JP7316994B2 - Optimized Preamble and Method for Interference Robust Packet Detection for Telemetry Applications - Google Patents

Description

Embodiments relate to receivers. A further embodiment relates to a method for receiving data packets. Some embodiments relate to optimized preambles. Some embodiments relate to interference robust detection. Some embodiments relate to preamble segmentation. Some embodiments relate to non-coherent correlation. Some embodiments relate to pilot signaling.

Systems are known for transmitting small amounts of data, eg sensor data, from a large number of nodes, eg heating, electricity or water meters, to a base station. A base station receives signals from, and possibly controls, multiple nodes. At the base station, receivers with higher computational power and more complex hardware, ie higher performance, are available. Only inexpensive crystals with frequency offsets of typically 10 ppm or more are available at the node.

[G. Kilian, H. Petkov, R. Psiuk, H. Lieske, F. Beer, J. Robert, and A. Heuberger, "Improved coverage for low-power telemetry systems using telegram splitting," in Proceedings of 2013 European Conference on Smart Objects, Systems and Technologies (SmartSysTech), 2013], coverage improvements for low-power telemetry systems using telegram splitting are shown.

[G. Kilian, M. Breiling, H. H. 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], an improvement in transmission reliability for telemetry systems using telegram splitting is shown.

[R. De Gaudenzi, F. Giannetti, and M. Luise, "Signal recognition and signature code acquisition in CDMA mobile packet communications," IEEE Transactions on Vehicular Tecohnology, vol. 47, no. 1, pp. 196-208, Feb. 1998], signal recognition and signature code acquisition in CDMA (code division multiple access) mobile packet communications is discussed.

[J. Block and E. W. Huang, "Packet Acquisition Performance of Frequency-Hop Spread-Spectrum Systems in Partial-Band Interference," in IEEE Military Communications Conference, 2007. MILCOM 2007, 2007, pp. 1-7], partially Packet acquisition performance of frequency-hopped spread spectrum systems in high-band interference is discussed.

WO2013/030303 shows a battery powered fixed sensor assembly with unidirectional data transmission.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a concept for improving communication between a transmitter and a receiver.

This object is solved by the independent claims.

Embodiments provide a receiver comprising a receiving unit and a synchronization unit. The receiving unit is configured to receive data packets including pilot sequences. The synchronization unit combines the pilot sequence with at least two partial reference sequences corresponding to the reference sequences for the pilot sequences of the data packet to obtain partial correlation results for each of the at least two partial reference sequences. configured to be independently correlated with a reference sequence; The synchronization unit is then configured to non-coherently add the partial correlation results to obtain a coarse correlation result for the data packet.

The idea of the present invention is to combine data packets (or pilot sequences of data packets) with pilot sequences contained in the data packets in order to obtain partial correlation results for each of at least two partial reference sequences. Synchronizing data packets by correlating with at least two partial reference sequences each shorter than the sequence. Here the partial correlation results are added non-coherently, thereby reducing the effect of the transmission channel where the data packets are reduced to improve synchronization performance.

A further embodiment provides a method, the method comprising:
- receiving data packets containing pilot sequences;
- the pilot sequence separately with at least two partial reference sequences corresponding to the reference sequence for the pilot sequence of the data packet, to obtain partial correlation results for the at least two partial reference sequences; a step of correlating;
- adding the partial correlation results non-coherently to obtain a correlation result for the data packet;
Prepare.

A further embodiment provides a receiver comprising a receiving unit and a synchronization unit. The receiving unit is configured to receive data packets (eg, at least two data packets). At least two of the data packets (eg, each of the at least two data packets) include one partial pilot sequence of the at least two partial pilot sequences. The synchronization unit is configured to individually correlate the partial pilot sequence with the at least two partial reference sequences to obtain partial correlation results for each of the at least two partial reference sequences. be. Thereby, the synchronization unit is configured to non-coherently add the partial correlation results to obtain a coarse correlation result for the two data packets.

A further embodiment provides a method, the method comprising:
- receiving at least two data packets, each of the at least two data packets comprising a partial pilot sequence of the at least two partial pilot sequences;
- individually correlating the partial pilot sequence with the at least two partial reference sequences to obtain a partial correlation result for each of the at least two partial reference sequences;
- non-coherently adding the partial correlation results to obtain a coarse correlation result for the two data packets;
Prepare.

A further embodiment provides a receiver comprising a receiving unit and a synchronization unit. The receiving unit is configured to receive data packets including pilot sequences. A synchronization unit is configured to correlate the reference sequence with the pilot sequence to obtain a correlation result. Thereby, the synchronization unit applies weight factors to symbols of a data packet, applies weight factors to symbols of a pilot sequence, or applies individual weight factors to each symbol of a pilot sequence. configured as

A further embodiment provides a receiver comprising a receiving unit and a synchronization unit. The receiving unit is configured to receive data packets including pilot sequences. A synchronization unit is configured to correlate the reference sequence with the pilot sequence to obtain a correlation result. The synchronization unit is thereby configured to use a correlation window for detecting data packets. Here, data packets are detected by detecting the highest of all correlation peaks exceeding a predetermined threshold within the correlation window.

Advantageous embodiments are described in the dependent claims.

In some embodiments, non-coherently summing the partial correlation results yields the post-correlation phase information, e.g., the absolute, squared, or approximate absolute value of the partial correlation results. Including discarding by adding.

In some embodiments, the synchronization unit is configured to add the partial correlation results non-coherently by adding absolute values, squared absolute values or approximate absolute values of the partial correlation results. can

In some embodiments, the at least two partial reference sequences can be at least two different portions of the reference sequence.

In some embodiments, a data packet can include at least two partial reference sequences as reference sequences.

In some embodiments, the receiving unit can be configured to receive at least two data packets. Here, only part of the at least two data packets contain pilot sequences, for example the receiving unit may be configured to receive data packets without pilot sequences.

In some embodiments, the receiving unit can be configured to receive at least two data packets. Here, each of the at least two data packets can include a pilot sequence. The synchronization unit corresponds respective pilot sequences of the at least two data packets to obtain partial correlation results for each of the at least two partial reference sequences for each of the at least two data packets. individually correlated with at least two partial reference sequences corresponding to the reference sequence for the pilot sequence of the data packet. Further, the synchronization unit non-coherently adds at least a portion of the partial correlation results for each of the at least two data packets to obtain coarse correlation results for each of the at least two data packets. and configured to combine at least a portion of the coarse correlation results of at least two data packets to obtain a combined coarse correlation result.

The synchronization unit is configured to combine the coarse correlation results of the at least two data packets by using an ideal Neyman-Pearson detector sum or approximation of the coarse correlation results of the at least two data packets. can

In some embodiments, the at least two data packets can be part of a telegram that can be sent separated into at least two data packets. The receiver may further comprise a data packet combining unit configured to combine at least two data packets to obtain a telegram.

The synchronization unit can be further configured to coherently add the partial correlation results to obtain a fine correlation for the data packet.

Further, if the combined coarse correlation result exceeds a predetermined threshold, the synchronization unit performs partial correlation for each of the at least two data packets to obtain fine correlation results for each of the at least two data packets. may be further configured to coherently sum the correlation results. For example, the synchronization unit may be configured to combine fine correlation results of at least two data packets to obtain a combined fine correlation result.

The synchronization unit is configured to normalize the coarse correlation results of the at least two data packets and combine the normalized coarse correlation results of the at least two data packets to obtain a coarse correlation result for the telegram. can

Further, the synchronization unit normalizes the fine correlation results of the at least two data packets and combines the normalized fine correlation results of the at least two data packets to obtain a combined fine correlation result. can be configured to

In some embodiments, the synchronization unit can be configured to estimate the frequency offset of data packets.

For example, the synchronization unit can be configured to estimate the frequency offset for large offsets (eg, above the data rate) by oversampling in the frequency domain and parallel correlation over multiple frequencies. The correlation result with the highest peak provides the coarse frequency offset.

Additionally, the synchronization unit can be configured to estimate the frequency offset for small offsets (eg, less than the data rate) based on the phase difference between adjacent symbols.

