Transmission of mu-mimo signals

Abstract

Mechanisms are provided for transmission of MU-MIMO signals. A method is performed by a network node. The method includes obtaining parameter values of statistics of commonly scheduled user equipments served by the network node. The statistics relate to at least mobility of each of the user equipments. The method includes dynamically selecting a multi-user transmission configuration from candidate multi-user transmission configurations. Each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beam steering mode selected from candidate analog beam steering modes. The multi-user transmission configuration is selected in accordance with the parameter values and a configuration mapping of the parameter values to the candidate multi-user transmission configurations. The method includes transmitting a MU-MIMO signal towards the user equipments using the selected multi-user transmission configuration.

Description

Technical Field

[0001] The embodiments presented herein relate to methods, network nodes, computer programs, and computer program products for transmitting multi-user multiple-input multiple-output (MU-MIMO) signals. Background Technology

[0002] The introduction of millimeter-wave (mmW) carrier frequencies has enabled beamforming using large antenna arrays (also known as massive MIMO) to increase link budget. Beamforming, performed at the transmit and receive points (TRPs) on the network side, ensures that desired energy is manipulated toward the desired user equipment (UE) on the user side, while suppressing or even neutralizing unwanted energy (used to interfere with the UE), through amplitude and phase control of the transmitted signal. To establish and maintain maximum spectral efficiency at each UE, it is desirable to continuously adapt the desired beam toward the UE being served. This is particularly important when mobility exists in the system due to movement either by the UE or by movement in a multipath environment (scatterer movement). Due to the short wavelength at mmW carrier frequencies, given a moderate level of Doppler spread (hundreds of Hz), beam updates need to occur on a timescale of milliseconds to microseconds. This places stringent requirements on the design of suitable beamforming architectures that can support multi-user functionality. Therefore, the processes facilitating beam generation and manipulation need to be of low complexity.

[0003] To achieve maximum flexibility and performance, beamforming for multiple UEs should be performed in the digital signal processing domain using amplitude and phase control from baseband processing algorithms. This approach not only enables the simultaneous formation of a large number of beams but also allows for precise manipulation (amplitude and phase control) of the formed beams. However, this necessitates the introduction of an active antenna system (AAS) at each element, where the element's backplane is interfaced with an up / down conversion radio frequency (RF) chain communicating with the baseband control unit. At mmW frequencies, equipping each antenna element (i.e., each AAS) of a massive MIMO antenna array with its own analog-to-digital converter (ADC), digital-to-analog converter (DAC), and baseband port connection is both energy-intensive and costly due to the increased power consumption and interconnect complexity of the mixed-signal circuitry in the RF chain. This complicates the use of fully digital beamforming at mmW frequencies.

[0004] This has led to the proposal of hybrid beamforming, under which most of the transceiver processing is shifted from the baseband control unit (in the digital signal processing domain) to the RF front end (in the analog signal processing domain). This reduces the processing load from the baseband control unit by creating a high-dimensional analog beamforming matrix using power splitting, phase shifting, and power combining circuitry close to the RF front end. Utilizing low-dimensional digital beamforming at the baseband allows for a significant reduction in the number of up / down conversion chains at the TRP. The inherent architecture of the hybrid beamformer complements the sparse speculative structure of the mmW physical radio propagation channel, where only a small number of RF chains are available to concentrate energy in the main direction of the physical radio propagation channel. Here, each antenna forms part of a joint hybrid beamformer, which has a significantly smaller number of ADC / DAC and baseband port connections relative to the total number of antenna elements.

[0005] Therefore, beamforming toward multiple UEs can be performed solely in the digital signal processing domain (hereinafter referred to as digital beamforming or digital precoding), solely in the analog signal processing domain (hereinafter referred to as analog beamforming or analog beam manipulation), or in a combination of digital and analog signal processing domains (hereinafter referred to as hybrid beamforming). For digital beamforming, each antenna has a dedicated RF signal and path, with per-element amplitude and phase control performed at the baseband. This results in high power consumption for the local oscillator distribution circuitry, data converters, and digital baseband processors, making it difficult to implement high-element-count antenna arrays. Conversely, analog beamforming can support a large number of antenna elements in a power-efficient manner, but is limited in algorithmic flexibility because only a single beam can be formed and manipulated around a single center departure angle (AOD). To strike the right balance between these two extremes, hybrid beamforming breaks this impasse by performing most of the spatial signal processing at the RF front-end and implementing additional spatiotemporal processing in the digital signal processing domain using a minimal number of RF down-conversion chains. Due to its design, the number of data streams supported by a hybrid beamforming system is capped by the number of down-conversion chains.

