FIELD

This invention relates to low power digital radio communications. It relates more specifically, although not exclusively, to establishing connections between devices in an Orthogonal Frequency Division Multiplexing (OFDM) radio system supporting Multiple-Input-Multiple-Output (MIMO) transmission modes.

BACKGROUND

OFDM is a form of radio transmission that is used in various radio protocols such as Long Term Evolution (LTETM), various IEEETM 802.11 standards, DABTM radio, DVB-T, and WiMAXTM. Rather than encoding data on a single carrier frequency, a data stream is spread over some or all of a radio channel containing multiple OFDM subcarriers. The OFDM subcarriers are typically closely spaced, at regular intervals, across the frequency spectrum, although this is not essential. The subcarriers are orthogonal to avoid mutual interference. OFDM can thereby provide good resilience to multipath fading and to external interference.

In some radio communication applications, the MIMO principle is employed whereby the transmitter and receiver are provided with multiple antennas which can be exploited in a number of ways. MIMO communications are dependent on the fact that transmission antennas are ‘separated’ between each other and receiver antennas also separated between each other. This separation enables frequency selectivity between transmission paths to be independent. Separation can be either physical (distance) or polarisation diversity (two antennas are orthogonally polarized) or combination thereof. One such MIMO mode is ‘beamforming’, in which a transmitter varies the phase and/or magnitude of signals transmitted from different antennas to generate regions of constructive and destructive interference, thereby allowing the transmitter to ‘direct’ the transmitted ‘single spatial stream’ towards a receiver. These signals are added in the radio channel and the receiver cannot deduce whether the transmission came from single antenna or more (and it does not need to). Since the receiver ‘sees’ such transmissions as though they had come from a single antenna, this is referred to as the transmission being one having a single effective antenna.

Another MIMO mode is ‘space-time block coding’ or just simply ‘transmit diversity’ which is similar to the above in that a single spatial stream is transmitted from multiple antennas but in which the signal is encoded to exploit independent fading from the multiple antennas to achieve space diversity. The gains achievable approach those of receive diversity with the maximum ratio combining gains as a result of the encoding applied. The receiver needs to know what encoding was used in transmission and needs to be able to estimate the radio channel between each TX-RX antenna pair. The number of effective antennas which are required depends on the space-time coding applied. For example one such space-time coding is 2 x 1 Alamouti coding which uses two effective antennas. If beamforming is also used, multiple physical antennas may be acting as each effective antenna.

Open loop spatial multiplexing’ or Open Loop MIMO’ is used to increase the data rate of the system up N times the single antenna data rate by sending different ‘spatial streams’ via each of N different effective transmit antennas. In Open Loop MIMO the transmitter has no channel knowledge.

‘Closed loop spatial multiplexing’ or ‘Closed Loop MIMO’ is similar to Open Loop MIMO, but the transmitter utilises channel information to enable simple spatial diversity or beamforming techniques to increase system data rate, increase signal to noise ratio (SNR) and even simplify receiver design. Closed Loop MIMO is dependent on the transmitter having knowledge of the channel state and both transmitter and receiver using known precoding (beamforming). The knowledge of the channel state mentioned above is usually ether provided as quantized feedback from the receiver to the transmitter or by exploiting the reciprocity inherent in time- division duplexing (TDD) whereby the same radio channel is used to transmit in both directions between the devices. This mode can adapt automatically to the radio channel conditions and provide either “beamforming” kind of boost or “spatial multiplexing” kind of boost.

In order for a receiving device to decode an incoming signal, it is necessary for it to know whether MIMO transmission had been employed, how many effective antennas are in use and which MIMO modes has been used. Depending on the radio protocol, some of this information may be fixed, but at least some of it needs to be communicated by the transmitter.

One OFDM radio protocol which supports MIMO communications is Long Term Evolution (LTE) - both in the form of legacy LTE and newer variants such as narrowband Internet of Things (NB-loT). In LTE the MIMO transmission mode is signalled in connection configuration messages - in other words establishing the transmission mode is part of the negotiation process that takes place between transmitter and receiver before any data can be exchanged. This involves for example the user equipment (UE) having to carry out blind decoding of a physical broadcast channel (PBCH) in order to distinguish between a single antenna transmission and a transmission using transmit diversity. Changing the transmission mode requires reconfiguration of the connection.

There is also support in IEEE 802.11 for MIMO. This specifies that the information on MIMO transmission mode should be included as part of the header of the data packets employed in this protocol. The header is always transmitted with a single antenna and thus transmit diversity cannot be used during transmission of the header. Additionally the header is relatively long, thus for short packets this kind of MIMO signalling requires an overly large overhead.

It has been proposed in ETSI TR 103514 V1.1.1 (2018-07) ‘Digital Enhanced Cordless Telecommunications (DECT) DECT-2020 New Radio (NR) interface Study on Physical (PHY) layer’ (the “DECT-2020 proposal”) to provide a packet structure in which multiple Channel Training Fields (CTFs) are provided corresponding to the number of antennas in use. However the Applicant has developed a different approach which is believed to have some advantages over this. SUMMARY

When viewed from a first aspect the invention provides a digital radio communication system comprising a transmitter and a receiver arranged to operate according to a predetermined communication protocol, wherein said transmitter is arranged to: transmit using an integer effective number Ntc ^qίί of antennas, wherein Ntc ^qίί is selected from a set of M possible pre-defined effective numbers of antennas; and transmit a data packet comprising a synchronisation sequence selected from a pre-defined group of M different synchronisation sequences such that the selected synchronisation sequence is indicative of the value of Ntc ^qίί ; and wherein the receiver is arranged to: receive said data packet; determine said synchronisation sequence; use said synchronisation sequence to determine a value of Ntc ^qίί ; and use said value of Ntc ^qίί to process part of said data packet.

