WO2017084714A1

WO2017084714A1 - A transmitter for multi-carrier communication

Info

Publication number: WO2017084714A1
Application number: PCT/EP2015/077121
Authority: WO
Inventors: Doron Ezri; Genadiy Tsodik; Shimi Shilo; Oded Redlich
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2015-11-19
Filing date: 2015-11-19
Publication date: 2017-05-26
Also published as: CN108293031A; CN108293031B

Abstract

The invention relates to a transmitter (100) for multi-carrier communication. The transmitter (100) comprises a processor (101) configured to generate a training signal (S), wherein the training signal (S) comprises a first subsignal (S1) and a second subsignal (S2). The first subsignal (S1) and the second subsignal (S2) are transmitted serially in time, and frequency components of the first subsignal (S1) are shifted in frequency relative to frequency components of the second subsignal (S2) by a pre-defined frequency shift. The transmitter (100) further comprises a communication interface (103) configured to transmit the training signal (S). The invention further relates to a receiver configured to perform a channel- distortion-correction, in particular an automatic gain control (AGC), based on the received training signal (S).

Description

DESCRIPTION

A transmitter for multi-carrier communication

TECHNICAL FIELD

The present invention relates to the field of communication technology, in particular to training signal generation and channel-distortion-correction. BACKGROUND

In communication technology, training signals are commonly used for characterization of a transmission channel between a transmitter and a receiver. Usually, training signals have a predetermined structure and content which are known at the transmitter and the receiver and thus allow for an adaption of the receiver to varying characteristics of the transmission channel, such as scattering, fading, and power decay over distance.

For accurately representing the characteristics of the transmission channel, the temporal and spectral characteristics of the training signals are of major importance. Particularly in multi- carrier communication, the length of communication symbols and the frequency spacing of subcarriers at a given overall bandwidth are interrelated. For example, an improved representation of frequency characteristics of the transmission channel can lead to an increased length of communication symbols and thus to a reduced representation of temporal characteristics of the transmission channel.

When performing channel-distortion-correction, in particular automatic gain control (AGC), at the receiver, an increased length of the communication symbols can further result in an increased processing delay and thus deteriorate the overall efficiency of communication between the transmitter and the receiver. These effects are particularly pertinent to multi- carrier communications according to the IEEE 802.1 1 ax communication standard.

SUMMARY

It is an object for the invention to improve communication over a non ideal communication channel.

This object is achieved by the features of the independent claims. Further implementation forms are apparent from the dependent claims, the description and the figures. The invention is based on the finding that a training signal can be provided, which comprises a first subsignal and a second subsignal being transmitted serially in time and differing in frequency components. In particular, frequency components of the first subsignal are shifted in frequency relative to frequency components of the second subsignal by a pre-defined frequency shift. This way, a finer sampling of a transfer function of the transmission channel between the transmitter and the receiver is realized, thus leading to an improved channel- distortion-correction at the receiver. The invention can be applied in any context of multi-carrier communication, in particular multi-carrier communication based on Orthogonal Frequency-Division Multiplexing (OFDM), Orthogonal Frequency-Division Multiple Access (OFDMA), Filter-Bank Multi-Carrier (FBMC), or Single-Carrier Frequency-Division Multiple Access (SC-FDMA). The invention can particularly be applied for multi-carrier communication according to the IEEE 802.1 1 ax communication standard.

According to a first aspect, the invention relates to a transmitter for multi-carrier

communication, comprising a processor configured to generate a training signal, wherein the training signal comprises a first subsignal and a second subsignal, wherein the first subsignal and the second subsignal are transmitted serially in time, and wherein frequency

components of the first subsignal are shifted in frequency relative to frequency components of the second subsignal by a pre-defined frequency shift, and a communication interface configured to transmit the training signal. In this regard, multi-carrier communication refers to OFDM-, OFDMA- , FBMC-, or SC-FDMA- based communication. The training signal can be used for channel-distortion-correction, in particular automatic gain control (AGC), at a receiver.