Further, the synchronization unit may estimate the frequency offset in the case of sufficiently large partial pilot sequences (e.g. dependent on the signal-to-noise ratio) based directly on these partial pilot sequences. can be configured.

Further, the synchronization unit can be configured to estimate the frequency offset based on the coarse correlation result to obtain the coarse frequency offset or based on the fine correlation result to obtain the fine frequency offset. .

The receiver extracts the header information from the data packet encoded at the phase shift of the pilot sequence by applying frequency correction of the data packet using the estimated frequency offset to estimate the phase shift of the pilot sequence. a header extraction unit configured to extract the

In some embodiments, the synchronization unit normalizes the symbols of the pilot sequence to obtain a normalized pilot sequence, and divides the normalized pilot sequence into at least two partial reference sequences. can be configured to individually correlate with .

In some embodiments, the synchronization unit calculates a variance of the partial correlation results for the data packet and detects the data packet if the variance of the partial correlation results for the data packet is less than or equal to a predetermined threshold. It can be configured to detect packets.

In some embodiments, the synchronization unit is configured to apply weight factors to the symbols of the data packet or apply individual weight factors to respective symbols of the at least two partial pilot sequences. configured to apply a separate weighting factor to each symbol of the at least two partial pilot sequences; or configured to apply a separate weighting factor to each symbol of the at least two partial pilot sequences; It may be configured to apply weighting factors, or it may be configured to apply individual weighting factors to each symbol of a data packet.

In some embodiments, the synchronization unit detects the main lobe and side lobes of the correlation and uses the known distance between the main lobe and the side lobes to determine the detected main lobe as the correlation result. It can be configured to provide a lobe.

In some embodiments, the synchronization unit can be configured to use a correlation window for detecting data packets. Here, data packets are detected by detecting the highest of all correlation peaks exceeding a predetermined threshold within the correlation window.

Embodiments provide computationally efficient frequency-focused detection of telegrams through the use of partial correlation of preambles within subpackets and combining across multiple subpackets.

A further embodiment has no or little impact on receiver performance by using a phase offset for the transmission of the partial preamble part of the pilot of the subpacket. Provide robust transmission of (additional) header information using non-provided synchronization pilots and detection.

Further embodiments provide interference robust detection.

Embodiments of the present invention are described herein with reference to the accompanying drawings.

FIG. 1 shows a schematic block diagram of a receiver, according to one embodiment. FIG. 2a is a schematic diagram of a data packet (subpacket), according to one embodiment. Figure 2b shows a schematic diagram of a data packet (sub-packet) according to a further embodiment. Figure 2c shows a schematic diagram of a data packet (sub-packet) according to a further embodiment. Figure 2d shows a schematic diagram of a data packet (sub-packet) according to a further embodiment. Figure 2e shows a schematic diagram of a data packet (sub-packet) according to a further embodiment. Figure 2f shows a schematic diagram of a data packet (sub-packet) according to a further embodiment. FIG. 3 shows a schematic diagram of the synchronization of data packets according to EP 2 914 039 A1. FIG. 4 shows a schematic diagram of synchronization of data packets, according to one embodiment. FIG. 5 shows the amplitude of the autocorrelation function of the Barker-7 code plotter over time. FIG. 6 shows the amplitude of the autocorrelation function of the Barker-7 code with higher side lobes caused by the interferometer plotter over time. FIG. 7a shows a schematic diagram of a data packet with two partial pilot and data sequences and a long interferor overlaying the data packet for three different time slots. FIG. 7b shows the wide normalization of telegrams plotted over time for each of the three different time slots or the normalized reception power for subpackets and the normalized reception power. FIG. 8a shows a schematic diagram of a data packet with two partial pilot and data sequences and a short interferor overlaying the data packet for three different time slots. FIG. 8b shows the reception power and normalized reception power for wide normalization or subpackets of telegrams plotted over time for each of three different time slots. FIG. 9a shows a schematic diagram of a data packet with two partial pilot and data sequences and a short interferor overlaying the data packet for three different time slots. FIG. 9b shows the reception power and normalized reception power for wide normalization of symbols plotted over time for each of three different time slots. FIG. 10 is a portion of a telegram sent separated into multiple data packets over a communication channel along with a schematic diagram of the calculation of variance over all (or at least some) data packets according to one embodiment. Some data packets are shown. FIG. 11 illustrates that each data packet has two partial pilot sequences and a separate weighting factor for each partial pilot sequence for each data packet, according to one embodiment. Fig. 3 shows the pilot sequence weighting performed by applying , in a schematic diagram of three data packets; FIG. 12 shows the amplitude of the correlation function plotted over time, according to one embodiment. FIG. 13 shows a schematic diagram of a detection window, according to one embodiment. FIG. 14 shows a flowchart of a method for detecting data packets using detection windows, according to one embodiment. FIG. 15 shows in three diagrams the detection window and threshold used to detect data packets and the amplitude of correlation results plotted over time for three different time slots, according to one embodiment; . FIG. 16 is a schematic diagram of a plurality of data packets that are part of a telegram that is sent separated into a plurality of data packets over a communication channel and a partial correlation across the three data packets, according to one embodiment; and FIG. 17 shows a flowchart of a method for receiving data packets. FIG. 18 shows a schematic block diagram of a receiver, according to one embodiment. FIG. 19 shows a flowchart of a method for receiving data packets, according to one embodiment.

Identical or equivalent elements having the same or equivalent function are denoted by the same or equivalent reference numerals in the following description.

In the following description, numerous details are set forth to provide a more thorough description of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form, rather than in detail, in order to avoid obscuring the embodiments of the invention. In addition, features of different embodiments described below may be combined with each other unless stated otherwise.

FIG. 1 shows a schematic block diagram of a receiver 100 according to one embodiment. Receiver 100 comprises a receiving unit 102 and a synchronization unit 104 . The receiving unit 102 is configured to receive data packets 106 including pilot sequences 108 .

For example, the receiving unit 102 is configured to receive and demodulate a signal transmitted from the transmitter over the communication channel to the receiver 100 and provide a data stream containing the data packets 106 thereon. there is

A data packet 106 includes a pilot sequence 108 and one or more data sequences positioned before, after, or between pilot sequence 108 (not shown in FIG. 1, see, eg, FIG. 2). 110 and . The data packet 106 can be part of a telegram that is separated into multiple data packets (or subpackets) and transmitted.

Synchronization unit 104 combines pilot sequence 108 with at least two partial reference sequences 112_1 through 112_n ( n can be a natural number greater than or equal to 2). Here, synchronization unit 104 is configured to non-coherently add partial correlation results 116_1 through 116_n to obtain coarse correlation result 118 for data packet 106 .

For example, synchronization unit 104 may be configured to independently correlate the data stream provided by receiving unit 102 with at least two partial reference sequences 112_1 through 112_n.

Each of the at least two partial reference sequences 112_1 through 112_n can be shorter than the pilot sequence 108 of the data packet.

At least two partial reference sequences 112_1 through 112_n may correspond to reference sequence 114 for pilot sequence 108 of data packet 106 . That is, at least two partial reference sequences 112_1 through 112_n can be part of the reference sequence 114 for the pilot sequence 108 of the data packet. Assuming an ideal communication channel between transmitter and receiver 100, reference sequence 108 and pilot sequence are identical. Each of the at least two partial reference sequences 112_1 through 112_n can be shorter than the reference sequence 114. FIG. For example, the reference sequence 114 can be split into at least two (or n) parts (or sets) to obtain at least two (or n) partial reference sequences 112_1 to 112_n. That is, the first portion of the reference sequence 114 is the first of the at least two partial reference sequences 112_1 to 112_n, and the second portion of the reference sequence 114 is the first of the at least two partial reference sequences 112_1 to 112_n. 2, and so on (if applicable).

Synchronization unit 104 may be configured to add the partial correlation results non-coherently by adding absolute values, squared absolute values, or approximate absolute values of the partial correlation results.

A pilot (or a sequence of pilot symbols (pilot sequence)) may be sent within a data packet or subpacket. Pilots may be used for at least one of packet, time synchronization and frequency synchronization detection.

As will become apparent from the following discussion of Figures 2a to 2f, there are different methods for pilot positioning within a subpacket.