Thus it will be seen by those skilled in the art that in accordance with the invention the effective number of transmit antennas Ntc ^qίί in use can be determined simply by establishing which of a predetermined set of synchronisation sequences has been used. This may allow this information to be ascertained in a far smaller packet size than would be the case with other protocols such as the IEEE 802.11 proposal which requires a separate packet header, which is decodable with single antenna reception, that signals the number of antennas. The actual number of transmit antennas Ntc in use by the transmitter may be greater than or equal to the effective number of transmit antennas Ntc ^qίί , wherein the effective number of transmit antennas Ntc ^qίί refers to the number of transmit antennas as seen from the point of view of the receiver. For example, if multiple antennas are used to perform beamforming on the transmitted signal - i.e. use phase and/or amplitude adjustments to generate regions of high and low signal intensity as a result of constructive and destructive interference in order to focus the energy at a particular receiver - from the point of view of the receiver only one signal is received and therefore the effective number of transmit antennas is one, despite the actual number of transmit antennas being greater than one. It will be seen therefore that in accordance with at least embodiments of the invention the transmitter can easily dynamically change the number of effective transmit antennas and so the MIMO mode used for the data channel - e.g. to account for current channel conditions.

The invention extends to a digital radio transmitter arranged to operate according to a predetermined communication protocol, wherein said transmitter is arranged to: transmit using an integer effective number Ntc ^qίί of antennas, wherein Ntc ^qίί is selected from a set of M possible pre-defined effective numbers of antennas; and transmit a data packet comprising a synchronisation sequence selected from a pre-defined group of M different synchronisation sequences such that the selected synchronisation sequence is indicative of the value of Ntc ^qίί .

The invention also extends to a digital radio receiver arranged to operate according to a predetermined communication protocol, wherein said receiver is arranged to: receive a data packet comprising a synchronisation sequence selected from a pre-defined set of M different synchronisation sequences indicative of the effective number of transmit antennas Ntc ^qίί in use; determine which of the set synchronisation sequences was received; use said synchronisation sequence to determine a value of Ntc ^qίί ; and use said value of Ntc ^qίί to process part of said data packet.

The invention also extends to a method of operating a digital radio communication system comprising a transmitter and a receiver according to a predetermined communication protocol, the method comprising: the transmitter transmitting using an integer effective number Ntc ^qίί of antennas, wherein Ntc ^qίί is selected from a set of M possible pre-defined effective numbers of antennas; the transmitter transmitting a data packet comprising a synchronisation sequence selected from a pre-defined group of M different synchronisation sequences such that the selected synchronisation sequence is indicative of the value of Ntc ^qίί ; the receiver receiving said data packet; the receiver determining said synchronisation sequence; the receiver using said synchronisation sequence to determine a value of

Nix ^eff ; and the receiver using said value of Ntc ^qίί to process part of said data packet.

The invention also extends to a method of operating a digital radio transmitter according to a predetermined communication protocol, the method comprising: transmitting using an integer effective number Ntc ^qίί of antennas, wherein Nix ^eff is selected from a set of M possible pre-defined effective numbers of antennas; and transmitting a data packet comprising a synchronisation sequence selected from a pre-defined group of M different synchronisation sequences such that the selected synchronisation sequence is indicative of the value of Ntc ^qίί .

The invention also extends to a method of operating a digital radio receiver according to a predetermined communication protocol, the method comprising: receiving a data packet comprising a synchronisation sequence selected from a pre-defined set of M different synchronisation sequences indicative of the effective number of transmit antennas Ntc ^qίί in use; determining which of the set synchronisation sequences was received; using said synchronisation sequence to determine a value of Ntc ^qίί ; and using said value of Ntc ^qίί to process part of said data packet.

The invention also extends to a non-transitory computer readable medium comprising instructions configured to cause a digital radio transmitter to operate in accordance with the method outlined above.

The invention also extends to a non-transitory computer readable medium comprising instructions configured to cause a digital radio receiver to operate in accordance with the method outlined above.

The set of M possible effective numbers of transmit antennas may be specified in the communication protocol. In one example the protocol may allow for 1, 2, 3, 4, 5, 6, 7 or 8 effective antennas and M may be 8 such that the transmitter is able to signal that it is using any of the allowed numbers of effective antennas. In another example the protocol may allow for 1 , 2, 4, or 8 effective antennas and M may be 4 such that the transmitter is able to signal that it is using any of the allowed numbers of effective antennas. Equally however M may be lower than the theoretical maximum if a given system is not using all the numbers of effective antennas allowed in the protocol. For example if a given system following the second example protocol mentioned above does not support 8 effective antennas, M may be 3 as it is only necessary for the transmitter to signal whether it is using 1 , 2 or 4 effective transmitters. In all cases the transmitter has a higher number of physical antennas available to it. If the number of physical antennas is in excess of the number of effective antennas the transmitter is allowed (or chooses) to use, the extra physical antennas may be employed to form the effective antennas e.g. by beamforming. For example a transmitter with 16 physical antennas may use 4 effective transmitters, each comprising 4 physical antennas transmitting the same signal with beamforming. Alternatively, all 16 physical antennas may be used to beamform all 4 streams, still resulting in there being 4 effective antennas (as there are 4 different streams) - the transmitter sums the 4 streams internally before being transmitted through all 16 antennas.

In accordance with the invention therefore, embodiments can use full gain multi antenna diversity from the beginning of a burst to the end. Moreover a payload or data field can be transmitted even in relatively short packet lengths without a large signalling overhead. It may also allow MIMO modes to be enabled dynamically in a random access transmission protocol rather than needing to reconfigure the connection.

In a set of embodiments the transmitter is arranged to transmit a data field after transmitting the synchronisation sequence, wherein the data field comprises: a reference signal to allow the receiver to perform channel estimation; a data channel comprising a payload of the data packet; and a control channel comprising side information required to decode the data channel; wherein the control channel comprises one or more bits dedicated to signalling an integer number of spatial streams N _ss used for the data channel, wherein the number of spatial streams N _ss is less than or equal to the effective number of transmit antennas Ntc ^qίί . As will be understood by those skilled in the art, a spatial stream refers to a single data stream transmitted from the transmitter to the receiver in MIMO wireless communications. When spatial multiplexing is used, the transmitter transmits multiple different data streams simultaneously, allowing for faster data rates if the receiver is able to receive and decode the different spatial streams successfully.