In a first implementation form of the transmitter according to the first aspect as such, the training signal has a pre-defined length, in particular 4 us. By limiting the training signal to the pre-defined length, an efficient processing of the training signal at a receiver is achieved.

In a second implementation form of the transmitter according to the first aspect as such or the first implementation form of the first aspect, the first subsignal comprises a plurality, in particular two, periods of a first type, wherein a period of a first type is a pre-defined subsubsignal, and/or the second subsignal comprises a plurality, in particular two, periods of a second type, wherein a period of a second type is a pre-defined subsubsignal. In an exemplary implementation, 3 periods of the first type and then 2 periods of the second type are used. In another exemplary implementation, 2 periods of the first type and then 3 periods of the second type are used.

By using a plurality of periods of a first type, identical subsubsignals within the first subsignal can be used. By using a plurality of periods of a second type, identical subsubsignals within the second subsignal can be used. The subsubsignals can be communication symbols, e.g. OFDM symbols. The periods can comprise the communication symbols.

In a third implementation form of the transmitter according to the second implementation form of the first aspect, the period of the first type has a pre-defined length, in particular 0.8 us, and/or the period of the second type has a pre-defined length, in particular 0.8 us. By limiting the period of the first type and/or the period of the second type to pre-defined lengths, an efficient processing of the training signal at a receiver is achieved. In an exemplary implementation, the pre-defined length is 0.8 us. In another exemplary implementation, the pre-defined length is 1 .6 us.

In a fourth implementation form of the transmitter according to the first aspect as such or any preceding implementation form, the processor is configured to assign a first set of

subcarriers to the first subsignal, and/or a second set of subcarriers to the second subsignal. By assigning a first set of subcarriers and/or a second set of subcarriers, efficient multi- carrier communication is realized.

In this regard, a subcarrier refers to a carrier used for multi-carrier communication, in particular OFDM-, OFDMA-, FBMC-, or SC-FDMA-based communication. A subcarrier can also be referred to as a tone. The first set of subcarriers and/or the second set of subcarriers can comprise only one subcarrier, respectively. The first set of subcarriers and/or the second set of subcarriers can be pre-stored in a memory of the transmitter.

In a fifth implementation form of the transmitter according to the fourth implementation form of the first aspect, the processor is configured to assign a first set of beamforming weights associated with the first set of subcarriers to the first subsignal, and/or a second set of beamforming weights associated with the second set of subcarriers to the second subsignal. By assigning beamforming weights, a transmission of the training signal using beamforming techniques is realized. The first set of beamforming weights and/or the second set of beamforming weights can be determined by the processor.

In a sixth implementation form of the transmitter according to the fourth implementation form or the fifth implementation form of the first aspect, the subcarriers within the first set of subcarriers are equally spaced in frequency, and/or the subcarriers within the second set of subcarriers are equally spaced in frequency. By using subcarriers being equally spaced in frequency, an efficient sampling of the transfer function of the transmission channel is achieved. Furthermore, this can mean that the time-domain signal is a repetition of identical periods. The subcarriers within the first set of subcarriers and the subcarriers within the second set of subcarriers can have identical pre-defined frequency spacings.

In a seventh implementation form of the transmitter according to the first aspect as such or any preceding implementation form, the processor is configured to generate a data packet comprising a training field, wherein the training field comprises the first subsignal and the second subsignal, wherein the communication interface is configured to transmit the data packet. By generating the data packet and by arranging the first subsignal and the second subsignal within the training field of the data packet, an efficient detection and processing of the training signal at a receiver is realized. The training field can be a High Efficiency Short Training Field (HE-STF). The data packet can be an IEEE 802.1 1 ax data packet.

According to a second aspect, the invention relates to a receiver for multi-carrier

communication, comprising a communication interface configured to receive a training signal from a transmitter according to the first aspect as such or any implementation form of the first aspect, wherein the training signal comprises a first subsignal and a second subsignal, wherein the first subsignal and the second subsignal are transmitted serially in time, and wherein frequency components of the first subsignal are shifted in frequency relative to frequency components of the second subsignal by a pre-defined frequency shift, and a processor configured to perform a channel-distortion-correction, in particular an automatic gain control (AGC), based on the received training signal, in particular the first subsignal and the second subsignal.

