GB2467146A - A user selection method in multi-user MIMO, based on quality of service, which enables communication with legacy devices - Google Patents

Abstract

A base station in a multi-user multiple-input-multiple-output environment (MU-MIMO) is operable to determine presence of local communications devices with which the base station can communicate. Some of the local communication devices may be designed to enable MU-MIMO communication, whilst others will be legacy devices not normally capable of MU-MIMO communication. The base station first assigns a target legacy device it is to communicate with, and then allocates one or more MU-MIMO enabled devices it can simultaneously communicate with, without causing unacceptable interference with the legacy device communications. This is achieved by sending a first pilot signal precoded using a first precoding to the target device and receiving from the target device a feedback signal indicating the quality of the first pilot signal. This information is used to establish a second precoding which is orthogonal to the first precoding. The second precoding is used to send a second pilot signal to a plurality of other devices, such as MU-MIMO enabled devices. These other devices send back to the base station information on the quality of the received second pilot signal, which is used to select one or more MU-MIMO devices for communication with the base station at the same time as the legacy device.

Description

Wireless Communications Methods and Apparatus The present invention is in the field of wireless communication and particularly, though not exclusively, the field of multiple input multiple output (MIMO) communications. It is more particularly applicable to multi-user MJMO (MU-MIMO).

MU-MIMO arrangements as currently presented in the art are advantageous as they allow for a direct gain in multiple access capacity, by allowing the spatial multiplexing gain at the base station (BS) to be obtained without the need for multiple antenna terminals. This permits the development of small and relatively inexpensive terminals, with processing power and unit cost is focused on the "infrastructure" part of a MU-MIMO network, i.e. the base station or the like.

In order to support MU-MIMO, precoding has been widely and conventionally employed. Precoding is often termed "beamforming" and will be readily visualised by the skilled reader in that context. Examples of precoding include linear precoding and non-linear precoding.

Linear precoding is a generalization of traditional SDMA (Space Division Multiple Access), in which each user is assigned a distinct precoding matrix at a transmitter. In the field of the invention, such matrices are themselves often known as "precoders".

Such precoders are designed on the basis of channel state information (CSI) with respect to the users, with a wide variety of specific approaches being employed.

In considering precoding performance, it has been claimed that non-linear precoding achieves higher performance than linear precoding. Fundamentally, non-linear precoding involves additional transmit signal processing to improve error rate performance.

However, any approach using linear or non-linear precoding can suffer high complexity and cost. The most substantial cost arises because MU-MIMO requires (but benefits from) CSI at the transmitter in order to properly serve the resultant spatially multiplexed users. This is described in "Shifting the MIMO Paradigm" (David Gesbert, Marios Kountouris, Robert W. Heath Jr., Chan-Byoung Chae, and Thomas Salzer, IEEE Signal Processing Magazine, September 2007).

Many MU-MIMO algorithms are currently developed based on full knowledge of CSI, which in fact has a severe impact on the deployment of MU-MIMO, and especially imposes restriction on the ability to support devices designed with an earlier, simpler communications implementation (so called "legacy support").

It would thus be desirable, having regard for this, to provide a MU-MIMO arrangement which reduces the need for significant information feedback and also supports legacy systems.