The number of spatial streams is less than or equal to the effective number of antennas as it is possible for the transmitter to transmit a different spatial stream from each effective antenna or transmit one spatial stream from multiple different antennas - e.g. with MIMO, transmit diversity or beamforming. If, for example, multiple effective antennas are used to transmit the same data stream with space- time block encoding, the physical signals transmitted from each antenna will appear different to the receiver as they comprise two different space-time streams.

However, these are not different spatial streams as the data transmitted from the antennas is the same, and therefore the signals transmitted from the multiple antennas as a group comprise a single spatial stream.

In a set of embodiments the transmitter is arranged to encode the control channel in a predetermined manner according to the effective number of transmit antennas in use. Thus it will be appreciated by those skilled in the art that the receiver is then able to decode the control channel simply by ascertaining the effective number of transmit antennas in use from the synchronisation sequence. In a set of such embodiments the control channel is coded such that it is decodable with a single receiver antenna. This implies that the control channel comprises only a single spatial stream. For example, in some embodiments the control channel may be encoded for single antenna transmission when the effective number of transmit antennas Ntc ^qίί is equal to one, and the control channel may be encoded according the Alamouti procedure when the effective number of transmit antennas Ntc ^qίί is equal to two or more.

In a set of embodiments the control channel further comprises a Cyclic Redundancy Check (CRC), wherein the transmitter is arranged to apply one or more predetermined bit masks to the CRC before transmission in order to indicate further information regarding the transmission mode used for the data channel. In a set of embodiments the transmitter is arranged to apply a specific bit mask to the control channel CRC to indicate the use of beamforming during transmission of the data channel and a different specific bit mask to the control channel CRC to indicate that the beamforming is closed loop rather than open loop. The transmitter may mask the control channel using one, both, or neither of these predetermined bit masks in order to indicate the use of various combinations of beamforming and open/closed loop transmission. For example, the transmitter may apply solely the beamforming mask to the control channel, indicating the use of open-loop beamforming transmission (as the closed-loop bit mask is not applied) for the data channel.

This is considered to be novel and inventive in its own right and thus when viewed from a second aspect the invention provides a digital radio transmitter arranged to operate according to a predetermined communication protocol, wherein said transmitter is arranged to transmit a data packet comprising a control channel and a data channel including a data payload, such that at least the data channel is transmitted using one of a predetermined set of beamforming transmission modes, wherein at least the control channel comprises a cyclic redundancy check field and said transmitter is further arranged to apply one or more masks to said cyclic redundancy check field in dependence upon which of the predetermined set of beamforming transmission modes is applied to said data channel.

This aspect of the invention also extends to a digital radio receiver arranged to receive and decode the data packet set out above and to a corresponding method of operating a digital radio transmitter and a corresponding method of operating a digital radio receiver and to a non-transitory computer readable medium containing instructions to cause a processor to carry out the respective methods.

In a set of embodiments of either aspect a or the receiver determines which (if any) bit masks have been applied to the transmitted control channel CRC by comparing the locally calculated CRC with that received. The receiver may apply various combinations of the same bit masks as those used by the transmitter to the locally calculated CRC and compare with the received CRC until a match is found, informing the receiver that the CRC is acceptable and which masks have been used by the transmitter (and therefore which of beamforming and open/closed transmission modes are used for the data channel). It will therefore be appreciated by those skilled in the art that the receiver is able to ascertain side information required to decode the data channel, including the number of spatial streams N _ss , the use of open/closed loop transmission, and whether or not beamforming is used, after decoding the control channel. Thus when the receiver has ascertained both the effective number of transmit antennas Ntc ^qίί and the number of spatial streams N _ss it can determine the encoding used for the data channel by the transmitter. For example, in some embodiments, if the effective number of transmit antennas Ntc ^qίί and the number of spatial streams N _ss are both determined to be equal to N and N is greater than one, the transmission mode for the data channel is determined to be N x N MIMO transmission.

In a set of embodiments the transmitter is arranged to encode the data channel and control channel using space-time block coding. In some embodiments, the transmitter is arranged to encode the control channel using the Alamouti procedure. In a set of embodiments the transmitter and receiver are arranged to operate using Orthogonal Frequency Division Multiplexing (OFDM).

In a set of embodiments the protocol is in accordance with DECT-2020 proposal to the extent that it does not conflict with the operation set out herein - for example having a frame of 24 slots, each slot comprising ten OFDM symbols.

As will be appreciated by those skilled in the art, it is typically necessary in a MIMO communication system for a receiver to be able to perform channel estimation for each effective transmitter in use to transmit a given packet - (i.e. for each space- time stream where space-time block coding is employed). In a set of embodiments of the present invention, the data packet comprises a reference signal for each effective transmitter in use as indicated by the synchronisation field. While it is possible to transmit the reference signal for each effective transmitter used to transmit the packet sequentially in time - i.e. time-multiplexed - the applicant has recognised that this may be inefficient in terms of radio resource usage. Instead the applicant proposes that the reference signals be frequency-multiplexed, with different OFDM sub-carriers being reserved for the transmission of the reference signal for each effective transmit antenna. The control channel and data channel may also be defined in the frequency domain of the data field, with different sub carriers being reserved for transmission of the control and data channels. In some embodiments, different sub-carriers may be reserved for different channels at different times.

In a set of embodiments the synchronisation sequences are designed such that the receiver can use them to correct one or more of receiver gain, frequency offset and timing offset relative to the transmitter. Typically, this means that the sequences should have good correlation properties. In a set of embodiments the group of synchronisation signals are orthogonal to each other.

Advantageously at least the synchronisation sequence is transmitted in a manner such that single antenna reception by the receiver is possible. For example, the synchronisation sequence may be transmitted with beamforming. In doing this, the receiver is able to calculate the effective number of transmit antennas Ntc ^qίί without requiring additional information to be transmitted in order to decode the synchronisation sequence. Once the effective number of transmit antennas Ntc ^qίί has been established, the control channel may be assumed to have been transmitted in whichever manner is applicable to the effective number of transmit antennas Ntc ^qίί in use. For example if there are two effective antennas it may be assumed that Alamouti encoding has been used. This allows the receiver to decode the control channel in order to establish the other parameters required to decode the data channel, which may be transmitted in any of a number of MIMO/MISO modes. For example where, as is preferred, the data packet comprises a data channel, the data channel may be transmitted according to the effective number of transmit antennas Ntc ^qίί indicated by the synchronisation signal and the number of spatial streams N _ss indicated by the control channel, and the MIMO parameters (e.g. beamforming, open/closed loop, etc...) indicated by the control channel - particularly the CRC therefor as previously outlined.