In this regard, channel-distortion-correction refers to an adaption of the receiver to

characteristics of the transmission channel based on the training signal. The channel- distortion-correction can comprise an automatic gain control (AGC) at the receiver. The channel-distortion-correction can further comprise an equalization (EQ) at the receiver.

In a first implementation form of the receiver according to the second aspect as such, the first subsignal comprises a plurality, in particular two, periods of a first type, wherein a period of a first type is a pre-defined subsubsignal, and/or the second subsignal comprises a plurality, in particular two, periods of a second type, wherein a period of a second type is a pre-defined subsubsignal, wherein the processor is configured to perform the channel-distortion- correction based on the plurality of periods, in particular the second period, of the first type, and/or perform the channel-distortion-correction based on the plurality of periods, in particular the second period, of the second type. By performing the channel-distortion- correction based on the plurality of periods, in particular the second period, of the first type and/or the second type, an efficient channel-distortion-correction based on subsubsignals is realized. The subsubsignals can be detected by the processor.

In a second implementation form of the receiver according to the second aspect as such or the first implementation form of the second aspect, the communication interface is configured to determine a first energy indicator indicating an energy of the first subsignal, and/or determine a second energy indicator indicating an energy of the second subsignal, wherein the processor is configured to perform the channel-distortion-correction based on the first energy indicator and/or the second energy indicator. By determining the first energy indicator and/or the second energy indicator, an efficient channel-distortion-correction, in particular automatic gain control (AGC), is realized. The first energy indicator and/or the second energy indicator can e.g. be indicative of an attenuation of the transmission channel between the transmitter and the receiver.

According to a third aspect, the invention relates to a communication system for multi-carrier communication, comprising a transmitter according to the first aspect as such or any implementation form of the first aspect, and a receiver according to the second aspect as such or any implementation form of the second aspect. The communication system allows for multi-carrier communication, in particular OFDM-, OFDMA-, FBMC-, or SC-FDMA-based communication, over a transmission channel. The communication system can allow for communication according to the IEEE 802.1 1 ax communication standard.

Further features of the communication system directly result from the features of the transmitter and the receiver.

According to a fourth aspect, the invention relates to a transmitting method for multi-carrier communication, comprising generating, by a processor, a training signal, wherein the training signal comprises a first subsignal and a second subsignal, wherein the first subsignal and the second subsignal are transmitted serially in time, and wherein frequency components of the first subsignal are shifted in frequency relative to frequency components of the second subsignal by a pre-defined frequency shift, and transmitting, by a communication interface, the training signal. The transmitting method can be performed by the transmitter, in particular the processor and the communication interface. Further features of the transmitting method directly result from the functionality of the transmitter. According to a fifth aspect, the invention relates to a receiving method for multi-carrier communication, comprising receiving, by a communication interface, a training signal, wherein the training signal comprises a first subsignal and a second subsignal, wherein the first subsignal and the second subsignal are transmitted serially in time, and wherein frequency components of the first subsignal are shifted in frequency relative to frequency components of the second subsignal by a pre-defined frequency shift, and performing, by a processor, a channel-distortion-correction, in particular an automatic gain control (AGC), based on the received training signal, in particular the first subsignal and the second subsignal. The receiving method can be performed by the receiver, in particular the processor and the communication interface. Further features of the receiving method directly result from the functionality of the receiver.

According to a sixth aspect, the invention relates to a computer program comprising a program code for performing one of the methods of the fourth aspect and/or the fifth aspect when executed on a computer. The program code can comprise machine-executable instructions for performing one of the methods.

The invention can be implemented in hardware and/or software.