FIG. 2a shows a schematic diagram of a data packet (subpacket) 106 according to one embodiment. Data packet 106 includes a pilot sequence 108 and two data sequences 110 positioned before and after pilot sequence 108 . The symbols of data packet 106 are shown using a complex vector diagram. That is, each arrow can depict one symbol of the modulation method used to transmit the data packet.

As shown in FIG. 2a, in an embodiment, pilot sequence 108 may include at least two partial pilot sequences 108_1 through 108_n. That is, pilot sequence 108 can be separated into at least two partial pilot sequences 108_1 through 108_n. Thereby, each of the partial pilot sequences 108_1 through 108_n can have a corresponding partial reference sequence 112_1 through 112_n. For example, first partial reference sequence 112_1 may have a corresponding first partial pilot sequence 108_1 (i.e., first partial pilot sequence 108_1 may be replaced by first partial pilot sequence 108_1). The correlation peak is maximized when correlating with the reference sequence 112_1.), the second partial reference sequence 112_2 can have a corresponding second partial pilot sequence 108_2 (i.e., the second 2 partial pilot sequence 108_2 with the second partial reference sequence 112_2, the correlation peak is maximized.), and so on (if applicable).

FIG. 2b shows a schematic diagram of a data packet (subpacket) 106 according to one embodiment. Data packet 106 includes two partial pilot sequences 108_1 and 108_2 and data sequence 110 positioned between the two partial pilot sequences 108_1 and 108_2. The symbols of data packet 106 are shown using a complex vector diagram. That is, each arrow can depict one symbol of the modulation method used to transmit the data packet.

FIG. 2c shows a schematic diagram of a data packet (subpacket) 106 according to one embodiment. Data packet 106 includes two partial pilot sequences 108_1 and 108_2 and data sequence 110 positioned between the two partial pilot sequences 108_1 and 108_2. Thereby, the second partial pilot sequence 108_2 is longer (eg, twice) than the first reference sequence 108_1. The symbols of data packet 106 are shown using a complex vector diagram. That is, each arrow can depict one symbol of the modulation method used to transmit the data packet.

FIG. 2d shows a schematic diagram of a data packet (subpacket) 106 according to one embodiment. Data packet 106 includes two partial pilot sequences 108_1 and 108_2 and three data sequences 110 positioned before, after, and between the two partial pilot sequences 108_1 and 108_2. . The symbols of data packet 106 are shown using a complex vector diagram. That is, each arrow can depict one symbol of the modulation method used to transmit the data packet.

FIG. 2e shows a schematic diagram of a data packet (subpacket) 106 according to one embodiment. Data packet 106 includes two partial pilot sequences 108_1 and 108_2 and three data sequences 110 positioned before, after, and between the two partial pilot sequences 108_1 and 108_2. . The symbols of data packet 106 are shown using a complex vector diagram. That is, each arrow can depict one symbol of the modulation method used to transmit the data packet. The symbols of data packet 106 are shown using a complex vector diagram. That is, each arrow can depict one symbol of the modulation method used to transmit the data packet.

FIG. 2f shows a schematic diagram of a data packet (subpacket) 106 according to one embodiment. Data packet 106 consists of pilot sequence 108 that is separated (or can be separated by receiver 100) into two partial pilot sequences 108_1 and 108_2. The symbols of data packet 106 are shown using a complex vector diagram. That is, each arrow can depict one symbol of the modulation method used to transmit the data packet.

Pilot 108 need not use the same modulation scheme as data part 110, but can. The pilot 108 of each data packet 106 can be split into at least two parts, here eg p1 (108_1) and p2 (108_2). The at least two parts p1 (108_1) and p2 (108_2) need not be separated in time, but can be separated in time. Signals over time p1 (108_1) and p2 (108_2) can be known to the receiver 100. FIG. A signal received at receiver 100 can be affected by channel impairments such as noise. Due to the offsets in the crystals used by the transmitter and receiver 100, the exact time, frequency and phase offsets of the received signal are initially unknown to the receiver 100.

To detect the signals, the receiver 100 can perform cross-correlation of the total signals p1 (108_1) and p2 (108_2) to the received signals. In the presence of frequency offset, this reduces correlation peaks.

EP 2 914 039 A1 proposes to use subpacket versions to reduce these effects, as will become apparent from the discussion of FIG.

In particular, FIG. 3 shows a schematic diagram of the synchronization of data packets 106 according to EP 2 914 039 A1. The received data packet 106 corresponds to the data packet 106 shown in Figure 2b. However, data packets 106 are affected by the frequency offset. This is illustrated by the rotation of the vectors used to describe the symbols of the data packets in FIG.

3, reference sequences (or correlation sequences) 112_1 and 112_2, correlation products 115_1 and 115_2 obtained by correlating reference sequences (or correlation sequences) 112_1 and 112_2 with data packet 106, and all products A correlation result 118 as a sum over , is shown. Hereby the length of the correlation peak is reduced by the frequency offset.

For larger frequency offsets, even the subpacket correlation peaks shown in Fig. 2a can be reduced in a significant way.

<Detection of Combined Partial Preamble Correlation>
In contrast to FIG. 3, the embodiment provides combined partial preamble correlation (cppc) detection. Thereby, a non-coherent combination of at least two received partial preamble parts (rp1, rp2, ...) within a small subpacket can be used.

For example, some embodiments propose non-coherent combining of code-matched filter outputs for CDMA detection. In a long stream of data, multiple matched filter outputs of a single CDMA symbol can be combined.

Further, embodiments propose different methods of non-coherent combination of single-hop correlation results for frequency-hop spread spectrum systems. Single-hop correlation results can be combined.

As will become clear from the discussion below, first, non-coherent combining on the subpacket (or HOP) level can be used (see FIG. 4), and second, the overall result Combining already combined subpacket level results into can be used.

FIG. 4 shows a schematic diagram of synchronization of data packets 106 according to one embodiment. The received data packet 106 corresponds to the data packet 106 shown in Figure 2b. However, data packets 106 are affected by frequency offsets, which is illustrated in FIG. 4 by the rotation of the vectors used to describe the symbols of the data packets.

Further, in FIG. 4, at least two partial reference sequences (or correlation sequences) rp1 (108_1) and rp2 (108_2) and at least two partial reference sequences (or correlation sequences) 108_1 and 108_2 are included in data packets. Summing the individual correlation products cp1(115_1) and cp2(115_2) with the correlation products cp1(115_1) and cp2(115_2) obtained by correlating with 106 (eg, the equation cp1=rp1*conj(p1 ) and cp2=rp2*conj(p2)), and the partial correlation results c1(116_1) and c2(116_2). A coarse correlation result spm (118) for the data packet 106 obtained by non-coherently adding is shown.

In other words, as shown in FIG. 4, a first partial reference sequence p1 (112_1) and a second partial reference sequence p2 (112_2) and a first partial pilot sequence rp1 (108_1) and the second partial pilot sequence rp2 (108_2) are performed respectively. This results in partial correlation results c1 (116_1) and c2 (116_2).

Furthermore, a non-linear operation such as abs(), an approximation of abs(), or any other non-linear operation may be applied to the partial correlation results c1 (116_1) and c2 ( 116_2), or any approximation of an ideal Neyman-Pearson detector. This results in values L1 and L2. Adding the values results in the subpacket preamble metric spm=L1+L2. spm=L1+L2 in the presence of frequency offsets is longer with direct correlation cdirect=abs(c1+c2), even for the subpackets shown in FIG. 2a.

This provides the following advantages. The method is robust to frequency offsets. In the presence of large crystal offsets between transmitter and receiver, fewer subbands have to be searched to find the preamble.

As already mentioned, the data packet 106 can be part of a telegram which is transmitted separated into multiple data packets (or sub-packets).