According to an aspect of the invention, there is provided wireless communications apparatus operable to act as a base station in a multi user environment, the apparatus comprising: a plurality of transmit antennas, device discovery means operable to determine presence of local communications devices with which the apparatus can communicate, device capability determination means operable to determine if one or more of the devices can support multi user communication, and device allocation means operable to interrogate devices to determine, in the event that one or more devices can support such multi user communication, which of said one or more devices can be allocated to communicate across the same channel, with regard to orthogonality of respective paths and interference at said devices and said apparatus, the device allocation means including: initial target device selection means operable to select one of said discovered devices as an initial target device, first precoding determining means operable to generate a first precoding operable to cause a signal emitted at the antennas to be directed substantially to said selected device, first pilot signal generation means operable to cause emission of a signal at the antennas precoded by the first precoding, said first pilot signal defining an enquiry for said device to send back information concerning quality of said first pilot signal, second precoding determining means operable on the basis of information fed back to said apparatus from said initial target device to detennine a second precoding which, when applied to a signal, is substantially orthogonal to said first precoding, and second pilot signal generation means operable to cause emission of a signal at the antennas precoded by the second precoding, said signal being such as to request a device in receipt of said signal to respond by measuring channel quality and feeding back information concerning said channel quality to said apparatus, wherein said device allocation means is operable to select from available fed back information a device capable of operating in a multi user environment which can be communicated with alongside the initial target device without causing undue interference either to communication between the apparatus and the initial target device or to communication between the apparatus and the selected device.

According to another aspect of the invention there is provided a method of scheduling devices in a multi user multiple input multiple output (IvlU-MIMO) environment comprising determining at a multiple antenna base station the presence of local communications devices with which the base station can communicate, determining if one or more of the devices can support multi user communication, and interrogating devices to determine, in the event that one or more devices can support such multi user communication, which of said one or more devices can be allocated to communicate across the same channel, with regard to orthogonality of respective paths and interference at said devices and said apparatus, the interrogating including: selecting one of said discovered devices as an initial target device, generating a first precoding operable to cause a signal emitted at the antennas to be directed substantially to said selected device, emitting a signal at the base station precoded by the first precoding, said first pilot signal defining an enquiry for said device to send back information concerning quality of said first pilot signal, generating a second precoding on the basis of information fed back to said base station from said initial target device which, when applied to a signal, is substantially orthogonal to said first precoding, and emitting a signal atthe antennas precoded by the second precoding, said signal being such as to request a device in receipt of said signal to respond by measuring channel quality and feeding back information concerning said channel quality to said apparatus, and selecting from available fed back information a device capable of operating in a multi user environment which can be communicated with alongside the initial target device without causing undue interference either to communication between the apparatus and the initial target device or to communication between the apparatus and the selected device.

According to another aspect of the invention, a base station in a multi user MIMO environment is operable to determine presence of local communications devices with which the base station can communicate. The station is then operable to determine if one or more of the devices can support multi user communication. In the event that one or more devices can support such multi user communication, the base station interrogates the devices to determine which of the devices can be allocated to communicate across the same channel, with regard to orthogonality of respective paths and interference at the devices and the base station.

Aspects of the invention can be delivered by was of application specific hardware, or by software on a suitable hardware configuration. For instance, a software product, which could be delivered on a carrier medium such as a storage medium or a signal medium, could be loaded into a suitable general purpose communications apparatus with multi-antenna MIMO capability, to deliver one or more of the above aspects of the invention.

Embodiments of the invention will now be described by way of example with reference to the accompanying drawings, in which: Figure 1 is a schematic diagram of a communications network employing an embodiment of the present invention; Figure 2 is a flow diagram of a process, in accordance with a specific embodiment of the invention, for seeking information for allocation and scheduling of devices to a MU-MIMO channel; Figure 3 is a diagram illustrating communications steps performed in accordance with the specific embodiment of the invention; and Figure 4 is a schematic diagram illustrating the embodiment of the invention in use.

In MU-MIMO, as mentioned above, there are two typical precoding types: linear precoding and non-linear precoding. The choice of approach trades off between performance and complexity. As discussed previously, non-linear precoding generally provides a higher performance than linear precoding, but at higher computational cost.

However, it has been found that linear precoding, with effective user selection, may approach the performance of non-linear precoding with much lower computational complexity. In the embodiments which will be described below, it will be seen that a hybrid mechanism can take full advantage of beamforrning and precoding, with both linear and non-linear precoding.

