Method and apparatus for interference mitigation in MU-MIMO wireless communication networks

By forming the first and additional beams in the MU-MIMO wireless communication network and adjusting the beam shape based on forward error correction and vector error calculation, the problem of dynamic changes in the interference source is solved, and interference with fast response and low signaling overhead is achieved Suppression improves communication quality.

CN112425086BActive Publication Date: 2024-11-01CAMBIUM NETWORKS
View PDF 10 Cites 0 Cited by

Patent Information

Application Number
CN201980047753.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2018-05-18
Filing Date
2019-05-16
Publication Date
2024-11-01
Estimated Expiration
2039-05-16

Smart Images

  • Figure CN112425086B_ABST
    Figure CN112425086B_ABST
Patent Text Reader

Abstract

In a multi-user multiple-input multiple-output MU‑MIMO wireless communication network, interference is mitigated. A first MU‑MIMO group is determined to allow corresponding transmission data to be received from wireless stations in the MU‑MIMO group at the same time at a first wireless station (1). A first beam (6) is formed for receiving transmission data from a given wireless station (2d) in the MU‑MIMO group. Based on determining that the number of members of the MU‑MIMO group is less than the number of beams that the beamforming weight matrix is ​​configured to receive, one or more additional beams (10a, 10b, 10c, 10d) are formed to receive transmission data from the given wireless station (2d), and the arrival angle of each of the additional beams is offset from the first beam (6) by a corresponding predetermined angle. A decoded data stream is generated based on the data stream received using the first beam (6) and the corresponding data streams received using the one or more additional beams (10a, 10b, 10c, 10d).
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention generally relates to interference mitigation in a multi-user multiple-input multiple-output (MU-MIMO) wireless communication network including a plurality of wireless stations, and particularly, but not exclusively, to an interference mitigation method for mitigating interference effects by forming a first receive beam and an additional receive beam and selecting between outputs of the beams. Background Art

[0002] Modern wireless communication networks are generally placed under high demands to provide high data capacity within the constraints of the allocated signal spectrum. In cellular wireless communication networks, capacity can generally be increased by reusing frequencies between cells according to a predetermined frequency reuse pattern. Fixed wireless access systems may include a base station (which may be referred to as an access point) that is typically mounted on an antenna tower and a plurality of subscriber modules installed at customer premises. The coverage area of ​​the access point may be divided into sectors, each sector being used for transmission and reception with a respective cell. The capacity within a cell may be further increased by directing individual beams to specific user devices (which may be referred to as subscriber modules) to allow communication between access points with improved gain and / or reduced interference reception compared to beams covering sectors. The access point may be equipped with an antenna array and a beamformer for each sector to form a beam for communication with each respective subscriber module. Capacity may be further increased by employing multi-user multiple input multiple output (MU-MIMO) beamforming, in which individual beams may be directed simultaneously to different subscriber modules selected for the MU-MIMO group, each beam carrying different data. The subscriber modules selected as members of the MU-MIMO group are sufficiently spaced apart so that each beam can be directed to one member of the group, while directing the null to other members of the group to avoid interference between members of the group. Alternatively, a peer-to-peer mesh network can be provided. In this case, there can be multiple wireless stations, which are not distinguished as specific access points and subscriber modules, and each station can communicate with several other stations using MU-MIMO beams.

[0003] A set of MU-MIMO beams may be formed by applying a beamforming weight set to a weighting matrix. The weighting matrix forms multiple output signals by applying appropriate amplitude and phase weights to the element signals received from each element of the antenna array, thereby forming individual MU-MIMO beams to receive individual signals from individual subscriber modules or additional wireless stations. The beamforming weight set may be calculated based on a channel estimate related to the amplitude and phase characteristics of the radio frequency propagation path through each receive chain and each array element of the antenna at the access point or first wireless station from each subscriber unit or additional wireless station. The channel estimate may be calculated based on a sounding symbol having predetermined phase and amplitude characteristics.

[0004] However, wireless communication networks may be subject to interference, especially if unlicensed frequency bands in which other wireless systems may operate are being used. Interference may be intermittent and may not be present when a sounding symbol is received, and thus it may be difficult to account for the effects of interference when calculating beamforming weight sets. Specifically, interference may vary dynamically from symbol to symbol, while updates to MU-MIMO weight sets may be performed on longer time scales (typically seconds or minutes), and frequently updating weight sets may result in heavy signaling overhead for channel sounding.

[0005] It is an object of the present invention to alleviate the problems of the prior art. Summary of the invention

[0006] According to a first aspect of the present invention, there is provided an interference mitigation method for a multi-user multiple-input multiple-output (MU-MIMO) wireless communication network including a first wireless station and a plurality of further wireless stations, the method comprising:

[0007] determining a MU-MIMO group of additional wireless stations to which respective beams are to be formed to allow for simultaneous reception of respective transmit data from respective additional wireless stations in the MU-MIMO group at the first wireless station;

[0008] forming a first beam for receiving transmit data from a second wireless station, which is one of the other wireless stations in the MU-MIMO group, based on the previously determined channel estimate;

[0009] Based on a determination that the number of members of the MU-MIMO group is less than the number of beams that the beamforming weight matrix is ​​configured to receive, forming one or more additional beams within the capacity of the beamforming weight matrix to receive transmit data from the second wireless station, the angle of arrival of each of the additional beams being offset from the first beam by a corresponding predetermined angle; and

[0010] A decoded data stream is generated based on a data stream generated by receiving the transmit data in the first beam and a corresponding data stream generated by receiving the transmit data in one or more additional beams.

[0011] The generation of the additional beam offset from the first beam provides a corresponding location of a null in the receive pattern, thus increasing the likelihood that an interferer falls on a null on at least one of the beams. The decoded data stream generated from the data streams from the first beam and the additional beam allows selection of data from the stream with optimal interference suppression, which may vary dynamically due to multiple intermittent interferers. The selection between beams may be performed faster and with less signaling overhead than recalculating a set of weights to adjust the beam pattern.

[0012] In a MU-MIMO wireless communication network, the first wireless station may be an access point and the further wireless stations may be subscriber modules.Alternatively, the first wireless station and the further wireless stations may be peers in a peer-to-peer mesh MU-MIMO wireless communication network.

