WO2006052774A2

WO2006052774A2 - A transmitter and a receiver for communicating a signal from multiple antennas using a preamble

Info

Publication number: WO2006052774A2
Application number: PCT/US2005/040076
Authority: WO
Inventors: Markus Muck; Stephanie Rouquette-Leveil; Marc De Courville
Original assignee: Motorola, Inc.
Priority date: 2004-11-09
Filing date: 2005-11-04
Publication date: 2006-05-18
Also published as: WO2006052774A9; WO2006052774A3

Abstract

A transmitter comprises a transmit processor (105), preamble inserters (107, 109) and transmit units (113, 117) which are operable to transmit sub-signals from the antennas (115, 119). A preamble set comprising an individual preamble for each antenna generated by a preamble generator (111). Each of the individual preambles comprise first and second sections. The first section comprises a first data sequence, different for each antenna, and is selected from a predetermined set of preamble sequences. The first data sequence may comprise only time domain symbols belonging to the alphabet [1,-1, i,-i,0]. The second section comprises repetitions of a predetermined second data sequence where each repetition is weighted by a coefficient that is different for each antenna. Weighting may be by coefficients of different rows of a Walsh Hadamard matrix. The second data sequence may comprise only frequency domain symbols belonging to the alphabet of [1,-1,0].

Description

A TRANSMITTER AND A RECEIVER FOR COMMUNICATING A SIGNAL FROM MULTIPLE ANTENNAS USING A PREAMBLE

Field of the invention

The invention relates to a transmitter and a receiver for transmitting a signal from multiple antennas of the transmitter to the receiver using a preamble and in particular, but not exclusively, to a communication system using a plurality of transmit and receive antennas.

Background of the Invention

In recent years, wireless data communication in domestic and enterprise environments have become increasingly commonplace and an increasing number of wireless communication systems have been designed and deployed. In particular, the use of wireless networking has become prevalent and wireless network standards such as IEEE 801.11a and IEEE 801. Hg have become commonplace.

The requirement for increasing data rates, communication capacity and quality of services has led to continued research and new techniques and standards being developed for wireless networking. One such standard is the IEEE 801. Hn standard which is currently under development. IEEE 801. Hn is expected to operate in the 5GHz frequency spectrum and promises data rates of around lOOMbps and above on top of the MAC layer. IEEE 801. Hn will use many techniques which are similar to the earlier developed IEEE 801.11a and IEEE 801.Hg standards. The standard is to a large extent compatible with many of the characteristics of the earlier standards thereby allowing reuse of techniques and circuitry developed for these. For example, as in the previous standards IEEE 801.11a and IEEE 801. Hg, IEEE 801. Hn will use Orthogonal Frequency Division Multiplex (OFDM) modulation for transmission over the air interface.

Furthermore, in order to improve efficiency and to achieve the high data rates, IEEE 801. Hn is planned to introduce a number of advanced techniques. For example, IEEE 801. Hn communication is expected to typically be - based on a plurality of transmit and receive antennas. Furthermore, rather than merely providing diversity from spatially separated transmit antennas, IEEE 801.Hn will utilise transmitters having at least partially separate transmit circuitry for each antenna thus allowing different sub-signals to be transmitted from each of the antennas. The receivers may receive signals from a plurality of receive antennas and may perform a joint detection taking into account the number and individual characteristics associated with each of the plurality of transmit antennas and receive antennas. Specifically, IEEE 801.Hn has seen the likely introduction of a

Multiple-Transmit-Multiple-Receive (MTMR) antenna concept which exploits Multiple-Input-Multiple-Output (MIMO) channel properties to improve performance and throughput.

In order to acquire characteristics of a signal, it is known to transmit known data, typically referred to as a pre-amble or training data. The receiver may determine characteristics of the received signal and the propagation channel by evaluating the distortions introduced to the training data by the transmission.

In order to optimise performance of the communication, it is important to use preambles which have the desired properties. For example, in order to generate accurate timing information and to reduce the probability of erroneous detection, it is desirable that a preamble has a very narrow auto-correlation.

Preambles have been specified for communication systems such as IEEE 801.11a and IEEE 801. Hg. However, currently, no preambles have been standardised for IEEE 801. Hn although a number of different preambles have been proposed as part of specific system proposals. An example of a current proposal for IEEE 801. Hn may be found in IEEE P802.ll Wireless LANs TGn Sync Proposal Technical Specification, IEEE document number IEEE802.il- 04/0889r04, August 13, 2004.

Thus, due to the different technical requirements for IEEE 801. Hn (and in particular the MTMR antenna concept) , existing WLAN standards and recent publications do not propose preamble sequences which are suitable for this application. Furthermore, the preambles currently proposed for MTMR do not provide optimal performance but suffer from one or more disadvantages. For example, current IEEE802.11n proposals suggest the use of preambles which are similar to IEEE802.11a preambles. Specifically, currently proposed preambles have non-constant amplitude and phase characteristics and require high complexity time-domain correlators in the receivers. Hence, an improved system having improved preamble performance would be advantageous and in particular a preamble allowing increased flexibility, improved performance, reduced complexity and/or improved suitability for multiple transmit antenna systems and in particular for IEEE 801.Hn systems would be advantageous .

Summary of the Invention

Accordingly, the Invention seeks to preferably mitigate, alleviate or eliminate one or more of the above mentioned disadvantages singly or in any combination.

According to a first aspect of the invention there is provided a transmitter comprising: means for transmitting a signal as sub-signals from a plurality of antennas using a preamble set comprising an individual preamble for each antenna; and means for generating the preamble set with each of the individual preambles comprising a first section and a second section wherein the first section comprises a first data sequence including at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprises repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna.

The invention may provide improved performance and/or facilitate implementation. In particular, the invention may provide improved performance communication performance and/or reduced complexity of receivers. The preamble set is for example suited for systems allowing individual transmissions from several antennas to a single receiver and may specifically provide improved performance in a communication system using Multiple- Transmit-Multiple-Receive (MTMR) antenna concepts exploiting Multiple-Input-Multiple-Output (MIMO) channel properties. The invention may allow transmissions to be made using low complexity preambles which are substantially orthogonal between different antennas while retaining desirable properties for each individual preamble, such as e.g. a low time domain peak-to-average- power ratio (PAPR) .