In order to demonstrate the application of embodiments of the invention, figure 1 illustrates a network 10 employing an embodiment of the invention. This shows a MU-MIMO configured base station 20 in accordance with an embodiment of the invention, a number of legacy configured mobile devices 30 (i.e. devices not expected to be able to comply with the requirements imposed by the communications protocols developed for the performance of the specific embodiment of the invention) and mobile devices 40 which are configured to be able to interact with the base station 20 in a manner not

contemplated in the prior art.

It is a practical assumption, and one made for the purpose of exemplifying the present invention, that none of the legacy mobile devices 30 support the functioning of MU-MIMO.

In comparison and contrast, a conventional MU-MIMO implementation might be only able to support MTJ-MIMO capable mobile terminals. Such a network would not be able to support legacy terminals. Furthermore, a conventional IvfU-MIMO could only be operated on one precoding mode, that is either linear precoding or non-linear precoding. As will become clear from the following, the presently described embodiment is able to provide a mechanism of MTJ-MIMO operation to support both new and legacy mobile terminals.

In designing a new communications system with the intention of offer legacy support, there might be several ways to fulfil this requirement. One way is to set up a specific legacy zone in which only those legacy users are allowed and supported. Another way is to support legacy users within a new system frame but dynamically allocate legacy resources. In OFDMA (Orthogonal Frequency Division Multiple Access), the dynamic resource allocation might be one or several data bursts, or say, subchannels. It is reasonable to assume that legacy transmission is mainly single user SISO/MIMO.

Furthermore, even for new users, some transmissions might also be single user SISO/MIMO, depending on transmission conditions and/or QoS (Quality of Service) requirements.

In fact, for those terminals employing SISO transmission (either permanently, in the case of legacy devices, on from time to time in the case of new devices), such a transmission would, strictly speaking, be in the MISO format, since the base station 20 comprises multiple antennas. Even though MIISO transmission can provide diversity gain, it fundamentally reduces radio resource efficiency. It is possible to take advantage of multiple-antenna configuration on the base station 20 in order to perform multiple user transmission alongside the SISO/MISO transmission.

It is worthwhile to emphasize here that the SISO/MISO transmission might always exist in either legacy system andlor new system, depending on transmission conditions and the required Q0S. Consequently, in such a case, the system in accordance with the present embodiment may have already achieved the basic service at a required Q0S, merely by using SISO/MISO transmission. With the multi-user transmission facility provided additionally in the present embodiment, system capacity naturally increases, as long as the impact of the introduced multi-user transmission on the underlying SISO/MISO is minimized.

Furthermore, even if a terminal is capable of communication with the base station in MIMO format, the terminal in question might not need to occupy the full spatial domain, which leaves freedom to deploy more terminals in space domain not required by that terminal. Consequently, the present embodiment can be characterised as follows, and as illustrated in figure 2: 1. Terminals are allocated to the base station, including new terminals and legacy terminals, in a conventional manner, such as by time/frequency multiplexing.

2. It is then determined whether any of the terminals have the potential for freedom in the spatial domain. That is, whether any of the terminals can be configured in accordance with a specific embodiment of the present invention.

3. If the above step identifies that the terminals are all legacy devices, or for some other reason, MU-MIMO cannot be supported, then any legacy SU-MIMO transmission regime is employed.

4. On the other hand, if one or more of the terminals can support MU-MIMO, then a legacy terminal is selected as a "target user" MST.

5. A beam is produced consistent with transmission with the selected target user MST.

6. CQI feedback is then requested from the MST if this is necessary for scheduling or other purposes, as this provides information about the strength of the beam.

Each other allocated user could also calculate CQI or other information from the beam, which would constitute interference to that other user if the target user were to join the transmission.

7. Based on the previously produced beam (or, as will be seen below, beams), a beam orthogonal to that or those previously produced is itself produced. This new beam is transmitted exclusively, as a test transmission.

