the present invention.

[0072] In the following description of the embodiments of the present invention, identical or equivalent elements in the figures are provided with the same reference numerals so that their description is interchangeable.

[0073] Before describing in section 4 embodiments of a data receiver or a correlator which make it possible to detect pilot sequences which are spaced apart in time and which are subject to a larger time-varying frequency drift, an exemplary underlying radio transmission system is first described in section 1 and exemplary embodiments of a data receiver or correlator to be expanded / modified are first described in sections 2 and 3. 1. Telegram-splitting based radio transmission system 1.1 Overview

[0074] Fig. 1shows a schematic block diagram of a system with a data transmitter 100 and a data receiver 110. The data transmitter 100 can be configured to transmit a signal 120, wherein the signal 120 has at least two separate partial data packets 142. The data receiver 110 can be configured to receive the signal 120 (or a version of the signal 120 modified by the transmission channel)) which has the at least two separate partial data packets 142.

[0075] As in Fig. 1 As can be seen, the at least two separate partial data packets 142 are separated or spaced apart from one another in time and / or frequency. The distribution of the at least two separate partial data packets 142 in time and / or frequency can be carried out according to a hopping pattern 140.

[0076] In embodiments, the data transmitter 100 may include a transmitting device (or transmitting module, or transmitter) 102 configured to transmit the signal 120. The transmitting device 102 may be connected to an antenna 104 of the data transmitter 100. The data transmitter 100 may further include a receiving device (or receiving module, or receiver) 106 configured to receive a signal. The receiving device 106 may be connected to the antenna 104 or to another (separate) antenna of the data transmitter 100. The data transmitter 100 may also include a combined transceiver.

[0077] In embodiments, the data receiver 110 may include a receiving device (or receiving module, or receiver) 116 configured to receive the signal 120. The receiving device 116 may be connected to an antenna 114 of the data receiver 110. Furthermore, the data receiver 110 may include a transmitting device (or transmitting module, or transmitter) 112 configured to transmit a signal. The transmitting device 112 may be connected to the antenna 114 or to another (separate) antenna of the data receiver 110. The data receiver 110 may also include a combined transceiver.

[0078] In embodiments, the data transmitter 100 may be a sensor node, while the data receiver 110 may be a base station. Typically, a communication system comprises at least one data receiver 110 (base station) and a plurality of data transmitters (sensor nodes, such as heating meters). Of course, it is also possible for the data transmitter 100 to be a base station, while the data receiver 110 is a sensor node. Furthermore, it is possible for both the data transmitter 100 and the data receiver 110 to be sensor nodes. Furthermore, it is possible for both the data transmitter 100 and the data receiver 110 to be base stations.

[0079] The data transmitter 100 and the data receiver 110 can be configured to send and receive data using a telegram splitting method. In this method, a data packet (or telegram) containing the data is split into a plurality of partial data packets (or sub-data packets) 142, and the partial data packets 142 are transmitted from the data transmitter 100 to the data receiver 110 in a time-distributed and / or frequency-distributed manner according to a hopping pattern 140. The data receiver 110 then reassembles (or combines) the partial data packets 142 to obtain the actual data packet. Each of the partial data packets 142 contains only a part of the data packet 120. The data packet can also be channel-coded, so that for error-free decoding of the data packet, not all of the partial data packets 142 but only a part of the partial data packets 142 are required.

[0080] The temporal distribution of the plurality of partial data packets 142 can, as already mentioned, be carried out according to a time and / or frequency hopping pattern 140.

[0081] A time hopping pattern can specify a sequence of transmission times or transmission time intervals at which the partial data packets are sent. For example, a first partial data packet can be sent at a first transmission time (or in a first transmission time slot) and a second partial data packet at a second transmission time (or in a second transmission time slot), whereby the first transmission time and the second transmission time are different. The time hopping pattern can define (or predetermine, or indicate) the first transmission time and the second transmission time. Alternatively, the time hopping pattern can specify the first transmission time and a time interval between the first transmission time and the second transmission time. Of course, the time hopping pattern can also only specify the time interval between the first time and the second transmission time. There can be transmission pauses between the partial data packets during which no transmission takes place.The partial data packets can also overlap in time.

[0082] A frequency hopping pattern can specify a sequence of transmission frequencies or transmission frequency hops with which the partial data packets are sent. For example, a first partial data packet can be sent at a first transmission frequency (or in a first frequency channel) and a second partial data packet at a second transmission frequency (or in a second frequency channel), wherein the first transmission frequency and the second transmission frequency are different. The frequency hopping pattern can define (or predefine, or indicate) the first transmission frequency and the second transmission frequency. Alternatively, the frequency hopping pattern can specify the first transmission frequency and a frequency spacing (transmission frequency hop) between the first transmission frequency and the second transmission frequency. Of course, the frequency hopping pattern can also only specify the frequency spacing (transmission frequency hop) between the first transmission frequency and the second transmission frequency.

[0083] Of course, the plurality of partial data packets 142 can also be transmitted from the data transmitter 100 to the data receiver 110 in a manner distributed both in time and frequency. The distribution of the plurality of partial data packets in time and frequency can occur according to a time-frequency hopping pattern. A time-frequency hopping pattern can be the combination of a time-frequency hopping pattern and a frequency hopping pattern, i.e., a sequence of transmission times or transmission time intervals at which the partial data packets 142 are transmitted, wherein transmission frequencies (or transmission frequency hops) are assigned to the transmission times (or transmission time intervals).

[0084] Fig. 2 shows a diagram of the occupancy of the transmission channel during the transmission of a plurality of partial data packets 142 according to a time-frequency hopping pattern. The ordinate represents the frequency and the abscissa represents time.

[0085] As in Fig. 2As can be seen, the data packet 120 can be divided into n = 7 partial data packets 142, for example, and transmitted from the data transmitter 100 to the data receiver 110 in a time-frequency hopping pattern distributed in time and frequency.

[0086] As in Fig. 2 As can be further seen, the majority of partial data packets 142 can contain, in addition to data (data symbols 146 in Fig. 2 ) also pilot sequences (pilot symbols (or synchronization symbols) 144 in Fig. 2 ) based on which the data receiver 110 detects the partial data packets 142 in a received signal 120 or received data stream.

[0087] However, when a large number of data packets divided into partial data packets are transmitted simultaneously or in a temporally overlapping manner by a large number of data transmitters, the computing power required in the data receiver to detect and decode the partial data packets increases considerably.

[0088] In order to reduce the computing power required for detection and decoding, in embodiments the detection and decoding of the partial data packets takes place separately or separately from one another, as explained below. 1.2 Group formation

[0089] Fig. 3 shows a schematic block diagram of a system with a data transmitter 100 and a data receiver 110. The data transmitter 100 is designed to transmit a signal 120 having a plurality of partial data packets 142_1 to 142_8, wherein at least two groups 148_1 and 148_2 of partial data packets 142_1 to 142_8 have the same relative group hopping pattern 140_1 and 140_2 in groups.

[0090] For example, as in Fig. 3As shown, a first group 148_1 of sub-data packets (e.g., sub-data packets 142_1 to 142_4) and a second group 148_2 of sub-data packets (e.g., sub-data packets 142_5 to 142_8) have the same relative group hopping pattern 140_1 and 140_2. In other words, the second group hopping pattern 140_2 may be a time-shifted and / or frequency-shifted version of the first group hopping pattern 140_1.

[0091] The data receiver 110 may be configured to receive the signal 120 (or a version of the signal 120 modified by a transmission channel between the data transmitter 100 and the data receiver 110), wherein the signal 120 comprises a plurality of partial data packets 142_1 to 142_8, wherein at least two groups 148_1 and 148_2 of partial data packets 142_1 to 142_8 have the same relative group hopping pattern 140_1 and 140_2 in groups.

[0092] In Fig. 3For example, it is assumed that the partial data packets 142_1 to 142_8 are transmitted using at least two frequency and time hopping patterns (i.e., combined frequency hopping patterns and time hopping patterns) 140_1 and 140_2. Of course, the partial data packets 142_1 to 142_8 can also be transmitted using only pure frequency hopping patterns or time hopping patterns.

[0093] As in Fig. 3 As can be seen, the second group hop pattern 140_2 can be a time-shifted version of the first group hop pattern 140_1. Alternatively, the second group hop pattern 140_2 can also be a frequency-shifted version of the first group hop pattern 140_1. Of course, the second group hop pattern 140_2 can also be a time- and frequency-shifted version of the first group hop pattern 140_1.

[0094] In Fig. 3The signal 120 comprises, for example, n = 8 partial data packets 142_1 to 142_n, which are transmitted using m = 2 time hopping patterns and / or frequency hopping patterns 140_1 to 140_m. In embodiments, a number n of the partial data packets can be an integer multiple of a number m of time hopping patterns and / or frequency hopping patterns, so that the partial data packets can be evenly distributed among the number m of time hopping patterns and / or frequency hopping patterns, wherein the number n of partial data packets 142_1 to 142_n is at least twice as large as the number m of time hopping patterns and / or frequency hopping patterns 140_1 to 140_m, so that at least two partial data packets are transmitted in each time hopping pattern and / or frequency hopping pattern 140_1 to 140_m.

[0095] The data can be transmitted in such a way that there are transmission pauses (pauses in which the data sender does not transmit) between the partial data packets 142_1 to 142_n.

[0096] The data may be a telegram divided into the plurality of partial data packets 142_1 to 142_m, wherein each of the plurality of partial data packets 142_1 to 142_m is shorter than the telegram.

[0097] Fig. 4 shows in a diagram an occupancy of a transmission channel during the transmission of a plurality of partial data packets 142_1 to 142_n, wherein three groups of partial data packets of the plurality of partial data packets have the same relative group hopping pattern 140_1, 140_2 and 140_3.

[0098] As in Fig. 4As can be seen by way of example, nine partial data packets 142_1 to 142_9 can be divided into three groups 148_1 to 148_3, so that each of the three groups 148_1 to 148_3 comprises three of the partial data packets 142_1 to 142_9. The second group hopping pattern 140_2 can be a time- and frequency-shifted version of the first group hopping pattern 140_1, wherein the third group hopping pattern 140_3 can be a time- and frequency-shifted version of the first group hopping pattern 140_1. The time intervals Δx1 and Δx2 and the frequency intervals between the partial data packets are the same in the three group hopping patterns 140_1 to 140_3. The partial data packets 142_1 to 142_9 or at least a part of the partial data packets can be provided with synchronization sequences or partial synchronization sequences (split synchronization sequence) for synchronization and / or detection at the data receiver.

[0099] In other words, Fig. 4shows a subdivision of partial data packets 142_1 to 142_n into groups 148_1 to 148_m. Fig. 4 This method is shown as an example for nine partial data packets 142_1 to 142_9, which were combined into three groups 148_1 to 148_3 with the size of three partial data packets. Within these groups 148_1 to 148_3, the group hopping pattern is the same for time and / or frequency. The pauses and frequencies can vary between groups 148_1 to 148_3.

[0100] However, when a large number of data packets divided into partial data packets are transmitted simultaneously or in a temporally overlapping manner by a large number of data senders, the computing power required in the data receiver to detect the partial data packets increases considerably.

[0101] In order to reduce the computing power required for detection and decoding, in embodiments a multi-stage correlation is carried out for the detection of the partial data packets 142, as explained below. 2. Examples of implementation of the data receiver (system description)

[0102] Fig. 5a shows a schematic block diagram of a data receiver 110 according to one embodiment. The data receiver 110 can be configured to receive a signal 120 comprising a plurality of partial data packets 142 distributed in time and frequency, for example, according to a hopping pattern, wherein the plurality of partial data packets 142 each comprise a part of a data packet.

[0103] As in Fig. 5aAs can be seen, the data receiver 110 (or a packet detector of the data receiver 110) may comprise a multi-stage correlator 122, which may be configured to perform a multi-stage correlation in order to detect the partial data packets 142 in the received signal 121, wherein a second correlation stage 128 of the multi-stage correlator 122 operates based on correlation results 125 of a first correlation stage 124 of the multi-stage correlator 122.

[0104] In embodiments, the multi-stage correlator 122 may be configured to perform a multi-stage correlation, in detail, a first correlation of the received signal 121 in the first correlation stage 124 and a second correlation of a processed version of the received signal 125 (= correlation results of the first correlation stage) in the second correlation stage 128.

[0105] In embodiments, the multi-stage correlator 122 may be configured to detect the partial data packets 142 based on their preambles in the received signal 121. Alternatively, the multi-stage correlator 122 may be configured to detect the partial data packets 142 using a blind estimation method.

[0106] Fig. 5b shows a schematic block diagram of a data receiver 110 according to another embodiment of the present invention. The data receiver 110 can be configured to receive a signal 120 comprising a plurality of partial data packets 142 distributed in time and frequency, for example, according to a hopping pattern, wherein the plurality of partial data packets 142 each comprise a part of a data packet.

[0107] As in Fig. 5bAs can be seen, the data receiver 110 (or a packet detector of the data receiver 110) can have a multi-stage correlator 122, which can be configured to perform a multi-stage correlation to detect the partial data packets 142 in the received signal 121. The multi-stage correlator 122 can have a first correlation stage 124, a second correlation stage 128, and a third correlation stage 129, wherein the second correlation stage 128 operates based on correlation results 125 of the first correlation stage 124, wherein the third correlation stage 129 operates based on correlation results 126 of the second correlation stage 128.

[0108] Embodiments relate to a preamble correlation and the detection of data packets in the receiver 110 of a radio transmission system. The German term "preamble" for the symbols used for correlation is used here independently of the arrangement of the preamble within the data packets and therefore includes the symbols referred to in the English-language literature as Preamble, Midamble and Postamble The procedure is explained below using the example of a preamble positioned exactly in the middle, but it applies equally to other arrangements.

[0109] Fig. 6shows a schematic block diagram of a packet detector 130 of a data receiver 110, according to an embodiment. The packet detector 130 can have an (optional) filter bank (e.g., a matched filter bank) 132, the multi-stage correlator 122 with the first correlation stage (e.g., preamble correlation) 124 and further correlation stages (e.g., sequence correlation) 127, as well as a packet detection 134. The further correlation stages 127 are composed either of a second correlation stage 128 according to Fig. 5a or from a second 128 and a third 129 correlation level according to Fig. 5b together.

[0110] In other words, as in Fig. 6 As can be seen, the packet correlator (multi-stage correlator) 122 is part of a packet detector 130. The packet detector 130 can, for example, have the following components: an (optional) matched filter bank 132 for decomposing a broadband signal 120 into channels to be processed in parallel; a preamble correlation 124 for channel-by-channel correlation with the known symbols of the preamble; a sequence correlation 127 for summarizing the results for packets consisting of several sub-packets, each with its own preamble; a packet detection 134 for detecting the packets.

[0111] The preamble correlation 124 and the sequence correlation 127 form the packet correlator 122.

[0112] As in Fig. 6 As can be seen, the received signal 121 may have a plurality of channels.

[0113] In embodiments, the data receiver 110 may, for example, include the filter bank 132 to, based on the signal 120 (e.g., a broadband signal in Fig. 6) to obtain a received signal 121 (e.g., signal to be processed) with a plurality of channels. Instead of the filter bank 132, the data receiver 110 may also comprise a group of narrowband receivers to obtain a received signal 121 with a plurality of channels based on the signal 120. For example, the reception frequencies of the plurality of narrowband receivers may be set such that the same signals result as at the output of the Fig. 6 shown filter bank 132.

[0114] In the following description, the plurality of channels of the received signal 121 (e.g., the signal to be processed) are referred to as subband signals. The subband signals can comprise different bands of the signal 120; how they are obtained is irrelevant.

[0115] Fig. 7shows in a diagram the occupancy of the transmission channel when transmitting data packets using four different transmission methods.

[0116] In this context, Fig. 7 the ordinate represents the frequency and the abscissa represents time. In other words, Fig. 7 shows four possible methods for transmitting a single packet,

[0117] A first transmission method (case 1) involves a continuous transmission of a data packet at a constant frequency.

