(19)
(11)EP 2 916 506 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
31.07.2019 Bulletin 2019/31

(21)Application number: 14158393.0

(22)Date of filing:  07.03.2014
(51)International Patent Classification (IPC): 
H04L 1/00(2006.01)
H04L 27/26(2006.01)

(54)

Walsh-Hadamard transformed GFDM radio transmission

Walsh-Hadamard-transformierte GFDM-Funkübertragung

Transmission radio GFDM transformée de Walsh-Hadamard


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(43)Date of publication of application:
09.09.2015 Bulletin 2015/37

(73)Proprietor: Vodafone GmbH
40549 Düsseldorf (DE)

(72)Inventors:
  • Mendes, Luciano
    01157 Dresden (DE)
  • Gaspar, Ivan
    01217 Dresden (DE)
  • Matthé, Maximilian
    01127 Dresden (DE)
  • Festag, Andreas
    01277 Dresden (DE)
  • Fettweis, Gerhard
    01326 Dresden (DE)

(74)Representative: KNH Patentanwälte Neumann Heine Taruttis PartG mbB 
Postfach 10 33 63
40024 Düsseldorf
40024 Düsseldorf (DE)


(56)References cited: : 
EP-A1- 2 200 244
EP-A1- 2 608 474
EP-A1- 2 348 655
  
  • BRANDES S ET AL: "Compensation of the Impact of Interference Mitigation by Pulse Blanking in OFDM Systems", GLOBAL TELECOMMUNICATIONS CONFERENCE, 2009. GLOBECOM 2009. IEEE, IEEE, PISCATAWAY, NJ, USA, 30 November 2009 (2009-11-30), pages 1-6, XP031646222, ISBN: 978-1-4244-4148-8
  • BIANCHI T ET AL: "Analysis of the effects of carrier frequency offset on filterbank-based MC-CDMA", GLOBAL TELECOMMUNICATIONS CONFERENCE, 2004. GLOBECOM '04. IEEE DALLAS, TX, USA 29 NOV.-3 DEC., 2004, PISCATAWAY, NJ, USA,IEEE, PISCATAWAY, NJ, USA, vol. 4, 29 November 2004 (2004-11-29), pages 2520-2524, XP010757981, DOI: 10.1109/GLOCOM.2004.1378460 ISBN: 978-0-7803-8794-2
  • RIERA-PALOU F ET AL: "On the design of uplink and downlink group-orthogonal multicarrier wireless systems", IEEE TRANSACTIONS ON COMMUNICATIONS, IEEE SERVICE CENTER, PISCATAWAY, NJ. USA, vol. 56, no. 10, 10 October 2008 (2008-10-10), pages 1656-1665, XP011236197, ISSN: 0090-6778, DOI: 10.1109/TCOMM.2008.060497
  
Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


Description


[0001] The invention relates to a radio transmission system using the generalized frequency divisional multiplexing (GFDM) method. In particular the invention relates to applying a Walsh-Hadamard transformation to a block of GFDM data symbols.

[0002] Current cellular networks are based on orthogonal frequency multiplexing OFDM. EP 2 348 655 A1 discloses a wireless communication system including a transmitter device and a receiver device used for transmission of wireless signals having a plurality of sub-carriers, the transmitter device performs error correction coding on transmitting data and selects at least one of the plurality of sub-carriers used for transmission of error-correction coded data as a null sub-carrier having zero amplitude, thus transmitting the error-correction coded data, whilst the receiver device receives signals having the plurality of sub-carriers from the transmitter device so as to retrieve original transmitted data by implementing error correction decoding on received signals, thus determining whether an interference wave occurs in the sub-carrier upon detecting reception power exceeding a predetermined threshold in the sub-carrier serving as the null sub-carrier.

[0003] Further, in "Compensation of the Impact of Interference Mitigation by Pulse Blanking in OFDM Systems" by Brandes S. et al., IEEE GLOBECOM winter 2009 a OFDM system using pulse blanking is described, wherein for an estimation of the transmitted data symbols and the channel coefficients of each subcarrier an iterative receiver structure is deployed.

[0004] The invention is defined by the appended claims; embodiments not falling under the scope of the claims should be interpreted as examples useful for understanding the invention.

[0005] The so-called GFDM system is a recent physical layer (PHY) scheme proposed to address the challenges for 5th generation cellular systems with opportunistic use of frequency spectrum and relaxed synchronization. The GFDM system provides a fast, flexible scalable and content aware physical layer scheme which at the same time is robust and reliable, while using vacant frequency bands, e.g. so-called TV white spaces. The basic GFDM system has been described in "GFDM - Generalized Frequency Division Multiplexing" by Fettweis et al., IEEE VTC spring 2009 and has been also disclosed in EP 2 200 244 A1.

[0006] Though the GFDM system shares some properties with well-known OFDM systems, there are significant differences. While both systems may divide an available frequency band into at least two adjoining sub-bands, the GFDM system may deploy individual transmission parameters for each of the at least two sub-bands while in OFDM the transmission parameters deployed in the sub-bands are similar or identical, i.e. uniform.

[0007] In contrast thereto the Generalized Frequency Division Multiplexing, GFDM, is a multicarrier scheme that uses pulse-shaping per subcarrier to achieve low out-of-band emissions OOBE, i.e. low out-of-band interference. Typically a prototype filter is used to pulse-shape a number of K subcarriers, wherein the filter impulse response is circularly shifted by K samples to obtain the pulse-shape for each time slot. This technique is known as tail-biting and eliminates the filtering tails, thus allowing a block-based data structure, where a number of M · K data symbols are transmitted using a number of K subcarriers, each subcarrier carrying a number of M data symbols per block.