8. The base station then sends a request to the target user MST for a CQI relating to the orthogonal beam. The CQI is sent back from MST. Similarly, the base station sends an enquiry to all MSs already identified as potential users to be added "on top" of the existing legacy system, for a channel quality indication (CQI) related to the orthogonal beam. In this embodiment, it may be useful to request Signal to Interference and Noise Ratio (SINR) to be included in the fed back CQI, as this includes a measure of interference which will be useful in subsequent steps.

The purpose of this enquiry in step 8 is to seek such information (and, generally, if an MS seeks connection to the channel, such information should be provided) but, if the enquiry is not responded to for whatever reason, this does not jeopardise operation of the network. Indeed, the base station can surmise from this that conditions are not suitable for transmission to the MS concerned. Another possibility can be that the MSs are aware of the signalling scheme and expect another beam which they may measure, or not, as the case may be. MSs are required to send back their CQI if they decide that the transmission conditions are suitable, for example if the interference from the previously measured first beam to MST (item 6) is not too high for useful and practical transmission to be conducted. It may also be that feeding back CQI information is mandatory in certain circumstances and networking arrangements, such as to improve scheduling at the BS.

Regarding the nature of the CQI, this can include one or more items of information such as Signal to Noise Ratio (SNR), SINR as previously mentioned, signal strength, and so on.

9. Then, a determination is made, from the measures received, as to the extent to which the introduction of yet another MS, beyond the MS1, would affect performance from the perspective of interference. This can be conveniently carried out by setting an "interference budget" and then determining if this would be exhausted by allocation of a further MS.

10. The identified MS with the most orthogonality from other devices currently planned to be assigned to the network, is thus assigned to the network.

While the interference budget is not exhausted, a further attempt is made to add further MSs to the network. In computer programming terms, a WHILE loop is formed.

11. Once the above steps have been carried out, the resultant multiple beams are refined with power control and an interference control mechanism, which can be implemented using standard approaches.

12. Finally, the data to be sent to the respective terminals can be precoded onto the multiple beams, to achieve MU-MIMO transmission.

Any further extension of this procedure is possible and extendable for an experienced engineer, such as performing detection/measurement and feedback simultaneously amongst terminals, including target terminal and other potential multiple terminals, etc. Also note that, any existing mechanisms, such as CQI measurement including CNR (Carrier to Noise Ratio), CINR (Carrier to Interference plus Noise Ratio), etc and CQI feedback, can be directly adopted in this mechanism with or without modifications.

The mechanism identified above is a hybrid opportunistic multi-user access scheme to enable MU-MIMO transmission. With the target MST setup, the system is confident to achieve, at least, the basic system capacity andlor quality of service. On top of this as described, if the support of the MS1-still has potential spatial freedom, the mechanism enables more user access on the same radio resource but based on spatial separation.

The system capacity and efficiency are improved with higher expected Q0S.

In the above procedure, step 1 specifies the basis of transmission for legacy transmission. This approach requires the system to operate on a specified manner of transmission (such as in accordance with a predetermined standard, such as an International Standard) without any modification. In this type of transmission, some mobile terminals with single antenna, or say, fewer antennas than that on the BS, might be allocated for transmission. In this case, the mobile terminals have the potential to use less spatial freedom, which can be identified by the BS.

However, these MSs are to be allocated according to system requirements and requested QoS. By supporting these MSs, the system will already meet the design requirements.

As described previously, the presently described embodiment provides a further facility to enable improvement of system performance/capacity on top of these MSs' transmission.

On item 4, by performing beamforming on the first target MS (MST), more spatial freedom is achieved. It enables the system to take advantage by deploying extra transmission through the freed spatial domain, which eventually and virtually realises the MU-MIMO transmission. For the beamforming operation of MST, there are two typical application scenarios: 1. The target MS (MS1) is a MS of the legacy system In this scenario, the target mobile terminal is a MS of the legacy system which has no MU-MIMO capability, and therefore no channel feedback is available. However, as has been extensively discussed in the literature, legacy systems support beamforming on link transmission. Therefore, the newly proposed mechanism takes advantage of this to perform beamforming on the target legacy user, such as by code book. Since the BS has full knowledge of the formed beam pattern, then, based on the formed beam, the BS is able to form other beams on the null space of the formed beam.

