ATTENUATION

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application 61/171,328, filed April 21, 2009, whose disclosure is incorporated herein by reference.

FIELD OF THE DISCLOSURE

[0002] The present invention relates generally to communication systems, and particularly to methods and systems for communication using multiple antennas.

BACKGROUND

[0003] Some communication systems transmit data from a transmitter to a receiver over multiple communication channels, using multiple transmit antennas and multiple receive antennas. Multiple-channel transmission is used, for example, in spatial multiplexing schemes that achieve high throughput, in beam-forming schemes that achieve high antenna directivity and in spatial diversity schemes that achieve high resilience against channel fading and multipath. These schemes are often referred to collectively as Multiple-Input Multiple-Output (MIMO) schemes.

[0004] MIMO schemes are contemplated, for example, for use in Evolved Universal Terrestrial Radio Access (E-UTRA) systems, also referred to as Long Term Evolution (LTE) systems. The Third Generation Partnership Project (3GPP) E-UTRA standards specify MIMO schemes for use by E-UTRA User Equipment (LTE) and base stations (eNodeB's). These schemes are described, for example, in 3GPP Technical Specification 36.211, entitled "Technical Specification Group Radio Access Network; Evolved Universal Terrestrial Radio Access (E-UTRA); Physical Channels and Modulation (Release 8)," (3GPP TS 36.211), version 8.6.0, March, 2009, which is incorporated herein by reference. The 3GPP is currently in the process of specifying an extension of the E-UTRA specification, which is referred to as LTE-Advanced (LTE-A).

[0005] Opportunistic beam-forming is a communication technique in which a transmitter transmits a pattern of directional transmission beams that alternates over time. The transmitter determines a suitable scheduling for transmitting to a particular receiver, for example the optimal scheduling, based on feedback from the receiver. Opportunistic beam- forming schemes are described, for example, by Viswanath et al., in "Opportunistic Beamforming Using Dumb Antennas," IEEE Transactions on Information Theory, volume 48, No. 6, June, 2002, pages 1277-1294, and by Sharif and Hassibi, in "On the Capacity of MIMO Broadcast Channels with Partial Side Information," IEEE Transactions on Information Theory, volume 51, No. 2, February, 2005, pages 506-522, which are incorporated herein by reference. [0006] The description above is presented as a general overview of related art in this field and should not be construed as an admission that any of the information it contains constitutes prior art against the present patent application.

SUMMARY

[0007] An embodiment that is described herein provides a method for communication. The method includes receiving at a receiver from a group of two or more transmitters multiple Radio Frequency (RF) transmission beams that alternate in time and space and include at least first and second transmission beams. The method identifies that the first transmission beam causes interference to reception of the second transmission beam. Feedback is sent from the receiver to one or more of the transmitters, so as to cause the transmitters to attenuate the first transmission beam during transmission of the second transmission beam.

[0008] In some embodiments, the first and second transmission beams alternate in time and space in accordance with a pattern that is coordinated among the transmitters. In an embodiment, the first transmission beam is identified as causing the interference after identifying the second transmission beam as being preferable for receiving subsequent transmissions to the receiver. In a disclosed embodiment, the method includes identifying the second transmission beam as being preferable for receiving subsequent transmissions to the receiver, and sending the feedback includes sending a request to receive the subsequent transmissions over the second transmission beam. [0009] In some embodiments, identifying that the first transmission beam causes the interference includes measuring signal quality on at least some of the received transmission beams, and detecting the interference responsively to the measured signal quality. In an embodiment, measuring the signal quality includes receiving pilot signals on the at least some received transmission beams, and measuring the signal quality over the pilot signals. [0010] In an embodiment, identifying that the first transmission beam causes the interference includes predicting the interference for a future occurrence of the first and second transmission beams, and sending the feedback includes causing the transmitters to attenuate the first transmission beam during the future occurrence. In a disclosed embodiment, the method includes selecting a preferred time interval for receiving the signals on the second transmission beam, and sending the feedback includes causing the transmitters to attenuate the first transmission beam during the preferred time interval.