The invention may for example be used in an expected IEEE 801. Hn standardised communication system and may provide increased performance and/or reduced complexity such as improved timing synchronisation, signal power estimation, frequency estimation, phase estimation and/or channel estimation.

The invention may in some embodiments allow individual design of a first section which is suited for efficient but low complexity determination of initial characteristics, such as coarse timing and amplitude estimations, and a second section which is suitable for a more exacting determination of more critical characteristics, such as a channel estimate. The approach may allow individual optimisation of preamble properties suitable for processing in one domain (e.g. the time domain) for the first section and in another domain (e.g. the frequency domain) for the second section. Hence, each section may be individually optimised, and for example, the weighting of each repetition in the second domain may be such that orthogonality between antennas is achieved in the frequency domain whereas the data sequences of the first section may independently be selected to provide substantial orthogonality in the time domain. This may for example allow a low complexity hardware implementation e.g. using simple time-domain correlators for the first section combined with frequency domain processing of the second section.

It will be appreciated that the word pre-amble is used herein in the broad interpretation of including any training data or known data irrespective of how such data is distributed in a transmission. Thus, the term preamble includes e.g. mid-ambles or post-ambles or any other location or distribution of training data in a transmission. The term preamble is thus used equivalently to the terms training data or known data. The first and second data sequences comprise known training data and in particular comprise known training data symbol (s) .

According to an optional feature of the invention, the coefficient sequence for each antenna comprises a different row of a Walsh-Hadamard matrix. This may improve performance while maintaining a low complexity and may result in substantially orthogonal second sections of the individual preambles .

According to an optional feature of the invention, the first data sequence comprises symbols belonging to a limited alphabet in the time domain. This may facilitate processing in the time domain and may reduce receiver complexity. The limited alphabet will comprise fewer symbols than the alphabet which corresponds to the sampling granularity of (user) data symbols.

According to an optional feature of the invention, the limited alphabet comprises only the symbols [1,-1, i,- i,0] . This may substantially reduce complexity and may in particular allow for time domain correlation by a receiver without requiring high complexity multiplications to be performed.

According to an optional feature of the invention, the first data sequence for each individual preamble of each antenna is a different time domain sequence from the group consisting of time shifted and phase shifted sequences of the set comprising:

1) 1 1 -1 1 -1 -1 1 -1 -1 0 1 -1 -1 -1 1 1,

2) 0 -1 -1 1 -1 -1 1 -1 1 1 1 1 -1 -1 -1 1,

3) i i i i i -i 1 -1 -1 1 0 1 -1 -1 1 -i, and

4) -i -i -i -i -i i 1 -1 -1 1 0 1 -1 -1 1 i.

The time shift could e.g. be a zero time shift or could be a cyclic time shift. The phase shift may be a multiplication of the coefficient by a complex number having a modulus of one, such as in particular a multiplication by a coefficient comprised in the set [-1, i, -i] . The sequences of this feature provide particularly advantageous performance and in particular provide attractive autocorrelation, cross- correlation and spectral properties for many embodiments. According to an optional feature of the invention, the second data sequence comprises symbols belonging to a limited alphabet in the frequency domain. This may facilitate processing in the frequency domain and may reduce receiver complexity. The limited alphabet will comprise fewer symbols than the alphabet which corresponds to the sampling granularity of (user) data symbols .

In some embodiments, the first data sequence comprises symbols belonging to a limited alphabet in the time domain and the second data sequence comprises symbols belonging to a limited alphabet in the frequency domain. This may allow facilitated processing of the first section in the time domain at the same time as facilitating processing in the frequency domain for the second section. Thus, the complexity of the different processes using the different sections of the preamble may individually be reduced resulting in a more efficient and lower complexity operation.

According to an optional feature of the invention, the limited alphabet comprises only the symbols [1,-1,0] . This may substantially reduce complexity and may in particular allow for frequency domain processing by a receiver without requiring high complexity multiplications to be performed.

According to an optional feature of the invention, the second data sequence comprises symbols corresponding to a 56 sub-carrier, optionally phase shifted, frequency domain data sequence comprising a first symbol sequence of: {-1, -1, -1, 1, -1, 1, -1, -1, -1, 1, 1, 1, 1, -I, -1,

1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, 1, -1} and a second symbol sequence of:

{-1, -1, -1, 1, -1, 1, -1, -1, -1, 1, 1, 1, 1, -1, -1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, 1, -1}.

The phase shift may be a multiplication of the coefficient by a complex number having a modulus of one, such as in particular a multiplication by a coefficient comprised in the set [-1, i, -i] . The phase shift may specifically be a zero phase shift. The sequence provides particularly advantageous performance and in particular provides attractive time domain peak-to-average-power ratio (PAPR) properties. In particular, improved performance may be achieved in an OFDM transmitter using a Discrete Fourier Transform (DFT) such as a 64 point Inverse Fast Fourier Transform (IFFT) .

According to an optional feature of the invention, the second data sequence comprises symbols corresponding to a 52 sub-carrier, optionally phase shifted, frequency domain data sequence comprising a first symbol sequence of:

{-1, 1, -1, 1, -1, -1, 1, -1, 1, -1, -1, 1, 1, 1, -1, 1, 1, 1, 1, 1, -1, 1, 1, 1, 1, 1} and a second symbol sequence of:

{-1, 1, -1, -1, 1, 1, -1, -1, -1, 1, 1, -1, 1, 1, 1, -

1, 1, -1, -1, -1, -1, -1, 1, 1, 1, -1}.

The phase shift may be a multiplication of the coefficient by a complex number having a modulus of one, such as in particular a multiplication by a coefficient comprised in the set [-1, i, -i] . The phase shift may be a zero phase shift. The sequence provides particularly advantageous performance and in particular provides attractive time domain peak-to-average-power ratio (PAPR) properties. In particular, improved performance may be achieved in an OFDM transmitter using Discrete Fourier Transform (DFT) such as a a 64 point Inverse Fast Fourier Transform (IFFT) and requiring a substantial guard band. The sequence may be compatible with some receive/ transmit processing of systems such as IEEE 801.11a and IEEE 801. Hg systems wherein 52 sub-carriers out of 64 sub-carriers are used for preambles.