[0013] In an embodiment of the present invention, generating a decoded data stream includes:

[0014] demodulating and applying forward error correction to each of the data streams; and

[0015] A decoded data stream is generated by combining segments of the respective data streams selected based on forward error correction of the respective data streams.

[0016] This provides an efficient method of selecting which data to include in the decoded data stream.

[0017] In an embodiment of the invention, generating the decoded data streams comprises combining segments of the respective data streams by selecting data from segments determined to have no uncorrected errors.

[0018] This provides an efficient method of combining data streams. Data determined to have no uncorrected errors can be taken from any data stream, so the selection of which data to include is quite straight forward.

[0019] In an embodiment of the present invention, selecting data determined to have no uncorrected errors includes: selecting OFDM symbols determined to have no uncorrected errors.

[0020] This allows reception from different beams to be selected on a symbol-by-symbol basis, thereby allowing flexible adaptation to multiple intermittent interferers that may fall into the nulls of different beams.

[0021] In an embodiment of the present invention, the method comprises:

[0022] generating a combined demodulated data stream from the data streams;

[0023] applying forward error correction to the combined demodulated data stream; and

[0024] A decoded data stream is generated based on the forward error correction of the data stream received in the first beam, the forward error correction of the data stream received in one or more beams, and the forward error correction of the combined demodulated data stream.

[0025] Generating a combined demodulated data stream provides an additional source in the form of demodulated data from which decoded data can be selected, which may provide better interference suppression than an additional beam alone, for example if the interference source falls between the nulls of the additional beam, or if the noise source is distributed in azimuth with an origin.

[0026] In an embodiment of the present invention, generating a combined demodulated data stream includes:

[0027] Each bit of the combined demodulated data stream is selected by comparing the corresponding bit of each of the data streams to be combined.

[0028] This allows the selection of the bits to be selected from the different streams on a bit-by-bit basis. Each bit may be associated with a corresponding subcarrier of the OFDM symbol at a different frequency, and thus the origin of the interference may vary between bits, thus allowing a different reception pattern to be selected per bit to reject the corresponding interference.

[0029] In an embodiment of the present invention, selecting each bit of the combined data stream comprises: selecting a majority value for each bit value of a corresponding bit of each of the data streams to be combined.

[0030] This allows an efficient method of selecting bits.

[0031] In an embodiment of the present invention, generating a combined demodulated data stream includes:

[0032] Respective corresponding bits of each of the data streams to be combined are weighted according to the respective quality metric.

[0033] This allows efficient combining of data streams.

[0034] In an embodiment of the invention, the quality metric comprises a forward error correction error rate and / or a per-tone vector error calculation for the corresponding stream.

[0035] This can provide improved performance by allowing optimal combining of the streams, which can be effective even if the error demodulator varies from data point to data point.

[0036] In an embodiment of the invention, the method comprises iteratively adjusting the offset of the angle of arrival of one of the selected additional beams from the corresponding predetermined angle based on a quality metric of the data stream received in the selected beam.

[0037] This may provide further improvement by allowing the position of the null of one or more additional beams to be adjusted to better align with the long term position of the interferer. The quality metric of the data stream received in the selected beam may include: the aggregate vector error of the received symbols, or for example include the error rate from a forward error correction decoder. The angle of arrival may be an azimuth and / or an elevation angle.

[0038] In an embodiment of the present invention, the selected beam is a beam selected based on a low quality metric of a data stream received in the beam when offset from the first beam by a corresponding predetermined angle.

[0039] This allows the beam with the worst interference suppression to be iteratively adjusted to a position with better interference suppression.

[0040] In an embodiment of the invention, the first beam and each additional beam have a respective sinc beam shape.

[0041] This provides a convenient implementation.

[0042] In an embodiment of the invention, the beam shape of the first beam has a sidelobe level that is different from a sidelobe level of a corresponding beam shape of the additional beam.

[0043] This may allow improved interference suppression at certain angles of arrival.

[0044] In an embodiment of the invention, a first sidelobe of at least one of the additional beams has a lower gain than a gain of a corresponding sidelobe in the sinc beam shape.

[0045] This may allow improved suppression of interference at angles of arrival between nulls in the beam.

[0046] In an embodiment of the present invention, the first wireless station is an access point and the further wireless stations are subscriber modules in a MU-MIMO wireless communication network.

[0047] This may allow further improved suppression of interference at angles of arrival between nulls in the beam.

[0048] According to a second aspect of the present invention, there is provided a first wireless station for a multi-user multiple-input multiple-output MU-MIMO wireless communication network comprising a first wireless station and a plurality of further wireless stations, the first wireless station comprising a processor, the processor being configured to cause the first wireless station to:

[0049] determining a MU-MIMO group of additional wireless stations to which to form respective beams to allow for simultaneous reception of respective data streams at the first wireless station from respective further wireless stations in the MU-MIMO group;

[0050] causing a beamforming weight matrix to form a first beam for receiving a data stream from a second wireless station, the second wireless station being one of the other wireless stations in the MU-MIMO group, based on the previously determined channel estimate;

[0051] Based on a determination that the number of members of the MU-MIMO group is less than the number of beams that the beamforming weight matrix is ​​configured to receive, causing the beamforming weight matrix to form one or more additional beams to receive transmit data from the second wireless station, the angle of arrival of each of the additional beams being offset from the first beam by a corresponding predetermined angle; and

[0052] A decoded data stream is generated based on a data stream generated by the transmit data received in the first beam and corresponding data streams generated by the transmit data received in one or more additional beams.