The BS forms virtually orthogonal beams and requests feedback from MSs, which will include the target legacy MST. In particular, when a search is carried out to determine if another user can share the resource block with the target user, the result is that the legacy MST will have been asked twice to report CQI. The first CQI reports the received signal strength of its beamformed channel, and the second CQI indicates the signal strength of the second (substantially orthogonal) beam, which actually represents the interference power that would arise at the legacy MST if the second beam were to be used for an additional MS. Essentially, this means that a measure of interference has been obtained from the legacy device without requiring the device in question to have any specific functionality to be able to report SINR or the like.

This is because, although the second beam would, in theory at least, be virtually orthogonal to the first, imposition of that beam might not guarantee that the legacy MST is free from interference. A variety of unpredictable reasons may give rise to an anomalous result at the MST, despite the robustness of the method by which the orthogonality of the second beam is determined. For instance, multipath effects might be unpredictable in a real set of circumstances. The feedback of the two CQI reports might be fed back simultaneously, or in different transmission slots, depending on implementation.

By not relying on the theoretical case, and by including this simple additional CQI report for the pilot signal on the second (virtually orthogonal beam), a more suitable MS scheduling allocation can be established.

Based on these two CQI reports for the MS-r, the BS will determine whether it is possible to utilize the second beam for an additional user. On the other hand, CQIs fed back from other MSs indicate their respective SINR, in which the interference from the legacy MST is included. Therefore, the MST is only required to feed back signal strength and other MSs are required to feed back SINR. The signalling between base station and users is summarized in Figure 3.

2. MST is a MS of the new system In this scenario, the target mobile terminal is a MS of the new system which has full MTJ-MIMO functionality. In this case, the beamforrning to the MST can be performed as in scenario 1. However, since the MS1 can in this case be assumed to have certain further capabilities, in particular that it is able to feed back full channel information, it becomes feasible for the system to produce other random channels which are orthogonal to the channel of the MST. Consequently, the BS can perform proper precoding on the known MIMO channels, including the one for the target MS.

For scenario 2, the BS has already formed a beam to the MST. Hence, the MST is not required to feed back any further information. However, other MSs are required to feed back information, but it is only necessary for them to feed back their SINR values.

It should be noted by the reader that the references above to other MSs' may include MSs of new and legacy systems, depending on the available features of legacy systems.

Consequently, following the procedure set out above and as illustrated in figure 2, the QoS of the MST is guaranteed. Transmissions by all other additional MSs can be accommodated without interference power exceeding the interference budget of the MST. Thus, any additional users' transmission will boost the system capacity on top of the MS1, and therefore improve the system efficiency.

More detailed explanation of the procedure and operation will be presented in due course when examples of implementation of the specific embodiment are described.

The main challenge of a new system design with l4TJ-MIMO concerns its ability to interact with legacy devices. In this case, the obstacle to providing legacy support by a system designed to overlay MTJ-MIMO capability, is that a legacy device is capable only of feeding back limited information. For example, no channel feedback is available for a legacy mechanism. The presently described embodiment of the invention overcomes such problems by enabling MU-MIMO on both legacy and new users with hybrid opportunistic access based on conventional user/users allocation and assignment.

The maj or difference between the presently described embodiment and those of the prior art legacy devices, concerns system operation. For a conventional MU-MIMO, the BS needs to know CSI on all the users in the MU-MIMO operation, which might be impossible for a legacy user to provide. For a conventional opportunistic beamforming scheme, the BS performs entirely random beamforming, which does not promise any QoS guarantees, and which might also not be feasible to apply directly on a legacy zone. The described hybrid opportunistic access approach takes full advantage of conventional MU-MIMO and opportunistic beamforming to achieve high system capacity and efficiency with less complexity and low feedback.