According to an optional feature of the invention, the first section comprises repetitions of the first data sequence. At least one repetition may in some embodiments be multiplied by a complex coefficient. This may improve performance in some embodiments and provide for a low complexity implementation. The complex coefficient may in particular.be a coefficient comprised in the set [-1, i, -i] .

According to an optional feature of the invention, the second section comprises a guard interval between the repetitions. The guard interval may for example be a cyclic repetition of the last samples of the first data sequence in the time domain. The guard interval may ensure that the symbols are cyclically convolved by the channel. As a consequence, this property may allow the use of e.g. a Walsh Hadamard weighting - e.g. if one guard interval is added prior to each repetition.

According to an optional feature of the invention, the transmitter is an OFDM (Orthogonal Frequency Division Multiplex) transmitter. The invention may provide particularly advantageous performance in a system using OFDM modulation such as for example an IEEE 801.Hn wireless network system.

According to an optional feature of the invention, the means for transmitting is operable to modulate the individual preambles using a 64 point Discrete Fourier Transform (DFT) such as a 64 point IFFT (Inverse Fast Fourier Transform) . The DFT may be a forwards DFT or an Inverse Discrete Fourier Transform (IDFT) . This may allow improved performance, low complexity and/or compatibility with many existing communication systems, such as for example an IEEE 801. Hn wireless network system.

According to an optional feature of the invention, the means for transmitting is operable to modulate user data using a 128 point Discrete Fourier Transform (DFT) such as a 128 point IFFT (Inverse Fast Fourier Transform) . The DFT may be a forwards DFT or an Inverse Discrete Fourier Transform (IDFT) . The invention may in particular allow a system wherein a preamble, for example modulated using a 64 point IFFT in a given bandwidth, may be used for e.g. channel estimation for user data modulated by a 128 point IFFT in the same bandwidth.

According to an optional feature of the invention, the coefficient sequence for each antenna is such that the second sections of the individual preambles are orthogonal in the frequency domain. This may provide improved performance. According to an optional feature of the invention, the plurality of antennas is four antennas, the second section comprises four repetitions and the coefficient sequence for each antenna is a different row of a four by four Walsh-Hadamard matrix. This may provide advantageous performance and may for example be compatible with IEEE 801.Hn.

According to an optional feature of the invention, the plurality of antennas is four antennas, the second section comprises four repetitions and the coefficient sequence for each antenna is a repeated different row of a two by two Walsh-Hadamard matrix. This may provide advantageous performance and may for example be compatible with IEEE 801.Hn.

According to a second aspect of the invention, there is provided a receiver comprising: means for receiving a signal transmitted as sub-signals from a plurality of antennas, the signal comprising a preamble set comprising an individual preamble for each antenna; each of the individual preambles comprising a first section and a second section wherein the first section comprises a first data sequence comprising at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprises repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna; and means for performing an initial acquisition in response to the first sections; and means for determining a channel estimate in response to the second sections. According to an optional feature of the invention the initial acquisition comprises timing synchronisation.

According to an optional feature of the invention, the means for determining the channel estimate is operable to determine the channel estimate by frequency domain processing. In some embodiments, time domain processing may be performed in addition to the frequency domain processing.

According to a third aspect of the invention, there is provided a preamble set comprising an individual preamble for each of a plurality of antennas; wherein each of the individual preambles comprises a first section and a second section, the first section comprising a first data sequence comprising at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprising repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna.

According to a fourth aspect of the invention, there is provided a method of transmitting a signal as sub-signals from a plurality of antennas, the method comprising: means for generating a preamble set comprising an individual preamble for each antenna, each of the individual preambles for each antenna comprising a first section and a second section wherein the first section comprises a first data sequence comprising at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprises repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna; and transmitting the signal as sub- signals from the plurality of antennas using the preamble set.

According to a fifth aspect of the invention, there is provided a method of receiving a signal transmitted as sub-signals from a plurality of antennas, the method comprising: receiving the signal, the signal comprising a preamble set comprising an individual preamble for each antenna; each of the individual preambles comprising a first section and a second section wherein the first section comprises a first data sequence comprising at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprises repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna; performing an initial acquisition in response to the first sections; and determining a channel estimate in response to the second sections.

These and other aspects, features and advantages of the invention will be apparent from and elucidated with reference to the embodiment (s) described hereinafter.

Brief Description of the Drawings Embodiments of the invention will be described, by way of example only, with reference to the drawings, in which

FIG. 1 illustrates a communication system incorporating some embodiments of the invention;

FIG. 2 illustrates an example of a preamble set in accordance with an embodiment of the invention;

FIG. 3 illustrates a spectral property of an example sequence of a preamble set in accordance with an embodiment of the invention;

FIG. 4 illustrates a spectral property of an example sequence of a preamble set in accordance with an embodiment of the invention;

FIG. 5 illustrates a spectral property of an example sequence of a preamble set in accordance with an embodiment of the invention;-

FIG. 6 illustrates a spectral property of an example sequence of a preamble set in accordance with an embodiment of the invention;

FIG. 7 illustrates a cyclic correlation property of an example sequence of a preamble set in accordance with an embodiment of the invention;

FIG. 8 illustrates a cyclic correlation property of an example sequence of a preamble set in accordance with an embodiment of the invention; FIG. 9 illustrates a cyclic correlation property of an example sequence of a preamble set in accordance with an embodiment of the invention;

FIG. 10 illustrates a cyclic correlation property of an example sequence of a preamble set in accordance with an embodiment of the invention;

FIG. 11 illustrates an example of a first section of preambles for four antennas in accordance with an embodiment of the invention;

FIG. 12 illustrates a Maximum Square Error (MSE) as a result of a typical channel estimation algorithm for different preambles;

FIG. 13 illustrates a Maximum Square Error (MSE) for different preambles having different numbers of non-zero sub-carriers;

FIG. 14 illustrates a time domain LTS preamble word in accordance with an embodiment of the invention; and

FIG. 15 illustrates an example of a preamble set in accordance with an embodiment of the invention.

Detailed Description of Embodiments of the Invention

The following description focuses on embodiments of the invention applicable to an OFDM communication system and in particular to an expected standardisation of an IEEE 801. Hn communication system. However, it will be appreciated that the invention is not limited to this application but may be applied to many other communication systems and modulation techniques.