BRIEF DESCRIPTION OF EMBODIMENTS

Embodiments of the invention will be described with respect to the following figures, in which: Fig. 1 shows a diagram of a transmitter for multi-carrier communication; Fig. 2 shows a diagram of a receiver for multi-carrier communication; Fig. 3 shows a diagram of a communication system for multi-carrier communication;

Fig 4 shows a diagram of a transmitting method for multi-carrier communication;

Fig 5 shows a diagram of a receiving method for multi-carrier communication; Fig. 6 shows diagrams illustrating relations between frequency components and communication symbols in multi-carrier communication; Fig. 7 shows diagrams illustrating the generation of a training signal comprising a first subsignal and a second subsignal;

Fig. 8 shows a diagram illustrating the generation of a training signal comprising a first subsignal and a second subsignal;

Fig. 9 shows a diagram illustrating an RSSI estimation performance in multi-carrier communication; and

Fig. 10 shows a diagram illustrating an RSSI estimation performance in multi-carrier communication.

DETAILED DESCRIPTION OF EMBODIMENTS

Fig. 1 shows a diagram of a transmitter 100 for multi-carrier communication according to an embodiment. The transmitter 100 comprises a processor 101 configured to generate a training signal S, wherein the training signal S comprises a first subsignal S1 and a second subsignal S2. The first subsignal S1 and the second subsignal S2 are transmitted serially in time, and frequency components of the first subsignal S1 are shifted in frequency relative to frequency components of the second subsignal S2 by a pre-defined frequency shift. The transmitter 100 further comprises a communication interface 103 configured to transmit the training signal S.

The transmitter 100 can be configured to perform multi-carrier communication, in particular OFDM-, OFDMA-, FBMC-, or SC-FDMA-based communication.

Fig. 2 shows a diagram of a receiver 200 for multi-carrier communication according to an embodiment. The receiver 200 comprises a communication interface 201 configured to receive a training signal S from a transmitter 100, wherein the training signal S comprises a first subsignal S1 and a second subsignal S2. The first subsignal S1 and the second subsignal S2 are transmitted serially in time, and frequency components of the first subsignal S1 are shifted in frequency relative to frequency components of the second subsignal S2 by a pre-defined frequency shift. The receiver further comprises a processor 203 configured to perform a channel-distortion-correction, in particular an automatic gain control (AGC), based on the received training signal S, in particular the first subsignal S1 and the second subsignal S2.

The receiver 200 can be configured to perform multi-carrier communication, in particular OFDM-, OFDMA-, FBMC-, or SC-FDMA-based communication.

Fig. 3 shows a diagram of a communication system 300 for multi-carrier communication according to an embodiment. The communication system 300 comprises a transmitter 100 and a receiver 200. The transmitter 100 and the receiver 200 communicate over a

transmission channel 301. The transmission channel 301 can be a wireless transmission channel.

The transmitter 100 comprises a processor 101 configured to generate a training signal S, wherein the training signal S comprises a first subsignal S1 and a second subsignal S2. The first subsignal S1 and the second subsignal S2 are transmitted serially in time, and frequency components of the first subsignal S1 are shifted in frequency relative to frequency

components of the second subsignal S2 by a pre-defined frequency shift. The transmitter 100 further comprises a communication interface 103 configured to transmit the training signal S. The receiver 200 comprises a communication interface 201 configured to receive the training signal S from the transmitter 100, wherein the training signal S comprises the first subsignal S1 and the second subsignal S2. The receiver 200 further comprises a processor 203 configured to perform a channel-distortion-correction, in particular an automatic gain control (AGC), based on the received training signal S, in particular the first subsignal S1 and the second subsignal S2.

Consequently, a finer sampling of a transfer function of the transmission channel 301 between the transmitter 100 and the receiver 200 is realized, thus leading to an improved channel-distortion-correction, in particular automatic gain control (AGC), at the receiver 200. The communication system 300 can be configured to perform multi-carrier communication, in particular OFDM-, OFDMA-, FBMC-, or SC-FDMA-based communication, over the transmission channel 301.