[0011] In an embodiment, receiving the transmission beams includes receiving signals conforming to a Long Term Evolution (LTE) specification. In another embodiment, receiving the transmission beams includes receiving at least one transmission beam that is transmitted jointly by two or more of the transmitters. In yet another embodiment, sending the feedback includes causing the transmitters to transmit data to at least one other receiver over the attenuated first transmission beam. [0012] There is additionally provided, in accordance with an embodiment that is described herein, a communication apparatus including a receiver and a processor. The receiver is configured to receive from a group of two or more transmitters multiple Radio Frequency (RF) transmission beams that alternate in time and space and include at least first and second transmission beams. The processor is configured to identify that the first transmission beam causes interference to reception of the second transmission beam, and to cause sending of feedback to one or more of the transmitters, so as to cause the transmitters to attenuate the first transmission beam during transmission of the second transmission beam. In an embodiment, a mobile communication terminal includes the disclosed communication apparatus. In an embodiment, a chipset for processing signals in a mobile communication terminal includes the disclosed communication apparatus.

[0013] There is also provided, in accordance with an embodiment that is described herein, method for communication. The method includes transmitting to a receiver from two or more transmitters multiple transmission beams that alternate in time and space and include at least first and second transmission beams. Feedback from the receiver is received at the transmitters. The feedback indicates that the first transmission beam causes interference to reception of the second transmission beam. Responsively to the feedback, the first transmission beam is attenuated during transmission of the second transmission beam.

[0014] There is further provided, in accordance with an embodiment that is described herein, a communication system including two or more transmitters. The transmitters are configured to transmit to a receiver multiple transmission beams that alternate in time and space and include at least first and second transmission beams, to receive from the receiver feedback indicating that the first transmission beam causes interference to reception of the second transmission beam, and, responsively to the feedback, to attenuate the first transmission beam during transmission of the second transmission beam. [0015] The present disclosure will be more fully understood from the following detailed description of the embodiments thereof, taken together with the drawings in which:

BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Fig. 1 is a block diagram that schematically illustrates a communication system that employs opportunistic beam-forming with selective beam attenuation, in accordance with an embodiment of the present disclosure;

[0017] Fig. 2 is a flow chart that schematically illustrates a method for communication using opportunistic beam-forming with selective beam attenuation, in accordance with an embodiment of the present disclosure; and [0018] Figs. 3 and 4 are graphs showing transmission protocols that use coordinated patterns of transmission beams, in accordance with embodiments of the present disclosure.

DETAILED DESCRIPTION OF EMBODIMENTS

[0019] Embodiments that are described hereinbelow provide improved methods and systems for opportunistic beam-forming. In some embodiments, a communication system comprises two or more transmitters that transmit to a plurality of receivers. The transmitters transmit multiple directional Radio frequency (RF) transmission beams, which alternate in time and space. In some embodiments, the transmission beams alternate in accordance with a pattern that is coordinated among the transmitters.

[0020] In some embodiments, each receiver receives the transmission beams and measures signal quality on the received transmission beams. Based on the signal quality measurements, the receiver identifies one or more preferred transmission beams over which it prefers to receive data, and one or more interfering beams that cause interference to the preferred beams. Typically although not necessarily, the receiver first identifies a preferred beam, and then identifies one or more beams that cause interference to the preferred beam. [0021] The receiver transmits feedback to the transmitters, which indicates the preferred and interfering beams. In an embodiment, the receiver transmits a request to attenuate the one or more interfering beams. Alternatively, the receiver reports the signal quality measurements as feedback, so as to enable the transmitters to select over which beam to transmit and which beam to attenuate. [0022] In some embodiments, the transmitters take various actions and apply various policies in response to the feedback received from the receivers. In an embodiment, the transmitters select over which beam or beams to transmit to a given receiver. As another example, based on received feedback, the transmitters decide to attenuate one or more of the interfering beams while transmitting one or more of the preferred beams. As yet another example, the transmitters select a preferred time at which to transmit to a given receiver over a certain preferred beam. Various other transmitter decisions and decision criteria are described herein. [0023] Unlike some opportunistic beam-forming schemes, in accordance with an embodiment of the disclosure, the transmitters proactively attenuate transmission beams that cause interference based on feedback relating to one or more received signals and/or specific requests from the receiver to attenuate one or more of the downlink signals. As a result, the disclosed techniques provide high signal quality, high throughput and small latency. The disclosed techniques are suitable for communication systems whose performance is limited by interference, such as some cellular networks. Performance gains achieved by embodiments of the present disclosure are achieved with a relatively small increase in communication overhead and system complexity.