[0008] The GFDM can be properly parameterized to address scenarios currently foreseen for the fifth generation of mobile communications, i.e. 5G. Single shot transmissions as required for the so-called tactile internet and random channel access as well as machine-to-machine communications and other low latency applications are challenging scenarios. In particular low-latency requirements prevent the use of retransmission protocols. Instead the data shall be reliably transmitted from source to sink using a single transmission burst, i.e. no retransmission, which at the same time avoids the use of long channel codes or interleavers. However, at least one drawback of using short single-shot transmission using GFDM occurs in frequency selective channels. Deep notches in the frequency response of a channel may lead to unrecoverable burst errors at the receiver side.

[0009] The described method and system at least partially solves the above described problem in GFDM transmission systems.

[0010] The proposed system will be understood more clearly when considering the figures wherein
Fig.1
depicts a block diagram of a Walsh-Hadamard-GFDM transmitter;
Fig. 2
depicts a block diagram of a Walsh-Hadamard-GFDM receiver;
Figs. 3a,b
illustrate simulation curves of coded GFDM for comparison to other encoding schemes.


[0011] The generalized frequency divisional multiplexing system defines a system architecture allowing opportunistically exploiting vacant frequency bands, for example TV white spaces, for wireless data communications. The system defines a concept for a physical layer and transmitters and receivers adapted and configured for using said physical layer concept as described below. Due to the opportunistic use of vacant frequency bands, i.e. bands that are not allocated for the GFDM system but currently are vacant, the signal generation in the GFDM system should ensure ultra-low out of band radiation to strictly avoid harmful interference to signals in adjoining frequency bands. Furthermore the receivers should exhibit high sensitivity in order to explore vacant frequency bands, i.e. to sense even very weak radio signals in the frequency bands that are candidates for use.

[0012] Vacant frequency bands for use may be detected by any arbitrary method or device. The transmitter and/or the receiver of the GFDM system may cooperatively detect vacant frequency ranges by scanning frequencies for radio signals. In that case the scanned frequency obviously is in use and consequently cannot be used by GFDM system. If at least one of the transmitter and the receiver detects a vacant frequency band, the detecting station may inform its communication partner via another established communication link about said detected vacant frequency band. The partner station may then sense that band in order to ensure that there is no signal detectable. If both communication partner stations consider the frequency band as vacant said band may be used subsequently by the communication partners. In this way transmitter and receiver may try to detect and agree on using vacant frequency bands in addition to the established communication link.

[0013] The described GFDM system may use a plurality of at least two frequency bands, wherein the carrier frequencies may have different bandwidths. In each of the two frequency bands a respective carrier frequency is individually modulated, i.e. the applied modulation scheme may vary from frequency band to frequency band, wherein some form of QAM modulation may be used.

[0014] Note that in one embodiment the frequency bands may be divided by a frequency band occupied by another transmission, i.e. at least two frequency bands used by the described system are not adjacent, but separated by a frequency band occupied by another transmission. So in contrast to OFDM the described GFDM system does not rely on adjacent subcarriers.

[0015] Each carrier is divided into time slots of equal duration, i.e. within each frequency band time slots of equal duration are defined. However, time slots of different frequency bands may differ in duration, i.e. symbol duration used for transmission in a first frequency band may differ from symbol duration used in a second frequency band. Consequently the number of time slots may vary between frequency bands. In frequency bands of larger bandwidth a higher number of time slots, i.e. shorter in duration, can be used within a transmit block than in frequency bands of smaller bandwidth. As a consequence a higher number of time slots are possible and used in a frequency band of larger bandwidth while achieving the same block size when combining said frequency band with a band of smaller bandwidth. Note that below described application of a transformation matrix to data symbols requires a common number of time slots, i.e. a common number of symbols. Consequently, below described transformation may be applied to data symbols in subcarriers having a common number of data symbols, i.e. time slots. However, a transmit symbol may comprise two groups of subcarriers, wherein each group of subcarriers exhibits a common number of time slots/data symbols, but wherein the groups exhibit different numbers of subcarriers, and wherein the groups require individual transform and modulation matrices.

[0016] Figure 1 depicts a block diagram of a GFDM transmitter 100 adapted for transmitting data bits through a channel according to the invention. A corresponding receiver 200 coupled to the transmission channel is depicted in figure 2. The system schematically depicts a GFDM transmission system adapted and configured for implementing the invention, wherein figures provide an overview of the processing.

[0017] The processing blocks of transmitter 100 may be all digital except for a digital-to-analog converter that converts the transmit signal from a digital representation to an analog signal prior to transmitting the signal by a transmit antenna and except for processing blocks arranged in the chain of processing blocks behind that converter. Similarly the receiver may comprise digital processing blocks only except for an analog-to-digital converter at the radio front end of the receiver and processing blocks arranged in the processing chain before said analog-to-digital converter such as a low-noise amplifier.

[0018] Note that in figure 1 the number of time slots per GFDM symbol is denoted by M, wherein an individual time slot is indexed by m with 0 ≤ mM - 1. Furthermore one GFDM symbol is transmitted using a number of K subcarriers, wherein an individual subcarrier is indexed by 0 ≤ kK - 1.