Fig. 4 shows a diagram of a transmitting method 400 for multi-carrier communication according to an embodiment. The transmitting method 400 comprises generating 401 a training signal S, wherein the training signal S comprises a first subsignal S1 and a second subsignal S2. The first subsignal S1 and the second subsignal S2 are transmitted serially in time, and frequency components of the first subsignal S1 are shifted in frequency relative to frequency components of the second subsignal S2 by a pre-defined frequency shift. The transmitting method 400 further comprises transmitting 403 the training signal S.

The transmitting method 400 can be performed by the transmitter 100.

Fig. 5 shows a diagram of a receiving method 500 for multi-carrier communication according to an embodiment. The receiving method 500 comprises receiving 501 a training signal S, wherein the training signal S comprises a first subsignal S1 and a second subsignal S2. The first subsignal S1 and the second subsignal S2 are transmitted serially in time, and frequency components of the first subsignal S1 are shifted in frequency relative to frequency

components of the second subsignal S2 by a pre-defined frequency shift. The receiving method 500 further comprises performing 503 a channel-distortion-correction, in particular an automatic gain control (AGC), based on the received training signal S, in particular the first subsignal S1 and the second subsignal S2.

The receiving method 500 can be performed by the receiver 200.

Further embodiments of the invention are described in the following in more detail. The further embodiments are illustrated in the context of multi-carrier communication according to the IEEE 802.1 1 ax communication standard. However, the invention can evidently be applied in the context of any communication standard.

When communicating according to the IEEE 802.1 1 ax communication standard, a

transmitter 100 can generate a data packet, in particular an IEEE 802.1 1 ax data packet, which comprises a training field, in particular a High Efficiency Short Training Field (HE-STF).

The HE-STF within the IEEE 802.1 1 ax data packet, like the HTA HT-STF in the IEEE

802.1 1 η and IEEE 802.1 1 ac communication standards, respectively, can be designed to allow a receiver 200 to adjust the settings of an AGC according to data pre-coding and/or variable bandwidth at the transmitter 100. The adjustment of the AGC can include gain estimation, gain settling and direct current (DC) offset estimation.

In the IEEE 802.1 1 a/g/n/ac communication standards, the STF is usually transmitted on every 4-th subcarrier. This can result in replicas of 0.8 us duration. In the IEEE 802.1 1 n/ac communication standard, there can be 5 replicas of 0.8 us duration. The IEEE 802.1 1 ax communication standard, however, may introduce a smaller frequency spacing of the subcarriers, e.g. 1/4 of the frequency spacing as used in IEEE 802.1 1 a/g/n/ac, i.e. 78.125 kHz vs. 312.5 kHz, and therefore a communication symbol may be 4x longer in time, i.e. 13.2 us vs. 3.2 us. Spectral patterns of the STF with a periodicity in time of 0.8/1 .6/2.4 us are further intended.

Fig. 6 shows diagrams 601 -607 illustrating relations between frequency components and communication symbols in multi-carrier communication. In diagram 601 and diagram 605, subcarriers are shown in frequency domain. In diagram 603 and diagram 607,

communication symbols are shown in time domain. The diagrams 601 -607 are interrelated by an inverse fast Fourier transform (IFFT) or a fast Fourier transform (FFT), respectively. F_s denotes a sampling frequency in time domain. The diagrams 601 -607 further illustrate different cyclic prefix (CP) requirements.