[0053] Further features of the present invention will become apparent from the following description of preferred embodiments of the invention, which description is given by way of example only. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a schematic diagram showing reception of individual MU-MIMO beams at an access point from subscriber modules that are members of a MU-MIMO group, wherein the number of members of the MU-MIMO group is the same as the number of available MU-MIMO beams;

[0055] Figure 2 is a schematic diagram illustrating a receive architecture for an access point having beamforming weights for simultaneously receiving multiple MU-MIMO beams;

[0056] Figure 3 is a schematic diagram showing receiving individual beams at an access point from a subscriber module that is a member of a MU-MIMO group, wherein the number of members of the MU-MIMO group is less than the number of available MU-MIMO beams, and forming several closely spaced additional beams to a subscriber module that is subject to interference from a noise source that may be intermittent;

[0057] Figure 4 An architecture is shown in accordance with an embodiment of the present invention, wherein selection between data streams is made based on forward error correction results, which may be on a symbol-by-symbol basis;

[0058] Figure 5 An architecture according to an embodiment of the present invention is shown, wherein a combined demodulated data stream is formed from individual data streams from receive beams, and selection is made between the data stream and the individual data streams based on forward error correction results;

[0059] Figure 6An architecture is shown according to an embodiment of the present invention, which shows a feedback path for adjusting beamforming weights to iteratively move a beam in azimuth based on forward error correction or vector error results;

[0060] Figure 7 is a schematic diagram showing a beamforming weight matrix;

[0061] Figure 8 is a schematic diagram showing a receive chain;

[0062] Fig. 9 is a flow chart of a method according to an embodiment of the present invention, the method illustrating allocating additional beams on a symbol-by-symbol basis with spare beam capacity;

[0063] Fig.10 shows interference suppression as a function of azimuth angle for a system with additional beams according to an embodiment of the present invention compared to interference suppression for a single MU-MIMO beam;

[0064] Fig.11 illustrates beam shapes and combined interference suppression in an embodiment of the present invention;

[0065] Fig.12 is a flow chart of a method according to an embodiment of the present invention in which a combined decoded data stream is generated from data streams from received beams;

[0066] Fig.13 is a flow chart of a method according to an embodiment of the present invention, in which a feedback loop is provided to adjust the position of the beam in azimuth; and

[0067] Fig.14 is another flow chart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0068] By way of example, an embodiment of the present invention will now be described in the context of a fixed wireless access system that operates a time division duplex system based on the IEEE 802.11 standard at a carrier frequency typically between 5 GHz and 6 GHz. However, it should be understood that this is exemplary only, and other embodiments may involve other wireless systems and frequencies, and that the embodiments are not limited to a specific operating band or a specific standard, and may involve operating in a licensed or unlicensed band. In the context of the described fixed wireless access system, the present system includes so-called "access points" and "subscriber modules". However, the embodiments are not limited to applications that provide access points and subscriber modules. For example, a peer-to-peer mesh network may be provided. In this case, there may be multiple wireless stations that are not distinguished as specific access points and subscriber modules, each of which may communicate with several other stations using MU-MIMO beams. Therefore, a reference to an "access point" may be interpreted as a reference to a "first wireless station", and a reference to a "subscriber module" or multiple "subscriber modules" may be interpreted as a reference to "another wireless station" or "another wired station", respectively.

[0069] Figure 1 Schematic diagram of an access point 1 according to an embodiment of the present invention in a multi-user multiple input multiple output (MU-MIMO) point-to-multipoint wireless communication network including access points and subscriber modules 2a-2g. In a peer-to-peer mesh network, a first wireless station 1 and further wireless stations 2a-2g are to be set. In the example shown, the access point is mounted on a tower and the subscriber modules are fixed on a building to, for example, provide data services (such as Internet access to homes and businesses). The access point 1 has an array of antenna elements, each element being arranged to receive and / or transmit a signal, which is appropriately weighted in amplitude and phase to form seven MU-MIMO beams 3-9 in the MU-MIMO operating mode in this example. In the data reception mode, a beamforming weight matrix is ​​used to form a corresponding beam for each of the multiple data streams. Thus, for example, different payload data can be received from each subscriber module 2a-2g at the same time. Typically each beam is arranged to form a null point in the direction of the other beams, thereby reducing interference from other subscriber units. The appropriately weighted signals from each antenna element are combined by superposition in a beamforming matrix to form a combined weighted signal for reception from each subscriber module. The signals may be combined separately to form a beam at each polarization. Figure 1In the illustration, for clarity, only beams at one polarization are shown (e.g., vertical polarization V), but another set of beams at a second orthogonal polarization (e.g., horizontal polarization H) may also be formed. In addition, the polarization in the beams may be controlled by appropriately weighting the signals of antenna elements at different polarization states, so that beams may be formed at two polarizations orthogonal to each other, rather than at polarizations that completely correspond to the polarizations of the antenna elements. Likewise, in the example shown, beams are shown as being formed at azimuth angles, but references to "azimuth angles" may be interpreted as covering azimuth angles and / or elevation angles (i.e., angles of arrival).

[0070] Figure 1 The subscriber modules 2a-2g shown may be subscriber modules that have been selected for MU-MIMO group membership based on that they have sufficient spatial separation (each beam may be directed to one member of the MU-MIMO group, and the nulls may be directed to other members of the group to avoid interference between members of the group). The selection of MU-MIMO group membership is also based on the subscriber modules having data to transmit, and the selection may be on a temporary basis, and may be per symbol, for example.

[0071] exist Figure 1 In the case of , the MU-MIMO group has as many members as the maximum number of beams that the system is configured to form. In this case, there is no spare capacity in terms of the number of beams, since there are active subscriber modules in each orthogonal MU-MIMO beam. The maximum number of beams that can be formed may be limited by the number of demodulator channels provided and the number of weights in the beamforming weight matrix. The maximum number of orthogonal MU-MIMO beams that can be formed may be limited by the angular separation of the nulls in the array, since each beam typically falls within the nulls of the other beams. This may also be limited by the number of elements in the antenna array. The hardware provision for the number of beams that can be formed simultaneously is typically arranged to correspond to the number of MU-MIMO beams that can be formed.

[0072] Figure 21 is a schematic diagram showing an architecture for reception of MU-MIMO beams at an access point. An array of antenna elements 15, typically patch antenna elements, receives transmissions from subscriber modules in a MU-MIMO group. The received signal is passed to a receiver chain 14, which downconverts the received signal from radio frequency (RF) to a typical baseband and converts the signal to the digital domain. The architecture can be used to receive payload data using a MU-MIMO beam in a data transmission mode in which the payload data is mapped from received orthogonal frequency division multiplexing (OFDM) tones to a beamforming weight matrix 12 to form a MU-MIMO beam. The OFDM tone value represents the amplitude and phase of a tone (also known as a subcarrier) of an OFDM symbol. This can typically be a vector with in-phase and quadrature values. Data can be encoded in the tone values ​​using QAM (quadrature amplitude modulation) modulation. After beamforming, the tone vectors are typically demodulated in a modulator 16 to convert to a binary stream, and forward error correction (FEC) decoding is then applied, typically on a symbol-by-symbol basis, to detect and correct errors in the binary stream.