Additionally, if an MS were provided having full knowledge of the embodiment of the invention, for example the signalling exchange and timing of the beams, the interference powers could be estimated at the BS and at new MSs by taking into consideration the active beams only. The other available beams, not in use, would not be considered when the SINR is calculated at potential users. Therefore, the BS could in these circumstances schedule additional users for MU-MIMO downlink transmission using a greedy scheduling algorithm.

Furthermore, multiple BSs can cooperate to avoid strong interference to users in neighbour cells by sharing information on their respective generated precoding matrices.

In addition to all advantages described previously, the hybrid opportunistic access approach identified above and forming an exemplary embodiment of the invention is also able to make flexible deployment of all existing beamforminglprecoding algorithms/architectures, which enables system optimisation on both legacy and new system.

Secondly, the feedback requirement of this new way of supporting MU-MIMO is minimal. With the system providing more feedback, the mechanism can further improve system capacity.

Furthermore, the new technology as described above enables legacy systems to adopt advanced technology to achieve higher system capacity.

Finally, the new mechanism simplifies scheduling, which it will be clear to the skilled reader has an impact on system efficiency. In this arrangement, a single user is scheduled as a conventional system without any extra, or say, advanced scheduling.

Two practical applications of the proposed opportunistic MU-MIMO access mechanism will now be presented for further clarifications and explanations.

In a first example of the specific embodiment, the MST is a "new system" MS. The example assumes that the MS allocated for transmission has fewer antennas than the base station. It is now possible to set, as MST, the MS which employs less spatial freedom than the freedom of the system. For simplicity, the example has the BS with two antennas and all of the MSs (including the MS1) with only one, as shown in figure 4.

Because the MST is a "new system" MS, it could be possible for the MS1 to feed back (or estimate) the full CSI. Consequently, in this example, it is assumed that the full CSI of the MS1 is known to the BS as indicated in the figure.

In the downlink, illustrated in figure 4, the received signal at the k-th receiver can be written as Yk = Hkx + flit, for k 1,2,3, (1) where Hk is the channel of k-th receiver, X is the transmit signal and is the Gaussian distributed noise.

To explain the concept of linear precoding, the reader should consider the scenario where this time Sk and k denote the k-th transmit symbol vector (for beamforming scenario, 5k is a scalar symbol), and the additive white Gaussian noise vector. The actual transmitted signal vector for user k is then given by WIcSk, where Wk denotes the beamforming matrix for the k-th user.

It is assumed that service will be provided to a set of K selected users (among all active ones). Scheduling algorithms as discussed in the sequel can be applied to perform this selection across possible subsets. The received signal vector at the k-th user is Yk HkWksk + Hk + k 1Usk,forkl,2,3, (2) It is normally assumed that each user has Mk antennas and will decode the Sk Mk streams that constitute its data. The goal of linear precoding is to design [Wk} and [Wjll!=k based on the channel matrix knowledge, so a given performance metric is maximized for each stream.

Conventionally, the approach for fmding the precoder is to pre-multiply the transmitted signal by a suitably normalized ZF (Zero Forcing) or MMSE (Minimum Mean Square Error) inverse of the multi-user matrix channel. However this requires knowledge of all transmit channels of all MSs, which is not available in reality.

For simplicity, in this example only 2 users (Kr=2) are provided and one of the two users is MST. It is assumed that the full CSI HT of MST is known at the BS. Based on equation (2), the received signal vector at the MST is YT HTWTST + HW0So + (3) Consequently, to maximize the SNR performance of yT, the beam W is set to be the maximal ratio transmit beam, t,t'r. The second beam is orthogonal to the channel of the target user, such that HTWO 0.

Clearly, this requires the beams to be orthogonal but this orthogonality is only to the target MS. Physically, based on the known H1, there are two main ways for the BS to produce orthogonal beams with the multiple-antenna configuration. One way of doing so is to form a beam on null space as indicated in the above equation.