[0019] In one embodiment a data source may provide a stream of data bits. The data bits are multiplexed by a serial-to-parallel converter 110 to a plurality of parallel substreams. The data bits are organized in M · K sets, wherein each set is mapped by a mapper 120 to a complex data symbol dk,m, wherein the mapping may be conventional. The mapping may be a conventional J-QAM mapping, wherein J is the order of the mapping. Thus data bits are mapped to a number of M · K data symbols dk,m. Note that in the following a group of data symbols assigned to a number of K subcarriers in one time slot m may be termed a subsymbol, thus there are K data symbols in the m-th subsymbol

[0020] Each subsymbol consists of a number of NK samples, wherein a sample is a complex value. Consequently, the mapping of sets of data bits to data symbols leads to a block structure of M · N samples.

[0021] Then in block 130 a transformation is applied all K data symbols of a subsymbol, i.e. all data symbols of an m-th time slot, that spreads the information of a single data symbol to the data symbols of all subcarriers of said m-th time slot. In particular said transformation is applied to all subsymbols of the transmit symbol. As a consequence the information of a single data symbol is spread to the data symbols of all subcarriers of the time slot corresponding to the one subsymbol. Since the transformation is applied to all data symbols of a time slot, each subcarrier carries at least a portion of the information of each other data symbol in said time slot.

[0022] Said spreading of information can be effected by applying a matrix transformation using a spreading matrix Ω for which the condition Ω×ΩH= I holds, wherein I is the identity matrix.

[0023] In one embodiment a Walsh-Hadamard matrix can be used as spreading matrix in block 130, wherein the Walsh-Hadamard transformation, WHT, requires K to be a power of two. As a consequence the Walsh-Hadamard transformation matrix is a square matrix with dimension K × K wherein K is a power of 2. In an alternative embodiment a Cazac matrix can be used as spreading matrix.

[0024] In the next processing step the transformed samples are upsampled in block 140.

[0025] The up-sampled samples, i.e. the block of samples, are pulse shaped by gk,m[n] by a pulse-shaping filter block 150. Note that for each subcarrier an individual pulse shaping filter may be provided, i.e. each of the K subcarriers may undergo an individual pulse shaping.

[0026] The pulse-shaped samples of all K subcarriers then are summed in block 160 to form a single WHT-GFDM symbol xΩ[n], to which a cyclic prefix CP and/or a cyclic suffix CS can be added in block 170 in order to avoid inter-symbol interference, ISI.

[0027] Lastly the symbol including the optional cyclic prefix and cyclic suffix, i.e. the complex-valued samples reflecting the symbol are forwarded to a digital-to-analog converter and subsequent analog processing and amplifying stages, which are not shown in the figure, before it is transmitted, i.e. radiated, through a channel.

[0028] Figure 2 depicts a receiver 200 adapted and configured for receiving and processing the signals transmitted by transmitter 100 as described above.

[0029] Receiver 200 receives analog signals transmitted by transmitter 100, wherein the signals are attenuated and distorted by the channel. The received analog signals as received by an antenna may be amplified and processed before it is digitized, i.e. sampled and quantized thus producing complex-valued samples representing a received symbol. The samples of a received signal can be represented by a receive vector ye.

[0030] Based on the receive vector the channel properties, i.e. the channel impulse response CIR, may be estimated and the vector may be synchronized in block 210, before samples of a cyclic prefix and cyclic suffix are removed in block 220 from the receive vector, wherein we assume that the cyclic prefix is larger than the channel length. Furthermore signal represented by the samples of the receive vector is equalized in block 230 before being passed to detector 240.

[0031] Basically detector 240 can be any conventional detector, e.g. implementing a minimum mean squared error algorithm, MMSE, or a matched filter algorithm, MF, or preferably a zero-forcing algorithm, ZF.

[0032] Detector 240 outputs a plurality of subsymbols, wherein a subsymbol comprises transformed data symbols. The transformed subsymbols are forwarded to transformation block 250 where a transformation, i.e. an inverse of the transformation applied to the data symbols at the transmitter side, is applied to the symbols to reverse the transformation performed at the transmitter side. Note that an optional narrow band interference detector and processing block 245 can be arranged between detector 240 and transformation block 250 for identifying subcarriers exhibiting excess narrow band interference as described in more detail below.

[0033] The transformed data symbols then are then passed to a parallel-to-serial converter 260 that outputs a serial stream of symbols.

[0034] The serial stream of symbols is then passed to a slicer 270 that processes the serial stream soft symbols, each soft symbols comprised of a data symbol and interference and noised added by the channel, determines the most probable data symbol corresponding to the soft symbol. Slicer 270 furthermore de-maps the determined data symbol, i.e. converts the serial stream of determined data symbol into a stream of bits thus outputting a stream of estimated data bits.

[0035] A transmit symbol xΩ[n] can be generated at the transmitter side by performing matrix operations. The subsymbol cm transmitted in the m-th time slot, i.e. the m-th subsymbol consisting of the data symbols assigned to all K subcarriers, can be calculated as

wherein m is the index of a time slot, K is the number of subcarriers and wherein ΩK is the transform matrix, i.e. the spreading matrix, wherein in this exemplifying calculation a Walsh-Hadamard transformation matrix is considered:

wherein Ω1 = 1.