The receiving unit 102 can be configured to receive at least two data packets 106 . wherein each of the at least two data packets 106 includes a pilot sequence 108 and the at least two data packets 106 are part of a telegram transmitted separated into at least two data packets 106 . Synchronization unit 104 generates partial correlation results c1 (116_1) and c2 (116_2) for each of at least two partial reference sequences p1 (112_1) and p2 (112_2) for each of at least two data packets 106. ), at least two partial reference sequences p1 (112_1) corresponding to the reference sequences for the pilot sequences of the data packets 106 corresponding to the respective pilot sequences 108 of the at least two data packets and It can be configured to correlate with p2 (112_2) separately. Further, the synchronization unit 104 performs partial correlation results c1 (116_1) for each of the at least two data packets 106 to obtain coarse correlation results spm (118) for each of the at least two data packets 106 and c2 (116_2) can be configured to add non-coherently. Further, the synchronization unit 104 can be configured to combine the coarse correlation results spm (118) of at least two data packets 106 to obtain a coarse correlation result for the telegram.

In other words, the subpacket's coarse correlation result spm(118) (which can also be based on only one partial correlation) is combined with the telegram preamble metric (or the coarse correlation result for the telegram) tpm. can be done. Combining can be performed, for example, by a single summation or by other approximations of an ideal Neyman-Pearson detector.

This has the advantage that less computational power is required.

For example, 30 with two partial preambles in each subpacket, such as 15 times subpacket version a) shown in FIG. 2a and 15 times subpacket version b) shown in FIG. subpackets can be used. Using a unique sum at each time step, 60 additions, ie 30 subpackets multiplied by two partial preambles, are required. If, as suggested, two consecutive sums are used, computational power can be reduced. At each time step, the spm over subpacket version a) and one sum over subpacket version b) can be calculated. The resulting spm a) and spm b) can be saved in memory. Then the precomputed spm
The sum over a) and spm b) can be calculated according to the values stored in memory. In that case, only 2 additions for the precomputed ones and 30 additions for the final sum are required.

the synchronization unit 104 for each of the at least two data packets 106 to obtain a fine correlation result for each of the at least two data packets 106 if the coarse correlation result for the telegram exceeds a predetermined threshold; It can be further configured to coherently add the partial correlation results c1 (116_1) and c2 (116_2). Further, the synchronization unit 104 can be configured to combine fine correlation results of at least two data packets to obtain fine correlation results for telegrams.

In other words, a first search (or stage) with non-coherent addition can be combined with a second search (or stage) with coherent addition.

The technique described above with non-coherent addition of at least two synchronous parts of one subpacket can be used. A sum over all subpackets can then be calculated. This value can be compared to a threshold and if the value exceeds the threshold a second correlation can be performed.

A second stage can compute the correlation with coherent summation of all parts inside a subpacket or across all hops of a telegram. This is done as a hypothesis test for a number of different hypothetical frequency offsets. Values resulting from coherent summation of semi-correlation results are also compared against a threshold. If the value is within the detection range, the beginning of the packet is detected. The first stage (non-coherent summation) produces a coarse frequency offset, which is required for the second stage. The second stage provides a more accurate frequency offset, which can be used for subsequent decoders.

This technique requires two stages of detection. The second correlation is much more frequency sensitive than the first, thus resulting in more calculations for the different frequency offsets required. To reduce computational power, the second correlation is performed only when the first stage detects a packet. Therefore, the increase in computing power is very low.

This technique provides a fine estimated frequency offset, which is beneficial to the decoder. A decoder saves computational power. because it does not need to calculate the frequency offset again.

<Use of pilots for signaling header information>
In the following, embodiments using pilots for signaling header information are described.

Data packets 106 may include header information encoded in phase shifts of pilot sequences 108 . The receiver 100 extracts header information from the data packets by applying a frequency correction to the data packets using the estimated frequency offset of the data packets 106 and estimating the phase shift of the pilot sequences. A header extraction unit may be provided configured to:

When combined partial preamble correlation (cppc) or other schemes are used, the performance of the preamble detector is either completely insensitive or the transmitted partial pilot sequence p1 (108_1 ) and p2(108_2) in a patient manner to phase rotations.

The transmitter can add an arbitrary phase shift phi in the range [-pi, pi] to the partial pilot sequences p1 (108_1) and p2 (108_2).

Other shift schemes have been proposed.
- p1'=p1*exp(2*pi*phi), p2'=p2*exp(-2*pi*phi), ie p1 and p2 are shifted in opposite directions.
- p1'=p1, p2'=p2*exp(-2*pi*phi), ie only p2 is shifted.
- p1'=p1*exp(2*pi*phi), p2'=p2, ie only p1 is shifted.
where p1' is the phase-shifted version of p1 and p2' is the phase of the phase-shifted version of p2.

Furthermore, combinations of the described schemes are possible. All differential phase modulation schemes can be used. Phase-shifting all or a subset of the partial pilot sequences/subpackets by encoding the header bits to be transmitted with Forward Error Correction (FEC) resulting in transmitter code symbols c can be calculated for Golay Codes, BCH Codes, Convolutional codes or Turbo Codes or LDPC Codes or other codes can be used. A code symbol may be mapped to a phase shift phi_i with index i for partial pilot sequence/subpacket i.

(Generation of phase offset from p1 to p2 when preamble is MSK/GMSK modulated)
Next, generation of phase offsets for partial pilot sequences p1 (108_1) and p2 (108_2) of MSK (MSK = minimum shift keying) or GMSK (GMSK = Gaussian filtered minimum shift keying) modulated preamble 108 is described. be.

If the system uses MSK or GMSK modulation for packets, the transmitter can be easily employed to introduce a phase offset for the partial pilot sequences p1 (108_1) or p2 (108_2). Further, we concentrate on p2.

If differential MSK/GMSK is used, the first bit of p2 can be inverted and/or the first symbol of the data portion after p2, if present, is inverted. can be

If pre-encoded MSK/GMSK is used, all symbols of p2 can be inverted.

(decoding received phase shift)
Receiver 100 (or synchronization unit 104) may be configured by: a. :
1. 2. Performing a rough estimation of the frequency offset f_r of the received signal by inspection of the partial preamble (eg, the phase difference of the received symbols at cp1 and cp2 can be analyzed); 3. Perform approximate frequency corrections rp1'=rp1*exp(-2*p*f_r) and rp2'=rp2*exp(-2*p*f_r); By estimating the phase offset phi' between rp1' and rp2' (e.g., calculating phi'=arg(c1*conj(c2)), the approximate frequency correction is sufficient and phi' is Note that the designs of p1 and p2 can be preformed so that they can be estimated without phase ambiguity in most cases)
4. 5. Computing the logarithmic likelihood IIr_i or a simple estimate of the transmitted phi_i; Decoding the header bit vector h_e transmitted from within IIr_i by the channel decoder

(removal of transmitted phase offset in preamble)
When vector h_e is decoded at the receiver, it can be encoded again. This gives a list of phase offsets phi_e_i.

This phase offset phi_e_i can be used to remove the phase shift of the received partial preamble (here rp2) in the received signal. Accordingly, the decoder can continue to decode the received subpackets in a manner similar to that without sending header information.

<Interference Robust Detection>
The transmission is usually in an unlicensed band (eg ISM (ISM=industrial, scientific and medical) band) and/or the sensor nodes are not synchronized with the base station. Therefore, interference with other systems using the same time slot will occur. Interference with other sensor nodes also occurs when the system is not synchronized with the base station.

This interference negatively impacts detection performance at the receiver. On the one hand it can reduce the correlation result of the correlation of the main lobe, on the other hand it increases the unwanted side lobes. These side lobes are shown in FIG. 5 for a Barker code with a length of seven. The side lobes are peaks that are not in the middle of the autocorrelation function and are not equal to zero.

To avoid false positives on the side lobes, the threshold should be greater than the highest side lobe.

In the autocorrelation function 13 the values are calculated. As a result, one time slot equals one symbol time. It is also possible to use more time slots (e.g. 1 time slot equal to 1/2 symbol time) or less time slots (1 time slot equal to 2 symbol times). is.

If there is an interferer in the air with a strong capability at the receiver, the correlation result will be very high most of the time in this time slot and false positives can occur. This is illustrated in FIG. Interferors multiply the correlation result and produce "interference peaks." As a result, the value exceeds the predetermined threshold, leading to false positives.

Several techniques exist to reduce the number of false positives in the case of interferers and/or for non-ideal correlated sequences. and they are described below. They can be used standalone or they can be used in combination to achieve better results.