In the IEEE 802.1 1 ax communication standard, it may be intended to use 5 periods within the HE-STF for AGC processing. If the frequency spacing between the non-zero subcarriers of HE-STF is preserved, i.e. every 4th subcarrier, then the duration in time is 3.2 us; thus AGC processing may last 16 us, i.e. 5 x 3.2 us. If the frequency spacing or gap between the subcarriers of the IEEE 802.1 1 a/g/n/ac communication standard is preserved, i.e. 1.25 MHz, then the duration in time is 0.8 us and AGC processing may last 4 us, i.e. 5 x 0.8 us. In this case, the frequency spacing is every 16^th subcarrier. When applying OFDMA techniques in general, a single client may be allocated on a very narrow band, e.g. a single resource unit (RU), while the rest of the bandwidth may be divided between other clients. The clients can e.g. be transmitters and/or receivers. If beamforming techniques are e.g. applied at the transmitters, a single client may see a more flat channel at its RU, whereas it may see a more selective channel at all the RUs that are not allocated to it. This can be due to pre-coding being used across other RUs for other clients. In this case, a HE-STF with frequency spacing of 1.25 MHz, e.g. using 16 subcarriers, may not be enough for an accurate representation of a spectrum and may result in an erroneous setting of the AGC. On the other hand, a HE-STF with frequency spacing of 312.5 kHz, e.g. using 4 subcarriers, can result in a longer period in time domain; thus the overhead of HE-STF transmission can be increased. The trade-off can be between accurate representation of spectrum and overhead in time.

Fig. 7 shows diagrams 701 -709 illustrating the generation of a training signal S comprising a first subsignal S1 and a second subsignal S2 according to an embodiment. The first subsignal S1 and the second subsignal S2 are transmitted serially in time, and frequency components of the first subsignal S1 are shifted in frequency relative to frequency components of the second subsignal S2 by a pre-defined frequency shift. In diagram 701 and diagram 705, subcarriers are shown in frequency domain. In diagram 703, diagram 707 and diagram 709, communication symbols are shown in time domain. The diagrams 701 -707 are interrelated by an inverse fast Fourier transform (IFFT) or a fast Fourier transform (FFT), respectively. F_s denotes a sampling frequency in time domain. The diagram 709 depicts the training signal S comprising the first subsignal S1 and the second subsignal S2. The training signal S can form an HE-STF.

The HE-STF can be generated by combining a short time duration and an accurate spectrum representation. The 1 st and 2nd replicas of HE-STF, each lasting 0.8 us and both being identical, can be generated using a multi-carrier signal with frequency spacing of 1 .25 MHz. The 3rd, 4th and 5th replicas of HE-STF, each lasting 0.8 us and all being identical, can be generated using a multi-carrier signal with frequency spacing of 1.25 MHz shifted by 2 x 312.5 kHz, i.e. having a pre-defined frequency shift of 2 x 312.5 kHz. At each replica, beamforming weights corresponding to active subcarriers can be employed such that an IFFT operation can provide relevant time domain 0.8 us duration replicas. The replicas can be different both in their frequency components and possibly beamforming weights. Although replicas can be different, they may satisfy the condition that they are sampled 1 :4 such that an amplitude periodicity of 0.8 us is maintained for all. Since these replicas may be of 0.8 us duration, compatibility with AGC mechanisms can be achieved.

This approach allows for improving HE-STF based gain estimation by modifying the frequency location of subcarriers in the HE-STF between replicas, so that different replicas sample the frequency at different locations. By using all replicas, a finer sampling in frequency domain by the training signal is performed and gain estimation is enhanced.

In diagram 709, the generation of 5 replicas is shown. The 3rd, 4th and 5th replicas are shifted in frequency by 2 x 312.5 kHz = 625 kHz relative to the 1 st and 2nd replicas. The 2nd replica can be used for gain estimation without any performance degradation. When energy or power is collected only over the 2nd replica in the HE-STF, the performance of standard approaches and this approach may be identical. This means that this approach may not degrade energy or power estimation performance. When additional replicas in the HE-STF are used for energy or power estimation, such as for example the 4th replica, the energy or power estimation can be enhanced since each replica can sample at different subcarriers. In effect, this will sample more densely in frequency.

Chipset vendors wishing to enhance energy or power estimation at receivers can consequently do so with minor changes to the AGC mechanism. The adjustment of the AGC can be performed upon the basis of energy or power indicators indicating the energy or power of the first subsignal S1 and/or the second subsignal S2.