In a set of embodiments the transmitter and receiver are each able to function in the other role as well - e.g. they may operate in time division duplexing so that the two communicating devices alternate the transmitter and receiver roles. Since a receiver is able to make a good estimate of channel conditions from the transmissions it receives, it can account for this when it acts as a transmitter and can determine the MIMO mode. There are several ways in which the receiver can determine the particular synchronisation sequence which has been transmitted. In a set of embodiments the receiver performs cross-correlation between the received signal and each of the pre-defined group of M different synchronisation sequences. The selected sequence may then be determined to be which of the pre-defined group gives the strongest cross-correlation. The value of Ntc ^qίί may then be deduced from the synchronisation sequence selected - e.g. by means of a look-up table.

Conceivably the synchronisation sequence could be transmitted just once. In a set of preferred embodiments however the synchronisation sequence is periodic. For example where an OFDM protocol is employed the synchronisation sequence is made repetitive in the time domain by modulating every nth sub-carrier and allocating no power to the remaining subcarriers. In a set of such embodiments the receiver performs auto-correlation to identify the repetition of a particular sequence. Preferably the auto-correlation is also used to detect initial timing and frequency offset. The determined frequency offset is preferably used to correct the frequency of the received signal. Advantageously the timing offset may be used to give a shorter search window for the above-mentioned cross-correlation.

In a set of embodiments the synchronisation sequences comprise even lengths based on a quadrature phase shift keying (QPSK) sequence whose cyclical shifts provide a desired degree of cross-correlation. Alternatively, the synchronisation sequences may comprise even lengths based on a bipolar sequence whose cyclical shifts provide a desired degree of cross-correlation. In another set of embodiments odd length sequences are used. In such cases a QPSK or a bipolar sequence may be better depending on circumstances.

The transmitter and receiver could be used in any of a wide range of applications such as cordless telephones or wireless microphones but the applications are not limited to transmission of voice and can include e.g. sensors and actuators for smart buildings, factories or cities. BRIEF DESCRIPTION OF THE DRAWINGS

Certain embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which:

Fig. 1 is a schematic diagram illustrating a first radio communication system in accordance with the invention;

Fig. 2 is a schematic diagram illustrating a second radio communication system in accordance with the invention;

Fig. 3 is a schematic diagram of a first data packet structure in accordance with the invention;

Fig. 4 is a key showing the different subcarrier allocation types shown in Figs. 5 and 6.

Fig. 5 is a schematic diagram illustrating subcarrier allocation for transmission of a data packet using a single space-time stream and two subslots.

Fig. 6 is a schematic diagram illustrating subcarrier allocation for transmission of a data packet using four space-time streams and four subslots.

Fig. 7 is a graph of cyclic autocorrelation value against cyclic shift for the STF modulation sequence s.

Fig. 8 is a flow chart indicating operation of the transmitter in the radio system in accordance with an embodiment;

Fig. 9 is a flow chart indicating operation of the receiver in the radio system in accordance with a first embodiment;

Fig. 10 is a flow chart indicating operation of the receiver in the radio system in accordance with a second embodiment. Fig. 11 is a schematic diagram of a second data packet structure in accordance with the invention;

DETAILED DESCRIPTION OF THE DRAWINGS As is known in the art, there exist numerous modes of operation for radio transceivers with a plurality of antennas available. The simplest possible mode of operation is Single Antenna Input, Single Antenna Output (SISO), whereby a transmitter uses a single antenna to transmit a single signal which is then received through a single antenna by a receiver.

Another possible mode of operation is Multiple Antenna Input, Single Antenna Output (MISO), whereby a transmitter uses a plurality of antennas to transmit signals which are then received through a single antenna by a receiver.

A possible MISO mode of operation is beamforming, whereby a one dimensional signal is beamformed with a beamforming matrix that modifies the phases and amplitudes of the signal in order to ‘steer’ a signal towards a recipient using the principles of constructive and destructive interference. In this case, N _ss = N _STS = N _j f < n _TX .

Another possible MISO mode of operation is 2 x 1 Alamouti transmit diversity,. In the case of Alamouti transmit diversity, in order for a receiver to successfully receive the two space-time streams using a single antenna, it must know in advance that Alamouti transmit diversity has been used. As is known in the art, Alamouti transmit diversity provides significant diversity gain over single antenna transmission.

Fig. 1 shows a first MIMO (multiple input, multiple output) transmission mode equipped digital radio transceiver device 100 and a second MIMO transmission mode equipped digital radio transceiver device 102. The first radio transceiver device comprises three antennas 104a, 104b and 104c; and the second radio transceiver device comprises three antennas 106a, 106b and 106c. As will be well understood by those skilled in the art, a number of standard modules such as processors, oscillators, filters, amplifiers, digital to analogue converters and analogue to digital converters are provided in the radio transceivers 100, 102 but description of these is omitted for the sake of brevity.

Fig. 1 also shows the possible signal paths 108 from the first transceiver 100 when it is acting as a transmitter through each of its antennas 104a, 104b, 104c to the second transceiver 102 acting as a receiver through each of its antennas 106a, 106b, 106c. In this example, each of the transmit antennas 104a, 104b, 104c provides a separate spatial stream. As will be appreciated by those skilled in the art, many different MMO transmission modes are known in the art and could be employed depending on the application from transmit diversity to open loop spatial multiplexing, closed loop spatial multiplexing, cyclic delay diversity, beamforming etc. The transceivers are configured to operate using OFDM modulation as is known per se in the art.

Fig. 2 further illustrates the principle of effective antennas. In this embodiment a first radio transceiver 110 comprises four antennas 114a, 114b, 114c and 114d; and a second radio transceiver 112 comprises a single antenna 116. As before, a number of standard modules such as processors, oscillators, filters, amplifiers, digital to analogue converters and analogue to digital converters are provided in the radio transceivers 110, 112 but description of these is omitted for the sake of brevity.