[0024] Fig. 1 is a block diagram that schematically illustrates a wireless communication system 20 that employs opportunistic beam-forming with selective beam attenuation, in accordance with an embodiment of the present disclosure. System 20 operates in accordance with any suitable communication standard or protocol. In the present example, system 20 operates in accordance with the LTE or LTE-A specifications, cited above. Alternatively, the techniques described herein can be used in systems operating in accordance with any other suitable communication standard or protocol, such as IEEE 802.16 (also referred to as WiMAX), Wideband Code Division Multiplexing (WCDMA) and Global System for Mobile communications (GSM), for example.

[0025] System 20 comprises multiple transmitters, which transmit signals to multiple receivers. In the present example, the transmitters are embodied in LTE or LTE-A base stations (eNodeB's) 24, and the receivers comprise LTE or LTE-A compliant mobile terminals (UE's) 28. Although Fig. 1 shows three eNodeB's and a single UE for the sake of clarity, real-life systems typically comprise a large number of eNodeB's and UE's. Although the embodiments are described herein in the context of downlink transmissions from the eNodeB's to the UE's, in some communication systems the disclosed techniques are suitably adapted to the uplink.

[0026] Each eNodeB 24 comprises a downlink transmitter 32, which transmits downlink signals to UE's 28 via one or more antennas 36. (In Fig. 1, the internal eNodeB structure is shown in detail for only one of the eNodeB's, for the sake of clarity. The other eNodeB's in the system typically have a similar structure.) At any given time, the eNodeB's 24 in system 20 transmit M downlink Radio Frequency (RP) transmission beams, which convey respective downlink signals that carry data to UE' s 28. The transmission beams (referred to herein as beams for brevity) are typically directional.

[0027] In an embodiment, a directional transmission beam is produced by a single eNodeB using two or more antennas 36. Alternatively, a directional beam is produced by multiple eNodeB 's that transmit the same signal simultaneously, in which case each eNodeB uses one or more antennas 36. The group of eNodeB 's producing a given beam is referred to herein as a cluster. In the description that follows, each beam is regarded as being transmitted by a cluster of eNodeB 's. When a beam is transmitted by a single eNodeB, the cluster is regarded as containing only the single eNodeB.

[0028] Each transmission beam thus comprises a directional RF signal that carries data. Each beam is generated by transmitting the same data-carrying signal from a set of antennas 36, which belong to one or more eNodeB's, while applying respective multiplicative weights to the antennas in the set. In an embodiment, the direction in which a beam is transmitted is modified by modifying the weights applied to the antennas. This action is referred to as beam steering.

[0029] The beam steering applied to the / beam (i.e., the weights applied to the different antennas that participate in transmission of the / beam) is represented by a weight

vector denoted Wj . If the cluster of eNodeB's that generate the / beam is denoted Bj, and the

number of antennas in the k eNodeB in this cluster is denoted N^, then vector Wj has

Nfc elements. Each element of Wj is a multiplicative weight (typically a complex number having a magnitude and phase) that is applied to the signal transmitted via the respective antenna 36. In some embodiments, the weight vectors in a given cluster are

orthogonal to one another, i.e., Wj W j = O If B ₁ =B _j (iφj).