Another way of producing orthogonal beams is to use channel processing. In this approach, the BS may produce a relevant and orthogonal channel, say, with the same dimension as the H1 It should be noted by the reader that the orthogonal channel o is a virtual channel produced by the BS. Based on H1 and H0, a conventional precoding can be performed, either on linear or non-linear precoders. After the precoding, it is virtually orthogonal between the MST and the other users (which is not a known user, or say, an assigned user). It is initially guaranteed that the beam formed by the virtual channel is orthogonal to the channel of MST.

As shown in figure 4, beam-i' is the beam to MST and beam-2' is orthogonal to beam-i', which has no interference to MST. Further following the procedure of the illustrated mechanism, beam-2 is generated and broadcasted. Potential users send back their received SINR and the BS is then always capable of finding the mobile station which is most orthogonal to MS1 and thus has the strongest received signal.

Nevertheless, it might be possible that the additionally selected user terminal, while having received a strong and useful signal, also has received strong interference from the target user's transmission. To avoid this problem, a simple procedure is added to ensure that, once an extra terminal is selected, the interference due to target user's link is controllable. This procedure is described in the following.

First, the BS broadcasts a pilot signal precoded by the first beam, that is the beam directed to the target user MST. All potential additional users MS measure the received signal powers; such users will consider these signals as interference from the target user. Second, the pilot signal precoded by the second beam, that is the beam orthogonal to the target user's channel, is broadcast. All potential MSs (which may be all MSs in receipt of the message, or only those which actually are intended to participate in communication) again measure the received signals and compute S1NR, which is then sent back to the BS in the CQI via a fast feedback channel. In this way, the BS can select another user with acceptable S[NR for MIJ-MIMO transmission.

Further to this application, another typical application is that the MST is a legacy MS, that is, it is incapable of handling MTU-MIMO. This behaves in much the same way as with the above example, but for the fact that the legacy MS might not be able to provide full channel information to the BS. However, it is reasonable to presume that the legacy MS is able to perform single user precoding without channel feedback, such as codebook beamforming. Consequently, the BS forms a beam to the MST. With the known beam pattern for the MST, the BS is able to form other beam(s) on null space to support orthogonality. Hence, the rest of the procedure is similar to that presented in figure 2 as well.

It is useful to mention here that the MU-MIMO support is in addition to the single user transmission associated with the MST. If, and only if, the system ensures the capacity gain of the MU-MIMO, the MU-MIMO transmission will be formed. Otherwise, it operates only on the single user transmission.

Claims (18)