Fig. 2 also shows signal paths 118 from the first transceiver 110 when it is acting as a transmitter through each of its antennas 114a, 114b, 114c and 114d to the second transceiver 112 acting as a receiver through its antenna 116. In this example, each of the transmit antennas 114a, 114b, 114c and 114d transmit the same signal, but with various phase and magnitude adjustments to the output of each individual antenna in order to beamform the output signal in such a way that some regions have increased signal intensity and other regions have reduced signal intensity as result of constructive and destructive interference between the signals transmitted by different antennas. It will therefore be understood how manipulating the phase and magnitudes of the signals output by each antenna provides the ability to ‘steer’ the transmitted signal towards the receiver 112 to increase the chance of successful reception and decrease signal strength in other regions to reduce interference with other signals. The receiver 112 cannot directly determine that the signal it receives is the result of four separate signals subjected to beamforming. Rather the receiver antenna 116 receives a single signal which is a product of the radio channel conditions along the respective paths 118 between the transmit antennas 114a-d and the ‘weights’ applied to the respective signals from each antenna in order to achieve the beamforming mentioned above. The transmit antennas 114a-d therefore act as single ‘effective’ transmit antenna. In order to decode the signal it receives the receiver 112 must estimate the combined channel between the effective antenna and the receive antenna 116.

The principles of the arrangements show in Figs. 1 and 2 may be combined so that multiple space-time streams are transmitted. For example, although Fig. 2 shows four transmit antennas these could generate two space-time streams, with the additional transmit antennas used to beamform the transmission. Such an arrangement would have two effective transmit antennas.

Fig. 3 shows an example structure of a data packet 200 in accordance with the invention. The data packet comprises: a Guard Interval (Gl) 202, a Synchronisation Training Field (STF) 204 and a Data Field (DF) 206. The lower part of the diagram shows various of these fields in more detail.

The Gl 202 forms the initial part of the packet 200, providing separation between consecutive packets 200. Immediately following the Gl 202 is the STF 204, comprising seven repetitions of a signal pattern 208, with the purpose of enabling synchronisation between the transmitter 100, 110 and the receiver 102, 112 by allowing the receiver 102, 112 to correct for receiver gain, frequency and timing errors relative to the transmitter 100, 110. In accordance with the present invention, the STF 204 comprises a specific sequence, selected by the transmitter 100, 110, indicative of the effective number of antennas it is using according to a pre-defined scheme in the communication protocol.

Following the STF 204 is the DF 206, comprising a number of data field symbols 212 each preceded by a cyclic prefix 210, with the purpose of delivering the payload of the packet 200. The DF 206 is transmitted using OFDM, providing the ability to dedicate different subcarrier frequencies for different signals/channels, as will be explained in more detail below with reference to Figs. 4-6.

An alternative example of a data packet 250 in accordance with the present invention is shown in Fig. 11, wherein the individual parts of the packet are the same as that described with reference to Fig. 3, but the Gl 202 forms the final part of the packet 250 rather than the initial part. The Gl 202 in this embodiment has the same purpose as that shown in Fig. 3 - to provide separation between consecutive packets 250.

In an OFDM radio system of FFT size N _DFT - i.e. there are N _DFT subcarriers available - only N _occ subcarriers are used for transmission. The subcarrier of index of zero is not used due to DC-offset concerns and the subcarriers at the edge of the frequency band are not used to protect adjacent radio channels. This means that only the subcarriers k _occ = are used for transmission. In the examples shown in Figs. 5 and 6, N _DFT = 64 and N _occ = 56.

Fig. 4 is a key showing the different subcarrier allocation types shown in Figs. 5 and 6, illustrating the different shadings and which subcarrier allocation types they correspond to. Shown are the Guard/Empty field 202, the Synchronisation Training Field (STF) 204, the Physical Control Channel (PCC) 406, the Demodulation Reference Signal (DRS) 408 and the Physical Data Channel (PDC) 410. In this example the PCC 406, DRS 408 and PDC 410 make up the Data Field 206 shown in Fig. 3. The DRS 408 is a reference signal which allows the receiver 102, 112 to perform channel estimation; the PDC 410 is a data channel comprising the payload of the data packet 200; and the PCC 406 is a control channel comprising various information bits which are required to decode the data channel (PDC 410).

The receiver 102, 112 uses the periodicity of the STF 204 to synchronise to the transmitter 100, 110 by performing an autocorrelation based search of the received STF 204. The process by which this occurs will be described in more detail below. This autocorrelation-based synchronisation mechanism requires that the STF 204 be repetitive/periodic. This is accomplished in an OFDM system by modulating only every n ^th subcarrier and allocating no power to the remaining subcarriers for the STF 204. This results in the STF 204 comprising n repetitions of a long signal sequence.

As will also be explained in more detail below, when the receiver decodes the STF 204, it is able to determine which of a predetermined set of sequences was used by the transmitter. This allows the receiver to deduce the number of effective transmitters used by the transmitter from a pre-agreed mapping between sequences and effective numbers of transmitters.

The DRS 408 contains no payload data, instead it provides a reference signal that is known to the receiver and the receiver may use to estimate the transfer function or channel conditions for each space-time stream transmitted by the transmitter 100, 110.

The PCC 406 comprises specific parameters that indicate information about the PDC 410 such as how many spatial streams are used; how many space-time streams are used; and the length of the data packet 200 in a number of subslots or a number of slots. As used herein, the term subslot refers to the transmission of five consecutive OFDM symbols. A slot may be made up of two, four, eight or 16 subslots depending on the subcarrier scaling applied.

The PCC 406 also includes a CRC (not shown) which is masked to indicate whether beamforming is used and whether open-loop or closed-loop transmission is used. This will be described in more detail below. As will be appreciated, as the effective number of transmitters is conveyed by the STF and the CRC conveys information about the transmission mode, it is not necessary to allocate bits on the PCC 406 for this information which is beneficial as the PCC may be of restricted length.