[0030] The set of M beams, which are transmitted by system 20 at a given time, is referred to herein as a beam setting. In some embodiments, system 20 modifies the beam setting at periodic time intervals. In other words, eNodeB's 24 transmit multiple transmission beams, which alternate over time and space. In LTE or LTE-A systems, for example, the beam setting can be modified every Transmission Time Interval (TTI), every several TTIs, or at any other suitable period. [0031] In some embodiments, the eNodeB's alternate the beams in time and space according to a pattern that is coordinated among the eNodeB's. In alternative embodiments, each eNodeB alternates the beams without coordination with other eNodeBs. Further alternatively, the eNodeB's may not use any sort of pattern. Typically, however, the eNodeB's synchronize the OFDM symbol timing with one another.

[0032] As will be explained below, in some embodiments the pattern of beam settings is predefined, whereas in other embodiments the pattern is pseudo-random. When using a periodic pattern, each eNodeB may change its local beam settings (i.e., the weight vector elements corresponding to its own antennas 36) in a certain cycle period, which may be the same as or different from the cycle periods of other eNodeB's. Typically, the beam settings are chosen to be sufficiently diverse, so that each UE is likely to encounter beam settings that provide high signal strength and low interference. In some embodiments, a cluster of eNodeB's generates and alternates the beams using an optimization process. The optimization process is often unknown to the UEs. [0033] When preparing to transmit data to a given UE 28, in an embodiment system

20 selects over which beam and at what time to transmit the data. Moreover, in some embodiments system 20 decides to attenuate one or more of the beams so as to reduce interference to another beam. In accordance with an embodiment of the disclosure, these scheduling and attenuation decisions are based on feedback that is provided by the UE' s, using methods that are described in detail herein.

[0034] Each eNodeB 24 comprises an uplink receiver 40, which receives uplink signals from the UE's (either via antennas 36 or via separate receive antennas, not shown in the figure). In particular, receiver 40 receives from the UE's feedback, which enables the eNodeB's to make scheduling and beam attenuation decisions. In some embodiments, the feedback comprises an explicit request from the UE to receive transmissions over one or more beams and/or to attenuate one or more beams. Alternatively, the feedback comprises signal quality measures, as measured by the UE on one or more of the received beams. Using this information, the eNodeB's decide over which beams to transmit and which beams to attenuate.

[0035] Each eNodeB 24 further comprises a controller 44, which manages the operation of the eNodeB. Controller 44 comprises a beam pattern generator 48, which generates the pattern of beam settings to be transmitted from transmitter 32. In an embodiment, the beam pattern generators 48 of the different eNodeB's 24 coordinate and synchronize the pattern generation with one another. [0036] In some embodiments, controller 44 comprises a beam attenuation module 52, which configures transmitter 32 to attenuate one or more of the transmitted beams. The beams to be attenuated are selected based on the feedback received from UE 's 28. Controller 44 further comprises a scheduler 56, which schedules data for transmission over the different beams. In particular, scheduler 56 decides over which beam to transmit data to each UE, and at what time (e.g., during which TTI). In some embodiments, controller 44 (e.g., using scheduler 56) selects the appropriate modulation and error correction coding scheme to be used for transmitting to each UE.

[0037] Each UE 28 receives the downlink transmission beams using one or more antennas 60. A downlink receiver 64 receives and decodes the signals transmitted over the different beams, so as to reconstruct and output the downlink data. In addition, receiver 64 measures the signal quality on each beam. In an embodiment, receiver 64 measures, for example, the Signal to Noise Ratio (SNR) on each beam, or any other suitable signal quality measure. [0038] As seen in Fig. 1, in an embodiment UE 28 comprises a processor 72, which manages and controls the operation of the UE. In some embodiments, processor 72 comprises a feedback calculation module 76, which determines the feedback to be transmitted from the UE to the eNodeB's. In an example embodiment, Module 76 determines the feedback based on the signal quality measurements performed by downlink receiver 64 on the different beams. In some embodiments, the feedback indicates (1) one or more beams that the UE regard as preferable for receiving downlink transmissions, (2) one or more beams that cause interference to downlink reception at the UE, and/or (3) a specific request to attenuate one or more of the transmitted beams.