1. CLAIMS: 1. Wireless communications apparatus operable to act as a base station in a multi user environment, the apparatus comprising: a plurality of transmit antennas, device discovery means operable to determine presence of local communications devices with which the apparatus can communicate, device capability determination means operable to determine if one or more of the devices can support multi user communication, and device allocation means operable to interrogate devices to determine, in the event that one or more devices can support such multi user communication, which of said one or more devices can be allocated to communicate across the same channel, with regard to orthogonality of respective paths and interference at said devices and said apparatus, the device allocation means including: initial target device selection means operable to select one of said discovered devices as an initial target device, first precoding determining means operable to generate a first precoding operable to cause a signal emitted at the antennas to be directed substantially to said selected device, first pilot signal generation means operable to cause emission of a signal at the antennas precoded by the first precoding, said first pilot signal defming an enquiry for said device to send back information concerning quality of said first pilot signal, second precoding determining means operable on the basis of information fed back to said apparatus from said initial target device to determine a second precoding which, when applied to a signal, is substantially orthogonal to said first precoding, and second pilot signal generation means operable to cause emission of a signal at the antennas precoded by the second precoding, said signal being such as to request a device in receipt of said signal to respond by measuring channel quality and feeding back information concerning said channel quality to said apparatus, wherein said device allocation means is operable to select from available fed back information a device capable of operating in a multi user environment which can be communicated with alongside the initial target device without causing undue interference either to communication between the apparatus and the initial target device or to communication between the apparatus and the selected device.
2. 2. Apparatus in accordance with claim 1 wherein said device allocation means is operable to determine if yet further devices can be communicated with alongside said initial target device and said selected device.
3. 3. Apparatus in accordance with claim 1 or claim 2 wherein said device allocation means includes interference information storage means operable to store interference information and wherein said device allocation means is operable to cease attempts to select a further device when stored interference information exceeds an acceptable level.
4. 4. Apparatus in accordance with any one of the preceding claims wherein said device discovery means is operable to generate a signal at the antennas which contains an invitation for devices in receipt of said signal to respond with a signal bearing technical capability information.
5. 5. Apparatus in accordance with claim 4 wherein said initial target device selection means is operable to select an initial target device on the basis of technical capability information received for said discovered devices.
6. 6. Apparatus in accordance with claim 5 wherein said initial target device selection means is operable to sort discovered devices in terms of technical capability and to select a device as initial target device on the basis of having least technical capability of discovered devices.
7. 7. Apparatus in accordance with any preceding claim wherein said first pilot signal generation means is operable to generate a first pilot signal directed specifically at the initial target device.
8. 8. Apparatus in accordance with any preceding claim wherein said second pilot signal generation means is operable to generate a second pilot signal suitable for receipt and action by any discovered device.
9. 9. Method of scheduling devices in a multi user multiple input multiple output (MU-MIMO) environment comprising determining at a multiple antenna base station the presence of local communications devices with which the base station can communicate, determining if one or more of the devices can support multi user communication, and interrogating devices to determine, in the event that one or more devices can support such multi user communication, which of said one or more devices can be allocated to be communicated across the same channel, with regard to orthogonality of respective paths and interference at said devices and said apparatus, the interrogating including: selecting one of said discovered devices as an initial target device, generating a first precoding operable to cause a signal emitted at the antennas to be directed substantially to said selected device, emitting a signal at the base station precoded by the first precoding, said first pilot signal defining an enquiry for said device to send back information concerning quality of said first pilot signal, generating a second precoding on the basis of information fed back to said base station from said initial target device which, when applied to a signal, is substantially orthogonal to said first precoding, and emitting a signal at the antennas precoded by the second precoding, said signal being such as to request a device in receipt of said signal to respond by measuring channel quality and feeding back information concerning said channel quality to said apparatus, and selecting from available fed back information a device capable of operating in a multi user environment which can be communicated with alongside the initial target device without causing undue interference either to communication between the apparatus and the initial target device or to communication between the apparatus and the selected device.
10. 10. Method in accordance with claim 9 and including determining if yet further devices can be communicated with alongside said initial target device and said selected device.
11. 11. Method in accordance with claim 9 or claim 10 including storing interference information and ceasing attempts to select a further device when stored interference information exceeds an acceptable level.
12. 12. Method in accordance with any one of claims 9 to 11 and including generating a signal at the base station which contains an invitation for devices in receipt of said signal to respond with a signal bearing technical capability information.
13. 13. Method in accordance with claim 12 including selecting an initial target device on the basis of technical capability information received for said discovered devices.
14. 14. Method in accordance with claim 13 including sorting discovered devices in terms of technical capability and to select a device as initial target device on the basis of having least technical capability of discovered devices.
15. 15. Method in accordance with any one of claims 9 to 14 wherein said first pilot signal is directed specifically at the initial target device.
16. 16. Method in accordance with any one of claims 9 to 15 wherein said second pilot signal is suitable for receipt and action by any discovered device.
17. 17. Computer program product comprising computer executable instructions operable, when loaded on a computer controlled communications apparatus, to cause the communications apparatus to become configured in accordance with any one of claims ito 8.
18. 18. Computer program product comprising computer executable instructions operable, when loaded on a computer controlled communications apparatus, to cause the communications apparatus to perform a method in accordance with any one of claims 9 to 16.