(Normalization)
Distortion of the transmitted symbols can occur if an interferer occurs in the used band of the desired signal. Distortion in this case is any phase and amplitude offset on each symbol during the transmission time of the interferer.

Normalization is performed to reduce the effects of such interferers. This non-linear operation is equal to the capacity over one subpacket, telegram or each transmitted symbol.

In other words, for subpacket-wise normalization, for example, the average capacity over the length of one subpacket is calculated. This calculation is done separately for each time slot. Pmean[m]=sum(Pin)/N (Pin is the power of the symbol inside the subpacket length. N is the length of one subpacket in symbols. m is each is an index for the time slot.)

This value applies to all symbols of one subpacket length inside the time slot it follows. For example, the receive power of each symbol is divided by the average power of one subpacket (Pout[k]=Pin[k]/Pmean[m], where k=number of symbols inside the subpacket length).

FIG. 7a shows a schematic diagram of a data packet 106 with two partial pilot sequences 108_1 and 108_2 and a data sequence 110. FIG. Here, data packet 106 is overlaid (or superimposed) by long interferer 130 . FIG. 7b shows the receive power plotted over time and the normalized receive power for each of the three time slots of FIG. 7a.

In particular, Figures 7a and 7b show an example of this technique with three different time slots. The length of one subpacket in symbols for each time slot is cut. The second time slot shows perfection. Here all symbols of the subpacket are inside the cut area. The first and last are too early or too late.

For all three time slots, it is assumed that the interferer is active the whole time and the power of the interferer is much higher than the symbol power. The receive power (sum of signal plus interferer sum in the band used) is shown as line 132 in all three cases in Fig. 7b.

After cutting, the average capacity with the above equation is calculated for each time slot. In each time slot, each symbol is separated by this average power value described by the above equation. Therefore, the average capacity in each time slot is now equal to one. The average power after normalization is also equal to 1 in the absence of interferers during transmission over the air. The impact of fully interfered subpackets is now the same as without interferers. The normalized receive power is shown as line 134 in all three cases of Figure 7b.

For the calculation of the normalization value, it is also possible to cut longer than one subpacket length, for example two subpacket lengths. Also in this case, the length before and after the subpacket is cut by half. The longer the length used in calculating the normalization value, the better the results for short interferers. A short coherent enhances only a subset of the symbols inside its area and is used in the computation. The impact of these symbols is very low if only a small subset of the symbols are interfered.

This method works well when the interferer period is much larger than the period of one subpacket. This normalization produces unwanted results if the period is within the same domain as the sub-packet period or shorter than the sub-packet period. This problem is illustrated with examples in FIGS. 8a and 8b.

FIG. 8a shows a schematic diagram of a data packet 106 with two partial pilot sequences 108_1 and 108_2 and a data sequence 110. FIG. Data packet 106 is overlaid (or superimposed) by short interferer 130 . FIG. 8b shows the receive power plotted over time and the normalized receive power for each of the three time slots of FIG. 8a. The receive power (sum of signal plus interferer sum in the band used) is shown as line 132 for all three cases in Fig. 8b. The normalized receive power is shown as line 134 for all three cases in Figure 8b.

As shown in Figures 8a and 8b, the interferer 130 is only for a fraction of the active subpacket period and therefore not all symbols have the same reception capability. do not have.

A normalization factor is calculated over all symbols in this time slot. This factor is then applied to all symbols inside the subpacket length. Therefore, the interfered symbols have much higher amplitude after normalization.

In the first time slot, the interferers are only for a small subset of the active symbols, and the impact of interferer power on the normalization factor is very small. In both other cases, the influence of interferers is higher. Normalization reduces all symbols in this time slot to have an average power distribution of unity inside this time slot. Non-interfered symbols are also reduced as interfered symbols. The symbols that are correct in correlation are then suppressed below the interfered symbols. The output after normalization is shown by line 134 in FIG. 8b. False positives can occur when the interfered symbols have more influence on the correlation result.

If the interferer length is unknown or not much longer than the duration over a subpacket, symbol-wise normalization can be performed to solve the aforementioned problems.

Symbol-wise normalization works in the same way as subpacket-wise normalization, except for the normalization factor. These are computed separately for each symbol inside the subpacket length, not just for the entire subpacket length. Figures 9a and 9b illustrate this technique.

FIG. 9a shows a schematic diagram of a data packet 106 with two partial pilot sequences 108_1 and 108_2 and a data sequence 110. FIG. Data packet 106 is overlaid (or superimposed) by short interferer 130 . FIG. 9b shows the receive power plotted over time and the normalized receive power for each of the three time slots of FIG. 9a. The receive power (sum of signal plus interferer sum in the band used) is shown as line 132 for all three cases in Fig. 9b. The normalized receive power is shown as line 134 for all three cases in Figure 9b.

Each symbol is normalized to the same power, eg, by dividing by its own symbol power. Furthermore, the output of the correlation depends only on the received phase of the synchronization sequence.

In correlation, all symbols are suppressed equally and the effect of interfered symbols is smaller than without normalization.

As an alternative to normalizing before correlation, normalizing the correlation product is also possible.

A correlation product can be derived using cp1=rp1*conj(p1) at each time slot. rp1 is the received synchronization (or pilot) sequence, p1 is the known ideal synchronization (or pilot) sequence, and cp1 is the correlation result. This technique can be done with one correlation over the entire sequence, or it can be done with quasi-correlation as described above.

However, strong noise impulses can also lead to high levels of cp1, so the output signal cp1 cannot provide unambiguous information whether the signal is present or not. One possibility is therefore to do a normalization of the output signal by norm1=abs(rp1)*abs(p1).

The normalized output is given by cp1norm=cp1/norm1. The value of cp1norm can take values from 0 to 1 if the pilot sequence 108 has a certain power (assumed below). A value of 1 indicates perfect correlation. For signals that do not contain signal p1, the absolute value of cp1 is always smaller than norm1.

Alternatively, norm1 can be calculated as norm1=abs(rp1)*c. where c is a constant that can be adjusted so that cp1norm reaches a maximum value of one.

Alternatively, norm1 can be calculated as norm1=sqrt(abs(rp1̂2))*2, or norm1=sqrt(abs(rp1)*abs(p1))̂2).

A normalization of the input symbols can be performed. Normalization is a non-linear technique, eg absolute value or power is used. Different techniques exist depending on the following interference scenarios.
- subpacket-wise normalization - telegram-wise normalization - symbol-wise normalization

This has the advantage of reducing the influence of interferers on the correlation result. Therefore, the number of false positives is reduced. If a false positive occurs, the decoder tries to encode the packet, but the CRC (CRC = cyclic redundancy check) fails. If the number of false positives is reduced, the CPU time used is reduced and other applications can use the CPU time, or the power consumption of the device is lowered.

(dispersion)
As described above, packet detection computes correlations for all sync sequences, and summing the absolute values of all quasi-correlations yields the output. If only one sequence is used, the sequence can be split into sub-parts as described above. A new packet is detected if the correlation value is above a predetermined threshold. This technique works well when there are no interferers in the channel.

Another technique is based on the distribution of subpacket correlations. The variance for discrete finite lengths can be computed by var=1/n*sum((xi−μ) ² ). The average value can be calculated by μ=1/n*sum(xi). where n is the number of sub-correlations used, μ is the previously calculated mean, and xi is the correlation result of sub-correlation i.

The partial correlation results are normalized to the reception power and to the length of the correlation portion. Therefore, the correlation result of one semi-correlation is between 0 and 1.

When no noise or interferers are applied to the signal, the correlation of each sub-correlation at perfect time slots yields the same value, and no variance between the correlation results of the sub-correlations is observed. The optimal time slot lies in the middle of the autocorrelation function and the peak has the highest value. There is high variance caused by unknown data in other time slots.

The variance calculation is shown as an example for subpacket-wise correlation in the following figures.

In the presence of noise on the channel, the variance in perfect time slots increases with decreasing SNR. The maximum variance can be reached at the lowest possible SNR, where the packets can be corrected decoded. This value can be used as a threshold. A packet is detected if the calculated variance is below this threshold.