Fig. 8 shows a diagram 801 illustrating the generation of a training signal S comprising a first subsignal S1 and a second subsignal S2 according to an embodiment. The first subsignal S1 and the second subsignal S2 are transmitted serially in time, and frequency components of the first subsignal S1 are shifted in frequency relative to frequency components of the second subsignal S2 by a pre-defined frequency shift. In diagram 801 , communication symbols are shown in time domain. In this regard, diagram 801 corresponds to diagram 709 in Fig. 7. The training signal S can form an HE-STF.

Naturally, the number of identical replicas may differ between the first subsignal S1 and the second subsignal S2 within the HE-STF. For example, as shown in diagram 801 , 3 identical replicas at the beginning and afterwards 2 identical replicas can be used. Such design can allow improved initial gain estimation, e.g. by averaging over two consecutive replicas.

Fig. 9 shows a diagram 901 illustrating an RSSI estimation performance in multi-carrier communication. The diagram 901 depicts a cumulative distribution function (CDF) in dependence of an estimation error.

Fig. 10 shows a diagram 1001 illustrating an RSSI estimation performance in multi-carrier communication. The diagram 1001 depicts a cumulative distribution function (CDF) in dependence of an estimation error.

In diagram 901 and diagram 1001 , exemplary simulation results of two HE-STF designs are presented. Both designs use a frequency spacing of 1 .25 MHz between pilot subcarriers of the HE-STF. In the legend, "Identical Periods" refers to 5 periods of 0.8 us duration each, wherein all periods are identical, and "Suggested Scheme" refers to 5 periods of 0.8 us duration each, wherein the design is as described previously. In the considered cases, 2 identical periods of the first type and then 3 identical periods of the second type are used.

The error in energy estimation, i.e. received signal strength indication (RSSI), is compared between each design and a reference design with 1 X symbol duration, i.e. pilot subcarrier every 4th subcarrier.

The following parameters are assumed with respect to the simulation:

20 MHz bandwidth 9 clients, each allocated a single resource unit (RU)

2 antenna transmission (Tx)

Random beamforming weight per subcarrier outside the client's resource unit (RU) Urban Micro (UMi) channel

In diagram 901 , the RSSI estimation error between the two schemes for the UMi channel model is compared, when the RSSI estimation is carried out only over the 2nd period in the HE-STF. As shown, the performance of the suggested scheme is identical to the

performance of the existing scheme in this example.

In diagram 1001 , the RSSI estimation error between the two schemes for the UMi channel model is compared, when the RSSI estimation is carried out over the 2nd and 4th periods in the HE-STF. As shown, the suggested scheme outperforms the "Identical Periods" scheme by up to 1 dB in this example.

In summary, a design for a training signal, in particular for an HE-STF, is presented, wherein two groups of identical 0.8 us periods are provided. The design yields an improved RSSI estimation performance compared to 5 identical 0.8 us periods, because frequency sampling is denser using the suggested design. The suggested design may not incur any additional overhead.

Using the suggested design, the difference in energy or power between each period may be small enough for only the variable gain amplifier (VGA) to modify the reference energy or power. No modification of the low noise amplifier (LNA) may be performed, so the

implementation is straightforward. Since a modification of the transmitted training signal is performed, it may be straightforward to receive the transmitted training signal and analyze the frequency contents of each replica. If some of the replicas are shifted in frequency compared to other replicas, the training signal S can be detected. Although the invention is described with reference to specific features, implementation forms, and embodiments, it is evident that various modifications and combinations can be made thereto without departing from the spirit and scope of the invention. The description and the figures are, accordingly, to be regarded simply as an illustration of the invention as defined by the appended claims, and are contemplated to cover any and all modifications, variations, combinations, or equivalents that fall within the scope of the invention.

Claims

1 . A transmitter (100) for multi-carrier communication, comprising:

- a processor (101 ) configured to generate a training signal (S),

wherein the training signal (S) comprises:

- a first subsignal (S1 );

- a second subsignal (S2);

- wherein the first subsignal (S1 ) and the second subsignal (S2) are transmitted serially in time, and wherein frequency components of the first subsignal (S1 ) are shifted in frequency relative to frequency components of the second subsignal (S2) by a pre-defined frequency shift;

- a communication interface (103) configured to transmit the training signal (S).