[0118] A second transmission method (case 2) comprises a continuous transmission of a data packet in conjunction with a frequency hopping method.

[0119] A third transmission method (case 3) involves a discontinuous transmission of a data packet (telegram splitting) at a constant frequency.

[0120] A fourth transmission method (case 4) comprises a discontinuous transmission of a data packet (telegram splitting) in conjunction with a frequency hopping method.

[0121] Embodiments of the data receiver 110 are relevant for all four methods when a plurality of data packets are to be received, when the data receiver 110 must receive a plurality of data packets that are transmitted asynchronously by different data transmitters 100 and at different frequencies within an assigned frequency band. As a result, the broadband signal 120 at the input has a significantly higher bandwidth than the partial data packets 142.

[0122] Embodiments of the data receiver 110 become particularly relevant in case 4, which enables a particularly high degree of parallel asynchronous packet transmissions. In this case, each partial data packet 142 contains its own preamble. The temporal and frequency sequence of the partial data packets 142 is referred to below as the (partial data packet) sequence. The throughput of the transmission system can be further increased by having different data transmitters 100 use different sequences; this reduces the probability of collisions between the partial data packets 142 of different data transmitters 100.

[0123] In addition, for cost reasons, frequency generators with relatively high tolerance are used in the data transmitters 100. This results in a frequency offset between the data transmitter 100 and the data receiver 110, which can be a multiple of the symbol rate f sym of the partial data packets. Since this effect also reduces the probability of packet collisions, the maximum throughput of the transmission system can be further increased by deliberately adding a stochastic component to the transmission frequencies. As a result, the transmission frequencies in the data receiver 110 are essentially unknown.

[0124] The detection of the data packets in the data receiver can be carried out using the preambles in the partial data packets 142. Since the transmission frequencies in the data receiver 110 are unknown, in exemplary embodiments the assigned frequency band can be divided into overlapping channels to be processed in parallel using a filter bank (e.g., matched filter bank) 132; the spacing Δf MF between the center frequencies of the individual channels can be only a fraction of the symbol rate f sym of the partial data packets 142. Values ​​are, for example (e.g., typically) in the range Δf MF / f sym = 1 / 4 ... 1 / 8. This, in conjunction with the bandwidth B of the assigned frequency band and the bandwidth BT of a partial data packet 142, results in, for example, the number N CH of channels to be processed in parallel: N CH = B − B T / Δf MF = 4 … 8 ⋅ B − B T / f sym 2.1 Structure of a partial data packet

[0125] Fig. 8shows a schematic view of the structure of a partial data packet 142 according to one embodiment. The partial data packet 142 may include NP preamble symbols 144 and ND data symbols 146, wherein the preamble may be arranged centrally. The total number of symbols of a partial data packet 142 may be NT = NP + ND.

[0126] A data packet may consist of M sub-data packets 142 that are transmitted discontinuously. Fig. 9 shows in a diagram the temporal and frequency arrangement of the partial data packets 142 of a data packet in the assigned frequency band with the bandwidth B. The ordinate describes the frequency and the abscissa the time.

[0127] The M points in time [t 1 , t 2 , ... , t M ] and the N carrier frequencies [f 1 , f 2 , ... , f N ] can be freely selected. In practice, however, an equidistant grid with the step size Δf T can be used for the frequencies, for example, since this facilitates signal generation in the data transmitter 100. In contrast, the points in time are not equidistant. The number N of carrier frequencies can be less than or equal to the number M of partial data packets 142. If N < M, individual carrier frequencies are used multiple times. However, this is not a general restriction; the number N of carrier frequencies can also be greater than the number M of partial data packets 142. In this case, not all carriers are occupied during a transmission.

[0128] The distance f off between the lower end of the frequency band and the frequency f 1 is variable due to the inaccurate frequency generation in the data transmitters and the previously mentioned stochastic component in the transmission frequencies and can, for example, correspond to at least half the bandwidth BT of a partial data packet 142, so that the partial data packet with the carrier frequency f 1 still lies completely within the frequency band. The resulting distance f off,B between the frequency f N and the upper end of the frequency band can also, for example, correspond to at least half the bandwidth BT, so that the partial data packet 142 with the carrier frequency f N still lies completely within the frequency band. This results in, for example: min f off = B T / 2 max f off = B − B T / 2 − N − 1 ⋅ Δf T

[0129] The variation range of f off has, for example, a width of: Δf off = max f off − min f off = B − B T − N − 1 ⋅ Δf T

[0130] The sequence SP of a data packet can be defined, for example, by the sequence of the indices of the frequencies with respect to the times [t 1 , t 2 , ... , t M ]. In Fig. 9 For example: S P = 7 , 10 , 1 , 5 , N , 12 , 4 , 8 , … , 11

[0131] For N < M, individual indices occur multiple times.

[0132] A completely free choice of the sequence of indices in the sequence SP results in a high computational effort in the packet correlator (multi-stage correlator) 122. In exemplary embodiments, sequences can be used that are composed of similar, frequency-shifted groups. For this purpose, the number M of sub-data packets 142 can be represented as the product of the number MG of sub-data packets 142 in a group and the number NG of the groups: M = N G ⋅ M G

[0133] This allows the Fig. 6The sequence correlation 127 shown can be split into a group correlation (second correlation level 128) and a subsequent group sequence correlation (third correlation level 129). The MG indices of the group SG can assume values ​​in the range [1, ... , N - X]. The NG values ​​of the group sequence SPG can then assume values ​​in the range [0, ... , X], so that the addition of any value from the group SG and any value from the group sequence SPG always results in a value in the range [1, ... , N]. By appropriately selecting the parameters MG , NG and X, the computational effort in the packet correlator (multi-stage correlator) 122 can be adjusted. If possible, X = N G − 1 can be chosen; this allows the group sequence S PG to be chosen so that it contains all possible values ​​exactly once. For X < N G − 1 individual values ​​occur multiple times in the group sequence.

[0134] Example: N = 20 , M = 24 , M G = 3 , N G = 8 , X = 7 S G = 1 , 13 , 7 S PG = 0 , 5 , 3 , 6 , 1 , 7 , 4 , 2 min S G + min S PG = 1 max S G + max S PG = N = 20

[0135] For the standardized bandwidths of the group and the group sequence, for example: B G , norm = max S G − min S G = N − X − 1 B PG , norm = max S PG − min S PG = X

[0136] MG < NG can hold, meaning that the length of a group can be less than the number of groups. In this case, the computational effort decreases with decreasing X.

[0137] The distance Δf T between adjacent frequencies can be related to the symbol rate f sym: Δf T = M Δ ⋅ f sym

[0138] For example, an integer value can be chosen for M Δ so that the frequency generation in the transmitters and receivers can be designed as simply as possible.

[0139] The times [t 1 , t 2 , ... , t M ] can be related to the time t 1 : t P = Δt 1 , Δt 2 , … , Δt M = t 1 , t 2 , … , t M − t 1 = 0 , t 2 − t 1 , … , t M − t 1

[0140] The values ​​can be chosen so that the distances are multiples of the symbol duration T sym = 1 / f sym. This gives the standardized distances T P = t P / T sym = f sym ⋅ t P = 0 , n 2 , n 3 , … , n M integer values. This also contributes to the simplification of frequency generation in the transmitters and receivers (Note: Frequency generation includes carrier frequency generation and clock frequency generation, where clock frequency generation is meant here, and carrier frequency generation above, for which the term "frequency generation" is used as a generic term). The formation of similar groups again results in a split into a group TG with length MG and a group sequence TPG with length NG.

[0141] Example: M = 12 , M G = 3 , N G = 4 T P = 0 ¯ , 33 , 60 , 95 ¯ , 128 , 155 , 188 ¯ , 221 , 248 , 290 ¯ , 323 , 350 T G = 0 , 33 , 60 T PG = 0 , 95 , 188 , 290

[0142] To reduce the probability of collisions between partial data packets 142 from different data transmitters 100, the transmitters can be divided into NS transmitter groups that use different group sequences S PG,i and T PG,i with i = 1 ... NS. This results in NS-fold parallel group sequence correlation. In this case, a lower value for the parameter X leads to a particularly significant reduction in computational effort. 2.2 Structure of the correlator

[0143] Fig. 10 shows a schematic block diagram of a first correlation stage 124 of the multi-stage correlator 122 of the data receiver 110, according to an embodiment.

[0144] The first correlation stage 124 may be configured to correlate the received signal 121 or a version derived therefrom in the preamble section correlation 150 with K preamble sections that correspond (e.g., match) to different (e.g., overlapping or adjacent) sections of the preambles (= preamble sections) of the partial data packets 142 in order to obtain K section correlation results 152 (e.g., section correlation amplitudes), for example, one section correlation result (e.g., one correlation amplitude) per preamble section. Furthermore, the first correlation stage 124 can be designed to combine 154 (e.g., add or incoherently add (e.g., by absolute value formation)) the plurality of section correlation results 152 (e.g., per sample) in order to obtain a first set of correlation results 156 (= correlation results 125) of the first correlation stage 124 for the received signal 121.

[0145] As in Fig. 10 As indicated, P subband signals can be present at the input of the first correlation stage, wherein the first correlation stage 124 can be designed to correlate a subband signal (e.g. the subband signal 121_1) of P subband signals in the preamble section correlation 150 with the K preamble sections in order to obtain K section correlation results 152 (e.g. section correlation amplitudes) for the subband signal (e.g. the subband signal 121_1), and to combine 154 the K section correlation results 152 in order to obtain a set of correlation results 156 for the subband signal (e.g. the subband signal 121_1).

[0146] In embodiments, the first correlation stage 124 can be designed to correlate at least two subband signals of the P subband signals (e.g., a plurality of subband signals of the P subband signals or all subband signals of the P subband signals) each with the K partial preambles in order to obtain a subset of correlation results 158 (e.g., a one-dimensional array of (normalized) correlation amplitudes) for each subband signal of the at least two subband signals, wherein the first correlation stage 124 can be designed to provide, as correlation results 125 of the first correlation stage, a first set of correlation results 156 comprising the subsets of correlation results 158, for example, the first set of correlation results 156 can comprise the one-dimensional subsets of correlation results 158.

[0147] The first set of correlation results 156 of the first correlation stage 124 can thus be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results describes (e.g., a sequence of) sampling times of the subband signals, wherein a second dimension of the two-dimensional array of correlation results describes the plurality of subbands.

[0148] Fig. 11 shows a schematic block diagram of the second correlation stage 128 of the multi-stage correlator of the data receiver 110, according to an embodiment.

[0149] As already mentioned above, at least two groups of partial data packets 148_1 and 148_2 can have the same relative group hopping pattern 140_1, 140_2 in groups, for example so that partial data packets 142 of a first group of partial data packets have the same relative group hopping pattern as partial data packets 142 of a second group of partial data packets 148_2 (cf. Fig. 3 ).

[0150] The second correlation stage 128 may be configured to select groups of correlation results 160 from the first set of correlation results 156 (e.g., the two-dimensional array of correlation results) of the first correlation stage 124 based on a group correlation pattern 162 derived from the group jump pattern 140_1, 140_2 and to combine 164 (e.g., add) them in groups to obtain a second set of correlation results 166 of the second correlation stage 128.

[0151] Here, the group correlation pattern 162 can indicate time and frequency intervals of the correlation results of the two-dimensional array of correlation results 156 of the first correlation stage 124, which correspond to the relative time and frequency intervals of the group jump pattern 140_1, 140_2 of the groups of partial data packets 142.

[0152] As in Fig. 11 As can be seen, the second correlation stage 128 can be configured to select the groups of correlation results 160 from the first set of correlation results 156 of the first correlation stage 124 in the time and / or frequency direction based on the group correlation pattern 162.

[0153] The second set of correlation results 166 of the second correlation stage 128 may be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results 166 describes a temporal position of the group of partial data packets 142, wherein a second dimension of the two-dimensional array of correlation results 166 describes a frequency position of the group of partial data packets 142.

[0154] Embodiments of the multi-stage correlator 122, which has three correlation stages, are described below. Here, the third correlation stage can essentially correspond to the second correlation stage 128, with the difference that the third correlation stage groups correlation results of the second correlation stage based on a group sequence correlation pattern instead of a group correlation pattern, wherein the group sequence hopping pattern defines relative time and frequency intervals between the groups of partial data packets 148_1 and 148_2 (cf. Fig. 3 ) indicates.

[0155] Fig. 12 shows a schematic block diagram of a multi-stage correlator 122 of a data receiver 110, according to an embodiment. The multi-stage correlator 122 comprises three correlation stages, in detail, a first correlation stage 124 (cf. Fig. 10 ), a second correlation level 128 (cf. Fig. 11) and a third correlation stage 129. The first correlation stage 124 can perform a preamble correlation, while the second (128) and third (129) correlation stages can perform a sequence correlation (127).

[0156] The input signal 121 of the multi-stage correlator 122 may comprise P subband signals. The P subband signals may be obtained, for example, by the filter bank (e.g., matched filter) 132.

[0157] It should be noted that the P subband signals present at the input of the multi-stage correlator 122 are based on the broadband signal 120 (see Fig. 6 ), however, it is irrelevant how the broadband signal 120, which is present at the antenna of the data receiver 110 (e.g. "in the air"), is decomposed into the P subband signals. This can be done, for example, with the Fig. 12shown filter bank 132, but just as well with P narrowband receivers or a broadband receiver and a different type of decomposition.

[0158] The multi-stage correlator 122 may include a first buffer (e.g., ring buffer) 170, which may be configured to temporarily store the P subband signals (e.g., provided by the filter bank 132).

[0159] The first correlation stage 124 can be configured to correlate the P subband signals into which the broadband signal 120 is decomposed, each with the K preamble sections in the preamble section correlation 150, in order to obtain K section correlation results 152 (e.g., section correlation amplitudes) for the respective subband signal. Furthermore, the first correlation stage 124 can be configured to normalize the K section correlation results 152 as a function of a determined power p[n] (e.g., calculated across multiple subbands). For example, the first correlation stage 124 can be configured to normalize the section correlation results 152 by forming magnitude squares, dividing by the determined power p[n], and calculating the roots of the quotients. Furthermore, the first correlation stage 124 can comprise K queue buffers (e.g.,Ring buffers) 153, which can be configured to temporarily store the respective section correlation results 152, wherein the plurality of queue buffers 153 can have different storage lengths, wherein the storage lengths of the K queue buffers 153 can depend on the respective preamble sections of the preambles of the partial data packets 142. Furthermore, the first correlation stage 124 can be configured to combine (e.g., add) the K section correlation results 152 temporarily stored in the K queue buffers 153 in order to obtain a subset of correlation results 158 for each of the subband signals, and to provide a first set of correlation results 156 comprising the subsets of correlation results 158 for the P subbands. The first correlation stage 124 can further comprise a (e.g., two-dimensional) output queue buffer (e.g.,Ring buffer) 172, which may be configured to temporarily store the first set of correlation results 156 of the first correlation stage 124.

[0160] The second correlation stage 128 may include a group correlation 165, which may be configured to select groups of correlation results based on a group correlation pattern from the first set of correlation results 156' of the first correlation stage 124, which are buffered in the output queue buffer (e.g., ring buffer) 172, and to combine them group by group (e.g., add them) to obtain a second set of correlation results 166. The second correlation stage 128 may include a (e.g., two-dimensional) output queue buffer (e.g., ring buffer) 174, which may be configured to buffer the second set of correlation results 166 of the second correlation stage 128.

[0161] The third correlation stage 129 may include a group sequence correlation 180, which may be configured to select groups of correlation results based on a group sequence correlation pattern from the second set of correlation results 166' of the second correlation stage 128, which are buffered in the output queue buffer (e.g., ring buffer) 174, and to combine them group by group (e.g., add them) to obtain a third set of correlation results 182. Furthermore, the third correlation stage 129 may include a (e.g., two-dimensional) output queue buffer (e.g., ring buffer) 176, which may be configured to buffer the third set of correlation results 182 of the third correlation stage 129.

[0162] In embodiments, the third correlation stage 129 may be configured to provide the correlation results in a form suitable for subsequent packet detection 134.