[0006] However, while hybrid beamforming offers a good compromise between analog and digital beamforming, there may be scenarios where traditional hybrid beamforming techniques are insufficient to achieve optimal network performance. Summary of the Invention

[0007] The purpose of the embodiments described herein is to address the aforementioned problems by providing a technique for transmitting MU-MIMO signals using dynamically selected multi-user transport configurations.

[0008] According to a first aspect, a method for transmitting MU-MIMO signals is presented. The method is performed by a network node. The method includes obtaining parameter values ​​of statistical data of user equipment (UEs) co-scheduled by the network node. The statistical data relates at least to the mobility of each UE. The method includes dynamically selecting a multi-user transmission configuration from candidate multi-user transmission configurations. Each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beamforming mode selected from candidate analog beamforming modes. The multi-user transmission configuration is selected based on the parameter values ​​and a configuration mapping of the parameter values ​​to the candidate multi-user transmission configurations. The method includes transmitting MU-MIMO signals toward the UEs using the selected multi-user transmission configuration.

[0009] According to a second aspect, a network node for transmitting MU-MIMO signals is presented. The network node includes processing circuitry. The processing circuitry is configured to enable the network node to obtain parameter values ​​of statistical data of user equipment (UEs) co-scheduled by the network node. The statistical data relates at least to the mobility of each UE. The processing circuitry is configured to enable the network node to dynamically select a multi-user transmission configuration from candidate multi-user transmission configurations. Each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beamforming mode selected from candidate analog beamforming modes. The multi-user transmission configuration is selected based on the parameter values ​​and a configuration mapping from the parameter values ​​to the candidate multi-user transmission configurations. The processing circuitry is configured to enable the network node to transmit MU-MIMO signals toward the UEs using the selected multi-user transmission configuration.

[0010] According to a third aspect, a network node for transmitting MU-MIMO signals is presented. The network node includes an acquisition module configured to acquire parameter values ​​of statistical data of user equipment (UEs) co-scheduled by the network node. The statistical data relates at least to the mobility of each UE. The network node includes a selection module configured to dynamically select a multi-user transmission configuration from candidate multi-user transmission configurations. Each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beamforming mode selected from candidate analog beamforming modes. The multi-user transmission configuration is selected based on parameter values ​​and a configuration mapping from parameter values ​​to candidate multi-user transmission configurations. The network node includes a transmission module configured to transmit MU-MIMO signals toward the UEs using the selected multi-user transmission configuration.

[0011] According to the fourth aspect, a computer program for transmitting MU-MIMO signals is presented, the computer program including computer program code that, when run on a network node, causes the network node to perform the method according to the first aspect.

[0012] According to a fifth aspect, a computer program product is presented, the computer program product comprising a computer program according to a fourth aspect and a computer-readable storage medium on which the computer program is stored. The computer-readable storage medium may be a non-transitory computer-readable storage medium.

[0013] Advantageously, these aspects can be used to provide hybrid beamforming with optimal network performance.

[0014] In conventional hybrid beamforming systems, precoding and scheduling methods are typically fixed, and dynamic switching between different beamforming modes is not feasible. The aspects disclosed herein enable dynamic selection between different hybrid beamforming modes, depending, for example, on attributes such as user equipment mobility, but optionally on other factors, parameters, and conditions. Thus, the optimal multi-user transmission configuration can be selected for each multi-user scenario to produce best spectral efficiency performance.

[0015] Advantageously, the adaptive selection of digital precoding mode and analog beam manipulation mode can be combined with the adaptive selection of multi-user scheduling mode to enable the co-design of beamforming and scheduling.

[0016] Advantageously, the above aspects improve the trade-off between reduced digital processing and forwarding for hybrid beamforming architectures with typical radio conditions at millimeter wave frequencies and potential performance losses.

[0017] Advantageously, compared to the case where a fixed beamforming pattern is deployed at any given site, the above aspects can significantly enhance performance (e.g., for a given level of processing complexity in network nodes, UE SINR, spectral efficiency, and cell throughput), improve network resource utilization and user experience, regardless of network configuration, load, and mobility levels.

[0018] Advantageously, the above aspects enable the reduction of processing complexity while maintaining the desired performance, thereby reducing network device costs and power consumption by shifting a large amount of digital processing to the analog signal processing domain.

[0019] Advantageously, the above aspects can be applied to cluster scenarios, such as user devices located in dense urban centers, various indoor environments (shopping malls, airports, cafes and offices), and public transportation scenarios.

[0020] Advantageously, the above aspects can provide insights and guidance for network deployment at an early stage by optimizing the network's radio access components (e.g., selecting the most appropriate network equipment for a given or expected distribution of user equipment at a particular site or area).