Figs. 5 and 6 show examples of subcarrier allocation for the data packet 200. The rows correspond to the subcarriers 302 available to the transmitter 100, 110. The number of each subcarrier 302 corresponds to that subcarrier’s index, with index 0 corresponding to the central subcarrier frequency. In the examples given in Figs. 5 and 6 there are sixty-four subcarriers 302 available to the transmitter - i.e. the transmitted data packet 200 has a Fast Fourier Transform (FFT) of size 64. The columns correspond to the time-evolution of symbols 304 transmitted by the transmitter 110. The number of each symbol 304 corresponds to that symbol’s index, with index 0 being the first symbol transmitted by the transmitter 110.

Fig. 5 shows an example of a possible subcarrier allocation 400 for the data packet 200, wherein the number of space-time streams N _STS = 1 and the number of subslots N _sub = 2 . This is the situation described with reference to Fig. 2 with a single effective transmitter and would mean that both the PDC 410 and PCC 406 would be transmitted without the use of space-time encoding nor the use of MIMO or MISO transmission modes. Beamforming may still employed in the example shown here. However, if the transmitter 110 utilises its four antennas to beamform the single space-time stream, it is will still only be using a single effective antenna.

Guard allocations 306 correspond to the allocations for guard/empty fields 202. STF allocations 308 correspond to subcarrier allocations for transmitting the Synchronisation Training Field 204. DRS allocations 310 correspond to subcarrier allocations for transmitting the Demodulation Reference Signal 408 for the single space-time stream of index zero. PCC allocations 320 correspond to subcarrier allocations for transmitting the Physical Control Channel 406. PDC allocations 322 correspond to subcarrier allocations for transmitting the Physical Data Channel 408.

Transmission of the STF 204 is completed by the end of symbol index one, and subsequent symbols are used to transmit the Data Field 206. The DF 206 comprises an initial transmission of the DRS 408 and PCC 406, with each being allocated to different subcarriers. Following transmission of the PCC 408, the transmitter 110 transmits the PDC 410. A second set of DRS allocations 320 are transmitted one subslot after transmission of the first set of DRS allocations 320 so as to allow the receiver to perform up-to-date channel estimation.

Fig. 6 shows another example of a possible subcarrier allocation 600 for the data packet 200, wherein the number of space-time streams N _STS = 4 and the number of subslots N _sub = 4. This would require at least four effective transmit antennas. In one example, since the number of space-time streams N _STS = 4, the PDC 410 is transmitted using a combination of spatial multiplexing and space-time encoding that uses the four space-time streams. The PCC 406 is transmitted with just two space-time streams, as it is transmitted using Alamouti space-time encoding. The increased number of space-time streams N _STS over the example shown in Fig. 5 requires that three extra DRS allocations 311 , 312 and 313 be transmitted for the extra space-time streams, as one DRS allocation is required for each space-time stream used. The increased number of subslots N _sub over the example shown in Fig. 5 increases the length of the data packet, and increases the separation between the first and second transmission of the DRS 408 to two subslots rather than one as shown in Fig. 5.

DRS allocations 310, 311, 312 and 313 correspond to subcarrier allocations for transmitting the Demodulation Reference Signal 408 for the space-time streams of index zero, one, two and three respectively. As a result of the extra DRS allocations 311, 312 and 313 when compared to the example shown in Fig. 5, an equivalent number of PCC allocations 320 are displaced and transmitted at the same time as the PDC 410, such that there are the same number of PCC allocations 320 in both examples.

As in Fig. 5, transmission of the STF 204 is completed by the end of symbol index one, and subsequent symbols are used to transmit the Data Field 206 comprising the DRS 408, the PCC 406 and the PDC 410. DRS allocations 310, 311, 312 and 313 correspond to subcarrier allocations for transmitting the Demodulation Reference Signal 408 for the space-time streams of index zero, one, two and three respectively. As a result of the extra DRS allocation 311 when compared to the example shown in Fig. 5, an equivalent number of PCC allocations 320 are displaced and transmitted at the same time as the PDC 410, such that there are the same number of PCC allocations 320 in both examples.

It will be understood by those skilled in the art that example subcarrier allocations for the data packet 250 shown in Fig. 11 will be substantially identical to those shown in Figs. 5 and 6, but wherein Guard Allocations 306 will not precede the STF allocations 308 (i.e. form the initial part of the packet) but instead follow the PDC allocations 322 (i.e. form the final part of the packet).

In the embodiment of the present invention shown in Figs. 5 and 6, every fourth sub-carrier is used to transmit the STF 204, excluding the central and guard sub carriers: only the subcarriers of indices 28, 24, 20, 16, 12, 8, 4, -4, -8, -12, -16, -20, -24 and -28 are used for transmission of the STF. By allocating every (or almost every) fourth sub-carrier, the time-domain signal produced by taking the Inverse Fast Fourier Transform (IFFT) repeats every sixteen samples, resulting in four repeating periods of sixteen samples within an OFDM symbol period of sixty four samples. This periodicity allows the receiver to carry out auto-correlation on the received signal. The non-modulated subcarriers (denoted as guard allocations 306 in figs. 5 and 6) could be used for data, however, in order to guarantee a repetitive time-domain signal, they are not in this particular embodiment.

In accordance with the present invention, the particular sequence of subcarriers used for the STF 204 conveys information about the effective number of transmit antennas being used. The subcarriers forming these sequences are selected using the equations:

In the examples shown in Figs. 5 and 6, with an FFT of size 64, these equations give the STF subcarriers of [-28, -24, -20, -16, -12, -8, -4, 4, 8, 12, 16, 20, 24, 28]

It will be understood that the subcarriers allocated for transmitting the STF 204 shown in Figs. 5 and 6 are not limiting, but any selection of sub-carriers that produces a repeating time-domain signal may be used. It will also be understood that the sub-carrier allocations for the DRS 408, PCC 406 and PDC 410 are also not limiting, but any selection of sub-carriers may be used. In another embodiment, the subcarriers allocated for transmitting the STF 204 provide M candidate STF signals which are not periodic, but do possess good cross-correlation properties. In this case it is not possible to perform auto correlation to detect which candidate signal was received, instead the receiver 112 performs cross-correlation of the received signal with each of the M candidate signals as stored within the receiver 112. This embodiment, however, does increase the complexity of detecting which candidate signal was received by at least M times compared to detection through auto-correlation. The modulating values for each sub-carrier - i.e. the M possible sequences - are chosen to have good auto-correlation properties which allow an accurate estimate of timing and frequency errors to be obtained by the receiver 102, 112. The sequences are chosen also to have good cross-correlation properties between each other in order to allow the receiver 102, 112 to detect which of the M sequences was transmitted.