2. The transmitter (100) of claim 1 , wherein

the training signal (S) has a pre-defined length, in particular 4 us.

3. The transmitter (100) of one of the preceding claims, wherein

- the first subsignal (S1 ) comprises a plurality, in particular two, periods of a first type, wherein a period of a first type is a pre-defined subsubsignal; and/or

- the second subsignal (S2) comprises a plurality, in particular two, periods of a second type, wherein a period of a second type is a pre-defined subsubsignal.

4. The transmitter (100) of claim 3, wherein

- the period of the first type has a pre-defined length, in particular 0.8 us; and/or

- the period of the second type has a pre-defined length, in particular 0.8 us.

5. The transmitter (100) of one of the preceding claims, wherein

the processor (101 ) is configured to assign

- a first set of subcarriers to the first subsignal (S1 ); and/or

- a second set of subcarriers to the second subsignal (S2).

6. The transmitter (100) of claim 5, wherein

the processor (101 ) is configured to assign

- a first set of beamforming weights associated with the first set of subcarriers to the first subsignal (S1 ); and/or

- a second set of beamforming weights associated with the second set of subcarriers to the second subsignal (S2).

7. The transmitter (100) of claims 5 or 6, wherein

- the subcarriers within the first set of subcarriers are equally spaced in frequency; and/or

- the subcarriers within the second set of subcarriers are equally spaced in frequency.

8. The transmitter (100) of one of the preceding claims, wherein

the processor (101 ) is configured to

- generate a data packet comprising a training field, wherein the training field comprises the first subsignal (S1 ) and the second subsignal (S2);

wherein the communication interface (103) is configured to

- transmit the data packet.

9. A receiver (200) for multi-carrier communication, comprising:

- a communication interface (201 ) configured to receive a training signal (S) from a transmitter (100) according to one of the preceding claims,

wherein the training signal (S) comprises:

- a first subsignal (S1 );

- a second subsignal (S2);

- a processor (203) configured to perform a channel-distortion-correction, in particular an automatic gain control (AGC), based on the received training signal (S), in particular the first subsignal (S1 ) and the second subsignal (S2).

10. The receiver (200) of claim 9, wherein

- the second subsignal (S2) comprises a plurality, in particular two, periods of a second type, wherein a period of a second type is a pre-defined subsubsignal; wherein

the processor (203) is configured to

- perform the channel-distortion-correction based on the plurality of periods, in particular the second period, of the first type; and/or

- perform the channel-distortion-correction based on the plurality of periods, in particular the second period, of the second type.

1 1 . The receiver (200) of claims 9 or 10, wherein

the communication interface (201 ) is configured to

- determine a first energy indicator indicating an energy of the first subsignal (S1 ); and/or

- determine a second energy indicator indicating an energy of the second subsignal (S2); wherein the processor (203) is configured to

- perform the channel-distortion-correction based on the first energy indicator and/or the second energy indicator.

12. A communication system (300) for multi-carrier communication, comprising:

- a transmitter (100) according to one of the claims 1 to 8; and

- a receiver (200) according to one of the claims 9 to 1 1 .

13. A transmitting method (400) for multi-carrier communication, comprising:

- generating (401 ) a training signal (S),

wherein the training signal (S) comprises:

- a first subsignal (S1 );

- a second subsignal (S2);

- transmitting (403) the training signal (S).

14. A receiving method (500) for multi-carrier communication, comprising:

- receiving (501 ) a training signal (S),

wherein the training signal (S) comprises:

- a first subsignal (S1 );

- a second subsignal (S2);

- performing (503) a channel-distortion-correction, in particular an automatic gain control (AGC), based on the received training signal (S), in particular the first subsignal (S1 ) and the second subsignal (S2).

15. A computer program comprising a program code for performing one of the methods (400; 500) of claims 13 or 14 when executed on a computer.