[0021] Other objects, features, and advantages of the appended embodiments will become apparent from the following detailed disclosure, the appended dependent claims, and the accompanying drawings.

[0022] Generally, unless otherwise expressly defined herein, all terms used in the claims shall be interpreted according to their ordinary meaning in the art. Unless otherwise expressly stated, all references to “a / an / the said element, device, component, part, module, step, etc.” shall be interpreted openly as referring to at least one instance of the said element, device, component, part, module, step, etc. Unless expressly stated otherwise, the steps of any method disclosed herein need not be performed in the exact order disclosed. Attached Figure Description

[0023] The inventive concept will now be described by way of example, with reference to the accompanying drawings, wherein:

[0024] Figure 1 and Figure 2 This is a schematic diagram illustrating a communication network according to an embodiment;

[0025] Figure 3 This is a schematic diagram illustrating a network node according to an embodiment;

[0026] Figure 4 This is a flowchart of the method according to an embodiment;

[0027] Figure 5 The correlation as a function of angular separation between UEs is illustrated schematically according to the embodiment;

[0028] Figure 6 A first mobility scenario according to an embodiment is illustrated schematically;

[0029] Figure 7 The simulation results of a first mobility scenario according to an embodiment are illustrated schematically;

[0030] Figure 8 A second mobility scenario according to an embodiment is illustrated schematically;

[0031] Figure 9 The simulation results of the second mobility scenario according to the embodiment are illustrated schematically;

[0032] Figure 10 This is a schematic diagram illustrating the functional units of a network node according to an embodiment;

[0033] Figure 11 This is a schematic diagram illustrating the functional modules of a network node according to an embodiment;

[0034] Figure 12 An example of a computer program product including a computer-readable storage medium according to an embodiment is shown. Detailed Implementation

[0035] The inventive concept will now be described more fully below with reference to the accompanying drawings, in which certain embodiments of the inventive concept are illustrated. However, the inventive concept may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided by way of example so that this disclosure will be thorough and complete, and will fully convey the scope of the inventive concept to those skilled in the art. Throughout the description, the same numerals refer to the same elements. Any step or feature illustrated by dashed lines should be considered optional.

[0036] Figure 1 This is a schematic diagram illustrating a communication network 100 in which the embodiments presented herein can be applied. The communication network 100 may be a third-generation (3G) telecommunications network, a fourth-generation (4G) telecommunications network, a fifth-generation (5G) telecommunications network, or any evolution thereof, and supports any 3GPP, IEEE, or other telecommunications standards where applicable.

[0037] Communication network 100 includes network node 200 configured to provide network access to user equipment (e.g., represented by UEs 150a, 150b, 150c, 150d) in radio access network 110. Radio access network 110 is operatively connected to core network 120. Core network 120 is operatively connected to serving network 130, such as the Internet. This enables UEs 150a:150d to access the services of serving network 130 and exchange data with serving network 130 via network node 200.

[0038] Network node 200 includes a transmit and receive point (TRP) 140, is co-located with, integrated with, or operatively communicates with a transmit and receive point (TRP) 140. Network node 200 (via its TRP 140) and user equipment UEs 150a:150d are configured to communicate with each other in directional beams, as illustrated by reference numeral 160. In this respect, the directional beam, which can serve as both a transmit beam and a receive beam, will be simply referred to as a directional beam or simply a beam in the following text. Network node 200 is equipped with a hybrid beamformer that enables the formation of a beam 160 to each UE 150a, 150b, 150c, 150d.

[0039] Examples of network nodes 200 are radio access network nodes, radio base stations, base transceiver stations, node B, evolved Node B (eNB), gNB, access points, access nodes, and backhaul nodes. Examples of UE 150a:150d are wireless devices, mobile stations, mobile phones, handsets, wireless local loop phones, smartphones, laptops, tablets, network-enabled sensors, network-enabled vehicles, and so-called Internet of Things (IoT) devices.

[0040] As mentioned above, there may be scenarios where traditional hybrid beamforming techniques are insufficient to achieve optimal network performance.

[0041] More specifically, previous efforts to understand and characterize the optimal spectral and energy efficiency performance of hybrid beamforming in multi-user scenarios assumed a static scenario where the mobility of either UE 150a:150d or the mobility in the physical radio propagation channel (as a result of scatterer movement) is not considered. This has an impact on the level of inter-UE channel correlation because the spatial consistency of propagation parameters across large and small scales is modeled. This, in turn, affects the resulting spectral efficiency performance, as multi-user interference may remain unsuppressed for known hybrid beamforming techniques.