[0073] Figure 3 It is shown that in an embodiment of the present invention, the number of members of the MU-MIMO group is less than the number of available MU-MIMO beams. The number of available MU-MIMO beams depends on the capacity of the access point or wireless station to form beams. The capacity may be limited by the architecture of the beamforming weight matrix and / or the number of demodulation and error correction channels at the output of the beamforming weight matrix. The spare capacity of beamforming is used to form several closely spaced additional beams to the subscriber module, which is interfered by what may be an intermittent noise source. The gain of the additional beam toward the subscriber module may be slightly lower than the first beam, which is formed as a MU-MIMO beam based on channel sounding data. This is because the peak of the additional beam is slightly offset from the direction of the subscriber module. However, the zero point in the reception pattern of the additional beam is also offset from the zero point of the first beam. Therefore, by selecting data from one or another beam, interference suppression can be improved by obtaining data from beams in which intermittent interference falls within the zero point when receiving.

[0074] like Figure 3 As shown, based on the previously determined channel estimate, a first beam 6 is formed for receiving transmit data from the first subscriber module 2d in the MU-MIMO group. This is a beam with a peak generally directed toward the subscriber module 2d, which would be formed if all the capacity of the beam forming was used. Figure 3As shown, if it is determined that the number of members of the MU-MIMO group is less than the number of beams that the access point or wireless station (and specifically, the beamforming weight matrix) is configured to receive, one or more additional beams 10a, 10b, 10c, 10d may be formed within the capacity of the access point (specifically, the beamforming weight matrix) for receiving transmit data from the first subscriber module 2d. The additional beams 10a, 10b, 10c, 10d are each offset in azimuth relative to the first beam 6 by a respective predetermined angle. Typically, the predetermined angle may be less than the spacing between the MU-MIMO beams. For example, the spacing between the MU-MIMO beams may be 7 degrees, and the additional beams may be offset typically in increments of 1 degree. Other offsets may be selected, such as offsets with increments of 2 degrees or less than 2 degrees (e.g., increments between 1 degree and 2 degrees). Figure 3 In the example of , there is capacity to form 7 beams, but as shown, there are only three subscriber modules in the MU-MIMO group. The spare capacity for forming and receiving 4 beams is used to form four additional beams towards the first subscriber module 2d.

[0075] Figure 4 An architecture according to an embodiment of the present invention is shown, where selection between data streams is made based on forward error correction results, which may be on a symbol-by-symbol basis.

[0076] like Figure 4 As shown, in this example, a decoded data stream is generated by selecting the output of a decoded (e.g., error-corrected) forward error correction (FEC) circuit 17. The selection is between a data stream resulting from receiving transmit data in a first beam and a corresponding data stream resulting from receiving transmit data in one or more additional beams.

[0077] The generation of additional beams offset from the first beam provides various locations of nulls in the received pattern, so the likelihood of an interference source falling into a null on at least one beam increases. Generating a decoded data stream from data streams from the first beam and the additional beam allows data to be selected from the stream with optimal interference suppression, which interference can change dynamically due to multiple intermittent interference sources. Compared to recalculating weight sets to adjust the beam pattern, the selection between beams can be performed faster and with less signaling overhead. Typically, the weight set can be recalculated over a period of seconds or minutes, while the selection between beams can be, for example, every symbol (which can be every 12 microseconds or so). Therefore, it can be seen that if the interference is dynamic (such as multiple intermittent sources in an unregulated band), embodiments of the present invention provide a faster and simpler response than recalculating weights.

[0078] like Figure 4As shown, each data stream from the beamforming weight matrix is ​​demodulated 16 and forward error correction 17 is applied to each data stream. A decoded data stream 22 is generated by combining segments of the corresponding data streams, each segment corresponding to a received OFDM symbol, for example. The segment can be selected based on the forward error correction of the corresponding data stream. In general, data can be selected from segments of the segments that generally correspond to OFDM symbols that are determined to have no uncorrected errors. The method is generally easy to implement because data determined to have no uncorrected errors can be obtained from any data stream.

[0079] Figure 5 An architecture according to an embodiment of the present invention is shown in which a combined demodulated data stream 23 is formed from the individual data streams from the receive beams. As shown, the combined demodulated data stream is passed to the forward error correction block 20, and the decoded and error-corrected data stream is passed to the selection block 18 to select the output to be used as the decoded data stream 22, the selection between this data stream and the individual data streams being based on the forward error correction results for each beam. Generating the combined demodulated data stream provides an additional source of demodulated data from which decoded data can be selected, which may provide better interference suppression than the additional beam alone, for example, if the interference source falls between the nulls of the additional beams, or if the noise source is distributed at the origin in azimuth.

[0080] like Figure 5 As shown, generating the combined demodulated data stream 23 may be performed by a multiplexer / combiner function 19. In an embodiment of the present invention, each bit in the combined demodulated data stream 23 may be selected by comparing the corresponding bit of each data stream to be combined, so that bits may be selected from different streams bit by bit. For example, each bit may be associated with a corresponding subcarrier of an OFDM symbol at a different frequency, and thus the origin of interference may vary between bits, so this allows a different reception pattern to be selected per bit to suppress the corresponding interference. Selecting each bit in the combined data stream may include: selecting a majority value for each bit by comparing the value of the corresponding bit of each data stream in the data stream to be combined. Thus, for example, if the corresponding bit has a value of "1" in four data demodulated data streams and a value of "0" in one demodulated data stream, the majority value 1 will be selected in the combined demodulated data stream, which value will then be passed to the forward error correction function 20, which can then work on an input stream with fewer errors to be corrected.

[0081] Alternatively, the multiplexer / combiner function 19 may generate the combined demodulated data stream 23 by a process that includes weighting the respective corresponding bits of each data stream to be combined according to respective quality metrics. For example, the quality metrics may include forward error correction error rates and / or per-tone vector error calculations for the respective streams. This may provide improved performance by allowing optimal combining of the streams.