[0036] Transmission matrix A is defined by

where

with gk,m[n] being the impulse response, i.e. a filter for the k-th subcarrier of the m-th subsymbol. Transmit matrix A thus translates all M circularly shifted impulse responses to all K subcarriers.

[0037] And with the concatenation of a number of M Walsh-Hadamard-transformed sample vectors

the transmit vector can be written in matrix form as



[0038] Thus the transmitted signal is

i.e. the transmitted signal is the sum of all symbols dj,m that have been Walsh-Hadamard-transformed by multiplication with aj,k and that have been filtered and up-converted to the respective k-th transmit frequency by multiplication with

wherein K is the number of subcarriers, M is the number of timeslots per GFDM transmit symbol, dj,m is the j-th data symbol of the m-th time-slot, aj,k is the coefficient of the j-th row and k-th column of the spreading matrix Ω, and gm[n] is the filter impulse response circularly shift of the mN samples and N is the number of samples per time-slot.

[0039] The processing at the receiver side also can be described in mathematical terms. After synchronization, equalization, and removal of a cyclic prefix CP and/or a cyclic suffix CS a received vector of samples y is

wherein h is the channel impulse response vector CIR, w is the additive white Gaussian AWGN vector with variance σ2 = N0/2, and ⊛ denotes the circular convolution.

[0040] Under the assumption that the channel impulse response is known at the receiver side and the cyclic prefix CP is larger than the channel length, frequency domain equalization can be applied, yielding

wherein F is the Fourier matrix, X is the transmitted vector and W is the noise vector in the frequency domain and H is the channel frequency response.

[0041] Decoder 240, which may be a ZFR, then may calculate the received transformed sample vector as

wherein A+ is the inverse of A and the equivalent noise vector is



[0042] The recovered data symbols for each time-slot thus are given by

where wem is the equivalent noise vector containing the samples corresponding to the m-th subsymbol. Note that because the transformation combines the noise samples without changing the variance, each equivalent noise sample is a linear combination to the AWGN noise samples weighted by the channel frequency response and enhanced by the ZFR.

[0043] So by applying a transformation, e.g. in one embodiment by applying a Walsh-Hadamard matrix or a Cazac matrix or any other matrix where rows and columns are orthogonal, to the data symbols of all K subcarriers of one time slot m, the information of each data symbol is spread to all data symbols of said m-th timeslot. In other words the transformed data symbols of the m-th timeslot are a linear combination of K data symbols of the m-th time slot. As a consequence each data symbol in the m-th time slot contains at least some information of each other data symbol of the m-th time slot.

[0044] In case of a frequency selective channel, FSC, i.e. one subcarrier of the received plurality of subcarriers is found unusable, e.g. due to heavy interference by another signal or severely attenuated or erased, the information carried by the unusable subcarrier is lost. However, since the remaining subcarriers still carry a linear combination of all data symbols, the information of all data symbols can be reconstructed by applying an inverse Walsh-Hadamard transformation to the remaining encoded received symbols. Accordingly, if a subset of subcarriers is severely attenuated by the channel, the information provided by the remaining subcarriers allows reconstruction of the entire information. Though a subset of subcarriers may be attenuated, the burst transmission is successful.

[0045] In one embodiment the receiver may optionally comprise a narrow band interference detector and processing block 245 to clip subcarriers having an amplitude exceeding a predefined threshold value in order to prevent spreading of a heavily distorted subcarrier. Subcarriers showing an excess amplitude, i.e. an amplitude significantly higher than an average expected, most probably have been distorted by a narrow band interference, i.e. the payload signal of the subcarrier is overlaid by a strong interfering signal. If such a distorted subcarrier would be processed using the inverse transformation matrix, then the strong interfering signal would be spread across all healthy, i.e. non- or less-interfered subcarriers. The method performed at the receiver accordingly my optionally comprise the step of identifying subcarriers having an excess amplitude, i.e. subcarriers having an amplitude exceeding the predefined threshold, before applying the inverse of the transformation matrix. In one embodiment that identification can be performed based on analyzing the noise distribution and noise variance of the subcarriers. In case a subcarrier of excess amplitude is identified, the receiver may ignore signals carried by that subcarrier frequency in order to prevent spreading the heavily distorted subcarrier signals across other, non-infected subcarriers. In this way the method may prevent distortion, i.e. deterioration, of subcarrier signals that do not exceed the predefined threshold amplitude value. As a consequence the described method can be used in channels that may exhibit notches, i.e. where subcarriers are heavily attenuated, but also where subcarriers may be deteriorated by a strong, narrowband and impulsive interferer. The receiver may adapted correspondingly, i.e. the receiver may comprise a corresponding narrow band interference detector and processing block 245 for identifying and - as the case may be - set a filter for ignoring signals of the identified subcarrier.

[0046] As a consequence the proposed application may enhance the reliability of data communication when using a frequency selective channel FSC, and particularly in single shot transmissions, when low out-of-band radiation is required and spectrum fragmentation is present.

[0047] Further note that a Walsh-Hadamard transformation at the transmitter side is applied to data symbols. The transformation does not add any redundancy, consequently it does not affect the amount of data to be transmitted, but only increases the processing load both on the transmitter and on the receiver side of the transmission system by applying the Walsh-Hadamard transformation.

[0048] The bit error rate, BER, performance of the afore-described transmission system over frequency selective channels, FSCs, can be estimated as an average of the performance of each subcarrier considering the corresponding channel frequency response. Therefore the bit error probability can be approximated by

where H[k] is the flat channel frequency-response per subcarrier, Eb is the average energy per bit of the deployed J-QAM constellation, and N0 is the noise spectral density.