[0163] As in Fig. 12 As indicated, the third correlation stage can be configured to detect NS sequences of partial data packets. For this purpose, the group sequence correlation 180, the output queue buffer 176, and the packet detection N s compartment can be implemented.

[0164] In other words, Fig. 12 shows a schematic view of the structure of the packet correlator 122, including the embedding in the preceding matched filter bank 132 and the subsequent packet detection. The packet correlator 122 comprises the following three parts: 1. Preamble correlation (first correlation level 124) 2. Group correlation (second correlation level 128) 3. Group sequence correlation (third correlation level 129)

[0165] Ring buffers for storing intermediate results are located between the individual processing blocks. The depth of these ring buffers is determined by the processing length of the subsequent processing block. Therefore, the processing blocks themselves do not contain any additional buffers for intermediate results.

[0166] For preamble correlation, the preamble can be divided into K segments of length LK, each of which is individually normalized and then added. The segments can be selected to be overlapping or non-overlapping. For this purpose, a value in the range 1...LK can be chosen for the offset DK between the individual segments. For DK = LK, non-overlapping segments are obtained. For NP preamble symbols, the relationship is: L K + K − 1 ⋅ D K = N P

[0167] Fig. 13shows a schematic view of an exemplary division of a preamble 190 according to one embodiment. The preamble 190 comprises 12 preamble symbols 144, wherein the preamble 190 is divided into three non-overlapping sections 192 of four preamble symbols 144 each according to a first example, and wherein the preamble 190 is divided into five overlapping sections 192 of four preamble symbols 144 each according to a second example.

[0168] By dividing the signal into 192 sections and their separate standardization, immunity to impulse interference can be significantly improved. Furthermore, the division allows a reduction in the number of CP channels to be processed in parallel (e.g., by reducing the overlap of the subband signals, i.e., by reducing the oversampling MF in the frequency direction).

[0169] In the exemplary embodiment, the preamble correlation results in 124 normalized, real-valued correlation amplitudes, which result from the summation of the corresponding values ​​of the individual sections. The temporal shift of the results of the individual sections can be achieved by appropriately selecting the depths of the ring buffers following the normalization. Since the correlation occurs exclusively in the time direction, the number of CPs of the channels remains unchanged.

[0170] In exemplary embodiments, the normalized correlation amplitudes of the MG preambles of the partial data packets 142 of a group can be added in the group correlation 128. This can be done using the temporal structure of a group described by TG and the frequency structure described by SG. Since the correlation here also occurs in the frequency direction, the number of channels is reduced from CP to CG.

[0171] In exemplary embodiments, the normalized correlation amplitudes of the NG groups of a sequence can be added in the group sequence correlation. This can be done using the temporal structure described by T PG,i and the frequency structure described by S PG,i of the respective group sequence. Since correlation also occurs in the frequency direction here, the number of channels is reduced from CG to CS. 2.3 Number of channels

[0172] The number of channels in the preamble correlation can correspond to the number of relevant channels (subband signals), e.g. the relevant channels of a matched filter bank: C P = N CH = M F ⋅ B − B T / f sym

[0173] Where B is the bandwidth of the band, f sym the symbol rate, and MF the oversampling factor in the frequency direction. The factor MF can be adapted to the length LK of the preamble correlation sections. In order to achieve sufficient sensitivity even for packets that are unfavorably located with respect to the frequency grid of the subband signals, which can be provided, for example, by a matched filter bank 132, a conventional implementation M F ≥ 2 ⋅ L K A procedure for reducing the MF factor to lower values ​​is described below.

[0174] In group correlation, the number of channels can be reduced to: C G = M F ⋅ B − B T − B G / f sym

[0175] This can B G = B G , norm ⋅ Δf T = B G , norm ⋅ M Δ ⋅ f sym the range of the group. This results in, for example: C G = C P − M F ⋅ M Δ ⋅ B G , norm

[0176] In group sequence correlation, the number of channels is reduced to e.g.: C S = M F ⋅ B − B T − B G − B PG / f sym

[0177] BG can use the already mentioned range of the group and B PG = B PG , norm ⋅ Δf T = B PG , norm ⋅ M Δ ⋅ f sym the bandwidth of the group sequence. This results in, for example: C S = C P − M F ⋅ M Δ ⋅ B G , norm + B PG , norm

[0178] The relationship between the standardized bandwidths and the number N of carrier frequencies is: B G , norm + B PG , norm = N − 1

[0179] This results in, for example: C S = C P − M F ⋅ M Δ ⋅ N − 1

[0180] The number of channels CS can correspond to the value Δf off by which the carrier frequencies of the partial data packets 142 can vary without exceeding the assigned frequency band: Δf off = B − B T − N − 1 ⋅ M Δ ⋅ f sym = C S ⋅ f sym / M F

[0181] The following table (Table 1) contains the parameter values ​​for two examples. Regarding the structure of a data packet, the two examples differ only in the carrier spacing Δf T . In both examples, the relative bandwidth of a group was chosen such that the number of channels is significantly reduced by group correlation. This is particularly important in Example 2. parameter Formula symbols Example 1 Example 2 Bandwidth of the band B 100 kHz 725 kHz Bandwidth of a sub-packet BT 5 kHz 5 kHz Symbol rate f sym 2.5 kB 2.5 kB Relative distance of carrier frequencies M Δ 1 12 Distance between carrier frequencies Δf T 2.5 kHz 30 kHz Oversampling in frequency direction MF 8 8 Number of carrier frequencies N 24 24 Normalized bandwidth of a group BG, standard 16 16 Standard bandwidth of a group sequence B PG, standard 7 7 Number of channels in the preamble correlation CP 304 2304 Number of channels by group correlation CG 176 768 Number of channels after group sequence correction CS 120 96 Variation range of carrier frequencies Δf off 37.5 kHz 30 kHz 2.4 Correlation process

[0182] The correlation can be carried out with a temporal oversampling factor MT, ie the sampling rate of all signals is e.g.: f S = M T ⋅ f sym

[0183] For example, MT = 2 can be used.

[0184] Fig. 14 shows a schematic block diagram of a section of the Fig. 12shown first correlation stage 124 as well as the provision of the subband signals upstream of the first correlation stage 124, which here is carried out by way of example with a filter bank (e.g., matched filter bank) 132, and the buffer (e.g., ring buffer) 170, according to one embodiment. In other words, Fig. 14 shows the signals and the ring buffer 170 of the preamble section correlation 150. The ring buffer 170 may have the size CP × (MT × LK ).

[0185] Fig. 15 shows a schematic view of the preamble section correlation of the subband signals temporarily stored in the ring buffer, including the combination of the section correlation results, performed by the first correlation stage, according to an embodiment.

[0186] As in Fig. 15As can be seen, the correlation can take place in parallel across all CP channels; in the case of MT = 2, the odd or even columns can be evaluated alternately. The preamble symbols belonging to the respective section can be used in an LK-cyclic manner as reference symbols s 1 ,..., s 4 (corresponding to an embodiment with LK = 4), so that in conjunction with the cyclic data transfer of the sample values ​​of the subband signals, which can correspond, for example, to the output values ​​of a matched filter bank 132, into the ring buffer 170, the desired correlation results.

[0187] Alternatively, consecutive columns of the ring buffer can be combined into a column of MT · CP elements. The ring buffer 170 then has the size (MT · CP ) × LK . The correlation can now be performed in parallel across MT · CP channels and provide MT output values ​​for each channel. Accordingly, on the input side, MT samples of the subband signals, which can correspond, for example, to MT output vectors of a matched filter bank, can be combined into a column in the ring buffer.

[0188] The values ​​of the ring buffer can be stored in memory column by column, i.e., starting with the values ​​of the first column and ending with the values ​​of the last column. This makes the variants CP × (MT · LK ) and (MT · CP ) × LK equivalent in memory.

[0189] The Fig. 14The power calculation 151 shown follows the same process, but instead of multiplying with the reference symbols, the square of the values ​​is calculated.

[0190] The correlation signals c 1 [n],...,c K [n] can then be normalized by: the squares of the amounts are formed; a division by the powers p[n] is carried out; the roots of the quotients are calculated.

[0191] Fig. 16 shows a schematic block diagram of a section of the Fig. 12 shown first correlation stage 124, according to an embodiment. As shown in Fig. 12As can be seen, the first correlation stage 124 can be configured to perform a normalization 155 of the section correlation results 152 in order to obtain normalized section correlation results 152'. Furthermore, the first correlation stage 124 can be configured to temporarily store the normalized section correlation results 152' in the queue buffers 153 and to combine 154 (e.g., add) the normalized section correlation results 152' temporarily stored in the queue buffers 153.

[0192] In other words, Fig. 16 shows the normalization 155 and the subsequent addition of the results of the sections. Ring buffers 153 can be used as delay elements to delay the results according to the temporal structure of the sections. The ring buffer of the first section has D 1 = K − 1 ⋅ M T ⋅ D K the greatest delay, while the ring buffer of the last section only serves as an intermediate buffer without delay.

[0193] Fig. 17 shows a schematic block diagram of the second correlation stage 128 of the multi-stage correlator 122 of the data receiver, according to one embodiment. The second correlation stage 128 may be configured to perform a group correlation 165 of the first set of correlation results 156' buffered in the output queue buffer (e.g., ring buffer) 172 of the first correlation stage 124 to obtain a second set of correlation results 166. In other words, Fig. 17 shows the signals and the ring buffer of the group correlation.

[0194] Fig. 18shows a schematic view of a two-dimensional memory structure of the output queue buffer (e.g., ring buffer) 172 of the first correlation stage 124 and the group correlation 165 performed by the second correlation stage 128, in which groups of correlation results 160 are selected and combined 164 in groups from the first set of correlation results buffered in the output queue buffer (e.g., ring buffer) 172 of the first correlation stage 124 based on a group correlation pattern.

[0195] In other words, Fig. 18shows the sequence of group correlation 165. According to the ring buffer operation, the MGs move horizontally and cyclically through the ring buffer to form subcolumns with CG values ​​to be added. As already described, the number of channels is reduced from CP to CG during group correlation. This reduction is greater the larger the normalized bandwidth BG,norm of the group.

[0196] Fig. 19 shows a schematic block diagram of a section of the Fig. 12 shown third correlation stage 129, according to one embodiment. The third correlation stage 129 may be configured to perform a group sequence correlation 180 of the second set of correlation results 166' buffered in the output queue buffer (e.g., ring buffer) 174 of the second correlation stage 128 to obtain a third set of correlation results 182. As in Fig. 19As indicated, the third correlation stage 129 can be designed to detect NS sequences of partial data packets 142. For this purpose, the group sequence correlation 180 can be implemented N s -fold.

[0197] In other words, Fig. 19 shows the signals and the ring buffer of the group sequence correlation. The process is essentially the same as for group correlation, but here it refers to the NG groups of the respective group sequence with the parameters S PG,i and T PG,i with i = 1 ... NS . As already described, the number of channels can be reduced from CG to CS within the scope of group sequence correlation. This reduction is greater the larger the normalized bandwidth B PG,norm of the group sequence is. The NS output signals c S,1 [n],...c S,Ns [n] form the output signals of the packet correlator. 2.5 Using multiple groups

[0198] To further reduce the probability of collisions between the partial data packets 142 from different data senders, multiple groups can be used. The lower part of Fig. 12 The part shown with the group correlation 165 and the subsequent group sequence correlations 180 is present multiple times in this case. 2.6 Methods for reducing the number of channels

[0199] The distances between the center frequencies of the CP channels (subband signals), which can correspond to the channels of a matched filter bank, are, for example: Δf MF = f sym / M F

[0200] This allows the frequency offset Δf between the actual reception frequency of a partial data packet and the center frequency of the nearest subband signal, which may correspond, for example, to the nearest channel of a matched filter bank, to be limited to the range Δf = ± Δf MF / 2 = ± f sym / 2 ⋅ M F To avoid the error caused by this frequency offset in the correlation of the preamble sections becoming too large, M F ≥ 2 ⋅ L K apply. LK is the length of a section of the preamble. For the Fig. 13 The examples shown with LK = 4 result in a lower limit of MF = 8. The frequency offset in this case is in the range: Δf = ± f sym / 16

[0201] The limiting factor here is the error in the preamble correlation. With regard to matched filtering, which can be achieved with a matched filter bank, for example, a larger frequency offset in the range Δf = ± f sym / 8 or - in the case of reduced performance - be tolerated beyond this. Fig. 20 shows how this fact can be used to reduce the number of channels, ie the number of subband signals, which can be achieved, for example, by reducing the number of channels in a matched filter bank 132.

[0202] In the following, it is assumed that the subband signals are provided by a matched filter bank, as this is the preferred implementation in practice. In principle, however, the provision can be achieved using any method that can provide a set of equivalent subband signals for further processing. In other words, the method of providing the subband signals is irrelevant for processing.

[0203] In detail, Fig. 20 a schematic view of a reduction of the number of channels of a matched filter bank 132 from an f / 8 matched filter bank 132_1 to an f / 4 matched filter bank 132_2. As in Fig. 20As can be seen, every second filter of the f / 8 matched filter bank can be omitted, and the outputs of the resulting f / 4 matched filter bank can be shifted by ± f sym / 16 using two mixers each, resulting in a frequency grid similar to that of an f / 8 matched filter bank. With this measure, the computational effort in the matched filter bank 132 can be reduced by approximately half.

[0204] In conjunction with the subsequent preamble section correlation, the mixers after the f / 4 matched filter bank 132_2 can be eliminated by using two different preambles or their sections rotated with the corresponding mixing frequencies as reference symbols in the correlation. This is shown in Fig. 21for a single channel. Although this procedure is not equivalent to the complex-valued correlation result, further processing requires only the squared magnitude of the result, so the deviation has no effect. The rotation of the reference symbols occurs once during component initialization.

[0205] A further possibility for reduction arises from the realization that the high frequency resolution of the preamble correlation is not absolutely necessary for the subsequent group and group sequence correlation. Therefore, a maximum formation over adjacent channels can be performed between the preamble correlation and the group correlation, and the number of channels can be reduced accordingly. However, this measure leads to an increase in the false detection probability in the packet detection following the packet correlator, so that usually only two adjacent channels can be combined. This case is in Fig. 22 shown.

[0206] If both measures are combined - halving the number of channels at the input or in the matched filter bank 132 and halving the number of channels after the preamble correlation by maximum formation over two adjacent channels - the result is Fig. 23Preamble correlation shown with reduced number of channels.

[0207] Fig. 23shows a schematic block diagram of the first correlation stage 124, according to a further embodiment.The first correlation stage 124 is designed to shift a subband signal provided by the f / 4 matched filter bank by ± f sym / 16 by means of two mixers, and to carry out a preamble section correlation 150 for each of the ± f sym / 16 shifted versions of the subband signal in order to obtain preamble section results, and to carry out a normalization 155 of the preamble section results in order to obtain normalized preamble section results, and to temporarily store the normalized preamble section results in the queue buffers 153, and to combine 154 the temporarily stored normalized preamble section results in order to obtain an intermediate correlation result for the versions of each subband signal shifted by ± f sym / 16, and to carry out a maximum formation 157 of the intermediate correlation results in order to obtain a correlation result for each subband signal.

[0208] Compared to the Fig. 12 In the embodiment shown, all channel numbers (CP, CG, and CS) can now be reduced by a factor of 2 by using MF = 4 instead of MF = 8, without reducing the frequency resolution of the preamble correlation. Due to the parallel processing with rotated reference symbols, the effective channel number in the preamble correlation is now C' P = 2 · CP .

[0209] A further reduction in the number of channels before preamble correlation by choosing MF < 4 is generally not possible, as the frequency offset Δf in the matched filter bank can assume values ​​that result in very pronounced symbol distortion. This distorts the results so severely that performance decreases significantly. In special cases where compromises are unavoidable due to computational effort, this may be necessary.