[0039] This sort of feedback enables the eNodeB's to schedule downlink data over the preferred beams, and to attenuate interfering beams. Several examples of selection criteria and feedback schemes are described further below. UE 28 comprises an uplink transmitter 68, which transmits uplink signals to the eNodeB's. In particular, the uplink transmitter transmits the feedback produced by module 76.

[0040] The system configuration shown in Fig. 1 is a simplified example configuration, which is depicted for the sake of conceptual clarity. In alternative embodiments, any other suitable system configuration can also be used. In an embodiment, the different components of eNodeB's 24 and UE's 28 are implemented using dedicated hardware, such as using one or more Application-Specific Integrated Circuits (ASICs) and/or Field- Programmable Gate Arrays (FPGAs). Alternatively, in an embodiment, some eNodeB and UE components are implemented using software running on general-purpose hardware, firmware, or using a combination of hardware and software elements.

[0041] In some embodiments, controller 44 and processor 72 comprise general- purpose processors, which are programmed in software to carry out computer instructions to provide the functions described herein, although they too may be implemented on dedicated hardware. The software instructions may be downloaded to the processors in electronic form, over a network, for example. Alternatively or additionally, the software instructions are provided and/or stored on tangible media, such as magnetic, optical, or electronic memory. In some embodiments, some or all of the elements of UE 28, and/or some or all of the elements of eNodeB 24, are fabricated in a chip-set. UE and eNodeB elements that are not mandatory for explanation of the disclosed techniques, such as various Radio Frequency (RF) elements, have been omitted from Fig. 1 for the sake of clarity.

[0042] Fig. 2 is a flow chart that schematically illustrates a method for communication using opportunistic beam-forming with selective beam attenuation, in accordance with an embodiment of the present disclosure. The method begins at a beam transmission operation 80, with eNodeB's 24 transmitting a pattern of RF transmission beams. At a beam reception operation 84, UE 28 receives at least some of the beams using downlink receiver 64. At a quality reception operation 88, receiver 64 measures the signal quality on each of the received beams. Receiver 64 reports the signal quality measures of the different beams to processor 72.

[0043] At a beam identification operation 92, processor 72 identifies one or more preferable beams and/or one or more interfering beams, based on the signal quality measurements. At a feedback operation 96, uplink transmitter 68 transmits the feedback indicative of the preferred and interfering beams to the eNodeB's. At a scheduling and attenuation operation 100, eNodeB's 24 schedule subsequent transmissions to the UE based on the received feedback. In addition, the eNodeB's select and attenuate one or more of the beams based on the UE feedback. Note that the eNodeB's in a given cluster typically cooperate in making scheduling decisions, regardless of whether they coordinate the pattern of beam settings with one another. [0044] UE's 28 suitably calculate and report the feedback in various ways. For example, in an embodiment the UE calculates the SNR on each beam, assuming all the other beams are active (and therefore potentially cause interference). In addition, in an embodiment the UE calculates the SNR on each beam, assuming that one or more of the worst-interfering beams are turned off. In some embodiments, each UE is pre-allocated a set of beams, referred to as the serving beams of the UE. The serving beams of a given UE may be generated by a single eNodeB, by a cluster of eNodeB's, or by multiple clusters. In an embodiment, the UE measures signal quality and calculates the feedback while considering only its pre-allocated serving beams (i.e., disregards potential interference from beams other than the serving beams).

[0045] In an example embodiment, the UE indicates the most desirable beams in the feedback by reporting the m strongest beams, m≥\. In some embodiments, the UE reports the SNR measured on these beams, as well. In some embodiments, the UE identifies v≥l preferable transmission times over the preferable beams. In an embodiment, The UE reports the preferable transmission times in the feedback, possibly together with the achievable SNR at these transmission times.