[0082] Figure 6 An architecture according to an embodiment of the present invention is shown, which shows a feedback path for iteratively moving a beam in azimuth using a controller 21 to adjust beamforming weights based on forward error correction or vector error results. The controller can be a function executed by a digital processor at an access point, which can generally be configured to cause the access point to operate according to the method of an embodiment of the present invention. A selected one of the additional beams can be adjusted based on a quality metric of the data stream received in the selected beam, with a number of steps away from a corresponding predetermined angle. The quality metric of the data stream received in the selected beam may include: an aggregate vector error of the received symbols, or, for example, an error rate from a forward error correction decoder. This can provide further improvement by allowing the position of the null point of one or more additional beams to be adjusted to better align with the long-term position of the interference source.

[0083] In an embodiment of the present invention, the selected beam is a beam selected based on a lower quality metric of a data stream received in the beam when offset by a predetermined angle corresponding to the first beam. This allows the beam with the worst interference suppression to be iteratively adjusted to a position with better interference suppression, so that another beam can then be selected for position adjustment.

[0084] Figure 7 An example of the beamforming weight matrix 12 is shown. Figure 7 As shown, the input from each antenna element (received via a corresponding receive chain to be converted into an OFDM tone value stream) is weighted and combined with the weighted tone values ​​from other elements to produce a corresponding beam output 1A to nA. To form the first beam 1A, the OFDM tone value stream received from each antenna element in this example is weighted by the corresponding weight value W 11 -W 71 Similarly, to form beam nA, the OFDM tone value stream received from each antenna element is weighted by the corresponding weight value W 1n -W 7n The weighting is performed and the weighted tones are combined. The number of beams that can be formed is generally limited by the number of weights that can be applied by the beamforming weight matrix and the number of output channels provided.

[0085] Figure 8is a block diagram showing an example of components of a typical receive chain 14. A signal is received from an antenna element at a radio frequency (i.e., carrier frequency) by a low noise amplifier 32. The signal is typically downconverted to baseband consisting of in-phase (I) and quadrature (Q) components by a downconverter 33, which is provided with a local oscillator by an RF generator 35. The I and Q components are digitized in analog-to-digital converters 34a, 34b, and the cyclic prefix of the OFDM symbol is typically discarded 36a, 36b. The OFDM symbol is then converted from the time domain to a series of OFDM tones by a fast Fourier transform 37, each OFDM tone representing the amplitude and phase of each of the subcarriers. The series of OFDM tones, typically in I and Q form, are then fed to a beamforming weight matrix to form beam outputs 1A to 1n.

[0086] Fig. 9 is a flow chart of a method according to an embodiment of the present invention, which shows allocating additional beams on a symbol-by-symbol basis with spare beam capacity.

[0087] Fig.10 Interference suppression as a function of azimuth angle for an interference mitigation system with additional beams according to an embodiment of the present invention is shown as compared to interference suppression for a single MU-MIMO beam as shown as curve 40. In this example, the interference mitigation system forms 7 beams, namely a "first" beam formed at a nominal 0 degrees toward a subscriber module from which data is received based on channel sounding, and 6 additional beams using spare capacity in the additional beamforming weight matrix at nominal azimuth angles of -3, -2, -1, 1, 2, and 3 degrees. For maximum suppression of noise, the best of the seven beams is selected for reception because the noise (i.e., the interference source) sweeps at azimuth angles from -60 degrees to +60 degrees. The vertical scale shows the gain towards the noise source at each arrival angle. It can be seen that noise suppression is improved at most arrival angles compared to the sensitivity of a single beam to noise as shown in curve 40. In Fig.10In the example of , each beam has sinc (i.e. (sin x) / x beam shape) which represents the receive gain in azimuth and / or elevation with respect to the angle of arrival x. In the case of an antenna array, which is an array of antenna elements arranged along a horizontal line, and the beam pattern is in azimuth. The usual spacing between antenna elements is half a wavelength at the antenna operating frequency. The sinc beam shape is a well-known case, and its first sidelobe typically has -13dB compared to the gain at the center of the main beam. Synchronous beam shapes can be formed by uniform amplitude excitation across the array and provide good receive gain. For example, for a seven-element array, a synchronized beam shape centered on the boresight (i.e. 0 degrees azimuth) can be formed by a weight set {(1, 0), (1, 0), (1, 0), (1, 0), (1, 0), (1, 0)}, which represents the real weight and imaginary weight of the signal applied to each antenna element, respectively. The weighted received signals are combined to produce the corresponding beam. To deflect the beam away from 0 degrees in azimuth, a phase slope can be applied to the weights across the array in a well-known manner.

[0088] Fig.11 The use of an alternative beam shape is shown in which the first beam has a different beam shape from the beam shapes of the additional beams. This allows each beam to have a different spatial filter characteristic so that in the combined detector the probability of interference being present at a high level in all beams is very low. The beam shapes may be chosen so that at least one of them has a low gain at each angle of arrival of interest.

[0089] exist Fig.11 In the example of FIG. 4 , the main beam 43 (“Mid”) has a wider mainlobe and lower sidelobe levels and is formed by a beamformed with the weight set {(0.289952, 0), (0.853548, 0), (1.47878, 0), (1.75543, 0), (1.47878, 0), (0.853548, 0), (0.289952, 0)}. To offset the beam from 0 degrees in azimuth, a phase slope can be applied to the weights across the array. In this example, there are two additional beams 44, 45 “Super D+” and “Super D-”. In this example, the “Super D+” beam 44 is offset by -1 degrees in azimuth and the “Super D-” beam 45 is offset by +1 degrees in azimuth. The set of weights used to form the additional beams in this example is {(1.76866, 0), (-0.315833, 0), (1.76005, 0), (0.574241, 0), (1.76005, 0), (-0.315833, 0), (1.76866, 0)}, offset by applying the appropriate respective phase slopes.

[0090] from Fig.11 It can be seen that the first beam has a beam shape with sidelobe levels that differ from the sidelobe levels of the corresponding beam shapes of the additional beams. The shape of the first "Mid" beam has a wider main lobe and lower side lobes than the corresponding lobes of the additional "Super D+" or "Super D-" beams.