In the examples shown in Figs. 5 and 6, the sequences used are of length fourteen to correspond to the selection of subcarriers used for the STF 204. While it is possible to use Hadamard sequences, in which case the sequence length is typically a power of two, in accordance with the present invention the sequence s used comprises: s= {0— ly,0— ly,—1+ Oy,—1+ Oy,0+ ly,0— ly,0+ ly,

0- ly,0- ly,0+ ly,0+ ly,-1- Oy,0- ly,-1+ Oy} ^{1 J}

The approach described herein provides the ability to create unique sequences (i.e. which subcarriers are modulated and how) using unique rotations of this sequence, given by: is the number of effective transmit antennas in use, the value of which is obtained from the set = {1,2, 4, 8}. For example, one possible unique sequence comprises the subcarrier of index -28 being modulated with the first value of the sequence, the subcarrier of index -24 being modulated with the second value of the sequence, etc. The next possible unique sequence then comprises the subcarrier of index -28 being modulated with the third value of the sequence, the subcarrier of index -24 being modulated with the fourth value of the sequence, etc. Thus it will be understood by those skilled in the art that a set of M possible unique sequences, each with good autocorrelation and cross-correlation properties, can be generated simply by rotating the modulation across the subcarriers selected for the

STF 204. In accordance with the present invention, the base sequence s is used to signify that a single effective transmit antenna is in use by the transmitter 100, 110. A cyclic rotation the sequence s by two is used to signify that two effective antennas are in use; a cyclic rotation of the sequence s by four is used to signify that four effective antennas are in use; and a cyclic rotation of six is used to signify that eight effective antennas are in use.

This sequence provides good cyclic autocorrelation properties, as shown in Fig. 7, which shows a graph 700 of cyclic autocorrelation value against cyclic shift. Cyclic autocorrelation of this sequence provides a peak 702 at a cyclic shift of zero, as is desired to allow the receiver 112 to correct for frequency and timing errors, and a smaller, undesired peak at a cyclic shift of seven. The low autocorrelation values for the remaining cyclic shifts (1-6 and 8-13) mean that cross-correlating the sequence s with a cyclic shift of the sequence s for all but a cyclic shift of seven (peak 704) gives a low value, and therefore the receiver 112 is able to distinguish between the different sequences transmitted when the shifts that produce low autocorrelation values (i.e. 1-6 and 8-13) are used when performing cross-correlation, as described in further detail below.

Typical operation of the transmitter device 100, 110 will now be described with reference to Fig. 8. At step 800, the transmitter 100, 110 determines the number of effective antennas it will use for transmission. For example this will be from a predetermined set of four possible values of the number of effective antennas specified in the protocol - namely 1, 2, 4 or 8 effective antennas.

The transmitter then determines which sequence to use to modulate the STF 204 (step 802), as described above, based on the selected number of effective antennas by means of a lookup table. At step 804, the transmitter determines the number of spatial streams it will use for transmission of the PDC 410. This may be the same as the effective number or it may be fewer, depending on the transmissions mode used.

The PCC 406 comprises other information (apart from the number of effective transmitters) required for the receiver 102, 112 to determine the transmission mode used during transmission of the PDC 410, which is required in order for the receiver 102, 112 to decode the PDC 410. In accordance with the present invention, in addition to comprising information such as: packet length, network ID, transmitter identity, transmit power, etc., the PCC 406 further comprises up to log ₂ (N _{SS rnax} ) bits for signalling the number of spatial streams N _ss used in transmission of the PDC 410, as calculated at step 804. It will be understood by those skilled in the art that as the number of spatial streams N _ss is a positive integer, log ₂ (N _{SS rnax} ) bits are required to be able to signal all possible values of N _ss up to and including N _{SS Tnax} in binary. However, in some cases only select values of N _ss are allowed in the communication protocol, in which case fewer than log ₂ (N _{SS rnax} ) bits are needed. For example, if N _{SS Tnax} = 8, three bits are required to be able to signal all possible values of N _ss (i.e. 1, 2, 3, 4, 5, 6, 7 and 8). If, however, the only allowed values of N _ss are 1, 2, 4 and 8, for example, then only two bits are required to signal the allowed values of N _ss . The transmitter then proceeds to step 806 where it inserts the value of N _ss , as determined at step 804, into the PCC 406.

In accordance with the present invention, the PCC 406 further comprises a 16-bit Cyclic Redundancy Check (CRC) field in order to allow the receiver 102, 112 to perform a data integrity check on the data received. The applicant has recognised that by applying specific XOR masks to the CRC, further side information about the transmission mode used for the PDC 410 may be signalled in the PCC 406 without requiring additional data bits.

At step 808, the transmitter determines whether a closed-loop mode of transmission is used for transmission of the PDC 410, or if an open-loop mode of transmission is used. As used herein, the term closed-loop mode of operation refers to the transmitter 100, 110 altering the signal it transmits based on feedback from the receiver 102, 112 on instantaneous channel conditions. For example, if the transmitter applies beamforming to the PDC 410, then it adjusts the beamforming weights used according to this feedback when operating in a closed-loop mode.

The term open-loop mode of operation refers to the transmitter 100, 110 transmitting data packets 200 without the receiver providing feedback on instantaneous channel conditions. For example, if the transmitter applies beamforming to the PDC 410, then the beamforming weights used are not dependent on any feedback from the receiver when operating in an open-loop mode.

If the transmitter 100, 110 determines that a closed-loop mode of transmission is used for transmission of the PDC 410, then it proceeds to step 810 where it applies an XOR mask of 0x5555 to the 16-bit CRC in the PCC 406. If the transmitter determines that an open-loop mode of transmission is used for the PDC 410, then it proceeds to step 812 where it does not apply this mask to the CRC. The transmitter 100, 110 then proceeds to 814, where it determines whether beamforming is used for transmission of the PDC 410 or not. If beamforming is used, the transmitter 100, 110 proceeds to step 816, where it applies an XOR mask of OxAAAA to the 16-bit CRC in the PCC 406. If beamforming is not used, the transmitter 100, 110 instead proceeds to step 818 where it does not apply this mask to the CRC.