[0049] Considering now that the equivalent channel after the inverse Walsh-Hadamard-Transformation has a flat frequency response

above given bit error probability approximation can be used to estimate the bit error rate, BER, of the Walsh-Hadamard-Transformation GFDM over frequency selective channels, FSCs. Since the channel is flat, all subcarriers present the same performance, thus



[0050] Based on this bit probability one can conclude that the performance gain of the Walsh-Hadamard-Transformation GFDM system depends on the channel frequency response. A slight or even no performance gain is expected for a flat channel, while a high performance gain is expected for FSCs with narrow notches, because in this case the data symbols can be successfully recovered even if a subset of subcarriers is severely attenuated by the channel.

[0051] Figures 3a and 3b depict simulation curves of coded GFDM for comparison to other encoding schemes. Two codes are considered:
  1. i) a low-density parity-check, LDPC, code with a length of 64,800 bits and a code rate of ¾ and
  2. ii) a Reed-Solomon, RS, code (224, 204).


[0052] LDPC is a powerful code that achieves the best performance of the presented cases. However, LDPC uses a long code word, and the iterative decoding process requires knowledge of the signal-to-noise ratio to compute the log-likelihood ratio (LLR). This process increases both the latency and the complexity of the receiver and might be inappropriate for low-latency scenarios, power-limited devices and single burst transmissions. The Reed-Solomon code is an interesting solution for avoiding burst errors. The non-iterative decoding process introduces small latency and it does not require knowledge of the channel parameters. The code word can be shortened to match the GFDM frame size, which is appropriate for burst transmissions. However, energy limited low-cost devices might not be able to afford the complexity of the RS decode. Although RS-GFDM outperforms WHT-GFDM for channel A, both schemes have similar performance when the channel frequency response has deep notches, as shown in fig. 3b.


Claims

1. A method for receiving digital data using a generalized frequency divisional multiplexing, GFDM, multicarrier scheme comprising the steps of:

- receiving a GFDM signal representing a transmit symbol, said transmit symbol comprising a plurality of data symbols assigned to at least two subcarriers, and

- removing at least one of a cyclic prefix or a cyclic suffix from the transmit symbol

- de-spreading the information of each received data symbol of a time slot of the transmit symbol to the data symbols of each subcarrier of said time slot by applying an inverse of a transformation applied at the transmitter, and

- decoding the data symbols,

the method further comprising the step of identifying at least one subcarrier having an amplitude exceeding a predefined threshold value before performing the step of de-spreading of received data symbols and ignoring said identified subcarrier in the step of de-spreading.
 
2. The method of claim 1 wherein the step of applying an inverse of the transformation applied to data symbols at the transmitter comprises applying an inverse of a corresponding transformation matrix wherein rows and columns of the transformation matrix are orthogonal.
 
3. The method of claim 2 wherein the transformation matrix is a Walsh-Hadamard transformation matrix or a Cazac transformation matrix.
 
4. The method of any of claims 1-3 wherein the at least two subcarriers are separated by another frequency band occupied by another transmission.
 
5. A method for transmitting and receiving digital data in a transmit symbol using a generalized frequency divisional multiplexing, GFDM, multicarrier scheme, said transmit symbol comprising a plurality of data symbols assigned to at least two subcarriers, the method comprising the steps of

- mapping a plurality of data bits to data symbols and assigning data symbols to the at least two subcarriers, and

- spreading the information of each data symbol of a time slot of the transmit symbol across the data symbols of each subcarrier of said time slot by applying a transformation, and

- appending at least one of a cyclic prefix or a cyclic suffix to the transmit symbol, and

- transmitting a GFDM signal representing the transmit symbol, and

- receiving the digital data according to the method of any of claims 1-4.


 
6. The method of claim 5 wherein the step of spreading the information of each data symbol of one time slot of the transmit symbol to the data symbols of each subcarrier of said time slot comprises applying a transformation matrix to all data symbols of said time slot, wherein the rows and columns of the transformation matrix are orthogonal.
 
7. The method of claim 6 wherein the transformation matrix is a Walsh-Hadamard matrix or a Cazac matrix.
 
8. The method of any of claims 5-7 wherein the at least two subcarriers are separated by another frequency band occupied by another transmission.
 
9. The method of any of claims 5-8 wherein the step of spreading the information of each data symbol of a time slot of the transmit symbol to the data symbols of each subcarrier of said time slot is performed for each time slot of the transmit symbol.
 
10. A receiver (200) for receiving digital data using a generalized frequency divisional multiplexing, GFDM, scheme, said receiver (200) being adapted to:

- receiving a GFDM signal representing a transmit symbol, said transmit symbol comprising a plurality of data symbols assigned to at least two subcarriers, and

- removing at least one of a cyclic prefix or a cyclic suffix from the transmit symbol

- de-spreading the information of each received data symbol of a time slot of the transmit symbol to the data symbols of each subcarrier of said time slot by applying an inverse of a transformation applied at the transmitter, and

- decoding the data symbols,

the receiver further being adapted to:
identifying at least one subcarrier having an amplitude exceeding a predefined threshold value before performing the step of de-spreading of received data symbols and ignoring said identified subcarrier in the step of de-spreading.
 