[0091] Fig.11 It is shown that at least one additional beam, and in this example two additional beams 44, 45 ("Super D+" and "Super D-"), have a gain in a first sidelobe that is lower than the gain of the corresponding sidelobe in the sinc beam pattern 46. Fig.11 , the sinc pattern 46 is shown only for comparison purposes with other beam shapes and is not used in the system shown. However, in alternative embodiments, at least one beam shape may be a sinc pattern.

[0092] Likewise Fig.11 As shown, the first beam shape 43 ("Mid") has a gain in at least the second sidelobe that is lower than the gain of the corresponding sidelobe in the sinc beam pattern.

[0093] therefore, Fig.11 A composite curve (uninterrupted "Comp" line 42) is shown which shows the receive gain obtained by selecting the beam with the lowest receive gain at each angle of arrival (here shown as azimuth), and the composite curve shows the gain at each angle given that selection. This can be taken as the gain to the interferer at each angle of arrival. In practice, the selection can actually be performed by selecting symbols or portions of symbols from the corresponding beam based on a quality metric (e.g., detected error rate before error correction or vector error in the demodulated signal from the beam related to the interference level). When selecting one of the offset additional beams, the gain to the desired signal may be slightly reduced, but this is generally a small effect and is more than compensated by the reduction in the gain to the interferer. Therefore, the assumption that the beam with the lowest gain to the interferer is selected is a useful approximation to the result of the data selection process.

[0094] Reference Fig.11, and tracing the resultant "Comp" diagram in the direction of increasing azimuth starting from the 0 azimuth direction, which corresponds to the direction of the wireless station or subscriber module from which the desired signal is intended to be received, the selection of the lowest gain (i.e. the lowest interference at a given angle) first leads to the selection of the additional beam 44 "Super D+", which is offset by -1 degree. Due to the offset, this provides a lower gain than the other additional beam 45 "Super D-". Moving further from the 0 azimuth direction, the selection changes to another additional beam 45 "Super D-" because its first null becomes available and by selecting another additional beam, good interference suppression is obtained. Starting further from the 0 azimuth position, with the reduced sidelobe level, the selection moves back to the additional beam 44 "Super D+". Further outward, as shown by the solid line, the first beam shape 43 is selected with reduced second, and possibly third, fourth and other sidelobes. Due to the reduced sidelobe level between the nulls, this allows to improve the interference suppression at the arrival angles between the nulls in the beam.

[0095] Those skilled in the art will appreciate that the shape of the beam and the sidelobe level can be selected by varying the weights used in the weight set to generate the beam of the beamforming weight matrix applied to the corresponding beam. Various beam shapes and corresponding weight sets are known in the art.

[0096] exist Fig.11 In the example of FIG. 4 , there is one main beam "Mid" (43) shown centered at 0 azimuth, which is used in this example as the direction of the desired signal. The desired signal may come from other angles of arrival, in which case the center of the main beam will be offset by the appropriate angle. Two additional beams 44, 45 are shown, offset by -1 degree and 1 degree respectively. The additional beams may be super-directive beams as known in the art.

[0097] Fig.12 is a flow chart of a method according to an embodiment of the present invention, wherein according to steps S12.1 to S12.4, a combined decoded data stream is generated from data streams from receive beams.

[0098] Fig.13 is a flow chart of a method according to an embodiment of the present invention, in which a feedback loop is provided to adjust the position of the beam in azimuth, as also shown in FIG. Figure 6 Virtual channels (such as Fig.13 ) is a MU-MIMO beam (in this case the "first" beam) formed based on the channel estimation. Step S13.3 involves Figure 6The operation of the multiplexer / combiner 19 selects bits according to quality weights which are fed to a forward error correction function 20 (designated in this example as a Reed-Solomon engine) and then to a selector to form a decoded data stream 22. At step S13.4, the beam position of the beam having the worst vector error is iteratively adjusted.

[0099] Fig.14 is another flow chart of a method according to an embodiment of the present invention, according to steps S14.1 to S14.4, reference is made to the first and further wireless stations, rather than specifically to the access point and the subscriber module.

[0100] In an embodiment of the present invention, the wireless communication system may be a time division duplex system, so that the downlink signal sent from the access point and the uplink signal sent from the subscriber module are sent at the same frequency. Alternating fixed duration segments may be allocated for uplink and downlink transmissions, respectively. The uplink and downlink frames may form a time division duplex frame, which is typically divided into time slots, each of which is typically used to communicate with a subscriber module, or in the case of MU-MIMO operation, with a group of subscriber modules. The access point may switch from one operating mode to another operating mode between time slots, such as from a sector mode to a MU-MIMO mode. For the purpose of channel sounding for forming a MU-MIMO beam, the transmission of the sounding symbol may be performed within a time slot.

[0101] It will be understood that the method of an embodiment of the present invention can be implemented by an electronic system including one or more processors, which may include program code stored in a memory configured to cause the processor to perform the method. The one or more processors may include one or more digital signal processors and / or programmable logic arrays.

[0102] return Figure 1 , the subscriber modules 2a to 2g may have an antenna having an aperture, for example, defined by a reflector, and each antenna element may include a probe for receiving and / or emitting a corresponding polarization from the aperture. The antenna is typically mounted so that the peak of the transmit / receive radiation pattern is aligned in the direction of the access point 1, which is typically mounted on a tower. The command sent to each subscriber module may include a mapping indicating the scheduling of radio resources and / or polarization to the subscriber module as a function of time. The mapping may indicate a corresponding allocation to several subscriber units (typically all subscriber units served by the access point) as a function of time. The mapping may, for example, indicate the time, polarization and / or frequency allocation for transmission and / or reception. The scheduling of radio resources and polarization may be updated periodically, with the time period between updates being determined by the scheduler at the access point.

[0103] A specific example of an access point according to an embodiment of the present invention is given by a point-to-multipoint (PMP) access point (AP) with a seven-element dual-polarization adaptive array smart antenna and multi-user MIMO (MU-MIMO) capability. It will be understood that embodiments of the present invention are not limited to this example. The access point in this example relates to an AP for outdoor deployment as a sector coverage in a PMP network. The unit can be deployed in multiples to provide 360° coverage from a tower or rooftop. The access point can be a complete radio transceiver operating in the frequency range of 5150 MHz to 5925 MHz using time division duplex (TDD) separation of uplink and downlink directions.