[0210] In contrast, a further reduction in the number of channels after preamble correlation by maximizing more than two adjacent channels is certainly possible if the associated higher false detection probability can be tolerated. The decisive factor here is the relative computational effort in the individual components of the preamble correlator. In practice, the computational effort in preamble correlation is often significantly higher than in group and group sequence correlation. In this case, reducing the number of channels after preamble correlation would only marginally reduce the computational effort. 3. Further examples

[0211] In uncoordinated radio communication systems (such as the ALOHA method), the data sender transmits its packet at an arbitrary point in time. The receiver has no or only imprecise knowledge of the time at which transmission begins. This time must be determined by the receiver using detection. 3.1 Multi-stage detection with preamble splitting

[0212] Classical systems use the transmission preamble to detect data packets at the receiver. This preamble is usually transmitted in one piece and can therefore be relatively easily detected using classical correlation.

[0213] However, the telegram splitting procedure or a time or frequency hopping procedure typically divides the preamble into several sections.

[0214] If this split sequence is to be detected, it is advantageous to calculate the correlation jointly across all sequence parts, which entails a very high computational effort.

[0215] Embodiments of the present invention take a different approach, in which the correlation is divided into several partial correlations, and the partial results are then combined into an overall result. A prerequisite for this methodology may be, for example, that the sequence is the same in all subsections. If this is the case, the correlation can be divided into a preamble correlation, an (optional) group correlation, and a group sequence correlation, as described in Section 2.

[0216] By using group and group sequence correlation, it is possible to detect multiple different hop patterns, which reduces the susceptibility of the transmission to interference. If, in the case of multiple hop patterns, the sequence is chosen to be the same for all patterns in the subsections, only a single preamble correlator is required.

[0217] This methodology significantly reduces the detector's required computing power, allowing for either the use of more cost-effective hardware or an increase in the number of supported jump patterns.

[0218] In embodiments, the correlator consists not only of a single-stage correlator, but of at least two correlators, with the second correlator operating based on the results of the first correlator. The results in the stages can be buffered (e.g., in a database or a ring buffer).

[0219] In embodiments, a correlation can first be performed across the preamble sequence sections. These results can then be combined in a second correlator to produce a group result. Based on the group correlation, the group sequence correlation can then be performed, which provides the overall result for detection.

[0220] If the above-described requirement that all sub-sections have the same pilot sequence is not met, the procedure described above can still be applied if there are only very few sequences (relatively fewer sequences than sub-packets).

[0221] In this case, there are v-parallel preamble correlations, where v is the number of distinct sequences. It is not mandatory for the different sequences to be of equal length.

[0222] In the next step of partial correlation, the results from the memories of the various preamble correlations can now be loaded and combined according to the jump pattern. Further processing is analogous to the procedure described above.

[0223] In embodiments, the first correlation stage has at least two parallel correlaters.

[0224] In embodiments, in the second correlation step, the results from the multiple correlators of the first stage can be loaded and combined according to the jump pattern. 3.2 Optimized preamble correlation

[0225] The ideas described in the following subsections are based on the preamble correlation from Section 3.1. However, they are generally valid for all systems that use a preamble for detection, even if only one coherent preamble is present in the telegram / packet. 3.2.1 Normalization of correlation results for noise suppression

[0226] In Fig. 13 Using the example of a preamble with 12 symbols, it is shown how the performance of the correlation against frequency offsets can be increased by splitting the preamble into subsections that are correlated individually and then added incoherently.

[0227] This idea to improve the correlation results under frequency offsets has already been explained in detail in [3] and [4].

[0228] In a typical system without interference, the threshold can be selected after correlation based on the background noise. The length of the correlation performs additional noise averaging, which limits the number of false detections at a suitable threshold. All correlation values ​​above the threshold represent with a very high probability the beginning of a transmitted data packet. The higher the received power of the transmitted data packet at the data receiver, the higher the correlation value and thus the probability that a data packet was transmitted.

[0229] If interference from other participants (in the same or a different network) occurs during transmission, the approach described above is only of limited use, as the interference influences the correlation result, and the value at this point typically exceeds the threshold for subsequent packet detection. At these points, the data receiver therefore incorrectly assumes detection. This is particularly problematic when the interference source's reception power is significantly higher than the noise, as the correlation then also produces a relatively high result.

[0230] This can be remedied by normalizing the correlation results to the received (estimated) interference. For example, the magnitudes of the individual data packets can be weighted according to the estimated interference. This means that disturbed data packets have a smaller impact than unaffected data packets.

[0231] Generally speaking, a nonlinear function is required for normalization. This can, for example, represent the weighting of the magnitudes according to the estimated disturbance, as described above.

[0232] In some embodiments, the subpackets can be normalized to the estimated interference. This normalization can take place either before or after the correlation.

[0233] A more concrete example of this normalization is the normalization of the correlation results to the received signal power. For this, the magnitude squares are calculated for all symbols of the preamble and the sum is then calculated.

[0234] This sum is divided by the square of the correlation result, and the square root of the quotient is then taken, resulting in the normalized correlation result. As a result of this normalization, all packets arriving at the data receiver (e.g., base station) now have a correlation value of one (for an ideally received pilot sequence without noise or interference) or lower.

[0235] Instead of squaring the absolute value and then taking the root, an approximation can also be carried out, for example: 1. abs I + abs Q 2. Newton Rapson method 1 / sqrt(X) 3. Approximation of the absolute values ​​and adding the square

[0236] If interference occurs, the correlation result is also normalized to the received signal power. However, since the received symbols generally deviate from the expected preamble sequence during interference, the correlation result is significantly lower than for a non-interfered signal.

[0237] The normalization thus ensures that the correlation result is significantly below one and thus the probability of false detection is reduced.

[0238] As an alternative to calculating the square of the correlation result, the correlation result can also be divided directly by the square root of the determined signal power.

[0239] Furthermore, normalization can also occur before correlation. For this, the signal power is calculated in the same way as above and then the square root is taken. This result is applied to each input symbol by division.

[0240] In some embodiments, the correlation result can be normalized to the received signal power of the preamble. This can be achieved in several ways.

[0241] If data symbols occur before and / or after the received preamble, these data symbols can also be (partially) included in the power calculation. Thus, the number of symbols used to determine the power is greater than the number of preamble symbols used for correlation.

[0242] In embodiments, the determination of the received signal power may be performed via at least one data symbol.

[0243] The previous methods always have a correlation without subdivision into subsections as in Fig. 13To be able to perform standardization in this case, a separate standardization can be performed for each subsection. After standardization, the subsections can be added together as usual.

[0244] There are now two ways to determine the performance for the standardization of the sub-areas: 1. Separate determination of performance for each sub-area 2. Joint determination of performance for all sub-areas

[0245] In both variants, as above, either the same number of symbols as for the correlation can be used or neighboring symbols can be included again.

[0246] In some embodiments, separate normalization of the correlation sub-ranges can be performed. This can involve either a separate power determination for each sub-range or a joint power determination.

[0247] If a multi-channel detector is used, as in the case of Section 2, the normalization can be performed separately for each channel. If it is assumed that interference always occupies at least two of the channels, the power determination can also be performed jointly for at least two channels.

[0248] In embodiments, in a multi-channel receiver, the standardization can be carried out in parallel on all channels, whereby the power determination can also be carried out across several channels. 3.2.2 Delay structure with ring buffers

[0249] When using a split piecewise correlation, results from different time points can be added, depending on the temporal position of the sequence.

[0250] One way to achieve this is to calculate the correlation for all necessary time points before summing. However, under certain circumstances, this has the disadvantage that the preceding buffer structure (in this case, the output of the filter bank) must retain the input data for the entire correlation period.

[0251] One solution to avoid this problem is to create a buffer structure for the partial correlation results.

[0252] This means that only the data for the length of the partial correlation can be kept at the input.

[0253] At the output, n-ring buffers can be used for the n-partial correlations. The length of each ring buffer allows the temporal dependence between the partial correlations to be established. This means that the length of the buffer determines the duration of the delay.

[0254] To calculate the overall correlation result, the oldest entries of all ring buffers can be added together before they are discarded in the next step.

[0255] In some embodiments, instead of a large buffer at the input of the (partial) correlation, a buffer structure can be used at the output of the partial correlations. The length of the respective buffer determines the time delay (see Fig. 16 ) was realized. 3.2.3 Reduction of the number of channels at the input of the packet correlator

[0256] In systems where the frequency offset (random and / or systematic offset) between the data transmitter and data receiver can be a multiple of the symbol rate, it is necessary to use a multi-channel correlator.

[0257] In order to perform the parallel correlation on the channels, a filter bank can be used in advance to generate the symbols for each channel.

[0258] Due to the relationship between the (partial) correlation length and the maximum permissible frequency offset between two channels (see section 2), a large number of channels results that must be calculated and stored in the filter bank.

[0259] However, as described in Section 2, this restriction applies to the multichannel correlator and not to the preceding filter bank. This is illustrated graphically in Fig. 20 illustrated using the frequency response of the filter.

[0260] If the oversampling of the filter bank in the frequency direction is reduced to a certain extent and the frequency resolution is subsequently restored by frequency shifting the symbols before correlation, this has little to no impact on the correlation performance. However, the required computing power and memory consumption of the filter bank and the subsequent memory are reduced by the selected factor.

[0261] In some embodiments, the filter bank of the multi-channel correlator may have a different frequency oversampling than the subsequent correlator. To increase the frequency resolution in the correlator, the symbols of the filter bank can be multiplied by a complex exponential oscillation (corresponding to a digital frequency shift), whereby the choice of the exponential oscillation depends on the frequency offset.

[0262] Instead of multiplying the input data by the exponential oscillation, the reference sequence can also be multiplied by the exponential oscillation. This creates a separate reference sequence for each frequency offset, but eliminates the effort of multiplication in each calculation step.

[0263] In embodiments, a separate reference sequence can be used for each frequency line to be generated from the data of the filter bank, wherein the adjusted reference frequency is generated from the original reference sequence by multiplication with the corresponding exponential oscillation. 3.2.4 Reduction of the number of channels at the output of the packet correlator

[0264] Another possibility for reducing the number of channels arises from the realization that the high frequency resolution of the preamble correlation is not required for the subsequent group and group sequence correlation.

[0265] Therefore, a maximum calculation can be performed between the preamble correlation and the group correlation over adjacent channels, and the number of channels can be reduced accordingly. However, this measure leads to an increase in the false detection probability in the packet detection following the packet correlator, so that usually only two adjacent channels can be combined. This case is in Fig. 22 shown.

[0266] By calculating the maximum and discarding the smaller value(s), the number of channels after the first correlation stage can be reduced, which results in less computational effort and smaller memory.

[0267] In some embodiments, after calculating the preamble correlation, a maximum calculation can be performed across neighboring channels. The smaller value(s) is / are discarded for further processing. 4. Packet correlator for a radio transmission system with time-variant frequency drift

[0268] Fig. 24 shows a schematic block diagram of a radio transmission system with a data transmitter 100 and a data receiver 110, according to an embodiment of the present invention.

[0269] The data receiver 110 may be configured to receive a signal 120 from a data transmitter 100, wherein the signal 120 comprises at least two pilot sequences 144_1, 142_2 (and 144_3) distributed in time and optionally in frequency according to a (transmitter-side) pilot pattern, wherein at least a second pilot sequence 144_2 (and 144_3) of the at least two pilot sequences 144_1, 142_2 (and 144_3) has a time-varying frequency shift Δf (frequency drift) compared to a first pilot sequence 144_1 of the at least two pilot sequences 144_1, 142_2 (and 144_3).

[0270] The data receiver may comprise a detector 130 (e.g., packet detector) with a correlator 122 configured to detect the at least two pilot sequences 144_1, 142_2 (and 144_3) (e.g., in the received signal 121) based on a correlation pattern, wherein the correlation pattern is frequency-adjusted to the time-varying frequency shift to reduce (e.g., compensate) an influence of the time-varying frequency shift Δf.

[0271] In Fig. 24 For example, the signal 120 has two or three pilot sequences 144_1, 144_2 (and 144_3). In embodiments, the signal 120 may have up to n pilot sequences 144_1 to 144_n, where n is a natural number greater than or equal to two, n ≥ 2.

[0272] As in this case Fig. 24As indicated by way of example, the signal 120 can have a data packet 141, wherein the data packet 141 can have the at least two pilot sequences 144_1 and 144_2. The pilot pattern can indicate a time interval between the at least two pilot sequences 144_1 and 144_2.

[0273] Alternatively, the signal 120 can have at least two partial data packets 142_1, 142_2 (and 142_3) (cf. Section 1), wherein the at least two partial data packets 142_1, 142_2 (and 142_3) each have a pilot sequence 144_1, 142_2 (and 144_3), and wherein the at least two partial data packets 142_1, 142_2 (and 142_3) are distributed in time and optionally in frequency according to a hopping pattern 140. The pilot pattern can specify a distribution of the at least two pilot sequences 142_1, 142_2 (and 142_3) in time and optionally in frequency. For example, the pilot pattern can be the same as the hopping pattern 140. Alternatively, the pilot pattern can be derived from the hopping pattern 140, e.g.if the jump pattern 140 defines absolute times of the partial data packets 142_1, 142_2 (and 142_3) instead of relative times or the pilot sequences 142_1, 142_2 (and 142_3) are located at different positions within the partial data packets 142_1, 142_2 (and 142_3) than those defined by the jump pattern 140.

[0274] In embodiments, the correlation pattern used by the correlator 122 may be equal to a (transmitter-side) pilot pattern. Alternatively, the correlation pattern used by the correlator 122 may be derived from the (transmitter-side) pilot pattern, e.g., if the correlation pattern already takes into account transmission characteristics of the communication channel used (with the exception of the time-varying frequency shift Δf).

[0275] However, due to the time-varying frequency shift Δf, an expected distribution of the at least two pilot sequences 144_1, 144_2 (and 144_3) indicated by the correlation pattern does not correspond to the actual distribution of the at least two pilot sequences 144_1, 144_2 (and 144_3) of the received signal.

[0276] In embodiments, the data receiver 110 (e.g., the detector 130 of the data receiver 110 or the correlator 122 itself) may therefore be configured to adjust the correlation pattern in frequency to reduce (e.g., compensate) an influence of the frequency shift Δf.

[0277] For example, the data receiver 110 may be configured to estimate the time-varying frequency shift of the at least one second pilot sequence 144_2 relative to the first pilot sequence 144_1 and to adapt the correlation pattern in frequency based on the estimated time-varying frequency shift, e.g., to shift the correlation pattern in frequency by the estimated time-varying frequency shift or to apply the estimated temporal changes to the correlation pattern.

[0278] For example, given an estimated time-varying frequency shift of, e.g., +100 Hz / s, the data receiver 110 may adjust the correlation pattern in frequency by +100 Hz / s (e.g., apply a time-varying frequency shift of +100 Hz / s) to reduce or compensate for an influence of the time-varying frequency shift. Accordingly, given an estimated time-varying frequency shift of, e.g., -300 Hz / s, the data receiver 110 may adjust the correlation pattern in frequency by -300 Hz / s (e.g., apply a time-varying frequency shift of -300 Hz / s) to reduce or compensate for an influence of the time-varying frequency shift.

[0279] Alternatively, in embodiments, the data receiver 110 (e.g., the detector 130 of the data receiver 110 or the correlator 122 itself) may also be configured to detect the at least two pilot sequences 144_1, 144_2 (and 144_3) based on a correlation pattern selected from a set of correlation patterns, wherein the correlation patterns of the set of correlation patterns reduce or compensate for different influences of time-varying frequency shifts.

[0280] For example, the data receiver 110 (or the correlator 122) may be configured to estimate the time-varying frequency shift of the at least one second pilot sequence 144_2 relative to the first pilot sequence 144_1, and to select a correlation pattern from the set of correlation patterns depending on the estimated frequency shift to obtain the selected correlation pattern.

[0281] Instead of adapting the correlation pattern in frequency, in embodiments a suitable correlation pattern can also be selected from the set of correlation patterns in order to reduce an influence of the time-varying frequency shift.