This threshold can be used standalone for packet detection, or used in conjunction with standard detection as a second stage for determining whether the first stage detection was erroneous. be able to.

FIG. 10 illustrates a plurality of data packets (or subpackets or hops) 106 that are part of a telegram that is separated and transmitted in a plurality of data packets 106 over a communication channel and all (or at least some ) and a schematic diagram of the calculation of the variance over a data packet (subpacket) 106. FIG. In FIG. 10, the ordinate indicates frequency and the abscissa indicates time.

This algorithm can also be used to detect side lobes in the correlation that do not necessarily arise from interference. For example, they can be generated by non-ideal correlation sequences.

As an example for detection in two stages, first the correlation can be calculated with normalized symbols. When the first stage detects a packet, the correlation results of all sub-correlations in the detected time slot can be used to calculate the variance. If this variance is less than a threshold, packet detection is activated.

Typically, the first stage threshold can be chosen below the sidelobe peak. If a value above the threshold is detected, the variance can be calculated. A new packet can only be detected if both values are within the detection range.

The correlation for the entire packet can be split into sub-correlations. These sub-correlations can also be used when only one correlation sequence is in the entire packet. In this case, the preamble can be split for quasi-correlation. Across all semi-correlations, the variance can be calculated and compared to a threshold.

An advantage of this technique is that the number of falsely detected packets is reduced in the case of interference. Furthermore, the threshold can be reduced, which results in a better detection rate for low SNR (SNR=signal-to-noise ratio).

(weighted sync symbol)
Additionally, the preamble symbols (or pilot symbols) can be weighted before correlation. Three different techniques exist:
- a weight factor for all sync symbols - a weight factor for each subpacket 106 - a weight factor for each preamble part

Weighting can also be done after correlation over one subpacket or over part of a correlation sequence. A partial correlation is therefore performed and then multiplied by a weighting factor.

As an example, the weight factor can be calculated by the variance over the assumed sync symbol in the time slot. Alternatively, the weight factors can be derived from the power variance of all symbols inside the time slot or based on a determined signal-to-noise ratio.

Weight factors may be applied to the sync symbols before correlation is performed. Interfered sync symbols have a lower weight factor, so that these symbols have less influence on the correlation result.

FIG. 11 shows that each data packet 106 has two partial pilot sequences 108_1 and 108_2, and a separate way for each of the partial pilot sequences 108_1 and 108_2 for each data packet 106. The weighting of pilot sequences 108_1 and 108_2 performed by applying a weighting factor is shown in the schematic diagram of three data packets 106. FIG.

In other words, FIG. 11 illustrates this concept of preamble part-wise weighting. Factors are multiplied after summing and non-linear operations over the preamble part. If weighting is done before correlation, the values in the figure are multiplied by a factor before the absolute value calculation is done.

If only one correlation sequence is used, this sequence can be split into sub-sequences. Therefore every sub-sequence gets a unique weight factor.

A sync symbol can be multiplied by a weight factor. It is also possible that only the preamble part part is weighted instead of symbol-wise weighting. Weighting factors can be applied before correlation or after semi-correlation.

This has the advantage that the number of false positives in interfered channels can be reduced. Therefore, the power consumption of the receiver can be reduced.

(side lobe detection)
Side lobes occur in the correlation output due to non-ideal correlation sequences. These side lobes are deterministic and at a constant offset from the main lobe. The receiver can compute these positions if the correlation sequence is known (almost always known at the receiver).

This is illustrated in the following figure, where the main lobe and two side lobes are shown. These side lobes have lower peaks than the main lobe. To avoid false positives, the threshold is set higher than the largest side lobe peak.

False positives occur when the threshold is set below the highest side lobe peak. To avoid this false detection, the receiver searches at known side lobe temporal distances when higher peaks occur. If YES, side lobes have been detected, otherwise the receiver has already found the main lobe.

These side lobes can also produce different frequency offsets. The receiver obtains the side lobes by performing an autocorrelation function at different frequency offsets.

FIG. 12 shows the amplitude of the correlation output plotted over time. In other words, FIG. 12 shows a typical correlation output. Time is plotted on the horizontal axis and correlation output is shown on the vertical axis. The main lobe 136, two side lobes 138 and the noise floor 140 are shown in FIG.

The additional computational power is very low since the side lobe correlation values can be computed earlier in the correlation and stored in history (or in memory).

Side lobe detection can be performed. The correlation value at the side lobe distance is compared to the actual correlation value if a value above the threshold is found. A side lobe 138 is detected if the value in the side lobe distance is higher. Otherwise, the main lobe 136 exists within the actual time slot.

This has the advantage that the threshold for detection can be set below the highest peak of side lobe 138 . Hereby an improved detection rate can be achieved even for low signal-to-noise ratios. Comparing at the same threshold without side lobe detection, the number of false positives can be reduced.

(detection window)
As shown in FIG. 12, ideal correlation around main lobe 136 does not exist. This is caused by non-ideal correlation sequences, by splitting the correlation part into data parts, and by interference. Therefore, the threshold should be set over the highest value excluding the main lobe 136 to avoid false positives. Main lobe 136 produces poor detection performance in noisy channels. As the SNR decreases, the value of the correlation result becomes lower. Packet detection is assumed only if the correlation value is above a predetermined threshold.

To get better performance against noise, a detection window can be introduced. This window typically has the size of the area before and after the main lobe 136 . Instead of directly driving a new packet detection, the highest peak inside the window is searched when a value above the threshold is detected. The packet detection output can be blocked until the index of the highest peak inside the detection window gets a predetermined value (detection index). Packet detection can be activated if the correlation value is above the threshold and the index is correct at the predetermined value.

FIG. 13 shows such a detection window. In this example, the detection window has 11 elements. The detection index is set in the middle of the window.

FIG. 14 shows a flowchart of a method 160 for detecting data packets using a detection window, according to one embodiment. In a first step 162 the time slot (index) can be incremented. In a second step 164 correlations can be calculated for the actual time slots. In a third step 166 the (correlation) results can be inserted into the detection window. In a fourth step 168, the maximum value within the detection window can be determined. In a fifth step 170, it can be determined whether the maximum value is greater than the threshold. If the maximum value is not greater than the threshold, first step 162 through fifth step 170 are repeated. If the maximum value is greater than the threshold, then in a sixth step 172 the index of the maximum value is determined. In a seventh step 174 it is determined whether the index is equal to the detection index. If the index is not equal to the detection index, first step 162 through seventh step 174 are repeated. If the index is equal to the detection index, then in an eighth step 176 the new packet is detected.

In other words, FIG. 14 shows an overview of how detection works. Before detection begins, a window is created and set to initial values (eg all values are 0). Continuous detection is then started.

In a first step 162, the time slot index is updated. Afterwards (164) the correlation in the actual time slot is performed. For this correlation, the techniques described above can be used, or any other technique will work as well. Correlation results are stored in the detection window at the latest time index 166 . Therefore, the oldest is removed from the array (shifting all values one place to the right and inserting the new value on the left).

At step 168, the largest peak within this window is searched. At step 170 , if the maximum peak within the window is below the threshold, processing returns to the first step 162 . Otherwise, in step 172 the index of the maximum value is extracted and in step 174 it is compared with the detection index. If both values are the same, then at step 176 a new packet is detected.

FIG. 15 illustrates the detection window 172 and threshold 171 used to detect data packets and the amplitude of the correlation output 170 plotted over time for three different time slots, according to one embodiment. Illustrated.

In other words, Figure 15 shows this method in three different time slots. In the first part, values above the threshold can be detected, not at the detection index. False detections occur when packet detection is performed in this slot.

In the detection window 172 the highest value is obtained. It is now verified whether the highest value inside this window 172 is above the threshold.

This is the case for the first time slot in FIG. However, the highest valued index must be exactly the detection index. The detection index does not apply in the first case. Since the index is greater than the detection index, this peak lies within a few steps in the detection index. If this peak is there, it must be the highest value inside the window that drives packet detection. Other correlation values are added to the window until closer to the detection index. In this example, the maximum index is not equal to the detection index because it has a higher correlation value.

In the second case, the highest value is the correct detection index, the value is above the threshold and packet detection is assumed.