It will be understood by those skilled in the art that the CRC masks described above to signify closed-loop transmission and beamforming respectively are compatible, as both specifying both closed-loop transmission and beamforming may be signalled by the transmitter by applying both masks to the 16-bit CRC, thereby resulting in an XOR mask of OxFFFF being applied to the CRC. It will also be understood by those skilled in the art that by using XOR masks for the CRC, the actual data stored within the CRC remains recoverable by the receiver 102, 112 and is not lost by applying the masks.

Finally, the transmitter 100, 110 proceeds to step 820, where it encodes the data packet 200 according to the transmission mechanisms it determines as outlined above, before transmitting the data packet. The specific processes involved with encoding the data packet with open-loop/closed-loop transmission, beamforming, spatial multiplexing, space-time encoding, etc. will be understood by those skilled in the art so the description of these are omitted for the sake of brevity.

Operation of the receiver device 102, 112 will now be described with reference to Fig. 9. At step 900 the receiver 102, 112 receives, using its antenna 116, the first part of the data packet 200 which is the synchronisation training field 201. At step 902 the receiver 102, 112 then uses a standard auto-correlation technique on the received signal samples using the following formula: where p _n denotes the auto-correlation value, r _n denotes the received signal samples, L _REP is the length of a period and N _REP is the number of periods used for the process of autocorrelation. The synchronisation signal is declared to be found when

\Pn\ ² > T (6), where T is a threshold value chosen to keep the number of false detections acceptably small. As is well known per se in the art, the timing and frequency of the correlation threshold being reached can be used to give an estimate of the frequency and timing offset of the received signal relative to the receiver (step 904). The estimate of the frequency offset Af may be calculated from the maximum autocorrelation value ' p _n ^li max angle(p _nmax )

Af = (7), 2p T _S L _REP where n _max is the time instant at which the synchronisation signal was declared to be found, T _s is the sampling period of the system, and all other symbols have the same meaning as that denoted previously or their standard mathematical meaning.

The received signal is then corrected to account for the estimated frequency error Af and the timing error by the receiver (step 906). At step 908 this signal is then cross-correlated with each of the candidate sequences corresponding to the different potential numbers of effective antennas that the transmitter 100, 110 could have, and the candidate sequence with the largest correlation value is selected (step 910). The correlation value of the selected sequence is given by: where the sample index beginning n _bgn ranges through a small neighbourhood of Um _a x and S _{j k} is the j ^th synchronisation sequence. The arguments for the maximum correlation value define the estimated synchronisation sequence and refined timing

The receiver 102, 112 is thus able to determine which of the M possible sequences was transmitted by the transmitter 100,110 in the STF 204, and therefore determine the number of effective antennas used for transmission. At step 912 the receiver then proceeds to decode the PCC 406 using knowledge of the effective number of transmit antennas and using the DRS 408 to arrive at an estimate of the radio channel impulse response for each effective transmit antenna to receive antenna channel.

By decoding the PCC 406, the receiver is able to determine the transmission mode used for the PDC 410, including the number of spatial streams used, whether closed-loop or open-loop transmission is used, and whether beamforming is used or not, as described in more detail above. In order to perform the data integrity check required in decoding the PCC 406, and to determine which masks have been applied to the CRC, the receiver compares its locally generated CRC to the one received with all combinations of the XOR masks 0x5555 and OxAAAA applied to the locally generated CRC. Whichever of these matches the CRC received by the receiver informs the receiver of which masks were applied at transmission and therefore which of closed-loop transmission, open-loop transmission and beamforming are used in transmission of the PDC 410.

After decoding the PCC 406, the receiver 102, 112 proceeds to step 914, where it decodes the PDC 410 using the information signalled in the PCC 406 as described in more detail above.

As will be appreciated, this auto-correlation based receiver synchronisation procedure requires a repeated periodic sequence as provided by the sub-carrier patterns described above with reference to Figs. 5 and 6. However it gives the advantage of providing a shorter search window for the cross-correlation search through candidate sequences than if this step was not carried out.

Fig. 10 illustrates an alternative approach in accordance with another embodiment of the invention. In this embodiment the receiver 102, 112 again receives the signal at step 950 but directly cross-correlates this with each of the candidate sequences with the received signal (st where x _{j n} is the correlation value of the j ^th candidate sequence with the received signal, and all other symbols have the same meaning as that denoted previously. The j ^th candidate sequence is selected (step 954) and n _max is declared to be found when \c _},h \ ² > G (10), where T is, as before, a threshold value chosen to keep the number of false detections acceptably small.

The frequency and timing errors are then estimated (step 956). The frequency offset Af is estimated from the correct timing n _max through cross-correlation with the determined candidate sequence, using equations 7, 8 and 9: where s _{jde k} is the selected candidate sequence, and all other symbols have the same meaning as that denoted previously. The timing and frequency error estimates are then compensated for in subsequent decoding of the received signal (step 958). At step 960 the receiver 102, 112 then proceeds, as with the first embodiment, to decode the PCC 408 using knowledge of the number of effective antennas . As described in further detail with reference to the first embodiment, the receiver then uses the information in the PCC 408 to decode the PDC 410 (step 962). Once the data packet has been decoded, or a predetermined number of such data packets, have been received and decoded, the transmitter 100, 110 and receiver 102, 112 may switch roles - i.e. to provide time division duplexing.

It will be appreciated that the ability to determine the number of effective antennas and details of the transmission mode used in the PDC 410 from the STF 204 and PCC 406 respectively means that these do not need to be signalled elsewhere - e.g. by occupying valuable bits in the PCC. Moreover it means that these parameters can easily be changed by the transmitter dynamically; they do not need to be negotiated while a new connection is established. This allows each device 100, 110, 102, 112 to easily to change the transmission mode to suit the channel conditions it notices whilst in receive mode.