[0282] The correlation patterns of the set of correlation patterns can be derived from the same pilot pattern that describes the distribution of the at least two pilot sequences 144_1 and 144_2 in the data packet 141 and can be subjected to different, time-varying frequency shifts, so that a respective correlation pattern reduces or compensates for an influence of a respective time-varying frequency shift.

[0283] Of course, the correlation patterns of the set of correlation patterns can also be derived from the same jump pattern that describes the distribution in time and optionally in frequency of the at least two partial data packets and can be subjected to different, time-varying frequency shifts, so that a respective correlation pattern reduces or compensates for an influence of a respective time-varying frequency shift.

[0284] For example, a first correlation pattern of the set of correlation patterns can be subjected to a time-varying frequency shift of, for example, +100 Hz / s in order to reduce or compensate for the influence of a time-varying frequency shift of, for example, +100 Hz / s. A second correlation pattern of the set of correlation patterns can be subjected to a time-varying frequency shift of, for example, +200 Hz / s in order to reduce or compensate for the influence of a time-varying frequency shift of, for example, +200 Hz / s.

[0285] In embodiments, the time-varying frequency shift, e.g., in the case of a movable data transmitter 100 and / or data receiver 110, may be caused by a relative change in movement between the data transmitter 100 and the data receiver 110.

[0286] In embodiments, the time-varying frequency shift may also additionally or alternatively result from a time-varying mismatch between oscillators of the data transmitter and the data receiver.

[0287] In embodiments, one of the data transmitter 100 and the data receiver 110 may be a satellite or a relay, while the other of the data transmitter 100 and the data receiver 110 may be a stationary station.

[0288] In embodiments, the data transmitter 100 may include a transmitting device (or transmitting module, or transmitter) 102 configured to transmit the signal 120. The transmitting device 102 may be connected to an antenna 104 of the data transmitter 100. Optionally, the data transmitter 100 may include a receiving device (or receiving module, or receiver) 106 configured to receive a signal. The receiving device 106 may be connected to the antenna 104 or to another (separate) antenna of the data transmitter 100. The data transmitter 100 may also include a combined transceiver.

[0289] In embodiments, the data receiver 110 may include a receiving device (or receiving module, or receiver) 116 configured to receive the signal 120 to provide a received signal 121. The receiving device 116 may be connected to an antenna 114 of the data receiver 110. Optionally, the data receiver 110 may include a transmitting device (or transmitting module, or transmitter) 112 configured to transmit a signal. The transmitting device 112 may be connected to the antenna 114 or to another (separate) antenna of the data receiver 110. The data receiver 110 may also include a combined transceiver.

[0290] Fig. 25 shows a schematic block diagram of a correlator 122 according to an embodiment of the present invention.

[0291] As in Fig. 25As can be seen, the correlator 122 can have at least two parallel-operating correlation units 118_1 and 118_2, wherein a first correlation unit 122_1 of the at least two parallel-operating correlation units 118_1 and 118_2 can be configured to perform correlations of the at least two pilot sequences 144_1, 144_2 (and 144_3) with corresponding reference sequences based on the correlation pattern in order to obtain a set of correlation results 119_1 of the first correlation unit 118_1, wherein a second correlation unit 118_2 of the at least two parallel-operating correlation units 118_1 and 118_2 can be configured to perform correlations of the at least two pilot sequences 144_1, 144_2 (and 144_3) with corresponding reference sequences based on the correlation pattern in order to obtain a set of correlation results 119_2 of the second correlation unit 118_2.

[0292] The at least two correlation units 118_1 and 118_2 operating in parallel can be configured to use correlation patterns that are differently adapted in frequency in order to reduce or compensate for influences of different time-varying frequency shifts during detection.

[0293] For example, the data receiver 110 may be configured to differently adapt the frequency of the correlation pattern used in the at least two correlation units 118_1 and 118_2 operating in parallel in order to reduce or compensate for influences of different frequency shifts.

[0294] The detector 130 may be configured to generate a set of correlation results from the set of correlation results 119_1 of the first correlation unit 118_1, and the set of correlation results 119_2 of the second correlation unit 118_2, based on which the detection of the at least two pilot sequences 144_1, 144_2 (and 144_3) takes place, depending on values ​​of the respective set of correlation results 119_1 and 119_2.

[0295] For example, the detector 130 may be configured to detect the at least two pilot sequences 144_1, 144_2 (and 144_3) based on one of the set of correlation results 119_1 of the first correlation unit 118_1, and the set of correlation results 119_2 of the second correlation unit 118_2 having the highest values.

[0296] Fig. 26 shows a schematic block diagram of a correlator 122 according to another embodiment of the present invention.

[0297] As in Fig. 26As can be seen, the multi-stage correlator 122 can have a first correlation stage 124 and at least one second correlation stage 128 (and 129) following the first correlation stage 124, which operates based on correlation results 125 of the first correlation stage 124, wherein at least one correlation pattern used in at least one correlation stage from the at least one second correlation stage 128 (and 129) is frequency-adjusted in order to reduce or compensate for the influence of the time-varying frequency shift.

[0298] In embodiments, the multi-stage correlator 122 may include a first correlation stage 124 and a second correlation stage 128 operating based on correlation results 125 of the first correlation stage 124, wherein a correlation pattern used in the second correlation stage 128 is frequency adjusted to reduce or compensate for the influence of the time-varying frequency shift.

[0299] In embodiments, the multi-stage correlator 122 may include a first correlation stage 124, a second correlation stage 128 operating based on correlation results 125 of the first correlation stage 124, and a third correlation stage 129 operating based on correlation results 126 of the second correlation stage 128, wherein a correlation pattern used in the second correlation stage 128 and / or a correlation pattern used in the third correlation stage 129 is frequency adjusted to reduce or compensate for the influence of the time-varying frequency shift.

[0300] In embodiments, the first correlation stage 124 may be configured to perform preamble correlation, as described in detail in sections 2 and 3.

[0301] In detail, the first correlation stage 124 can be configured as described, for example, with respect to Fig. 10 explained, in order to correlate 150 the received signal 121 or a version derived therefrom with a plurality of pilot sequence sections corresponding to different sections of the pilot sequences of the at least two partial data packets 142_1, 142_2 (and 142_3) in order to obtain a plurality of section correlation results 152, wherein the first correlation stage 124 can be configured to combine 154 the plurality of section correlation results 152 in order to obtain a set of correlation results 156 or a subset of correlation results 158 as correlation results 125 of the first correlation stage 124.

[0302] In embodiments, the first correlation stage 124 can of course also be configured to detect at least two pilot sequences that are not contained in at least two partial data packets 142_1, 142_2 (and 142_3), but in a single data packet 141 (cf. Fig. 24 ). In this case, the first correlation stage 124 may be configured to correlate 150 the received signal 121 or a version derived therefrom with a plurality of pilot sequence sections corresponding to different sections of the at least two pilot sequences 144_1 and 144_2 to obtain a plurality of section correlation results 152, wherein the first correlation stage 124 may be configured to combine 154 the plurality of section correlation results 152 to obtain a set of correlation results 156 or a subset of correlation results 158 as correlation results 125 of the first correlation stage 124.

[0303] In embodiments, the second correlation stage 128 may be configured to perform a group correlation if the signal 120 comprises a plurality of partial data packets 142_1 to 142_n which, in groups, have the same relative group hopping pattern 140_1 and 140_2, as described in detail in sections 2 and 3.

[0304] In detail, the second correlation stage 128 can be configured as described, for example, with respect to Fig. 11 explained, in order to select groups of correlation results 160 from the set of correlation results 156 of the first correlation stage 124 based on a group correlation pattern 162 derived from a group jump pattern and to combine them group by group 164 to obtain a set of correlation results 166 of the second correlation stage 128.

[0305] In embodiments, the second correlation stage 128 can of course also be configured to detect a plurality of pilot sequences that are not contained in a plurality of partial data packets 142_1 to 142_n, but rather in a single data packet 141 and within the data packet 141, have the same relative group pilot pattern in groups. In this case, the second correlation stage 128 can be configured to select groups of correlation results 160 from the set of correlation results 156 of the first correlation stage 124 based on a group correlation pattern 162 derived from the group pilot pattern and to combine them in groups 164 to obtain a set of correlation results 166 of the second correlation stage 128.

[0306] In embodiments, the third correlation stage 129 may be configured to perform a group sequence correlation, provided that the at least two groups of partial data packets form a sequence and have a relative group sequence hopping pattern to each other, as described in detail in sections 2 and 3.

[0307] In detail, the third correlation stage 129 can be configured as described, for example, in relation to Fig. 12 explained, to select from the set of correlation results 166 of the second correlation stage 128 groups of correlation results based on a group sequence correlation pattern derived from the group sequence hopping pattern and to combine them group by group to obtain a set of correlation results 182 of the third correlation stage 129.

[0308] In embodiments, the third correlation stage 129 can of course also be configured to detect a plurality of pilot sequences that are not contained in a plurality of partial data packets 142_1 to 142_n, but rather in a single data packet 141 and have a relative group sequence pilot pattern within the data packet 141. In this case, the third correlation stage 129 can be configured to select groups of correlation results from the set of correlation results 166 of the second correlation stage 128 based on a group sequence correlation pattern derived from the group sequence pilot pattern and to combine them in groups to obtain a set of correlation results 182 of the third correlation stage 129.

[0309] As already mentioned, a correlation pattern used in the second correlation stage 128 and / or in the third correlation stage 129 can be adjusted in frequency in order to reduce or compensate for the influence of the frequency shift.

[0310] For example, the data receiver 110 (e.g., or the detector 130 of the data receiver 110 or the multi-stage correlator 122 itself) may be configured to frequency adjust a correlation pattern used in the second correlation stage 128 and / or in the third correlation stage 129 to reduce or compensate for the influence of the frequency shift.

[0311] For example, the data receiver 110 may be configured to frequency-adjust the group correlation pattern used in the second correlation stage 128 to reduce or compensate for the influence of the time-varying frequency shift. Further, or alternatively, the data receiver 110 may be configured to frequency-adjust the group sequence correlation pattern used in the third correlation stage 129 to reduce or compensate for the influence of the time-varying frequency shift.

[0312] In the following, detailed embodiments of the present invention are described in more detail, which are based in particular on the Fig. 26 shown multi-stage correlator 122. 4.1 System description

[0313] The exemplary embodiments described below relate to the correlation and detection of pilot sequences (e.g., of a data packet with multiple pilot sequences or of multiple data packets, each with at least one pilot sequence) in the data receiver 110, also referred to below as the receiver, of a radio transmission system. In the following, the German term "Präambel" is used to designate the pilot sequence, i.e., the symbols to be used for correlation, regardless of the arrangement of the preamble within the data packets. The term "Präambel" used below therefore encompasses the cases referred to in English-language literature as preamble, midamble, and postamble. In the following, the method is explained using the example of a preamble arranged exactly in the middle; however, it applies equally to other arrangements.

[0314] As already mentioned in Section 2 with regard to Fig. 6As explained, detector 130 (e.g. packet detector) can have an (optional) filter bank (e.g. a matched filter bank) 132, the multi-stage correlator 122 with the first correlation stage (e.g. preamble correlation) 124 and further correlation stages (e.g. sequence correlation) 127, as well as a packet detection 134. The further correlation stages 127 consist either of a second correlation stage 128 (cf. Fig. 5a ) or from a second 128 and a third 129 correlation level (cf. Fig. 5b ) together.

[0315] The multi-stage correlator 122 can thus be part of a packet detector 130, which can have the following components: an (optional) matched filter bank 132 for decomposing a broadband signal 120 into channels to be processed in parallel; a preamble correlation 124 for channel-by-channel correlation with the known symbols of the preamble; a sequence correlation 127 for summarizing the results for packets consisting of several sub-packets, each with its own preamble; a packet detection 134 for detecting the packets.

[0316] The preamble correlation 124 and the sequence correlation 127 form the packet correlator 122.

[0317] As already mentioned in Section 2, embodiments of the data receiver 110 can be applied to all four Fig. 7shown cases, e.g., when the data receiver 110 is to receive a plurality of data packets that are transmitted asynchronously by different data transmitters 100 and at different frequencies within an assigned frequency band. As a result, the broadband signal 120 at the input has a significantly higher bandwidth than the partial data packets 142_1-142_n.

[0318] Embodiments of the data receiver 110 become particularly relevant in case 4, which enables a particularly high degree of parallel asynchronous packet transmissions. In this case, each partial data packet 142_1-142_n contains its own preamble 144_1-144_n (see Fig. 24). The temporal and frequency sequence of the partial data packets 142_1-142_n is referred to below as the (partial data packet) sequence. The throughput of the transmission system can be further increased by having different data transmitters 100 use different sequences; this reduces the probability of collisions between the partial data packets 142_1-142_n of different data transmitters 100.

[0319] The detection of the partial data packets 142_1 to 142_n in the receiver is carried out using the preambles in the partial data packets 142_1 to 142_n. Since the transmission frequencies in the receiver 110 are unknown, the assigned frequency band is divided into overlapping channels to be processed in parallel using a matched filter bank 132. The spacing Δf MF between the center frequencies of the individual channels is only a fraction of the symbol rate f sym of the partial data packets 142_1-142_n. Typical values ​​are in the range Δf MF / f sym = 1 / 4...1 / 8. This, in conjunction with the bandwidth B of the assigned frequency band and the bandwidth BT of a partial data packet 142_1-142_n, results in the number N CH of channels to be processed in parallel: N CH = B − B T / Δf MF = 4 … 8 ⋅ B − B T / f sym

[0320] Embodiments of the present invention find application, for example, in satellite communications systems, but generally whenever frequency drifts are expected. Due to the movement of LEO satellites (LEO = low earth orbit), the reception frequency of the received signal changes during transmission in the receiver due to the so-called Doppler effect.

[0321] The maximum Doppler frequency depends on the speed of the satellite and the carrier frequency of the communication system; it is calculated as follows: f D , max = v c f c

[0322] Where fc = 868 MHz, c = 3 * 10 8< m / sec and v is the speed of the moving receiver. In general, a high Doppler frequency is a problem for the coherent detection of telegrams when the transmission time is long (e.g., as in [4]).

[0323] Fig. 27shows in diagrams, starting from the zenith at time T = 0, a distance of a LEO satellite in km plotted over time in s, a relative speed of the LEO satellite in m / s plotted over time, a Doppler shift of a signal of the LEO satellite in Hz plotted over time, and a delta-Doppler shift of a signal of the LEO satellite in Hz / s plotted over time, each for a carrier frequency fc of 1 GHz.

[0324] For an exemplary system with LEO satellites, Fig. 27 Frequency drifts (Delta Doppler shifts) of up to 200 Hz / s.

[0325] This results in frequency offsets (Doppler shift) and frequency drifts (Delta Doppler) as can be seen in Fig. 28a and 28b are shown.

[0326] In detail, Fig. 28aa Doppler shift in kHz plotted against an elevation angle of the LEO satellite, while Fig. 29b shows a delta Doppler shift in Hz / s plotted against an elevation angle of the LEO satellite, each for a carrier frequency fc of 1 GHz and an Earth orbit at an altitude of 700 km.

[0327] The starting point for the exemplary embodiments described below is the multi-stage correlator 122 described in sections 2 and 3, which performs packet detection using a multi-stage correlation. This is modified according to the exemplary embodiments described below such that reception of packets (e.g., data packets 141 or partial data packets 142_1-142_n, cf. Fig. 24 ) with the frequency drifts mentioned above is possible. 4.2 Improved detection of a telegram with time-variant frequency drift

[0328] The time-variant Doppler shift described above results in frequency shifts of up to 200 Hz / s. In a transmission system according to [5], in which a telegram (or data packet) to be transmitted is divided into at least 24 so-called radio bursts (= partial data packets 142_1-142_24), which form a so-called core frame, with a transmission duration of approximately four seconds for the so-called core frame, these frequency shifts lead to a frequency change of approximately 800 Hz between the beginning and the end of the core frame. According to [5], the pilot sequences are incorporated in each of the 24 radio bursts (partial data packets 142_1-142_24) in the core frame. The interval between the first and the last pilot sequence is therefore also approximately 4 seconds.