[0046] In some embodiments, the UE identifies one or more beams that cause considerable interference to the reception of the preferable beams. For example, in an embodiment the UE identifies one or more beams, whose attenuation (e.g., silencing) would cause the highest improvement in SNR over the preferable beams. In an example embodiment, the UE reports the u≥\ beams that cause the strongest interference. In another example embodiment UE also reports the SNR improvement that is expected to result from attenuating the interfering beams. As another example, the UE reports all interfering beams, whose attenuation would improve the SNR over some preferred beam by at least Δ dB. [0047] In some embodiments, a group of two or more beams is addressed to a given

UE, and the UE comprises a mechanism for canceling the interference between different beams in the group (e.g., by decoding the signals of the different beams and subtracting one signal from another). In these embodiments, the UE disregards interference between beams within the group when identifying the preferable and interfering beams, assuming it will be able to cancel this interference internally. Additionally or alternatively, the UE may identify and report the preferable and interfering beams in any other suitable way.

[0048] Receiver 64 in the UE may measure the signal quality on the different beams in any suitable way. For example, in an embodiment the downlink signal transmitted over a given beam comprises pilot symbols that are uniquely associated with that beam. Receiver 64 measures the signal quality on a given beam by estimating the achievable SNR using the pilot symbols of the beam. For example, the receiver may estimate the channel response between the eNodeB's and the UE using the pilot signals. Based on the estimated channel and the known beam steering vector (weight vector), the UE estimates the achievable SNR on that beam.

[0049] The eNodeB's 24 of system 20 receive the above-described feedback from the different UE's 28, and configure the transmission beams in various ways based on the feedback. The eNodeB's suitably decide, for example, over which beam or beams to transmit to each UE and at which times, which interfering beams to attenuate and at which times, the actual attenuation level to applied to each attenuated beam, which modulation and coding scheme to use in transmitting to a given UE, and/or any other relevant decision. When making these decisions, the eNodeB's may attempt to attain certain performance measures over each scheduling interval. Any performance criterion, such as maximum system throughput, proportional fairness among UE's or any other suitable criterion can be used for optimization.

[0050] In an embodiment, the eNodeB decisions are performed separately and independently at each cluster. In various embodiments, the eNodeB's of each cluster communicate with one another for sharing the feedback they receive and for coordinating the beam setting patterns and decisions. Alternatively, the decisions are made by a centralized processor (not shown in the figures) for the entire system 20. In these embodiments, the different eNodeB's typically communicate with the centralized processor for reporting the feedback they receive and for accepting configuration instructions. In an example embodiment, the centralized processor instructs the eNodeB's, for each beam, (1) whether the beam is to be transmitted or not, (2) the UE to which the beam is addressed (several beams can be addressed to the same UE), and (3) the modulation and coding scheme to be used in transmitting the beam.

[0051] In various embodiments, attenuation of a given beam is suitably performed in various ways. In some embodiments, the eNodeB's inhibit transmission of interfering beams altogether. In alternative embodiments, the eNodeB's reduce the transmission power of the beam below its normal intended level by a certain amount. In some embodiments, attenuation factors on the order of 10-30 dB are used. In alternative embodiments, smaller attenuation factors, e.g., between 3-6 dB, are suitable for enabling acceptable SNR on the preferred beams. Further alternatively, any other suitable attenuation factors can be used. In some embodiments, the eNodeB's select the actual level of attenuation to be applied to each attenuated beam.

[0052] In some embodiments, a beam that is being attenuated is still used for transmitting signals to other UE's. For example, a beam that is attenuated by 3-6 dB, or more, so as not to interfere with transmission to a first UE, is often sufficiently strong to be useful for signal transmission to a second UE. In some embodiments, the eNodeB's consider this effect when deciding on the desired attenuation factors of different transmission beams, so as to improve the overall system throughput.

[0053] In some embodiments, a given UE may refrain from reporting interfering beams to the eNodeB's. For example, in an embodiment the UE may choose not to report any interfering beam if the interference is relatively small. As another example, a UE may be requested by the eNodeB's not to report interfering beams (e.g., in a specific system configuration).