11. The receiver (200) of claim 10 further adapted to:
wherein the step of applying an inverse of the transformation applied to data symbols at the transmitter further comprises applying an inverse of a corresponding transformation matrix wherein rows and columns of the transformation matrix are orthogonal.
 
12. The receiver (200) of claim 11 wherein the transformation matrix is a Walsh-Hadamard transformation matrix or a Cazac transformation matrix.
 
13. A generalized frequency divisional multiplexing, GFDM, system comprising

a) a transmitter (100) adapted and configured to transmit digital data in a transmit symbol using a GFDM multicarrier scheme, wherein a transmit symbol comprises a plurality of data symbols assigned to at least two subcarriers, said transmitter (100) being adapted to:

- mapping a plurality of data bits to data symbols and assigning data symbols to the at least two subcarriers, and

- spreading the information of each data symbol of a time slot of the transmit symbol to the data symbols of each subcarrier of said time slot by applying a transformation, and

- appending at least one of a cyclic prefix or a cyclic suffix to the transmit symbol, and

- transmitting a GFDM signal representing the transmit symbol,

b) a receiver (200) according to any of claims 10-12.


 
14. The GFDM system of claim 13, wherein the transmitter (100) is further adapted, in the step of spreading the information of each data symbol of one time slot of the transmit symbol to the data symbols of each subcarrier of said time slot, to apply a transformation matrix to all data symbols of said time slot, wherein the rows and columns of the transformation matrix are orthogonal.
 
15. The GFDM system of claim 14 wherein the transformation matrix is a Walsh-Hadamard matrix or a Cazac matrix.
 


Ansprüche

1. Verfahren zum Empfangen von digitalen Daten unter Verwendung eines generalisierten Frequenzmultiplexing(GFDM)-Mehrträgerschemas, das die folgenden Schritte umfasst:

- Empfangen eines GFDM-Signals, das ein Sendesymbol repräsentiert, wobei das Sendesymbol eine Vielzahl von Datensymbolen umfasst, die mindestens zwei Unterträgern zugewiesen sind, und

- Entfernen von mindestens einem zyklischen Präfix oder einem zyklischen Suffix aus dem Sendesymbol

- Entspreizen der Informationen jedes empfangenen Datensymbols eines Zeitschlitzes des Sendesymbols zu den Datensymbolen jedes Unterträgers des Zeitschlitzes durch Anwenden einer Umkehrung einer Transformation, die am Sender angewendet wird, und

- Decodieren der Datensymbole,

wobei das Verfahren ferner den folgenden Schritt umfasst Identifizieren von mindestens einem Unterträger, der eine Amplitude aufweist, die einen vordefinierten Schwellwert überschreitet, bevor der Schritt des Entspreizens von empfangenen Datensymbolen durchgeführt wird, und Ignorieren des identifizierten Unterträgers im Schritt des Entspreizens.
 
2. Verfahren nach Anspruch 1, wobei der Schritt des Anwendens einer Umkehrung der Transformation, die am Sender auf Datensymbole angewendet wird, das Anwenden einer Umkehrung einer entsprechenden Transformationsmatrix umfasst, wobei Zeilen und Spalten der Transformationsmatrix orthogonal sind.
 
3. Verfahren nach Anspruch 2, wobei die Transformationsmatrix eine Walsh-Hadamard-Transformationsmatrix oder eine Cazac-Transformationsmatrix ist.
 
4. Verfahren nach einem der Ansprüche 1-3, wobei die mindestens zwei Unterträger durch ein anderes Frequenzband, das von einer anderen Übertragung belegt ist, getrennt sind.
 
5. Verfahren zum Senden und Empfangen von digitalen Daten in einem Sendesymbol unter Verwendung eines generalisierten Frequenzmultiplexing(GFDM)-Mehrträgerschemas, wobei das Sendesymbol eine Vielzahl von Datensymbolen umfasst, die mindestens zwei Unterträgern zugewiesen sind, wobei das Verfahren die folgenden Schritte umfasst

- Zuordnen einer Vielzahl von Datenbits zu Datensymbolen und Zuweisen von Datensymbolen zu den mindestens zwei Unterträgern und

- Spreizen der Informationen jedes Datensymbols eines Zeitschlitzes des Sendesymbols über die Datensymbole jedes Unterträgers des Zeitschlitzes durch Anwenden einer Transformation und

- Anhängen von mindestens einem zyklischen Präfix oder einem zyklischen Suffix an das Sendesymbol und

- Übertragen eines GFDM-Signals, das das Sendesymbol repräsentiert, und

- Empfangen der digitalen Daten gemäß dem Verfahren nach einem der Ansprüche 1-4.


 
6. Verfahren nach Anspruch 5, wobei der Schritt des Spreizens der Informationen jedes Datensymbols eines Zeitschlitzes des Sendesymbols zu den Datensymbolen jedes Unterträgers des Zeitschlitzes das Anwenden einer Transformationsmatrix auf alle Datensymbole des Zeitschlitzes umfasst, wobei die Zeilen und Spalten der Transformationsmatrix orthogonal sind.
 
7. Verfahren nach Anspruch 6, wobei die Transformationsmatrix eine Walsh-Hadamard-Matrix oder eine Cazac-Matrix ist.
 
8. Verfahren nach einem der Ansprüche 5-7, wobei die mindestens zwei Unterträger durch ein anderes Frequenzband, das von einer anderen Übertragung belegt ist, getrennt sind.
 