[0329] According to [5], the data rate of the radio transmission system is approximately 2400 sym / s. The frequency resolution of the multi-stage detector according to sections 2 and 3 is typically 1 / 8 of the symbol rate in Hz. In other words, the frequency oversampling during correlation is typically a factor of 8 higher than the symbol rate in Hz.

[0330] Under this assumption, the correlation results in a frequency jump of approximately 2400 Hz / 8 = 300 Hz. This means that telegrams can only be detected if the frequency drift is less than 300 Hz over the entire telegram (in [5] the core frame) and less than 75 Hz per second.

[0331] It should be noted, however, that embodiments of the present invention are not limited to the so-called telegram splitting method [5]. Rather, embodiments of the present invention can be generally applied to detecting packet transmissions with split or distributed pilot sequences. For example, there are systems with a preamble and a postamble. Then, the duration between the preamble and postamble serves as a reference for the time-variant frequency drift.

[0332] In order to be able to detect telegrams or data packets even in the presence of time-variant frequency drifts, the correlator 122 can be modified. In detail, when using the multi-stage correlator 122, only a part of the correlator can be modified (see Section 4.3).

[0333] Fig. 29shows a diagram of a correlation pattern 200 as used above in sections 2 and 3. The abscissa describes the time, divided into time correlation steps, which according to sections 2 and 3 can be, for example, half the symbol duration, while the ordinate describes the frequency, divided into frequency steps of the correlation, which, as mentioned above, can be, for example, 1 / 8 of the symbol rate.

[0334] Assuming, for example, a linear frequency drift of 150 Hz / s, a correlator frequency resolution of 300 Hz, a transmission duration of four seconds, and five split pilot sequences (or partial data packets), the average interval between the pilot sequences (or partial data packets) is one second. This means that the frequency of each received pilot sequence (or partial data packet) has changed by an average of 150 Hz compared to the previous pilot sequence (or partial data packet). The frequency difference between the first and last pilot sequences (or partial data packets) is 600 Hz.

[0335] With a correlation resolution of 300 Hz per step, the frequency offset is larger than the step size starting with the third pilot sequence (or partial data packet). Starting with the fifth pilot sequence (or partial data packet), the frequency offset is already two step sizes.

[0336] A detection of the telegram after Fig. 29is therefore not possible.

[0337] However, if the correlation pattern of the correlator 122 is modified such that, for example, the next higher frequency element is selected starting with the third pilot sequence (or partial data packet), detection of the telegram is again possible. Starting with the fifth pilot sequence (or partial packet), the next but one frequency element can then be selected, as shown in Fig. 30 is shown.

[0338] In detail, Fig. 30 in a diagram, starting from the one in Fig. 29 shown correlation pattern 200, a modified correlation pattern 202, according to an embodiment of the present invention.

[0339] The respective frequency lines are modified according to the expected frequency drift. If the frequency drift is larger, for example, the modified correlation pattern must be adjusted accordingly.

[0340] If a negative frequency drift is expected instead of the positive frequency drift mentioned above, the correlation pattern can be adjusted accordingly so that values ​​below the correlation pattern are Fig. 30 must be chosen.

[0341] In embodiments, when transmitting telegrams or data packets with distributed / divided pilot sequences and / or telegram splitting, the correlation pattern (or correlation pattern) can be modified according to an expected frequency drift so that detection of the telegram or data packet is possible.

[0342] Under an exemplary second assumption of a linear frequency drift of 50 Hz / s, a frequency resolution of the correlator of 300 Hz, a transmission duration of four seconds and five divided pilot sequences (or partial data packets), an average distance of one second between the pilot sequences (or partial data packets) also results.

[0343] The smaller change of 50 Hz / s results in a maximum change of 200 Hz over the four-second telegram duration (between the first and last pilot sequence (or partial data packet). This means that the correlation does not need to be modified for successful detection, since the maximum difference is smaller than the correlation step size.

[0344] In this case, however, it is still useful to adjust the correlation pattern to improve detection performance. The last pilot sequence (or partial data packet) has a frequency difference of approximately 200 Hz compared to the first pilot sequence (or partial data packet). This value is greater than half the step size of the correlator 122. Since the last pilot sequence (or partial data packet) does not lie exactly on a correlation line, the "energy" of the correlation is distributed among the neighboring lines. This is the Fig. 29 shown index, as well as the next following index.

[0345] Since the frequency offset is greater than half the correlation step size, more "energy" is located in the next higher index. Therefore, it is useful to adjust the correlation pattern even for frequency drifts that exceed half the correlation step size.

[0346] In embodiments, when transmitting telegrams or data packets with distributed / divided pilot sequences and / or telegram splitting, the correlation pattern can be modified according to an expected frequency drift so that detection of the telegram is improved. 4.3 Application to multilevel correlation according to sections 2 and 3

[0347] The starting point is again the system according to [5] with a symbol rate of approximately 2400 Sym / s, 24 partial data packets 142_1-142_24 and a transmission time of approximately four seconds.

[0348] For this system, an efficient multi-stage correlation 122 for detection was described in Sections 2 and 3. However, this type of telegram detection only works if the maximum frequency offset between the first and last partial data packets 124_1 and 124_24 is smaller than the frequency step size of the correlation (with the values ​​above, less than 300 Hz). As already calculated above, this only applies to frequency drifts of less than 75 Hz / s.

[0349] The operation of the multi-stage correlator 122 is briefly explained below (see sections 2 and 3). The multi-stage correlation takes place in several separate stages, each of which builds on the results of the previous stage.

[0350] The first stage is the so-called preamble correlation 124. This is based on the output signal of the optional matched filter bank 132, which divides the input signal 120 of the receiver 110 into overlapping subband signals. As the name suggests, the preamble correlation 124 correlates the symbols within a pilot sequence or a partial data packet in the necessary subbands. Thus, the preamble correlation only operates within a partial data packet, thus its duration is limited to approximately 15 ms.

[0351] The second stage is the so-called group correlation 128. This correlation stage exploits the special properties of the jump patterns described in [5] by subdividing the jump patterns into subpatterns (=group jump patterns), which are patterns of a basic pattern shifted in frequency and time. The basic pattern in [5] comprises three sub-data packets (or radio bursts). According to [5], the duration of a basic pattern is approximately 316 ms.

[0352] The third and final correlation level comprises the so-called group sequence correlation 129, in which the results of group correlation 128 are correlated across the entire core frame to produce an overall result. The duration thus encompasses the entire four seconds, as mentioned above.

[0353] If we now consider the maximum 200 Hz / s mentioned as an example in the system description (Section 4.1), this results in a maximum frequency deviation of 3.0 Hz over the duration of a partial data packet for the duration of the preamble correlation 124. For the group correlation 128, this results in a maximum frequency deviation of 63.2 Hz within the group. For the group sequence correlation 129, the maximum frequency deviation of 800 Hz applies over the entire core frame (i.e., the 24 partial data packets), as already calculated above.

[0354] Since the maximum frequency deviations in the preamble correlation 124 and the group correlation 128 are significantly smaller than the step size (or half the step size), it is not absolutely necessary to consider the frequency drift in the first two correlation stages 124 and 128. In some embodiments, it is sufficient to modify the group sequence correlation 129. Of course, in some embodiments, the group correlation 128 and the group sequence correlation 129 can also be modified.

[0355] The group sequence correlation can be modified as described in Chapter 4.1, where the intervals in time and frequency correspond to the distances between the groups (of sub-packets that have the same group hopping pattern in groups), whereby the duration of the group is also taken into account.

[0356] Fig. 31shows a schematic block diagram of a detector 130 with a multi-stage correlator 122, according to an embodiment of the present invention. In other words, Fig. 31 shows a sequence of correlation in the detector 130 with multi-stage correlation with modification for frequency drifts, where in the Fig. 31 In the embodiment shown, only the three-stage correlation is examined in more detail.

[0357] In the Fig. 31 In the embodiment shown, it is assumed that (only) the third correlation stage 129 (group sequence correlation) is modified, whereas the first two correlation stages 124 and 128 are not modified and are therefore used directly as in sections 2 and 3.

[0358] However, it should be noted that the principle of adapting the correlation pattern presented here can also be applied in the second correlation stage 128 (group correlation), provided that the frequency drift also has relevant effects on the subsequences of a group.

[0359] In exemplary embodiments, in a multi-level correlation, only those correlation levels are modified where the frequency drift has an impact on performance. This generally applies when the frequency deviation is greater than half the correlation step size. 4.4 Using multiple correlation patterns in parallel

[0360] In the embodiments described in sections 4.1 to 4.3, it was assumed that the frequency drift is known in advance at the receiver 110 and does not change or can be estimated by the receiver 110. However, this is taking into account Fig. 27 not given.

[0361] Since it is not known in advance when which frequency drift occurs, it is not possible to switch the detector according to the current frequency drift.

[0362] In order to be able to receive with and without frequency drift at any time, it is possible to run several correlations with different assumptions regarding the frequency drifts.

[0363] This is exemplified using the system from section 4.3 in Fig. 32 for a two-way parallel group sequence correlation 129_1 and 129_2. Since the multi-stage correlator 122 is used there, it is sufficient, as already explained above, to duplicate the third correlation stage.

[0364] In detail, Fig. 32a schematic block diagram of a detector 130 with a multi-stage correlator 122 having at least two third correlation stages (group sequence correlations) 129_1 and 129_2, which use correlation patterns differently adapted in frequency, according to an embodiment of the present invention. In Fig. 32 For example, it is assumed that either no frequency drift or only a specific frequency drift can be present. By duplicating the third correlation level (group sequence correlation), more correlation results are available, which also have to be processed separately in detection 134_1 and 134_2.

[0365] In embodiments, if the frequency drift is not constant and / or not known in advance, a parallel correlation 129_1 and 129_2 and detection 134_1 and 124_2 may be performed with different assumptions for the frequency drifts. Selector for managing multiple detection

[0366] Fig. 33 shows a schematic block diagram of a detector 130 with a multi-stage correlator 122 having five third correlation stages (group sequence correlations) 129_1-129_5 that use correlation patterns differently adjusted in frequency to compensate for different time-varying frequency shifts (-200 Hz / s, -100 Hz / s, 0 Hz / s, 100 Hz / s, 200 Hz / s), according to an embodiment of the present invention. The detector 130 may further comprise a selector 131 configured to select the correlation results of one of the five third correlation stages 129_1-129_5, based on which the detection is to be performed, and provide them to a packet detector 134.

[0367] In other words, shows a fivefold parallel correlation 129_1-129_5 with five different assumptions, where the frequency drifts in this case can range from -200 Hz / s to + 200 Hz / s. Since in this case the detection would also have to run five times, Fig. 33 In the embodiment shown, a selector 131 is inserted after the group sequence correlation 129_1-129_5. It often happens that the actual frequency drift does not exactly correspond to the assumptions. According to the description in Chapter 4.2 for the system from [5], up to 75 Hz / s can be tolerated without correction. The various assumptions are spaced 100 Hz / s apart. If, for example, the frequency drift is 60 Hz / s, the telegram will be correctly detected by both the 0 Hz / s and the +100 Hz / s correlation. The decoder would therefore have to process a packet twice. Selector Scenario 1: Avoiding Multiple Detection

[0368] Multiple detections are to be avoided with the help of selector 131, which compares the correlation results of the individual assumptions for each time point and the same jump pattern and passes only the highest value to the detection. Thus, detection can continue according to sections 2 and 3.

[0369] In embodiments, to avoid multiple detections, a selector 131 can only pass on the highest value per time step and jump pattern (e.g., to a packet detection 134). Selector scenario 2: Controlled forwarding of (multiple) detections

[0370] Fig. 34shows a schematic block diagram of a data receiver 110 with a detector 130 with a multi-stage correlator 122, which has three third correlation stages (group sequence correlations) 129_1-129_3, whose correlation results are used in three packet detections 134_1-134_3 to detect the pilot sequences, wherein by means of a selector only the detection results of one of the packet detections 134_1-134_3 are provided for a decoder, according to an embodiment of the present invention.

[0371] In other words, Fig. 34 shows multiple parallel group sequence correlations 129_1-129_3 and detections 134_1-134_3 for different frequency drifts with subsequent selector 131. In the Fig. 34In this scenario, a selector 131 is inserted after the detections 134_1-134_3. The task of the selector 131 is to detect multiple detections and to forward a) none or b) only a limited number of multiple detections for the subsequent decoding 133.

[0372] Case a) For this purpose, the feedback 135 of the decoder 133 is required. After successful decoding of a telegram or data packet, the decoder 133 transmits information (e.g., start time, end time) to the selector 131, which enables the selector 131 to clearly recognize and ignore multiple detections of a telegram or data packet.

[0373] Case b) The detection of multiple detections as described in a) can be used to improve decoder performance. The joint consideration of multiple detections can provide additional information that positively supports the decoding process. For this purpose, suitable multiple detections are selected and forwarded. Criteria for selecting suitable multiple detections can be implemented both dynamically, e.g., through appropriate feedback 135 from the decoder 133, or fixed through preconfiguration.

[0374] In embodiments, the selector 131 can clearly detect multiple detections through feedback 135 from the decoder 133. The selector 131 forwards multiple detections in a controlled manner for subsequent decoding.

[0375] The described extension of the correlations with different assumptions on the frequency drift has the disadvantage that the receiver implementation is adapted to the effect that new correlators (at least for the last correlation stage) are inserted. As in Fig. 33 If a selector 131 is used, it must also be included in the receiver implementation. If a modification is to be made to the receiver implementation, it must usually be tested and verified again later using sometimes complex tests.

[0376] As in Fig. 30As shown, the frequency indices of the individual pilot sequences or subpackets change due to frequency drift. Without frequency drift, the indices are directly related to the hop patterns or can be derived from them. Thus, as an alternative to modifying the receiver software, the hop patterns (or correlation patterns) can also be adjusted. This has the advantage that only the configuration parameters need to be adjusted, which is sometimes even possible dynamically.

[0377] If, as described above, several possible frequency drifts are to be investigated, this can be achieved by increasing the number of jump patterns (or correlation patterns) accordingly.

[0378] In embodiments, instead of modifying the receiver software, the hopping patterns may be adapted according to the frequency drift(s) to be analyzed. 4.5 Transmission of the determined frequency drift to the decoder

[0379] Previous ideas have always focused on detection. However, decoding is often more critical, as significantly fewer errors can be tolerated here. Therefore, in state-of-the-art systems, synchronization (time, frequency, and possibly phase) is typically performed after detection and before decoding.

[0380] The frequency drift can now also be included in the synchronization, but this increases the degrees of freedom and thus the complexity and computing power.

[0381] However, since the detection has already examined several assumptions regarding frequency drift, it makes more sense to use this assumption for the correction as well. This way, no additional computing power is required, and with sufficiently precise resolution of the assumptions, this estimate of the frequency drift is sufficient.

[0382] In embodiments, the estimation of the frequency drift by the various assumptions in the detector can be used directly for the correction in or before the decoder.

[0383] Fig. 35 shows a schematic block diagram of a data receiver 110 with a detector 130, a delta-Doppler compensator 137, and a decoder 133, according to an embodiment of the present invention. In other words, Fig. 35 shows an interplay of detection and decoding with compensation of the determined frequency drift.

[0384] Fig. 36 shows a schematic block diagram of a data receiver 110 with a detector 130, a Doppler compensator 137, and a decoder 133, according to an embodiment of the present invention. In other words, Fig. 36 shows an interplay of detection and decoding with compensation of the determined frequency drift.

[0385] Figs. 35 and 36show possibilities how the determined frequency drift can be compensated or corrected before decoding 133.

[0386] If the frequency drift is estimated by modifying the jump patterns, as described in Section 4.4, the frequency drift can be corrected directly in decoder 133. Detector 130 typically informs decoder 133 of the number of the detected jump pattern, as this extracts the symbols from the input signal. By modifying and, if necessary, extending the jump patterns, it is also known which frequency drift was successful in detection 130. From this, the frequency drift can be recalculated and corrected.