[0054] As noted above, eNodeB's 24 alternate the beam setting over time in accordance with a certain pattern. In some embodiments, the beam setting is modified relatively slowly. In some cases, modifying the beam setting at a rate that is slower than once per several TTIs (e.g., once per ten TTIs or slower) is considered slow. In other cases, modifying the beam setting at a rate that is slower than once per TTI is considered slow. In these embodiments, the system typically has sufficient time to complete a feedback cycle (i.e., measure signal quality and report feedback by the UE 's, transmit downlink signals in response to the feedback by the eNodeB's, and receive the downlink signals by the UE 's), all while the same beam setting is still valid.

[0055] In some embodiments, the pattern of beam settings is known in advance, so that a UE measure signal quality and report feedback on a given beam setting, which will occur again at a known future time. In other words, the UE can measure the interference inflicted by a certain beam on another beam in a current beam setting, and predict that this interference will also occur in a future occurrence of this beam setting. In an embodiment, the eNodeB's use this feedback for making scheduling decisions and attenuating beams at the future occurrence of this beam setting. This technique enables the system, for example, to compensate for the delays associated with the feedback process. A technique of this sort is suitably used when the beam settings change rapidly, e.g., every TTI, although it is also applicable for slowly-varying patterns.

[0056] Fig. 3 is a graph showing a transmission protocol that uses a coordinated pattern of transmission beams, in accordance with an embodiment of the present disclosure. The present example refers to two eNodeB's, each transmitting a single beam. In the present example, the two eNodeB's change the beam setting every TTI. The first eNodeB applies a periodic pattern of beam settings having a period of L _\ TTIs. The weight vectors in this pattern

are denoted VVj (I) ₅ WJ (2) ,..., Wj (Z/[) . The second eNodeB applies a periodic pattern of beam settings having a period of Z2 TTIs. The weight vectors in this pattern are denoted

^2 O) '^2 (^) '" •'— 2(^2 ) - ^A g ^ra P ^{n 104 m Fi} §- ³ shows the f ⁱ rst pattern of beam settings transmitted by the first eNodeB, and a graph 108 shows the second pattern of beam settings transmitted by the second eNodeB. [0057] In each TTI, each eNodeB transmits pilot signals that enable the UE 's to measure signal quality on the two beams and calculate the feedback, as explained above. Since the two patterns are periodic, any occurrence of a pair of ^we ight vectors will occur again at periodic intervals. Thus, in an embodiment, a given UE suitably measures signal quality and reports feedback to the eNodeB 's on a certain occurrence of a certain weight vector. The eNodeB 's apply the decisions taken with respect to this pair of weight vectors (e.g., scheduling or attenuating beams) in a future occurrence of this pair of weight vectors. Note, however, that the feedback typically remains valid for only a limited length of time, until the channel response changes considerably. In another embodiment, the UE can estimate the communication conditions in the k TTI based on the signal measured at the {k-L\) and (k-L2) TTIs (and possibly other past measurements).

[0058] In the present embodiment, a given UE may be configured to report various kinds of feedback, such as the identity of the best-performing TTI out of the next five TTIs in the pattern, the identity of the beam that achieves the highest SNR, the achievable SNR in the best-performing TTI on the best-performing beam, and/or the achievable SNR at the best- performing TTI on the best-performing beam assuming the other beam is attenuated. In an embodiment, based on this feedback, the two eNodeB 's cooperate to decide on the preferred scheduling policy. The policy suitably involves, for example, attenuating one of the two beams during one or more of the TTIs. If Zj and Z2 have no common divisor, for example, the two eNodeB' s select the scheduling policy over L _\ -Li possible beam settings. [0059] Fig. 4 is a graph showing a transmission protocol that uses a coordinated pattern of transmission beams, in accordance with an alternative embodiment of the present disclosure. In the present example, the system comprises three eNodeB 's, referred to as eNodeB 1, eNodeB2 and eNodeBβ. eNodeB 1 has two transmit antennas and transmits two beams. eNodeB2 and eNodeBβ have one transmit antennas each, and jointly transmit two beams. Thus, Bi=B2={ 1 }, B3=B4={2,3}, and all weight vectors are of length 2. [0060] In the example of Fig. 4, the eNodeB's change the beam settings (weight vectors) relatively slowly - every several TTIs. The time period during which the beam setting is kept constant is denoted Tβ. In this configuration, the pattern of beam settings need not be periodic or even deterministic. For example, the pattern may be a pseudo-random pattern. [0061] In the configuration of Fig. 4, during the first few TTIs of Tβ (marked