In the last case, the maximum index falls below the middle of the window.

If the maximum index is higher than the detection index, the time slot will be detected too early and too late for detection. If the value is lower than the detection index, packet detection was already driven earlier.

A detection window 172 can be introduced. Packet detection is not immediately triggered when a value above threshold 171 is detected. Alternatively, packet detection can be blocked until the maximum index inside the detection window 171 reaches a predetermined detection index.

This has the advantage that the threshold can be set low, resulting in better detection rate at low SNR with less false detection rate.

(partial correlation)
Instead of computing correlations over all sub-correlations, correlations can be performed over only a portion of all correlation sequences. This technique also works if only one correlation sequence is used. In this case, the correlation sequence can be divided into sub-parts as described above.

FIG. 16 shows multiple data packets (or subpackets or hops) 106 and three data packets (subpackets) 106 that are part of a telegram that is separated and transmitted in multiple data packets 106 over a communication channel. ) and a schematic diagram of the partial correlation over 106. FIG. In FIG. 16, the ordinate indicates frequency and the abscissa indicates time.

In other words, FIG. 16 gives an example of this technique with subpacket-wise correlation. Instead of computing correlations over all subpackets, correlations are performed over only three subpackets. Summing the subsets then yields the correlation output.

A threshold can be fitted to a smaller number of semi-correlations.

Unfortunately, the minimized correlation sequence has a higher probability of false positives caused by interferers or by noise. In order to obtain improved (or best) performance, a two stage decision is used. In a first stage, correlation can be performed over a subset of the correlation sequences. If a packet is detected in the first stage, then in the second stage correlation can be performed over all correlation portions. Packet detection can be triggered only if the second correlation is also above the threshold.

The first stage correlation output can be used for the overall correlation calculation. Therefore the correlation over the remaining correlation sequence is calculated and added to the result of the first stage.

Correlations can only be computed over a subset of the synchronization sequences. If packets are detected by this method, a second correlation over all sequences can be performed.

This has the advantage that the power consumption of the receiver can be reduced, since the algorithm must not compute the correlation of all parts. The overall correlation is calculated only if the semi-correlation detects a packet.

(Method)
FIG. 17 shows a flowchart of a method 200 of receiving data packets. The method includes step 202 of receiving a data packet containing a pilot sequence; and step 206 of non-coherently adding the partial correlation results to obtain a correlation result for the data packet. include.

Although some aspects are described in the context of an apparatus, it should be clear that these aspects also represent a description of the corresponding method. Here, blocks or devices correspond to method steps or features of method steps. Similarly, aspects described in the context of method steps also represent descriptions of corresponding blocks or items or corresponding apparatus features. Some or all method steps may be performed by (using) hardware devices such as microprocessors, programmable computers or electronic circuits. In some embodiments, one or more of the most important methods can be performed by such devices.

(further embodiment)
FIG. 18 shows a schematic block diagram of receiver 100 according to one embodiment. Receiver 100 comprises a receiving unit 102 and a synchronization unit 104 . The receiving unit 102 is configured to receive data packets 106 (eg, at least two data packets), wherein at least two of the data packets 106 (eg, each of the at least two data packets) ) includes partial pilot sequences of at least two partial pilot sequences (that the receiver may receive additional data packets that do not have partial pilot sequences, Please note).

For example, the receiving unit 102 is configured to receive and demodulate a signal transmitted from the transmitter to the receiver 100 over a communication channel and provide a data stream comprising at least two data packets 106 thereon. be able to.

a first data packet 106 of the at least two data packets 106 can include a first partial pilot sequence 108_1 of the at least two partial pilot sequences 108_1-108_n; A second data packet 106 may include a second partial pilot sequence 108_2 of the at least two partial pilot sequences 108_1-108. Additionally, the at least two data packets 106 may include one or more data sequences 110 positioned before or after the partial pilot sequences 108_1 and 108_2.

Synchronization unit 104 references at least two partial pilot sequences 108_1-108_n to obtain partial correlation results 116_1-116_n for each of the at least two partial reference sequences 112_1-112_n. It is configured to be individually correlated with the sequences 112_1-112_n. Here, the synchronization unit 104 is configured to non-coherently add the partial correlation results 112_1-112_n to obtain a coarse correlation result 118 for the two data packets 106. FIG.

For example, synchronization unit 104 combines partial pilot sequence 108_1 of first data packet 106 with a first partial and a partial correlation result 116_2 for the second partial reference sequence 112_2. pilot sequence 108_2 can be configured to correlate with the second partial reference sequence 112_2.

Synchronization unit 104 non-coherently converts partial correlation results 116_1-116_n by adding absolute values, squared absolute values, approximate absolute values, or any other non-linear operation of partial correlation results 116_1-116_n. can be configured to add to

The at least two partial reference sequences 112_1-112_n can be at least two different portions of the reference sequence 114. FIG. Here, the at least two partial pilot sequences 108_1-108_n can be at least two different portions of the pilot sequence 108.

Therefore, in comparison to the embodiments of receiver 100 described with respect to FIGS. , at least two data packets), at least two of the data packets 106 including the partial pilot sequences of the at least two partial pilot sequences (e.g., each of the at least two data packets) , is received. However, the functionality of synchronization unit 104 is substantially the same. That is, the partial correlation results 112_1-112_n can be non-coherently added to obtain the coarse correlation result 118. FIG. If at least two further data packets are received, in the same way the partial correlation results of the at least two further data packets obtain coarse correlation results for the at least two further data packets. can be added non-coherently for Further, the coarse correlation results for the at least two data packets and the at least two further data packets can be combined to obtain a combined coarse correlation result.

It is clear that the description of the receiver shown and described with respect to Figures 1 to 16 can also be applied to the receiver shown in Figure 18 and vice versa.

FIG. 19 shows a flowchart of a method 210 of receiving. Method 210 is a data packet method wherein at least two of the data packets (eg, each of the at least two data packets) include a partial pilot sequence of the at least two partial pilot sequences. (e.g., at least two data packets); and step 216 of non-coherently adding the partial correlation results to obtain a coarse correlation result for the two data packets.

(general)
Embodiments can be used in systems for transmitting small amounts of data, eg sensor data, from a large number of nodes, eg heating, electricity or water meters, to a known base station. A base station receives (and possibly controls) many nodes. At the base station, receivers with more computational power and more complex hardware, ie higher performance, are available. Only cheap crystals are available at the node. And cheap crystals typically have frequency offsets of 10 ppm or more. However, embodiments can also be applied to other application scenarios.

Embodiments provide multiple optimized preamble (or pilot sequence) partitions. Multiple optimized preamble (or pilot sequence) splitting improves interferor robustness.

Embodiments provide a correlation method. The correlation method is robust to frequency offsets. Thereby partial correlations are used and then non-coherently added. Non-coherent addition of partial correlations can be used to transmit additional information in the preamble, eg length information.

Embodiments provide several methods by which packet detection can be performed with good performance even when the communication channel is jammed. Some of these methods allow additional gain in terms of noise.

Depending on the requirements of a particular implementation, embodiments of the invention can be implemented in hardware or in software. Embodiments may be implemented using digital storage media such as floppy disks, DVDs, Blu-rays, CDs, ROMs, PROMs, EPROMs, EEPROMs or flash memory. The digital storage medium, in turn, has electronically readable control signals stored thereon for cooperating (or capable of cooperating) with the programmable computer system such that the respective method is performed. is). As such, the digital storage medium is computer readable.

Some embodiments according to the invention comprise a data carrier having electronically readable control signals, the data carrier being programmable such that one of the methods described herein is performed. It is capable of cooperating with a computer system.

Generally, embodiments of the invention can be implemented as a computer program product having program code. The program code operates to perform one of the methods when the computer program product runs on the computer. Program code may be stored, for example, on a machine-readable carrier.

Another embodiment comprises a computer program for performing one of the methods described herein. A computer program is stored on a machine-readable carrier.

In other words, an embodiment of the method of the present invention is therefore a computer program having program code for performing one of the methods described herein when the computer program runs on a computer. is.