[0387] In embodiments, the detected jump pattern number can be used to calculate back the frequency drift and correct it. 5. Further examples

[0388] Fig. 37shows a flowchart of a method 300 for receiving a signal, according to an embodiment of the present invention. The method 300 comprises a step of receiving a signal from a data transmitter, wherein the signal has at least two pilot sequences distributed in time according to a pilot pattern, wherein at least a second pilot sequence of the at least two pilot sequences has a time-varying frequency shift compared to a first pilot sequence of the at least two pilot sequences. Furthermore, the method comprises a step of detecting the at least two pilot sequences based on a correlation pattern, wherein the correlation pattern is frequency-adapted to the time-varying frequency shift in order to reduce an influence of the time-varying frequency shift.

[0389] Fig. 38shows a flowchart of a method 310 for receiving a signal, according to a further embodiment of the present invention. The method 310 comprises a step 312 of receiving a signal from a data transmitter, wherein the signal has at least two pilot sequences distributed in time according to a pilot pattern, wherein at least a second pilot sequence of the at least two pilot sequences has a time-varying frequency shift compared to a first pilot sequence of the at least two pilot sequences. Furthermore, the method comprises a step 314 of detecting the at least two pilot sequences based on a correlation pattern selected from a set of correlation patterns, wherein the correlation patterns of the set of correlation patterns reduce different influences of time-varying frequency shifts.

[0390] In the following, further embodiments of the present invention are described, which can be used alone or in combination with the embodiments described above.

[0391] Embodiments provide a data receiver, wherein the data receiver is designed to receive a signal having a plurality of partial data packets [e.g. distributed in time and frequency according to a hopping pattern], wherein the plurality of partial data packets each comprise a part of a data packet, wherein the data receiver has a multi-stage correlator which is designed to perform a multi-stage correlation [e.g. of the received signal (e.g. in a first correlation stage) and a version of the received signal conditioned (e.g. by the first correlation stage) (e.g. in a second correlation stage)] in order to detect the partial data packets [e.g. based on preambles thereof or by means of a blind estimation method] in the received signal, wherein a second correlation stage of the multi-stage correlator is designed to determine the partial data packets based on correlation results [e.g.based on the processed version of the received signal] of a first correlation stage of the multi-stage correlator.

[0392] In embodiments, the multi-stage correlator may be configured to detect the plurality of partial data packets based on preambles thereof in the received signal.

[0393] In embodiments, the plurality of sub-data packets may be distributed in time and frequency according to a hopping pattern, wherein the multi-stage correlator is configured to detect the plurality of sub-data packets [e.g., based on preambles thereof] in the received signal or a version derived therefrom [e.g., a plurality of sub-band signals].

[0394] In embodiments, the received signal may comprise a plurality of subband signals, wherein the plurality of subband signals comprise different [e.g., partially overlapping] subbands of the signal [e.g., broadband signal] [For example, the data receiver may be configured to obtain a received signal comprising the plurality of subband signals based on the signal (e.g., broadband signals).

[0395] In embodiments, the plurality of subband signals may be used directly for the correlation performed by the multi-stage correlator.

[0396] In embodiments, the multi-stage correlator may be configured to perform a multi-stage correlation of at least a subset of the plurality of subband signals in order to detect the plurality of sub-data packets in the subset of the plurality of subband signals [For example, a number of the provided subband signals and / or their sampling rate may not match the corresponding values ​​of the multi-stage correlator, so that the multi-stage correlator processes only a part of the plurality of subband signals and / or only a part of the sample values].

[0397] In embodiments, the data receiver may be configured to multiply the plurality of subband signals by a complex exponential waveform to increase the frequency resolution in the multi-stage correlator.

[0398] In embodiments, the multi-stage correlator may comprise a first correlation stage, which may be configured to correlate the received signal or a version derived therefrom [e.g., a filtered and / or stored version of the signal to be received (e.g., a subband signal of the plurality of subband signals)] with a plurality of preamble sections that correspond [e.g., match (e.g., in the case of an undisturbed transmission channel)] to different [e.g., overlapping or adjacent] sections of the preambles of the plurality of partial data packets, in order to obtain a plurality of section correlation results [e.g., section correlation amplitudes; e.g., one section correlation result (e.g., one correlation amplitude) per preamble section per sample], wherein the first correlation stage is configured to combine the plurality of section correlation results [e.g., per sample] [e.g.,to add or to add incoherently (e.g. by absolute value formation)] in order to obtain a set of correlation results [e.g. (normalized) correlation amplitudes; e.g. for the signal to be received] or a subset of correlation results [e.g. (normalized) correlation amplitudes or a one-dimensional array of (normalized) correlation amplitudes; e.g. for the subband signal of the plurality of subband signals of the received signal] of the first correlation stage as correlation results of the first correlation stage.

[0399] In embodiments, the first correlation stage may be configured to normalize the plurality of section correlation results [e.g., by forming magnitude squares].

[0400] In embodiments, the first correlation stage may be configured to normalize the plurality of section correlation results as a function of a determined (e.g., calculated) power (p[n]) of the received signal or the version derived therefrom [e.g., the filtered and / or stored version of the signal to be received (e.g., the subband signal of the plurality of subband signals)] [For example, the first correlation stage may be configured to normalize the section correlation results by forming magnitude squares, dividing by the determined power, and calculating the roots of the quotients].

[0401] In embodiments, the power for normalization can be determined over several subbands.

[0402] In embodiments, the power for the normalization can be determined based on synchronization symbols and at least one data symbol of the respective partial data packets.

[0403] In embodiments, the first correlation stage may be configured to separately normalize the plurality of section correlation results, wherein the power is determined separately for each preamble section or jointly for all preamble sections.

[0404] In embodiments, the first correlation stage may comprise a plurality of queue buffers (e.g., ring buffers) configured to temporarily store the respective section correlation results, wherein the plurality of queue buffers have different storage lengths, wherein the storage lengths of the plurality of queue buffers depend on the respective preamble sections of the preambles of the plurality of partial data packets.

[0405] In embodiments, the first correlation stage can be configured to correlate at least two subband signals of the plurality of subband signals [e.g., a plurality of subband signals of the plurality of subband signals or all subband signals of the plurality of subband signals] with the plurality of preamble sections in order to obtain a subset of correlation results [e.g., (normalized) correlation amplitudes or a one-dimensional array of (normalized) correlation amplitudes] for each subband signal of the at least two subband signals, wherein the first correlation stage is configured to provide, as correlation results of the first correlation stage, a set of correlation results comprising the subsets of correlation results [e.g., the set of correlation results can comprise the one-dimensional subsets of correlation results].

[0406] In embodiments, the set of correlation results of the first correlation stage may be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results [e.g., a sequence of] describes sampling times of the received signal [e.g., time direction], wherein a second dimension of the two-dimensional array of correlation results describes subbands of the received signal [e.g., frequency direction].

[0407] In embodiments, the first correlation stage may comprise an [e.g., multi-channel] output queue buffer [e.g., ring buffer] configured to temporarily store the set of correlation results of the first correlation stage.

[0408] In embodiments, the first correlation stage may be configured to perform a maximum formation over correlation results of adjacent subband signals and to discard the smaller values.

[0409] In embodiments, the plurality of partial data packets may have at least two different preambles, wherein the first correlation stage is configured to correlate the received signal with a second plurality of preamble sections that correspond to different [e.g. overlapping or adjacent] sections of a second preamble of the plurality of partial data packets [e.g. match (e.g. in an undisturbed transmission channel)] in order to obtain at least a second plurality of section correlation results [e.g. section correlation amplitudes; e.g. one section correlation result (e.g. one correlation amplitude) per preamble section per sample value], wherein the first correlation stage is configured to combine the second plurality of section correlation results [e.g. per sample value] [e.g. add them or add them incoherently (e.g.by magnitude formation)] to obtain a second set of correlation results [e.g. (normalized) correlation amplitudes; e.g. for the signal to be received] or a second subset of correlation results [e.g. (normalized) correlation amplitudes or a one-dimensional array of (normalized) correlation amplitudes; e.g. for the subband signal of the plurality of subband signals of the received signal to be processed].

[0410] In embodiments, the at least two preambles may have different lengths.

[0411] In embodiments, the plurality of partial data packets may have the same preamble.

[0412] In embodiments, the at least two partial data packets can be a plurality of partial data packets, wherein at least two groups of partial data packets of the plurality of partial data packets [e.g. the at least two groups of partial data packets are real [e.g. disjoint] subsets of the plurality of partial data packets], have the same relative group hopping pattern in groups [e.g. such that partial data packets of the at least two groups of partial data packets have the same relative time and frequency distance from one another, or in other words, such that partial data packets of a first group of partial data packets have the same relative hopping pattern (= group hopping pattern) as partial data packets of a second group of partial data packets], wherein the second correlation stage is designed to determine from the set of correlation results [e.g.a two-dimensional array of correlation results] of the first correlation stage, to select groups of correlation results based on a group correlation pattern [e.g. which specifies time and frequency intervals of the correlation results of the two-dimensional array of correlation results] derived from the group hopping pattern [e.g. which specifies relative time and frequency intervals of the group of sub-data packets], and to combine them in groups [e.g. to add them] to obtain a set of correlation results of the second correlation stage.

[0413] For example, a second data packet of the first group of data packets may have the same time and frequency spacing from a first data packet of the first group of data packets as a fourth data packet of the second group of data packets has from a third data packet of the second group of data packets.

[0414] In embodiments, the second correlation stage may be configured to select the groups of correlation results from the set of correlation results of the first correlation stage in the time and / or frequency direction based on the group correlation pattern.

[0415] In embodiments, the set of correlation results of the first correlation stage may be a two-dimensional array of correlation results, wherein the group correlation pattern indicates time and frequency intervals of the correlation results of the two-dimensional array of correlation results of the first correlation stage that correspond to the relative time and frequency intervals of the group hopping pattern of the groups of sub-data packets.

[0416] In embodiments, the set of correlation results of the second correlation stage may be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results describes a [e.g., relative] temporal position of the group of partial data packets [time direction], wherein a second dimension of the two-dimensional array of correlation results describes a [e.g., relative] frequency position of the group of partial data packets [frequency direction].

[0417] In embodiments, at least one dimension [e.g., frequency direction] of the two-dimensional array of correlation results of the second correlation stage may be smaller than the respective at least one dimension of the two-dimensional array of correlation results of the first correlation stage.

[0418] In embodiments, the second correlation stage may comprise an [e.g., two-dimensional] output queue buffer [e.g., ring buffer] configured to temporarily store the set of correlation results of the second correlation stage.

[0419] In embodiments, at least two further groups of partial data packets of the plurality of partial data packets can have, in groups, the same relative further group hopping pattern [e.g. so that partial data packets of the at least two second further groups of partial data packets have the same relative time and frequency distance from one another, or in other words, so that partial data packets of a third group of partial data packets have the same relative further hopping pattern (= further group hopping pattern) as partial data packets of a fourth group of partial data packets], wherein the second correlation stage is designed to generate further groups of correlation results from the set of correlation results [e.g. a two-dimensional array of correlation results] of the first correlation stage based on a further group correlation pattern [e.g.which indicates time and frequency intervals of the correlation results of the two-dimensional array of correlation results] derived from the further group hopping pattern [e.g. which indicates relative time and frequency intervals of the second group of partial data packets], and combine them group-wise [e.g. add them] to obtain a set of further correlation results of the second correlation level, wherein the group hopping pattern and the further group hopping pattern are different.

[0420] In embodiments, the at least two groups of partial data packets can form a sequence, wherein the at least two groups of partial data packets have a relative group sequence hopping pattern [e.g. relative time and frequency intervals between the groups] to one another, wherein the data receiver has a third correlation stage which is designed to select groups of correlation results from the set of correlation results [e.g. a two-dimensional array of correlation results] of the second correlation stage based on a group sequence correlation pattern [e.g. which indicates time and frequency intervals of the correlation results of the two-dimensional array of correlation results] derived from the group sequence hopping pattern and to combine them in groups [e.g. to add them] to obtain a set of correlation results of the third correlation stage.

[0421] In embodiments, the third correlation stage may be configured to select the groups of correlation results from the set of correlation results of the second correlation stage in the time and / or frequency direction based on the group sequence correlation pattern.

[0422] In embodiments, the set of correlation results of the second correlation stage may be a two-dimensional array of correlation results, wherein the group sequence correlation pattern indicates time and frequency spacings of the correlation results of the two-dimensional array of correlation results of the second correlation stage that correspond to the relative time and frequency spacings of the group sequence hopping pattern.

[0423] In embodiments, the set of correlation results of the third correlation stage may be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results describes a [e.g., relative] temporal position of the groups of partial data packets [time direction], wherein a second dimension of the two-dimensional array of correlation results describes a relative frequency position of the groups of partial data packets [frequency direction].

[0424] In embodiments, at least one dimension [e.g., frequency direction] of the two-dimensional array of correlation results of the third correlation stage may be smaller than the respective at least one dimension of the two-dimensional array of correlation results of the second correlation stage.

[0425] In embodiments, the third correlation stage may comprise an [e.g., multi-channel] output queue buffer [e.g., ring buffer] configured to temporarily store the set of correlation results of the third correlation stage.

[0426] In embodiments, the data receiver may be configured to pass the set of correlation results in a suitable form to a subsequent packet detection.

[0427] In embodiments, the at least two groups of partial data packets can form a further sequence [e.g. a first group and a second group of partial data packets form a first sequence, wherein a third group and a fourth group of partial data packets form a second sequence], wherein the at least two groups of partial data packets have a relative further group sequence hopping pattern [e.g. relative time and frequency intervals between the groups] to one another, wherein the data receiver has a third correlation stage which is designed to generate groups of correlation results from the set of correlation results [e.g. a two-dimensional array of correlation results] of the second correlation stage based on a further group sequence correlation pattern [e.g.which indicates time and frequency intervals of the correlation results of the two-dimensional array of correlation results] derived from the further group sequence hopping pattern and combining them group-wise [e.g. adding them] to obtain a further set of correlation results of the third correlation level, wherein the group sequence hopping pattern and the further group sequence hopping pattern are different.

[0428] In embodiments, the data packets may be distributed in time and frequency according to a hopping pattern, wherein the second correlation stage may be configured to select groups of correlation results from the set of correlation results [e.g., a two-dimensional array of correlation results] of the first correlation stage based on a correlation pattern [e.g., which indicates time and frequency intervals of the correlation results of the two-dimensional array of correlation results] derived from the hopping pattern of the partial data packets, and to combine them in groups [e.g., to add or coherently add] to obtain a set of correlation results of the second correlation stage.

[0429] In embodiments, the second correlation stage may be configured to select the groups of correlation results from the set of correlation results of the first correlation stage in the time and / or frequency direction based on the correlation pattern.

[0430] In embodiments, the set of correlation results of the first correlation stage may be a two-dimensional array of correlation results, wherein the correlation pattern indicates time and frequency intervals of the correlation results of the two-dimensional array of correlation results of the first correlation stage that correspond to the relative time and frequency intervals of the hopping pattern of the sub-data packets.

[0431] In embodiments, the set of correlation results of the second correlation stage may be a two-dimensional array of correlation results, wherein a first dimension of the two-dimensional array of correlation results describes a [e.g., relative] temporal position of partial data packets [time direction], wherein a second dimension of the two-dimensional array of correlation results describes a [e.g., relative] frequency position of partial data packets [frequency direction].

[0432] In embodiments, the first dimension and / or the second dimension of the two-dimensional array of correlation results of the second correlation stage may be smaller than the respective dimension of the two-dimensional array of correlation results of the first correlation stage.

[0433] In embodiments, the second correlation stage may comprise an [e.g., multi-channel] output queue buffer [e.g., ring buffer] configured to temporarily store the set of correlation results of the second correlation stage.

[0434] In embodiments, the data receiver may be configured to pass the set of correlation results in a suitable form to a subsequent packet detection.