"DATA" in the figure) the eNodeB's transmit both data and pilot symbols using the currently- valid beam setting. During a TTI marked "PLT", the eNodeB's continue to transmit data using the currently-valid beam setting, but switch to transmit pilot symbols using the beam setting of the next Tβ period. The pilot symbols used here suitably comprise dedicated pilot symbols. During this TTI, the UE' s are configured to perform signal quality measurements that are applicable to the next Tβ period, while at the same time continuing to receive data using the currently-valid beam setting. The UE 's typically measure the achievable SNR using each of the possible beams. In alternative embodiments, the pilots used in the "PLT" TTI suitably comprise common pilot symbols, assuming that all weight vectors are known to the UE. [0062] During a TTI marked "FBK", the eNodeB's still continue to transmit data using the currently-valid beam setting, and at the same time receive the feedback from the UE' s. This feedback that was calculated based on the measurements performed during the

"PLT" TTI, and is therefore applicable to the next Tβ period. The eNodeB's apply this feedback (e.g., make scheduling decisions and attenuate beams) in time for the next Tβ period. [0063] In the example of Fig. 4, the "FBK" TTI is the last TTI in the T _B period, and

"PLT" TTI precedes it. In alternative embodiments, the "PLT" and "FBK" TTIs are suitably positioned at any other suitable locations in the Tβ period, which allow sufficient time for calculating and transmitting the feedback, and for applying it in time for the next Tβ period.

[0064] In some embodiments, the feedback transmitted by the UE' s in the "FBK" TTI comprises, for example, the identity of the two best-performing beams out of the four possible beams, the achievable SNR using each pair of selected beams, the best achievable

SNR over the best-performing beam assuming one of the other three beams is attenuated, and/or the identity of a preferred beam to be attenuated.

[0065] After receiving this feedback from the UE's, the three eNodeB's cooperate to decide on the best scheduling scheme to be applied in the next Tβ period. In contrast to the example of Fig. 3 above, in which eNodeB cooperation was needed only for attenuation purposes, in the example of Fig. 4 eNodeB2 and eNodeB3 should perform the entire scheduling scheme together, since both eNodeB's transmit the same data over their two common beams. Note also that in the present example, several TTIs may be available for scheduling during the same Tβ period. Therefore, in the embodiment, the eNodeB's are configured to make several scheduling decisions using the same beam setting.

[0066] In some embodiments, the Tβ period is relatively long, so that changes in the communication channel during this period cannot be neglected. In these embodiments, the UE' s calculate and send additional feedback during the Tβ period, so as to allow better scheduling decisions. Additionally or alternatively, prediction techniques are suitably utilized to compensate for the feedback delay.

[0067] In some embodiments, the scheme of Fig. 4 is combined with other types of feedback, or with other channel information available to the eNodeB's. Inasmuch as each eNodeB is free to choose its weight vectors, it can use the additional channel information to produce transmission beams that are more likely to be favorable by some UE's. In an embodiment, the choice of weight vectors is performed with little or no degradation in downlink performance.

[0068] In embodiments described above, the disclosed techniques are carried out by all eNodeB's 24 of system 20 and over the entire spectrum allocated to system 20. Generally, however, the disclosed techniques are not limited to this sort of implementation by all eNodeB's. In alternative embodiments, for example, the disclosed techniques are carried out using only part of the eNodeB's in system 20, and/or over only part of the spectrum allocated to the system.

[0069] It is thus noted that the embodiments described above are cited by way of example, and that the present invention is not limited to what has been particularly shown and described hereinabove. Rather, the scope of the present invention includes both combinations and sub-combinations of the various features described hereinabove, as well as variations and modifications thereof which would occur to persons skilled in the art upon reading the foregoing description and which are not disclosed in the prior art.