A further embodiment of the method of the invention therefore comprises a data carrier (or digital storage medium) comprising a computer program recorded thereon for carrying out one of the methods described herein. media or computer readable media). Data carriers, digital storage media or recorded media are typically tangible and/or non-transitional.

A further embodiment of the method of the invention is therefore a sequence of signals or a data stream representing a computer program for performing one of the methods described herein. A data stream or sequence of signals, for example, may be arranged to be transferred over a data communication connection, such as the Internet.

Further embodiments comprise processing means, eg a computer or a programmable logic device. The processing means is configured or adapted to carry out one of the methods described herein.

A further embodiment comprises a computer having installed therein a computer program for performing one of the methods described herein.

A further embodiment according to the present invention relates to a device or device configured to transfer (e.g. electronically or optically) to a receiver a computer program for performing one of the methods described herein. Have a system. A receiver can be, for example, a computer, mobile device, memory device, or the like. The device or system may, for example, comprise a file server for transferring computer programs to receivers.

In some embodiments, programmable logic devices (eg, field programmable gate arrays) can be used to perform some or all of the functions of the methods described herein. can. In some embodiments, a field programmable gate array can cooperate with a microprocessor to perform one of the methods described herein. In general, the method is preferably performed by any hardware device.

The devices described herein can be implemented using a hardware device, using a computer, or using a combination of hardware devices and computers.

The apparatus described herein, or any component of the apparatus described herein, may be implemented at least partially in hardware and/or software.

The methods described herein can be performed using hardware devices, using computers, or using a combination of hardware devices and computers.

Any component of the methods described herein or the apparatus described herein can be performed, at least in part, by hardware and/or software.

The above-described embodiments are merely illustrative of the principles of the invention. It is understood that modifications and variations of the sequences and details described herein will be apparent to those skilled in the art. It is therefore intended that the invention be limited only by the scope of the appended claims and not by any specific details presented in the description and explanation of the embodiments herein.

Claims (19)

a receiving unit (102) configured to receive a data packet (106) comprising a pilot sequence (108);
a synchronization unit (104) configured to correlate said pilot sequence (108) with a reference sequence to obtain a correlation result;
A receiver (100) comprising:
The synchronization unit (104) is configured to use a correlation window to detect the data packet (106), the data packet (106) passing a predetermined threshold within the correlation window. detected by finding the highest correlation peak of all correlation peaks exceeding
the correlation window is divided into a plurality of time slots, each time slot having an index associated with it;
A highest correlation peak inside the correlation window is detected if a correlation peak above the predetermined threshold is detected, and data packet detection detects the index of the highest correlation peak inside the detection window. a predetermined detection index is set in the middle of said detection window , until a detection index of
Receiver (100). The synchronization unit (104) synthesizes the pilot sequence (108) into the at least two reference sequences (112_1-112_n) to obtain partial correlation results (116_1-116_n) for each of the at least two partial reference sequences (112_1-112_n). configured to individually correlate with one partial reference sequence (112_1-112_n);
The synchronization unit (104) is configured to non-coherently add the partial correlation results (116_1-116_n) to obtain a coarse correlation result (118) for the data packet (106). , a receiver (100) according to claim 1. The synchronization unit (104) may add the partial correlation results (116_1-116_n) to the partial correlation results (116_1-116_n) by adding the absolute value, or the square of the absolute value, or the approximate absolute value, or any other non-linear operation. The receiver (100) of claim 2, configured to add the correlation results (116_1-116_n) non-coherently. 2. The at least two partial reference sequences (112_1-112_n) are at least two different parts of a reference sequence (114) for the pilot sequence (108) of the data packet (106). Or a receiver (100) according to claim 3. Receiver (100) according to any of claims 2 to 4, wherein said data packet (106) comprises at least two partial reference sequences (108_1-108_n) as said reference sequences. said receiving unit (102) is configured to receive at least two data packets (106), each of said at least two data packets (106) comprising a pilot sequence (108);
said synchronization unit (104) for obtaining partial correlation results (116_1-116_n) for each of said at least two partial reference sequences for each of said at least two data packets; at least two portions corresponding to said pilot sequence (108) of each of two data packets (106) to a reference sequence (114) for said pilot sequences (108) of said at least two data packets (106) configured to individually correlate with a typical reference sequence (112_1-112_n);
The synchronization unit (104) performs the partial correlation results for each of the at least two data packets (106) to obtain coarse correlation results (118) for each of the at least two data packets. configured to non-coherently add at least a portion of (116_1-116_n);
the synchronization unit (104) is configured to combine at least a portion of the coarse correlation results (118) of the at least two data packets (106) to obtain a combined coarse correlation result; Receiver (100) according to any of claims 2-5. the synchronization unit (104) using an ideal Neyman-Pearson detector sum or approximation of the coarse correlation results (118) of the at least two data packets (106); 7. The receiver (100) of claim 6, configured to combine the coarse correlation results (118) of the at least two data packets (106) by a. Said at least two data packets (106) are part of a telegram transmitted separated into said at least two data packets (106), said receiver (100) obtaining said telegrams by: Receiver (100) according to claim 6 or claim 7, comprising a data packet (106) combining unit adapted to combine said at least two data packets (106). The synchronization unit (104) is further configured to coherently add the partial correlation results (116_1-116_n) to obtain a fine correlation result for the data packet (106), according to claim Receiver (100) according to any of claims 2-8. If the combined coarse correlation exceeds a predetermined threshold, the synchronization unit (104) performs a fine correlation on each of the at least two data packets (106). - further configured to coherently add the partial correlation results (116_1-116_n) for each of the packets (106);
the synchronization unit (104) is configured to combine the fine correlation results of the at least two data packets (106) to obtain a combined fine correlation result;
Receiver (100) according to claim 6. The receiver (100) of any of claims 2-10, wherein the synchronization unit (104) is configured to estimate a frequency offset of the data packets (106). said data packets (106) comprising header information encoded in a phase shift of said pilot sequence (108);
The receiver (100) applies a frequency correction to the data packets (106) using the estimated frequency offset to estimate the phase shift of the pilot sequence (108), the a header extraction unit configured to extract said header information from a data packet (106);
Receiver (100) according to claim 11. The synchronization unit (104) normalizes the coarse correlation results of the at least two partial reference sequences to obtain a combined coarse correlation result, and normalizes the coarse correlation results of the at least two partial reference sequences. A receiver (100) according to any of claims 2 to 12, arranged to combine the normalized coarse correlation results. The synchronization unit (104) normalizes symbols of the pilot sequence to obtain a normalized pilot sequence, and converts the normalized pilot sequence to the at least two partial references. A receiver (100) according to any of claims 2 to 12, adapted to individually correlate with the sequences (112_1-112_n). The synchronization unit (104) calculates the variance of the partial correlation results (116_1-116_n) for the data packets (106) and calculates the partial correlation results for the data packets (106). A receiver (100) according to any of claims 2 to 14, adapted to detect the data packet (106) if the variance of results is less than or equal to a predetermined threshold. The synchronization unit (104) is configured to apply a weight factor to symbols of the data packets (106) or the at least two partial pilot sequences (108_1- 108_n), or individually to each symbol of said at least two partial pilot sequences (108_1-108_n) according to claim 5. 16. A receiver (100) according to any of claims 2 to 15, adapted to apply a weighting factor of . The synchronization unit (104) detects a main lobe and a side lobe of the correlation and uses a known distance between the main lobe and the side lobe to detect as a correlation result. 17. A receiver (100) according to any of claims 2 to 16, adapted to provide the emitted main lobe. receiving a data packet (106) containing a pilot sequence (108);
correlating said pilot sequence (108) with a reference sequence to obtain a correlation result;
a method comprising
A correlation window is used to detect the data packet (106), the data packet (106) having the highest correlation among all correlation peaks exceeding a predetermined threshold within the correlation window. is detected by detecting the peak,
the correlation window is divided into a plurality of time slots, each time slot having an index associated with it;
A highest correlation peak inside the correlation window is detected if a correlation peak above the predetermined threshold is detected, and data packet detection detects the index of the highest correlation peak inside the detection window. wherein the predetermined detection index is set in the middle of the detection window, until a detection index of
Method. A computer program for performing the method of claim 18.

Images