5 FIG. 1 illustrates a communication system incorporating some embodiments of the invention. The communication system 100 comprises a transmitter 101 and a receiver 103. The transmitter 101 comprises a transmit processor 105 which receives user data to be transmitted and

10 processes this for transmission over the air interface. In the particular embodiment, the transmit processor 105 is capable of generating channel data to be transmitted for a plurality of antennas as known to the person skilled in the art from communication systems such as for

15 example IEEE 801. Hn. The transmit processor 105 thus comprises the functionality required or desired for interleaving and coding user data as well as for allocating this to the individual transmit antennas.

20 The transmit processor 105 is coupled to a first and second preamble processor 107, 109 which are operable to insert an individual preamble into the data stream for each transmit antenna. The first and second preamble processor 107, 109 are coupled to a preamble generator

25 111 which generates a preamble set comprising an individual preamble for each transmit antenna of the transmitter 101.

The first preamble processor 107 is coupled to a first 30 transmit unit 113 which is operable to transmit the data sequence from a first antenna 115 of the transmitter 101. Similarly, the second preamble processor 109 is coupled to a second transmit unit 117 which is operable to transmit the data sequence from a second antenna 119 of the transmitter 101. The first and second transmit units 113, 117 are capable of transmitting individual sub- signals to the receiver 103. The transmissions use the same transmission parameters, such as the same modulation scheme, bandwidth and carrier frequency. Thus, in order to receive the transmitted signal, the receiver 103 is capable of performing a joint detection as known from existing systems using Multiple-Transmit-Multiple-Receive (MTMR) techniques.

In order to perform such reception, the receiver must be able to determine various properties of the received signal and the individual communication channels between the individual transmit and receive antennas. For this purpose, the receiver 103 uses information of the known data of the preambles. It is thus important that the generated preambles are suited for estimation of these properties.

In the example of FIG. 1, the transmitter is an OFDM (Orthogonal Frequency Division Multiplex) transmitter and the first and second transmit units 113, 117 comprise functionality for performing an FFT on the received data to generate the appropriate sub-carrier symbols, as will be well known to the person skilled in the art. In accordance with communication systems of the specific example, the first and second transmit units 113,117 perform a 64 point IFFT. Furthermore, in some embodiments, the first and second transmit units 113, 117 may use a 64 point IFFT for the preamble data but a 128 point IFFT (in the same frequency bandwidth) for user data. The receiver 103 of FIG. 1 comprises a first and second receiver front end 121, 123 each of which is coupled to a receive antenna 125, 127. The first and second receiver front ends 121, 123 are operable to filter, amplify, down-convert and digitize the received signal. The first^' and second receiver front ends 121, 123 are coupled to a preamble extractor 129 which is operable to extract the data samples for received preambles . The preamble extractor 129 is coupled to an initial acquisition processor 131 which in the described example is operable to determine a coarse timing estimate and amplitude estimate based on the preamble data received from the preamble extractor 129. The initial acquisition processor 131 is in the example coupled to the first and second receiver front ends 121, 123 and is arranged to set a gain level and a sample timing of the front ends 121, 123 in response to the generated estimates .

The preamble extractor 129 is furthermore coupled to a channel estimator 133 which in the example is operable to generate a channel estimate for each channel between a transmit antenna 115, 119 and a receive antenna 125, 127 in response to the preamble data received from the preamble extractor 129.

The channel estimator 133 and the preamble extractor 129 are coupled to a receive processor 135. The receive processor 135 determines the information symbols of the received signals in response to the data samples received from the preamble extractor 129 and the channel estimates received from the channel estimator 133. In particular, the receive processor 135 performs an FFT corresponding to the IFFT of the transmitter 101 and performs a joint detection of the data taking into account the individual channel estimates from each of the transmit antennas 115, 119 to each of the receive antennas 125, 127.

In the example of FIG. 1 the preamble generator 111 generates a preamble set wherein each individual preamble comprises a first section and a second section. The first section is in the example shorter than the second section and is used by the initial acquisition processor 131 to determine the timing and amplitude estimate. The longer second section is, in the example, used by the channel estimator 133 to determine the channel estimates. The first section may specifically correspond to a Short Training Symbol (STS) and the second section to a Long Training Symbol (LTS) of an IEEE 801. Hn system.

The preamble generator 111 generates a first section (an STS) which comprises a first data sequence which has at least one sequence which is different for each antenna. The sequence is selected from a predetermined set of preamble sequences and a different sequence of the set is selected for each antenna. Specifically, the preamble generator 111 generates an STS which has a number of repetitions of the sequence, such as e.g. ten repetitions of a 16 sample sequence.

In some embodiments, improved performance may be obtained by multiplying one or more of the repetitions by different coefficients. For example, some of the repetitions may be inverted (multiplied by -1) . In some embodiments, this may improve detection and correlation at the receiver. Furthermore, the preamble generator 111 may specifically generate a second section (an LTS) comprising repetitions of a predetermined second data sequence where each repetition is weighted by a coefficient of a coefficient sequence that is different for each antenna. Thus, the same second data sequence, henceforth referred to as the LTS training word, may be used in the preambles of all antennas. Furthermore, the same number of repetitions may be used for all antennas. However, the repetitions are weighted differently for the different antennas and in accordance with the described embodiments the weighting varies such that the second sections of the preambles are substantially orthogonal in the frequency domain.

The weighting may specifically be performed by using a coefficient sequence which for each antenna comprises a different row of a Walsh-Hadamard matrix. For example, the repetitions of a first antenna are multiplied by the coefficients of the first row of a Walsh-Hadamard matrix, the repetitions of a second antenna are multiplied by the coefficients of the second row of the same Walsh-Hadamard matrix etc.

A property of a Walsh-Hadamard matrix W is that it provides an orthogonal basis. All elements of W are chosen from {-1, 1}. I.e. the Hermitian of W multiplied by W results in the identity matrix multiplied by the constant factor ^VN' assuming that W is of dimension N-by- N. Accordingly, by multiplying the repetitions by the coefficients of a Walsh-Hadamard matrix, an orthogonality in the frequency domain between the second sections of the preambles is achieved. FIG. 2 illustrates a specific example of a preamble set which may be generated for a two antenna transmitter. In the example, the LTS training word is repeated twice in the second section. The LTS repetitions are in this example multiplied by the first and second row respectively of the 2x2 Walsh Hadamard matrix

In accordance with some embodiments of the invention, the symbols of the first sequence belong to a limited alphabet in the time domain. The limited alphabet may be an alphabet which comprises only zero and some discrete points on the complex unity circle.