[0104] The access point can include an integrated dual-polarized seven-element adaptive array smart antenna. Seven identical dual-polarized antenna elements and 14 associated transceiver chains can be contained in a single rigid member, with each antenna element directly connected to two transceiver chains using printed conductors and wireless through-hole connections. The integration of the components ensures that the spacing and alignment of the antenna elements are known and constant.

[0105] Each antenna element may consist of a vertical column of eight radiating patches and separate passive feed networks for horizontal and vertical polarizations. A single element may have a relatively narrow beamwidth in the elevation direction (approximately 8°) and a wider beamwidth in the azimuth direction (approximately 80°). The gain of each antenna element (that is, eight patches per column) is approximately 14 dBi. The entire antenna element may contain 56 patches in an array that is seven elements wide (seven patches) and one element high (eight patches).

[0106] In this example, the maximum output power for a single transmitter chain is approximately 10 dBm, and for each dual-polarization chain pair, 13 dBm.

[0107] The associated Subscriber Module (SM) device may contain a directional dual-polarized antenna with two transceiver chains. The SM may use polarization diversity or polarization multiplexing to support a single data stream. In MU-MIMO operation, the AP may support several (e.g., seven) parallel data streams, where each stream is associated with a different SM device.

[0108] Application firmware in the access point can calculate the transmitter gains and amplitude and phase weights for the combining matrix to provide the desired MU-MIMO beam pattern. This operation can be based on an accurate model of smart antenna operation in which the composite signal strength at any azimuth angle is determined as the superposition of the signals radiated by the individual antenna elements. This model can automatically and inherently allow for array gain in smart antennas.

[0109] The beamforming weight matrix may be implemented using a programmable gate array or by programming a digital signal processor, or by other well-known digital signal processing implementation techniques.

[0110] Therefore, in an embodiment of the present invention, a MU-MIMO system may have the ability to receive, demodulate, and decode multiple data streams in parallel. The ability to do this may involve steering the RF beam at the source of the transmitted data. In order to mitigate the harmful effects of other transmitters, which may appear as noise (i.e., interference) being received, in an embodiment of the present invention, spare beams may be used to improve the system's tolerance to any noise or interference when all beams are not fully utilized. The main beams of the antenna, RF chain, and demodulator may be directed toward the transmitter source. From zero to "N", additional beams may also be directed toward the same source, but with a variable offset from the source. The number of additional beams may vary with the spare capacity on the MU-MIMO system. The demodulator associated with the additional beam may attempt to demodulate the same signal as the main beam, so there may be "N+1" demodulated streams available for independent decoding and error correction. In an embodiment of the present invention, a system is provided for multiplexing / combining the outputs of available demodulators to provide one or more additional streams for error correction algorithms to work, where additional FEC blocks may be added as needed. Examples of multiplexed / combined outputs may include weighted voting streams where each bit of data is passed to an additional error correction block and a vote is taken based on the outputs of the N+1 demodulators. For example, where N+1 is an odd number and the weights are equal, a simple majority vote is taken on the n+1 streams as to whether the bit to be passed to the additional FEC should be a 1 or a 0. Methods and systems for utilizing spare capacity and optimizing data throughput may include selecting the number and direction of additional beams, a weighted sum method for combining multiple streams into an additional FEC component, and a feedback system that continuously optimizes the selection of the number and direction of additional beams, and a weighted sum method for combining multiple streams into an additional FEC component.

[0111] In an embodiment of the present invention, the main beam can be optimized to point at the transmitter. This beam can have the highest signal strength. Additional beams can be directed to the sides of the main beam and therefore have lower signal strength. The beams can have side lobes and different levels of noise suppression (i.e., nulls) that vary with azimuth away from the center zero degrees. The pattern of available beams with each additional beam added offset from the center of the main beam allows for making it more likely that a noise source will fall into the null of one or more beams. If the noise source is located in a deep null, the demodulated stream will most likely not contain errors introduced from the noise source and will pass the correct transmitted data through FEC. In Fig.10An example of noise suppression when using several beams compared to a single beam is given in FIG. It can be seen that a single beam 40 is more likely to see a higher noise level from an interference source at an unknown azimuth position. If the azimuth position of the interference is known, the null can be driven directly to that position. If the noise source is not directly within the null of the demodulator, or the source point is wide, rather than a point source, then the process of combining several demodulator streams to form an optimized stream to the new FEC will further reduce the impact of the noise source and increase the likelihood of correctly receiving the transmitted data. The noise source received on multiple beams will have a different impact on the way each demodulator determines the original transmitted data. For example, in a QPSK system, each demodulator slice decision used to determine the data tone value will be affected by noise. Because there is a distribution of nulls across multiple demodulators, it is unlikely that all are affected, so that noise causes QPSK slice determinations to be made incorrectly for all the tone values ​​on the same data tone. Therefore, even if the erroneous demodulator changes from data point to data point, the combination / voting system can provide the best output stream. Analysis and feedback mechanisms can provide the best method to combine the outputs of the demodulators to the FEC in order to recover the original transmitted data. Metrics used to determine the weighting scheme for demodulator combining can include FEC error rate calculations or per-tone vector error calculations. Simulations of the combining schemes show that a 2-3dB improvement can be achieved over the single beam case under a variety of useful and noise sources.

[0112] In an embodiment of the present invention, data loss in an RF system can be minimized by utilizing the spare capacity of a MU-MIMO system, where spare capacity is available when each possible MU-MIMO receive stream does not receive data from an independent source. In addition to the main beam toward the transmitting source, an additional beam can also be directed offset in that direction. The offset can be fixed or adjusted to optimize the possibility of receiving data correctly. All beams can be demodulated independently, and their outputs are directed to a forward error correction function (FEC). One or more additional streams can be formed from the combined output and fed to one or more FECs. There may be a weighting scheme for the additional FEC or a combination of multiple FECs, which may be fixed or dynamic, such as varying from symbol to symbol. The scheme may be based on locally available data at the demodulator level or an additional system based on metric optimization, or a combination of these.

[0113] The above embodiments should be understood as illustrative examples of the present invention. It should be understood that any feature described with respect to any one embodiment can be used alone or in combination with other features described, and can also be used in combination with one or more features of any other embodiment or any combination of any other embodiment. In addition, without departing from the scope of the present invention defined by the appended claims, equivalents and modifications not described above may also be adopted.