9. Verfahren nach einem der Ansprüche 5-8, wobei der Schritt des Spreizens der Informationen jedes Datensymbols eines Zeitschlitzes des Sendesymbols zu den Datensymbolen jedes Unterträgers des Zeitschlitzes für jeden Zeitschlitz des Sendesymbols durchgeführt wird.
 
10. Empfänger (200) zum Empfangen von digitalen Daten unter Verwendung eines generalisierten Frequenzmultiplexing(GFDM)-Mehrträgerschemas, wobei der Empfänger (200) zu Folgendem angepasst ist:

- Empfangen eines GFDM-Signals, das ein Sendesymbol repräsentiert, wobei das Sendesymbol eine Vielzahl von Datensymbolen umfasst, die mindestens zwei Unterträgern zugewiesen sind, und

- Entfernen von mindestens einem zyklischen Präfix oder einem zyklischen Suffix aus dem Sendesymbol

- Entspreizen der Informationen jedes empfangenen Datensymbols eines Zeitschlitzes des Sendesymbols zu den Datensymbolen jedes Unterträgers des Zeitschlitzes durch Anwenden einer Umkehrung einer Transformation, die am Sender angewendet wird, und

- Decodieren der Datensymbole,

wobei der Empfänger ferner zu Folgendem angepasst ist: Identifizieren von mindestens einem Unterträger, der eine Amplitude aufweist, die einen vordefinierten Schwellwert überschreitet, bevor der Schritt des Entspreizens von empfangenen Datensymbolen durchgeführt wird, und Ignorieren des identifizierten Unterträgers im Schritt des Entspreizens.
 
11. Empfänger (200) nach Anspruch 10, der ferner zu Folgendem angepasst ist:
wobei der Schritt des Anwendens einer Umkehrung der Transformation, die am Sender auf Datensymbole angewendet wird, ferner das Anwenden einer Umkehrung einer entsprechenden Transformationsmatrix umfasst, wobei Zeilen und Spalten der Transformationsmatrix orthogonal sind.
 
12. Empfänger (200) nach Anspruch 11, wobei die Transformationsmatrix eine Walsh-Hadamard-Transformationsmatrix oder eine Cazac-Transformationsmatrix ist.
 
13. Generalisiertes Frequenzmultiplexing(GFDM)-System, das Folgendes umfasst

a) einen Sender (100) der angepasst und dazu ausgelegt ist, digitale Daten in einem Sendesymbol unter Verwendung eines GFDM-Mehrträgerschemas zu übertragen, wobei ein Sendesymbol eine Vielzahl von Datensymbolen umfasst, die mindestens zwei Unterträgern zugewiesen sind, wobei der Sender (100) zu Folgendem angepasst ist:

- Zuordnen einer Vielzahl von Datenbits zu Datensymbolen und Zuweisen von Datensymbolen zu den mindestens zwei Unterträgern und

- Spreizen der Informationen jedes Datensymbols eines Zeitschlitzes des Sendesymbols zu den Datensymbolen jedes Unterträgers des Zeitschlitzes durch Anwenden einer Transformation und

- Anhängen von mindestens einem zyklischen Präfix oder einem zyklischen Suffix an das Sendesymbol und

- Übertragen eines GFDM-Signals, das das Sendesymbol repräsentiert,

b) einen Empfänger (200) nach einem der Ansprüche 10-12.


 
14. GFDM-System nach Anspruch 13, wobei der Sender (100) ferner angepasst ist, im Schritt des Spreizens der Informationen jedes Datensymbols eines Zeitschlitzes des Sendesymbols zu den Datensymbolen jedes Unterträgers des Zeitschlitzes auf alle Datensymbole des Zeitschlitzes eine Transformationsmatrix anzuwenden, wobei die Zeilen und Spalten der Transformationsmatrix orthogonal sind.
 
15. GFDM-System nach Anspruch 14, wobei die Transformationsmatrix eine Walsh-Hadamard-Matrix oder eine Cazac-Matrix ist.
 


Revendications

1. Procédé permettant de recevoir des données numériques en utilisant un schéma multiporteuse de multiplexage par répartition en fréquence généralisée, GFDM, comprenant les étapes suivantes :

- recevoir un signal GFDM représentant un symbole de transmission, ledit symbole de transmission comportant une pluralité de symboles de données affectés à au moins deux sous-porteuses, et

- supprimer au moins un élément parmi un préfixe cyclique ou un suffixe cyclique à partir du symbole de transmission

- désétaler les informations de chaque symbole de données reçu d'un intervalle de temps du symbole de transmission dans les symboles de données de chaque sous-porteuse dudit intervalle de temps en appliquant l'inverse d'une transformation appliquée au niveau du transmetteur, et

- décoder les symboles de données, le procédé comprenant en outre l'étape suivante

identifier au moins une sous-porteuse ayant une amplitude excédant une valeur seuil prédéfinie avant d'exécuter l'étape de désétalement des symboles de données reçus et ignorer ladite sous-porteuse identifiée dans l'étape de désétalement.
 
2. Procédé selon la revendication 1, dans lequel l'étape d'application de l'inverse de la transformation appliquée aux symboles de données au niveau du transmetteur comprend appliquer l'inverse d'une matrice de transformation correspondante dans laquelle les lignes et les colonnes de la matrice de transformation sont orthogonales.
 
3. Procédé selon la revendication 2, dans lequel la matrice de transformation est une matrice de transformation Walsh-Hadamard ou une matrice de transformation Cazac.
 