Specifically, the STS sequence may comprise only symbols corresponding to the zero value and QPSK data values, i.e. the symbols are selected from the set of

[1,-1, i,-i,0] .

This restriction for the first data sequence results in a time domain signal which is highly suitable for time domain processing. In particular, the initial acquisition processor 131 of the receiver 103 may perform a time domain correlation of the first sections of the received preambles to a local replica without requiring high complexity multiplications. Rather the required multiplications may be achieved by simple sign inversions and/or switching of real and imaginary values. Furthermore, the first sequence may be selected to provide a high degree of autocorrelation and cross correlation to other preambles of the set thereby improving detection performance.

Accordingly, a preamble being particularly well suited to the operation of the initial acquisition may be achieved.

In accordance with some embodiments of the invention, the symbols of the second sequence alternatively and/or additionally belong to a limited alphabet in the frequency domain.

The limited alphabet may be an alphabet which comprises only zero and some discrete points on the complex unity circle. Specifically, the second data sequence may be determined in order to comprise only frequency domain symbols corresponding to the zero value and QPSK data values, i.e. the symbols are selected from the set of

[1,-1, i,-i,0] .

or may use only BPSK data values, i.e. the frequency domain symbols are selected from the set of

[1,-1,0] .

The corresponding LTS training word may be determined e.g. by an inverse FFT of the selected frequency domain training word. This restriction for the second data sequence results in a frequency domain preamble section which is highly- suitable for frequency domain processing. In particular, the channel estimator of the receiver 131 may perform channel estimation partially or wholly in the frequency domain with reduced complexity.

Specifically, frequency domain sub-carrier channel estimates may be determined from a simple division of the received signal by a local replica of the second section. Hence, using the limited alphabet set of the example, a required division may be achieved by a simple sign inversion of the received symbols in each sub-carrier.

Furthermore, even if only partial frequency domain channel estimation is used, the initial frequency processing may be facilitated.

The described preamble design may in particular lead to a preamble set providing high performance in an MTMR system while allowing the processing of initial acquisition and channel estimation to be individually improved taking into account whether processing is performed in the time domain or the frequency domain.

The predetermined set of preamble sequences used for the first sequence, in the following referred to as the STS sequence, is preferably determined to have a high auto¬ correlation, cross correlation and desirable spectral properties.

In the following a design process for determining a suitable set of preamble sequences is described with specific reference to an application suitable for an expected standardised IEEE 801. Hn system.

In the example, a set of four sequences having a length of 16 samples are determined by the following quasi-exhaustive approach:

1. The goal is to look for four sequences A,B,C,D with 16 samples each. One sample of the sequence is chosen to be the zero-element (this improves the spectral properties) . Since a multiplication of a sequence by a constant factor (of module 1) does not change the correlation properties and a cyclic shift also keeps the correlation properties unchanged, it is possible to reduce the number of combinations to be evaluated by imposing two elements per sequence:

We choose to fix the position of the zero element and we choose to fix one element being ^xl' (which is part of the considered alphabet) . For the remaining fields, all combinations of the {1,-1,1,- i} alphabet are tested. It can be shown that the sequences missed by imposing the ^λl' element can be achieved by any sequence that is considered multiplied by a suitable element of the alphabet. Since this does not change the correlation properties, the corresponding sequences must not be considered separately. 2. Define Ψ to be the ensemble of all possible permutations of the sequences for A,B,C,D: AeΨ,5eΨ,CeΨ,DeΨ . The first step is to define a spectral cost function for each of the permutation:

J^STS, ^SPECTRUM _{= max (} | _seq_f (_Out_band_frequencies) |^Λ2) + max( ( I seq_f (in_band_frequencies) | ^A2- mean( | seq_f (in_band_frequencies) | ^A2) | )

seq_f is the 16-samples STS sequence in the frequency domain. For the 4-element alphabet, 4^Λ14 = 268.435.456 permutations are possible. All corresponding values of the cost function are stored. 3. For all sequences, eliminate the ones with bad auto-correlation properties, i.e. all correlations of the sequence with its cyclically shifted version should be below a given threshold. For the optimization this threshold was fixed to ^λ2' . 4. Sort all remaining sequences corresponding to their cost function value J^{STS/ SPECTRUM}. Among all permutations, a set Θ of the best ones is selected (i.e. a set of the ones where the cost function JSTS, SPECT^RUM _{is as c]}__{ose to zero as} possible₎ . In the optimization procedure, the 256 best ones have been considered.

5. Compare the correlation properties of (A,B,C,D)eΘ*and choose the ensemble which leads to the minimum cost function value of j^STS<CORR _:

∑ 2 Σ Σ CyclicShifl,_ι {STS^"#s₁}®CyclicShift_u {STS^"#s₂} s,=i i,=i ;,=o ;,=o

Hereby, JV_OTS=4 is the number of STS sequences , L₃Ts=IS is the length of the STS sequence in number of time domain samples. The operator ®indicates the correlation operation. β. The resulting four sequences (A,B,C,D) are finally selected.

Following this optimisation, the following set of sequences has been found to have particularly advantageous properties :

#1 1 1 -1 1 - 1 -1 1 -1 0 1 -1 -1 1 1

#2 0 - 1 -1 1 - 1 -1 1 - 1 1 1 1 1 -1 - 1 - 1 1

#3 i i i i i -i 1 - 1 -1 1 0 1 -1 _2_^ 1 -i

#4 -i -i -i -i i 1 -1 1 0 1 -1 -1 1 i

As the sequences keep the same correlation properties if any of the sequences is multiplied by a constant factor of module 1, the first sequence for each antenna may be selected as one of the above or may be a corresponding sequence multiplied by a complex coefficient with modulus 1 (i.e. by a phase shift) . Moreover, since all sequences can be cyclically shifted in the same way without changing the correlation properties (although the correlation peak may be shifted) , the selected sequences may correspond to time shifted versions of the above sequences.

The spectral properties of the four sequences are shown in FIGs. 3 to 6. As can be seen, the out-of-band radiation for an IEEE 801.Hn system specification is kept low. This avoids a loss of preamble energy due to low-pass filtering in the transmitter and/or receiver. The in-band signals cover the whole band, preventing that selective fading degrades performance disproportionally.