Claims

1. An interference mitigation method for a multi-user multiple-input multiple-output (MU-MIMO) wireless communication network, the MU-MIMO wireless communication network comprising a first wireless station and a plurality of additional wireless stations, the method include: selecting wireless stations from the plurality of additional wireless stations for members of a MU-MIMO group of the additional wireless stations, the selection being based on them having sufficient spatial separation such that when a corresponding beam is directed to each respective member of the group, nulls are directed to other members of the group to avoid interference between members of the MU-MIMO group to allow for simultaneous reception at the first wireless station of respective transmit data from respective additional wireless stations in the MU-MIMO group; forming a first beam for receiving the transmit data from a second wireless station based on a previously determined channel estimate and directing nulls to other members of the MU-MIMO group, the second wireless station being one of the other wireless stations in the MU-MIMO group; Based on determining that the number of members of the MU-MIMO group is less than the number of beams that the beamforming weight matrix is ​​configured to receive, forming one or more additional beams within the capacity of the beamforming weight matrix for receiving the transmit data from the second wireless station, the angle of arrival of each of the additional beams being offset from the first beam by a corresponding predetermined angle; and A decoded data stream is generated based on a data stream generated by receiving the transmit data in the first beam and a corresponding data stream generated by receiving the transmit data from one or more of the additional beams.

2. The method according to claim 1, in, Generating the decoded data stream comprises: demodulating each of the data streams and applying forward error correction to each of the data streams; and The decoded data stream is generated by combining segments of the respective data streams selected based on the forward error correction of the respective data streams.

3. The method according to claim 2, in, Generating the decoded data stream includes combining the segments of the corresponding data stream by selecting data from segments determined to have no uncorrected errors.

4. The method according to claim 3, in, Selecting the data determined to have no uncorrected errors includes selecting OFDM symbols determined to have no uncorrected errors.

5. The method according to any one of claims 2 to 4, include: generating a combined demodulated data stream from the data streams; applying forward error correction to the combined demodulated data stream; and The decoded data stream is generated based on the forward error correction of the data stream received in the first beam, the forward error correction of the data stream received in one or more beams, and the forward error correction of the combined demodulated data stream.

6. The method according to claim 5, in, Generating the combined demodulated data stream comprises: Each bit in the combined demodulated data stream is selected by comparing the corresponding bit of each of the data streams to be combined.

7. The method according to claim 6, in, Selecting each bit in the combined data stream comprises selecting a majority value for each bit of a corresponding bit value of each of the data streams to be combined.

8. The method according to claim 5, in, Generating the combined demodulated data stream comprises: Respective corresponding bits of each of the data streams to be combined are weighted according to the respective quality metric.

9. The method according to claim 8, in, The quality metric comprises a forward error correction error rate of the corresponding stream.

10. The method according to claim 8 or claim 9, in, The quality metric includes a per-tone vector error calculation.

11. The method according to claim 1, include: The offset of the angle of arrival of a selected one of the additional beams from the corresponding predetermined angle is iteratively adjusted based on a quality metric of the data stream received in the selected beam.

12. The method according to claim 11, in, The quality metric of the data stream received in the selected beam comprises an aggregate vector error of received symbols.

13. The method according to claim 11, in, The quality metric of the data stream received in the selected beam comprises an error rate from a forward error correction decoder.

14. The method according to claim 11, in, The selected beam is a beam that is selected based on a quality metric of the data stream received in the beam being low when offset from the first beam by the corresponding predetermined angle.

15. The method according to claim 1, in, The first beam and each of the additional beams have a respective sinc beam shape.

16. The method according to claim 1, in, The beam shape of the first beam has a sidelobe level that is different from a sidelobe level of a corresponding beam shape of the additional beam.

17. The method according to claim 16, in, At least one of the additional beams has a lower gain in a first sidelobe than a corresponding sidelobe in a sinc beam shape.

18. The method according to claim 16, in, The first beam has a beam shape having a gain in at least a second sidelobe that is lower than a gain of a corresponding sidelobe in a sinc beam pattern.

19. The method according to claim 1, in, The first wireless station is an access point and the further wireless stations are subscriber modules in the MU-MIMO wireless communication network.

20. The method according to claim 1, in, The first wireless station and the further wireless station are in a peer-to-peer mesh MU-MIMO wireless communication network.

21. A first wireless station for a multi-user multiple-input multiple-output MU-MIMO wireless communication network, the MU-MIMO wireless communication network comprising a first wireless station and a plurality of additional wireless stations, the first wireless station comprising a processor, the processor being configured to cause the first wireless station to: select wireless stations from the plurality of additional wireless stations as members of a MU-MIMO group for the additional wireless stations, the selection being based on them having sufficient spatial separation such that when a corresponding beam is directed to each respective member of the group, nulls are directed to other members of the group to avoid interference between members of the MU-MIMO group to allow simultaneous reception at the first wireless station of respective data streams from respective further wireless stations in the MU-MIMO group; causing a beamforming weight matrix to form a first beam for receiving a data stream from a second wireless station, one of the additional wireless stations in the MU-MIMO group, and to steer nulls to other members of the MU-MIMO group based on a previously determined channel estimate; Based on determining that the number of members of the MU-MIMO group is less than the number of beams that the beamforming weight matrix is ​​configured to receive, causing the beamforming weight matrix to form one or more additional beams for receiving transmit data from the second wireless station, the angle of arrival of each of the additional beams being offset from the first beam by a corresponding predetermined angle; and A decoded data stream is generated based on a data stream generated by receiving the transmit data in the first beam and a corresponding data stream generated by receiving the transmit data in the one or more additional beams.

Citation Information

Patent Citations

  • Apparatus and method for transmitting signal in a wireless communicaiton system

    CN103999421A

  • Multi-input and multi-output communication method in large-scale antenna system

    CN104247289A

  • Method and apparatus for transmitting information simultaneously to multiple destinations over shared wireless resources

    CN105245264A

  • Simultaneous communication with multiple wireless communication devices

    CN107005322A

  • MU-MIMO system and MU-MIMO communication method

    JP2015046839A