[0435] In embodiments, at least two further groups of partial data packets of the plurality of partial data packets can have, in groups, the same relative further group hopping pattern [e.g. so that partial data packets of the at least two second further groups of partial data packets have the same relative time and frequency distance from one another, or in other words, so that partial data packets of a third group of partial data packets have the same relative further hopping pattern (= further group hopping pattern) as partial data packets of a fourth group of partial data packets], wherein the second correlation stage is designed to generate further groups of correlation results from the set of correlation results [e.g. a two-dimensional array of correlation results] of the first correlation stage based on a further group correlation pattern [e.g.which indicates time and frequency intervals of the correlation results of the two-dimensional array of correlation results] derived from the further group hopping pattern [e.g. which indicates relative time and frequency intervals of the second group of partial data packets], and combine them group-wise [e.g. add them] to obtain a set of further correlation results of the second correlation level, wherein the group hopping pattern and the further group hopping pattern are different.

[0436] Further embodiments provide a method for receiving a signal, wherein the signal comprises a plurality of partial data packets (e.g., distributed in time and frequency according to a hopping pattern), wherein the plurality of partial data packets each comprise a part of a data packet. The method comprises a step of performing a multi-stage correlation (e.g., of the received signal (e.g., in a first correlation stage) and a conditioned version of the signal to be processed (e.g., in a second correlation stage)) in order to detect the plurality of partial data packets (e.g., based on preambles thereof) in the received signal, wherein a second correlation stage of the multi-stage correlation is performed based on correlation results (e.g., based on the conditioned version of the received signal) of a first correlation stage of the multi-stage correlation.

[0437] Further embodiments provide a data receiver which is designed to receive a signal which has at least one data packet, wherein the data packet has a preamble, wherein the data receiver has a correlation stage which is designed to correlate the received signal or a version derived therefrom [e.g. a filtered and / or stored version of the received signal] with a plurality of preamble sections [e.g. which correspond to different (e.g. overlapping or adjacent) sections of the preamble of the data packet (e.g. coincide (e.g. in an undisturbed transmission channel))] in order to obtain a plurality of section correlation results [e.g. section correlation amplitudes; e.g. a section correlation result (e.g.a correlation amplitude) per preamble section per sample], wherein the first correlation stage is designed to combine the plurality of section correlation results [e.g. per sample] [e.g. to add them or to add them incoherently (e.g. by forming the magnitude)] in order to obtain a set of correlation results [e.g. (normalized) correlation amplitudes; e.g. for the signal to be received], wherein the first correlation stage is designed to normalize the plurality of section correlation results [e.g. by forming magnitude squares], wherein the first correlation stage is designed to normalize the plurality of section correlation results as a function of a determined (e.g. calculated) power or interference power (p[n]) of the signal to be received or the version derived therefrom [e.g. the filtered and / or stored version of the signal to be received].

[0438] In embodiments, the correlation stage may be configured to correlate the signal to be processed or a version derived therefrom with a plurality of preamble sections that correspond [e.g., match (e.g., in an undisturbed transmission channel)] to different [e.g., overlapping or adjacent] sections of the preamble of the data packet, in order to obtain the plurality of section correlation results [e.g., section correlation amplitudes; e.g., one section correlation result (e.g., one correlation amplitude) per preamble section per sample].

[0439] In embodiments, the first correlation stage may be configured to normalize the section correlation results by forming magnitude squares, dividing by the determined power, and calculating the roots of the quotients.

[0440] Further embodiments provide a method for receiving a signal, wherein the signal comprises at least one data packet, wherein the data packet comprises a preamble. The method comprises a step of correlating the received signal or a version derived therefrom [e.g., a filtered and / or stored version of the received signal] with a plurality of preamble sections that correspond [e.g., match (e.g., in an undisturbed transmission channel)] to different [e.g., overlapping or adjacent] sections of the preamble of the data packet, in order to obtain a plurality of section correlation results [e.g., section correlation amplitudes; e.g., one section correlation result (e.g., one correlation amplitude) per preamble section per sample].The method further comprises a step of normalizing the plurality of section correlation results, wherein the plurality of section correlation results are normalized as a function of a determined (e.g., calculated) power or interference power (p[n]) of the received signal or the version derived therefrom [e.g., the filtered and / or stored version of the received signal]. The method further comprises a step of combining the plurality of normalized section correlation results [e.g., per sample] [e.g., adding them or adding them incoherently] to obtain a set of correlation results [e.g., (normalized) correlation amplitudes; e.g., for the received signal].

[0441] Although some aspects have been described in connection with a device, it should be understood that these aspects also represent a description of the corresponding method, so that a block or component of a device can also be understood as a corresponding method step or as a feature of a method step. Analogously, aspects described in connection with or as a method step also represent a description of a corresponding block, detail, or feature of a corresponding device. Some or all of the method steps can be carried out by (or using) a hardware apparatus, such as a microprocessor, a programmable computer, or an electronic circuit. In some embodiments, some or more of the key method steps can be carried out by such an apparatus.

[0442] Depending on specific implementation requirements, embodiments of the invention may be implemented in hardware or software. The implementation may be performed using a digital storage medium, such as a floppy disk, a DVD, a Blu-ray Disc, a CD, a ROM, a PROM, an EPROM, an EEPROM, or a FLASH memory, a hard disk, or other magnetic or optical storage device storing electronically readable control signals that can interact or cooperate with a programmable computer system to perform the respective method. Therefore, the digital storage medium may be computer-readable.

[0443] Some embodiments according to the invention thus comprise a data carrier having electronically readable control signals capable of interacting with a programmable computer system such that one of the methods described herein is carried out.

[0444] In general, embodiments of the present invention may be implemented as a computer program product having program code, wherein the program code is operable to perform one of the methods when the computer program product is run on a computer.

[0445] The program code can, for example, also be stored on a machine-readable medium.

[0446] Other embodiments include the computer program for performing one of the methods described herein, wherein the computer program is stored on a machine-readable carrier.

[0447] In other words, an embodiment of the method according to the invention is thus a computer program which has a program code for carrying out one of the methods described herein when the computer program runs on a computer.

[0448] A further embodiment of the method according to the invention is thus a data carrier (or a digital storage medium or a computer-readable medium) on which the computer program for performing one of the methods described herein is recorded. The data carrier, the digital storage medium, or the computer-readable medium are typically physical and / or non-perishable or non-transitory.

[0449] A further embodiment of the method according to the invention is thus a data stream or a sequence of signals which the computer program uses to

[0450] Performing one of the methods described herein. The data stream or sequence of signals may, for example, be configured to be transferred via a data communication connection, for example via the Internet.

[0451] A further embodiment comprises a processing device, for example a computer or a programmable logic device, which is configured or adapted to carry out one of the methods described herein.

[0452] A further embodiment comprises a computer on which the computer program for performing one of the methods described herein is installed.

[0453] A further embodiment according to the invention comprises a device or a system designed to transmit a computer program for performing at least one of the methods described herein to a recipient. The transmission can be electronic or optical, for example. The recipient can be, for example, a computer, a mobile device, a storage device, or a similar device. The device or system can, for example, comprise a file server for transmitting the computer program to the recipient.

[0454] In some embodiments, a programmable logic device (e.g., a field-programmable gate array, an FPGA) may be used to perform some or all of the functionality of the methods described herein. In some embodiments, a field-programmable gate array may interact with a microprocessor to perform any of the methods described herein. In general, in some embodiments, the methods are performed by any hardware device. This may be general-purpose hardware, such as a computer processor (CPU), or method-specific hardware, such as an ASIC.

[0455] The devices described herein may be implemented, for example, using a hardware apparatus, or using a computer, or using a combination of a hardware apparatus and a computer.

[0456] The devices described herein, or any components of the devices described herein, may be implemented at least partially in hardware and / or in software (computer program).

[0457] The methods described herein may be implemented, for example, using a hardware apparatus, or using a computer, or using a combination of a hardware apparatus and a computer.

[0458] The methods described herein, or any components of the methods described herein, may be implemented at least in part by hardware and / or by software.

[0459] The above-described embodiments are merely illustrative of the principles of the present invention. It is understood that modifications and variations of the arrangements and details described herein will be apparent to others skilled in the art. Therefore, it is intended that the invention be limited only by the scope of the following claims and not by the specific details presented in the description and explanation of the embodiments herein. Bibliography

[0460] [1] G. Kilian, M. Breiling, HH Petkov, H. Lieske, F. Beer, J. Robert, and A. Heuberger, "Increasing Transmission Reliability for Telemetry Systems Using Telegram Splitting," IEEE Transactions on Communications, vol. 63, no. 3, pp. 949-961, Mar. 2015. [2] DE 10 2011 082098 B4 [3] WO 2017 / 167366 A1 [4] DE 10 2017 206248 A1 [5] ETSI TS 103357, ETSI Technical Specification

Claims

1. A data receiver (110), wherein the data receiver (110) is configured to receive a signal (120) from a data transmitter (100), wherein the signal (120) comprises at least two pilot sequences (144_1-144_n) which are distributed in time according to a pilot pattern (140), wherein at least one second pilot sequence (144_2) of the at least two pilot sequences (144_1-144_n) comprises a temporally varying frequency shift relative to a first pilot sequence (144_1) of the at least two pilot sequences (144_1-144_n), wherein the data receiver comprises a detector (130) with a correlator (122) which is configured to detect the at least two pilot sequences (144_1-144_n) based on a correlation pattern, wherein the correlation pattern corresponds to the pilot pattern (140) and describes a distribution of the at least two pilot sequences (144_1-144_n) in time and frequency, wherein the correlation pattern is provided with a temporally varying frequency shift in order to reduce an influence of the temporally varying frequency shift of the at least one second pilot sequence (144_2) relative to the first pilot sequence (144_1).

2. The data receiver (110) according to the preceding claim, wherein at least one of the data transmitter (100) and the data receiver (110) moves relative to the other of the data transmitter (100) and the data receiver (110), wherein the temporally varying frequency shift results from a relative movement change between the data transmitter (100) and the data receiver (110).

3. The data receiver (110) according to any of the preceding claims, wherein the data receiver (110) is configured to estimate the temporally varying frequency shift of the at least one second pilot sequence (144_2) relative to the first pilot sequence (144_1) and to adapt the correlation pattern in frequency based on the estimated temporally varying frequency shift.

4. The data receiver (110) according to any of claims 1 to 3, wherein the correlator (122) comprises at least two correlation units (118_1, 118_2) operating in parallel, wherein a first correlation unit (118_1) of the at least two correlation units (118_1, 118_2) operating in parallel is configured to perform, based on the correlation pattern, correlations of the at least two pilot sequences (144_1-144_n) with reference sequences in order to obtain a set of correlation results (119_1) of the first correlation unit (118_1), wherein a second correlation unit (118_2) of the at least two correlation units (118_1, 118_2) operating in parallel is configured to perform, based on the correlation pattern, correlations of the at least two pilot sequences (144_1-144_n) with reference sequences in order to obtain a set of correlation results (119_2) of the second correlation unit (118_2), wherein the data receiver (110) is configured to adapt the correlation pattern used in the at least two correlation units (118_1, 118_2) operating in parallel differently in frequency in order to reduce influences of different temporally varying frequency shifts, wherein the data receiver (110) is configured to select a set of correlation results from the sets of correlation results (119_1, 119_2) of the at least two correlation units (118_1, 118_2) operating in parallel, based on which the detection of the at least two pilot sequences (144_1-144_n) takes place, depending on values of the respective set of correlation results.

5. The data receiver (110) according to any of claims 1 to 3, wherein the correlator (122) is a multi-stage correlator which comprises a first correlation stage (124) and at least one second correlation stage (128, 129) following the first correlation stage, which second correlation stage operates based on correlation results (125) of the first correlation stage (124), wherein the detector (130) is configured to detect the at least two pilot sequences (144_1-144_n) in the received signal (121), wherein the data receiver (110) is configured to adapt at least one correlation pattern, which is used in at least one correlation stage from the at least one second correlation stage (128, 129) of the multi-stage correlator (122), in frequency in order to reduce the influence of the temporally varying frequency shift.

6. The data receiver (110) according to any of the preceding claims, wherein the data receiver (110) comprises a decoder (133) which is configured to decode data comprising the received signal, wherein the decoder (133) is configured to take into account or correct the temporally varying frequency shift when decoding the data.

7. A method (300), performed by a data receiver (110), comprising the following steps: receiving (302) a signal from a data transmitter, wherein the signal comprises at least two pilot sequences which are distributed in time according to a pilot pattern, wherein at least one second pilot sequence of the at least two pilot sequences comprises a temporally varying frequency shift relative to a first pilot sequence of the at least two pilot sequences, and detecting the at least two pilot sequences based on a correlation pattern, wherein the correlation pattern corresponds to the pilot pattern and describes a distribution of the at least two pilot sequences in time and frequency, wherein the correlation pattern is provided with a temporally varying frequency shift in order to reduce an influence of the temporally varying frequency shift of the at least one second pilot sequence (144_2) relative to the first pilot sequence.

8. A data receiver (110), wherein the data receiver (110) is configured to receive a signal (120) from a data transmitter (100), wherein the signal (120) comprises at least two pilot sequences (144_1-144_n) which are distributed in time according to a pilot pattern (140), wherein at least one second pilot sequence (144_2) of the at least two pilot sequences (144_1-144_n) comprises a temporally varying frequency shift relative to a first pilot sequence (144_1) of the at least two pilot sequences (144_1-144_n), wherein the data receiver (110) comprises a detector (130) with a correlator (122) which is configured to detect the at least two pilot sequences (144_1-144_n) based on a correlation pattern selected from a set of correlation patterns, wherein the correlation patterns of the set of correlation patterns correspond to the pilot pattern (140) and describe a distribution of the at least two pilot sequences (144_1-144_n) in time and frequency, wherein the correlation patterns of the set of correlation patterns are provided with different temporal frequency shifts in order to reduce influences of different temporally varying frequency shifts of the at least one second pilot sequence (144_2) relative to the first pilot sequence (144_1).

9. The data receiver according to claim 8, wherein the detector (130) is configured to estimate the temporally varying frequency shift of the at least one second pilot sequence (144_2) relative to the first pilot sequence (144_1), wherein the detector (130) is configured to select a correlation pattern from the set of correlation patterns depending on the estimated frequency shift in order to obtain the selected correlation pattern.

10. The data receiver (110) according to claim 8 or 9, wherein the data receiver (110) comprises a decoder (133) which is configured to decode data comprising the signal (120), wherein the decoder (133) is configured to take into account or correct the temporally varying frequency offset, which is reduced by the correlation pattern selected for the detection, when decoding the data.

11. The data receiver (110) according to any of claims 8 to 10, wherein the signal (120) comprises a plurality of partial data packets (142_1-142_n) which are distributed in time and frequency according to a hopping pattern (140), wherein the at least two pilot sequences (144_1-144_n) are a plurality of pilot sequences (144_1-144_n), wherein the plurality of partial data packets (142_1-142_n) each comprise a pilot sequence of the plurality of pilot sequences (144_1-144_n), wherein the set of correlation patterns is derived from the hopping pattern (140).

12. The data receiver (110) according to any of claims 8 to 11, wherein the set of correlation patterns is derived from at least one hopping pattern defined in ETSI TS 103 357.

13. A method (310), performed by a data receiver (110), comprising the following steps: receiving (312) a signal from a data transmitter, wherein the signal comprises at least two pilot sequences which are distributed in time according to a pilot pattern, wherein at least one second pilot sequence of the at least two pilot sequences comprises a temporally varying frequency shift relative to a first pilot sequence of the at least two pilot sequences, and detecting (314) the at least two pilot sequences based on a correlation pattern selected from a set of correlation patterns, wherein the correlation patterns of the set of correlation patterns correspond to the pilot pattern and describe a distribution of the at least two pilot sequences in time and frequency, wherein the correlation patterns of the set of correlation patterns are provided with different temporally varying frequency shifts in order to reduce influences of different temporally varying frequency shifts of the at least one second pilot sequence relative to the first pilot sequence.

14. A computer program, comprising instructions which, when the program is executed by a computer, cause the same to carry out the method according to claim 7 or 13.