The cyclic correlation properties of the four sequences are shown in FIGs. 7 to 10. As can be seen, all four sequences are well de-correlated. It can furthermore be shown that the cross-correlation between the sequences is very low.

The characteristics of the identified sequences are particularly suitable for determining amplitude and timing estimates. In particular, the correlation properties can be exploited for both timing synchronization and amplitude estimates for automatic gain control.

It will be appreciated that in some embodiments cyclically shifted versions of the identified sets may be used. Similarly, it will be appreciated that any sequence corresponding to the identified sequences and obtained by multiplying any of these sequences by any complex value can be used. The complex value will be identical within one sequence but it can be different from one sequence to another.

The first section of an individual preamble is preferably obtained by a weighted combination of the sequences of the predetermined set. FIG. 11 illustrates an example of a first section of preambles for four antennas. In the example, the first identified sequence is denoted nSTSl, the second identified sequence nSTS2, the third identified sequence nSTS3 and the fourth identified sequence nSTS4. In the example, the chosen weights are for any sequence nSTSx chosen from the set {nSTSl, nSTS2, nSTS3, nSTS4}:

÷nSTSx, +nSTSx, -nSTSx, ÷nSTSx, +nSTSx, -nSTSx, +nSTSx, - nSTSx, ÷nSTSx, -nSTSx

i.e. the weights are [1, 1,-1, 1, 1, -1, 1, -1, 1, -I] .

This may provide improved performance.

It will be appreciated that if less than four transmit antennas are used, any distinct nSTSx sequences may be chosen for transmission from the set of {nSTSl, nSTS2, nSTS3, nSTS4}.

In the following, a design process for determining a suitable LTS (second sections of the preambles) is described with specific reference to an application suitable for an expected standardised IEEE 801.Hn system.

In the specific example, the LTS is selected to have a 20 MHz bandwidth. This is suitable for IEEE 801.Hn systems wherein the channel bandwidth is expected to be standardised as 20 MHZ. Furthermore, IEEE 801. Hn provides for a modulation mode for user data using 64 point IFFTs and an optional mode using a 128 point IFFT. Selecting a single LTS word in a 20MHz bandwidth allows the LTS word to be used for channel estimation for both the mandatory OFDM modulation (0.8μs guard interval duration, 64-point IFFT) and also for the optional OFDM modulation (l.βμs guard interval duration, 128-point IFFT) of expected IEEE 801.Hn systems . The optional modulation is considered more robust when the time domain length of the channel (including transmit and receive filtering effects) is potentially superior to 0.8μs, e.g. in outdoor environments. The LTS word suggested herein is based on the 64-point IFFT based OFDM modulation, with an extended guard interval of l.βμs to cope with long channels. In that configuration, channel estimates for optional modes based on 128-point IFFT can be determined by using an interpolation based algorithm.

The exemplary design process for the LTS word for a transmitter having N_τx = 2, 3 or 4 transmit antennas is based on the minimization of the Mean Square Error (MSE) in the frequency domain.

When channel estimation for a channel of Nt coefficients is performed in the time domain, the MSE can be expressed as MSE=σ²(i®F_n)pinv(∑[i®F_n]) where

• Pl denotes the number of data and pilots of the OFDM modulation: Pl = 52 for the mandatory modulation, and Pl = 112 for the optional one. P2 is the number of non-zero sub carriers in the LTS word, which can be different from 52 if we want to increase the number of observations .

• F_p1XS the Pl x Nt truncated Fourier matrix, where the Pl rows of this matrix correspond to the Pl sub carriers that are used among the 64 or 128 sub carriers of the modulation, and where we only keep the first Nt columns of the Fourier matrix. SimilarlyF_p2denotes the P2 x Nt truncated 64- point Fourier matrix.

• The matrixXis a block matrix of size (Ns x P2) x

(N_τx x P2) , where Ns is the number of OFDM symbols in the Long Training Sequence. The (s, t) block of this matrix, 1 ≤ s ≤ Ns and 1 ≤ t ≤ N_Tχ, is a P2 x P2 diagonal matrix whose {p,p) element is the p^th non-zero symbol of the s^th OFDM symbol transmitted on the t^th antenna. • J is the identity matrix of size N_Tχ x N_Tχ/ andσ²the variance of the additive Gaussian noise.

• ^λ ® ' is the Kronecker multiplication and ^λpinv(X) ' denotes the Moore-Penrose pseudo-inverse matrix of ^λX' , i.e. '

. x"denotes the hermitian matrix of the matrix X.

For example, let's consider the design of a LTS with two OFDM symbols for two transmit antennas.

• If we choose an orthogonal design based on the Walsh Hadamard matrix of size 2x2, we get

X= where X is a diagonal matrix whose

elements are the symbols of the training symbol normalized so that the transmit power does not depend on the number of antennas.

• Using a cyclic shift to separate in time the channels, the matrix X can be expressed as

The matrix X is defined as

previously, whereas the matrix Φ is a diagonal matrix whose ip,p) element is exp(-j2τdD/64) , where 1 is the rank of the subcarrier among the 64 subcarriers, and D*50ns the value in time of the delay, e.g. 1600ns.

As previously mentioned, low cost channel estimation algorithms can be implemented at the receiver when the second section is orthogonal in the frequency domain, e.g. when using a Walsh Hadamard structure for the coefficients. Indeed, in that configuration, channel estimation can be performed directly in the frequency domain for the mandatory OFDM modulation; however interpolation is still required for the optional modes based on 128-point IFFT.

FIG. 12 compares simulated results for the Maximum Square Error (MSE) for the following three scenarios:

1. A cyclic shift based LTS with channel estimation in the time domain in accordance with current proposals for IEEE 801. Hn.;

2. A Walsh-Hadamard based LTS with channel estimation in the time domain in accordance with some embodiments of the current invention; and

3. A Walsh-Hadamard based LTS with channel estimation in the frequency domain in accordance with some embodiments of the current invention.

As clearly illustrated the performance of the Walsh- Hadamard based LTS is significantly better than the cyclic shift based LTS for a wide range of channel coefficients (taps of the channel estimate) . Furthermore, when using a Walsh-Hadamard based LTS and performing the channel estimation in the time domain this may significantly improve the accuracy of the channel estimates.