4. Procédé selon l'une quelconque des revendications 1-3, dans lequel lesdites deux sous-porteuses sont séparées par une autre bande de fréquences occupée par une autre transmission.
 
5. Procédé permettant de transmettre et de recevoir des données numériques dans un symbole de transmission en utilisant un schéma multiporteuse de multiplexage par répartition en fréquence généralisée, GFDM, ledit symbole de transmission comportant une pluralité de symboles de données affectés à au moins deux sous-porteuses, le procédé comprenant les étapes suivantes

- mapper une pluralité de bits de données à des symboles de données et affecter des symboles de données auxdites deux sous-porteuses, et

- étaler les informations de chaque symbole de données d'un intervalle de temps du symbole de transmission à travers les symboles de données de chaque sous-porteuse dudit intervalle de temps en appliquant une transformation, et

- ajouter au moins un élément parmi un préfixe cyclique ou un suffixe cyclique au symbole de transmission, et

- transmettre un signal GFDM représentant le symbole de transmission, et

- recevoir les données numériques selon le procédé de l'une quelconque des revendications 1-4.


 
6. Procédé selon la revendication 5, dans lequel l'étape d'étalement des informations de chaque symbole de données d'un intervalle de temps du symbole de transmission dans les symboles de données de chaque sous-porteuse dudit intervalle de temps comprend appliquer une matrice de transformation à tous les symboles de données dudit intervalle de temps, les lignes et les colonnes de la matrice de transformation étant orthogonales.
 
7. Procédé selon la revendication 6, dans lequel la matrice de transformation est une matrice Walsh-Hadamard ou une matrice Cazac.
 
8. Procédé selon l'une quelconque des revendications 5-7, dans lequel lesdites deux sous-porteuses sont séparées par une autre bande de fréquences occupée par une autre transmission.
 
9. Procédé selon l'une quelconque des revendications 5-8, dans lequel l'étape d'étalement des informations de chaque symbole de données d'un intervalle de temps du symbole de transmission dans les symboles de données de chaque sous-porteuse dudit intervalle de temps est exécutée pour chaque intervalle de temps du symbole de transmission.
 
10. Récepteur (200) permettant de recevoir des données numériques en utilisant un schéma de multiplexage par répartition en fréquence généralisée, GFDM, ledit récepteur (200) étant conçu pour :

- recevoir un signal GFDM représentant un symbole de transmission, ledit symbole de transmission comportant une pluralité de symboles de données affectés à au moins deux sous-porteuses, et

- supprimer au moins un élément parmi un préfixe cyclique ou un suffixe cyclique à partir du symbole de transmission

- désétaler les informations de chaque symbole de données reçu d'un intervalle de temps du symbole de transmission dans les symboles de données de chaque sous-porteuse dudit intervalle de temps en appliquant l'inverse d'une transformation appliquée au niveau du transmetteur, et

- décoder les symboles de données, le récepteur étant en outre conçu pour :
identifier au moins une sous-porteuse ayant une amplitude excédant une valeur seuil prédéfinie avant d'exécuter l'étape de désétalement des symboles de données reçus et ignorer ladite sous-porteuse identifiée dans l'étape de désétalement.


 
11. Récepteur (200) selon la revendication 10 conçu en outre pour : dans lequel l'étape d'application de l'inverse de la transformation appliquée aux symboles de données au niveau du transmetteur comprend en outre appliquer l'inverse d'une matrice de transformation correspondante dans laquelle les lignes et les colonnes de la matrice de transformation sont orthogonales.
 
12. Récepteur (200) selon la revendication 11, dans lequel la matrice de transformation est une matrice de transformation Walsh-Hadamard ou une matrice de transformation Cazac.
 
13. Système de multiplexage par répartition en fréquence généralisée, GFDM, comprenant :

a) un transmetteur (100) conçu et configuré pour transmettre des données numériques dans un symbole de transmission en utilisant un schéma multiporteuse GFDM, un symbole de transmission comportant une pluralité de symboles de données affectés à au moins deux sous-porteuses, ledit transmetteur (100) étant conçu pour :

- mapper une pluralité de bits de données à des symboles de données et affecter des symboles de données auxdites deux sous-porteuses, et

- étaler les informations de chaque symbole de données d'un intervalle de temps du symbole de transmission à travers les symboles de données de chaque sous-porteuse dudit intervalle de temps en appliquant une transformation, et

- ajouter au moins un élément parmi un préfixe cyclique ou un suffixe cyclique au symbole de transmission, et

- transmettre un signal GFDM représentant le symbole de transmission,

b) un récepteur (200) selon l'une quelconque des revendications 10-12.


 
14. Système GFDM selon la revendication 13, dans lequel le transmetteur (100) est en outre conçu, dans l'étape d'étalement des informations de chaque symbole de données d'un intervalle de temps du symbole de transmission dans les symboles de données de chaque sous-porteuse dudit intervalle de temps, pour appliquer une matrice de transformation à tous les symboles de données dudit intervalle de temps, les lignes et les colonnes de la matrice de transformation étant orthogonales.
 
15. Système GFDM selon la revendication 14, dans lequel la matrice de transformation est une matrice Walsh-Hadamard ou une matrice Cazac.
 




Drawing














Cited references

REFERENCES CITED IN THE DESCRIPTION



This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Patent documents cited in the description




Non-patent literature cited in the description