FIG. 12 illustrates an example wherein the LTS preamble word corresponds to 52 non-zero sub-carriers in the frequency domain. The performance of the time domain estimators may in some embodiments be improved by increasing the number of non-zero sub carriers for the LTS preamble word. FIG. 13 illustrates the MSE for different numbers of non-zero sub-carriers.

As can be seen, significant performance improvement is achieved for up to 56 non-zero sub-carriers whereas the improvement thereafter reduces significantly. Hence, in order to facilitate filter designs and reduce the impact of non-ideal filtering in the transmitter and receiver, it is desirable in some embodiments to use a LTS preamble word which results in 56 non-zero sub-carriers.

The mean square errors are independent of the symbols of the LTS preamble word. Accordingly, it is desirable to select an LTS preamble word which results in advantageous properties. In particular, it may be advantageous in many embodiments to select an LTS preamble word which has the desired characteristics in the frequency domain, i.e. the desired number of non¬ zero sub carriers and using only the symbols of the limited alphabet. Performing a search over all possible values for a 56 non-zero sub-carrier frequency representation of an LTS preamble word using only symbols of the alphabet [1,-1,0] and selecting the word having the lowest Peak-to Average Power Ratio (PAPR) results in the frequency domain LPT preamble word of

LTS(#-28...#+28) = {-1, 1, -1, 1, 1, 1, 1, -1, -1, 1, 1, 1, -1, 1, 1, -1, -1, -1, -1, 1, 1, -1, 1, 1, -1, 1, - 1, -1, 0, -1, -1, -1, 1, -1, 1, -1, -1, -1, 1, 1, 1, 1, -1, -1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, 1, -1}

i.e. the frequency domain LPT preamble word of will comprise a first symbol sequence of {-1, -1, -1, 1, -1, 1, -1, -1, -1, 1, 1, 1, 1, -1, -1,

1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, 1, -1} and a second symbol sequence of

{-1, -1, -1, 1, -1, 1, -1, -1, -1, 1, 1, 1, 1, -1, -1,

1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, 1, -1}.

It will be appreciated that any (frequency domain) phase shifted version of this preamble word may equally be used.

FIG. 14 illustrates the corresponding time domain LTS preamble word. This word has a low time domain PAPR of 3.408dB. As a comparison, the preamble specified in IEEE 802.11a has a PAPR of 3.964dB. The relative gain is approx. 0.6dB which typically will keep the preamble in the linear region of the PA and thus lead to improved channel estimation results. Performing a search over all possible values for a 52 non-zero sub-carrier frequency representation of an LTS preamble word using only symbols of the alphabet [1,-1,0] and selecting the word having the lowest Peak-to Average Power Ratio (PAPR) results in the frequency domain LPT preamble word of

LTS(#-2β...#+26) = {-1, 1, -1, 1, -1, -1, 1, -1, 1, -1,

-1, 1, 1, 1, -1, 1, 1, 1, 1, 1, -1, 1, 1, 1, 1, 1, 0, -1, 1, -1, -1, 1, 1, -1, -1, -1, 1, 1, -1, 1, 1, 1, -

1, 1, -1, -1, -1, -1, -1, 1, 1, 1, -1} i.e. the frequency domain LPT preamble word of will comprise a first symbol sequence of ,

{-1, 1, -1, 1, -1, -1, 1, -1, 1, -1, -1, 1, 1, 1, -1, 1, 1, 1, 1, 1, -1, 1, 1, 1, 1, 1} and a second symbol sequence of

{-1, 1, -1, -1, 1, 1, -1, -1, -1, 1, 1, -1, 1, 1, 1, -

1, 1, -1, -1, -1, -1, -1, 1, 1, 1, -1}.

This word has a lower PAPR than the IEEE 802.11a preamble. For comparison, the identified sequence has a PAPR of 3.346dB, which represents a relative gain of approx. 0.6dB compared to 802.11a preamble.

In some specific embodiments, a preamble for four antennas may be used wherein the second section comprises four repetitions of the LTS preamble word and where the coefficient sequence for each antenna is a different row of a four by four Walsh-Hadamard matrix. In some other specific embodiments, a preamble for two antennas may be used wherein the second section comprises four repetitions and the coefficient sequence for each antenna is a repeated different row of a two by two Walsh-Hadamard matrix.

Specifically, for an expected IEEE 801. Hn system an LTS in 20MHz bandwidth may be based on the identified low PAPR preamble word to be sent on 56 sub carriers, using 64-point IFFT. The length of the guard interval (l.βμs) may be the same as for the optional OFDM modulation proposed to cope with long channels.

The transmit scheme of the multiple antennas is in this example based on a Walsh-Hadamard structure in order to provide channel estimates with high accuracy (time domain estimation) and/or from a low cost algorithm (frequency domain estimation) . The length of the LTS is in the example constant whatever the number of antennas is, and corresponds to the transmission of 4 OFDM symbols on each antenna.

FIG. 15 illustrates such an example wherein the LTS preamble word is denoted by S.

It will be appreciated that the above description for clarity has described embodiments of the invention with reference to different functional units and processors. However, it will be apparent that any suitable distribution of functionality between different functional units or processors may be used without detracting from the invention. For example, functionality illustrated to be performed by separate processors or controllers may be performed by the same processor or controllers. Hence, references to specific functional units are only to be seen as references to suitable means for providing the described functionality rather than indicative of a strict logical or physical structure or organization.

The invention can be implemented in any suitable form including hardware, software, firmware or any combination of these. The invention may optionally be implemented at least partly as computer software running on one or more data processors and/or digital signal processors. The elements and components of an embodiment of the invention may be physically, functionally and logically implemented in any suitable way. Indeed the functionality may be implemented in a single unit, in a plurality of units or as part of other functional units. As such, the, invention may be implemented in a single unit or may be physically and functionally distributed between different units and processors.

Although the present invention has been described in connection with some embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the scope of the present invention is limited only by the accompanying claims. Additionally, although a feature may appear to be described in connection with particular embodiments, one skilled in the art would recognize that various features of the described embodiments may be combined in accordance with the invention. In the claims, the term comprising does not exclude the presence of other elements or steps . Furthermore, although individually listed, a plurality of means, elements or method steps may be implemented by e.g. a single unit or processor. Additionally, although individual features may be included in different claims, these may possibly be advantageously combined, and the inclusion in different claims does not imply that a combination of features is not feasible and/or advantageous . Also the inclusion of a feature in one category of claims does not imply a limitation to this category but rather indicates that the feature is equally applicable to other claim categories as appropriate. Furthermore, the order of features in the claims do not imply any specific order in which the features must be worked and in particular the order of individual steps in a method claim does not imply that the steps must be performed in this order. Rather, the steps may be performed in any suitable order. In addition, singular references do not exclude a plurality. Thus references to "a", "an", "first", "second" etc do not preclude a plurality.

Claims

1. A transmitter comprising: means for transmitting a signal as sub-signals from a plurality of antennas using a preamble set comprising an individual preamble for each antenna; and means for generating the preamble set with each of the individual preambles comprising a first section and a second section wherein the first section comprises a first data sequence including at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprises repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna.

2. The transmitter of claim 1 wherein the coefficient sequence for each antenna comprises a different row of a Walsh-Hadamard matrix.

3. The transmitter of any previous claim wherein the first data sequence comprises symbols belonging to a limited alphabet in the time domain.

4. The transmitter of claim 3 wherein the limited alphabet comprises only the symbols [1,-1, i,-i,0] .

5. The transmitter claimed in claim 4 wherein the first data sequence for each individual preamble of each antenna is a different time domain sequence from the group consisting of time shifted and phase shifted sequences of the set comprising:

5) 1 1 -1 1 -1 -1 1 -1 -1 0 1 -1 -1 -1 1 1,

6) 0 -1 -1 1 -1 -1 1 -1 1 1 1 1 -1 -1 -1 1, 7) i i i i i -i 1 -1 -1 1 0' 1 -1 -1 1 -i, and 8) -i -i -i -i -i i 1 -1 -1 1 0 1 -1 -1 1 i.

6. The transmitter claimed in any of the previous claims wherein the second data sequence comprises symbols belonging to a limited alphabet in the frequency domain.

7. The transmitter claimed in claim 6 wherein the limited alphabet comprises only the symbols [1,-1,0] .

8. The transmitter claimed in claim 7 wherein the second data sequence comprises symbols corresponding to a 56 sub-carrier phase shifted frequency domain data sequence comprising a first symbol sequence of: {-1, -1, -1, 1, -1, 1, -1, -1, -1, 1, 1, 1, 1, -1, -1,

1, 1, 1, 1, 1, 1, I₇ 1, 1, -1, -1, 1, -1} and a second symbol sequence of :

{-1, -1, -1, 1, -1, 1, -1, -1, -1, 1, 1, 1, 1, -1, -1,

1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, 1, -1}.

9. The transmitter claimed in claim 7 wherein the second data sequence comprises symbols corresponding to a 52 sub-carrier phase shifted frequency domain data sequence comprising a first symbol sequence of: {-1, 1, -1, 1, -1, -1, 1, -1, 1, -1, -1, 1, 1, 1, -1, 1, 1, 1, 1, 1, -1, 1, 1, 1, 1, 1} and a second symbol sequence of: {-1, 1, -1, -1, 1, 1, -I₁ -1, -1, 1, 1, -1, 1, 1, 1, -

I, 1, -1, -1, -1, -1, -1, 1, 1, 1, -1}.

10. The transmitter claimed in any previous claim wherein the first section comprises repetitions of the first data sequence.

II. The transmitter of claim 10 wherein at least a one repetition is multiplied by a complex coefficient.

12. The transmitter claimed in claim 11 wherein the second section comprises a guard interval between the repetitions.

13. The transmitter claimed in any previous claim wherein the transmitter is an OFDM (Orthogonal Frequency Division Multiplex) transmitter.

14. The transmitter claimed in claim 13 wherein the means for transmitting is operable to modulate the individual preambles using a 64 point Discrete Fourier Transform (DFT) .

15. The transmitter claimed in claim 13 wherein the means for transmitting is operable to modulate user data using a 128 point Discrete Fourier Transform (DFT) .

16. The transmitter claimed in any previous claim wherein the coefficient sequence for each antenna is such that the second sections of the individual preambles are orthogonal in the frequency domain.

17. The transmitter claimed in any previous claim wherein the plurality of antennas is four antennas, the second section comprises four repetitions and the coefficient sequence for each antenna is a different row of a four by four Walsh-Hadamard matrix.

18. The transmitter claimed in any of the previous claims 1 to 16 wherein the plurality of antennas is two antennas, the second section comprises four repetitions and the coefficient sequence for each antenna is a repeated different row of a two by two Walsh-Hadamard matrix.

19. A receiver comprising: means for receiving a signal transmitted as sub- signals from a plurality of antennas, the signal comprising a preamble set comprising an individual preamble for each antenna; each of the individual preambles comprising a first section and a second section wherein the first section comprises a first data sequence comprising at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprises repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna; and means for performing an initial acquisition in response to the first sections; and means for determining a channel estimate in response to the second sections. ■

20. The receiver of claim 19 wherein the initial acquisition comprises timing synchronisation.

21. The receiver claimed in any of the claims 19 or 20 wherein the means for determining the channel estimate is operable to determine the channel estimate by frequency domain processing.

22. A preamble set comprising an individual preamble for each of a plurality of antennas; wherein each of the individual preambles comprises a first section and a second section, the first section comprising a first data sequence comprising at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprising repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna.

23. A method of transmitting a signal as sub-signals from a plurality of antennas, the method comprising: means for generating a preamble set comprising an individual preamble for each antenna, each of the individual preambles for each antenna comprising a first section and a second section wherein the first section comprises a first data sequence comprising at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprises repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna; and transmitting the signal as sub-signals from the plurality of antennas using the preamble set.

24. A method of receiving a signal transmitted as sub- signals from a plurality of antennas, the method comprising: receiving the signal, the signal comprising a preamble set comprising an individual preamble for each antenna; each of the individual preambles comprising a first section and a second section wherein the first section comprises a first data sequence comprising at least one sequence different for each antenna and selected from a predetermined set of preamble sequences, and the second section comprises repetitions of a predetermined second data sequence, each repetition being weighted by a coefficient of a coefficient sequence that is different for each antenna; performing an initial acquisition in response to the first sections; and determining a channel estimate in response to the second sections.