[0042] In the following text, it is assumed that at least UE 150a and 150b will be jointly scheduled.

[0043] Previous hybrid beamforming algorithms also do not cater to tightly positioned UEs 150a and 150b, where the angular separation across multiple UEs 150a and 150b is relatively small in comparison. In such scenarios, most of the previously proposed hybrid beamforming algorithms produce suboptimal results for various combinations of analog and digital beamforming techniques.

[0044] For efficient spectrum operation, scheduling strategies (e.g., selecting UEs 150a and 150b for priority and co-scheduling) and beamforming choices (e.g., achieving the right balance between maximizing power and neutralizing interference) have a major impact on performance (e.g., signal-to-interference-and-noise ratio (SINR) per UE, spectral efficiency, and the resulting cell throughput).

[0045] Given the mobility and spatial distribution of UEs 150a and 150b, and considering the propagation conditions, previous hybrid beamforming techniques do not allow for low-complexity methods of configuring suitable beamforming algorithms at the network side. This coupling with the performance of UE scheduling mechanisms typically deployed in network nodes at the network side is also uncharacterized.

[0046] Therefore, when using hybrid beamforming to form multiple beams, there is a need to address the performance optimization of mmW systems by combining at least mobility and possibly other factors, parameters, or conditions (such as multi-user scheduling).

[0047] Therefore, the embodiments disclosed herein relate to mechanisms for the transmission of MU-MIMO signals. To obtain such a mechanism, a network node 200, a method executed by the network node 200, and a computer program product are provided, the computer program product comprising, for example, code in the form of a computer program, which, when run on the network node 200, causes the network node 200 to execute the method.

[0048] consider Figure 2 The scenario described in the text. Figure 2 It shows Figure 1 The communication network, but without radio access network 110, core network 120, and service network 130. Network node 200 simultaneously (on the same time-frequency resource block) serves L single-antenna UEs 150a, 150b, 150c, and 150d within a 120° field of view (marked by dotted lines) of an antenna array geographically distributed at TRP 140. The scattering between TRP 140 and UEs 150a, 150b, 150c, and 150d (for clarity, ...) Figure 2 (Not shown in the diagram) can be moving, which contributes to the overall Doppler spread of the physical radio propagation channel. Assume UE 150a is the desired UE, while other UEs 150b, 150c, and 150d receive interference. For illustration, the angular separation in the azimuth domain between UE 150a and UE 150b is shown as α. Each UE 150a, 150b, 150c, and 150d has an associated, individually estimated 1×M downlink physical radio propagation channel vector h, which is indexed by the UE index, from a to d in the diagram, and generally from 1, 2, ..., L for a number of L UEs 150a and 150b.

[0049] Next, we will disclose aspects of digital precoding and analog beam manipulation.

[0050] Digital precoding (also known as digital beamforming) in this disclosure refers to applying appropriate weights to the output signals of different baseband ports to control the overall radiation pattern and the lobe (and in some cases, zero) of the signal transmitted to each co-scheduled UE 150a, 150b. In the context of hybrid beamforming, the weighted baseband port outputs are fed to a set of multiple antenna elements, where analog beam manipulation is additionally applied (see below). The output signal to each port can be the sum of signals from multiple users or multiple feeds where users are separate. Some non-limiting examples of digital precoding modes will be disclosed below.

[0051] For matched filtering (MF), the downlink beamforming vector for each UE 150a / 150b is designed from network node 200 by performing a Hermitian transpose on the estimated propagation channel vector from UE 150a / 150b to TRP 140. This technique focuses on maximizing the received power of the served UE 150a / 150b and thus suffers from multi-user interference, especially in high signal-to-noise ratio (SNR) conditions. For massive MIMO scenarios, it has recently been shown that MF works optimally as a digital beamforming technique within constraints as the number of network node 200 antennas serving a fixed number of UE 150a / 150b increases.

[0052] Unlike Multi-User Forcing (MF), Zero Forcing (ZF) focuses on neutralizing current multi-user interference. This is accomplished by computing the left pseudo-inverse of the estimated propagation channel from given UEs 150a and 150b to TRP 140. Since this involves placing the channels of unwanted UEs 150c and 150d into the null space of the channels of desired UEs 150a and 150b, interference can be neutralized. The algorithm offers superior performance at high SNR.

[0053] For Regularized ZF (RZF), a regularization factor proportional to the link SNR is introduced to improve the low SNR suboptimality of the ZF algorithm. By regularizing the left pseudo-inverse of the estimated propagation channel from specific UEs 150a, 150b to TRP 140, the inverse can be better tuned, and any potential rank deficiencies or poor matrix adjustment that may exist in the inverse process can be removed. The algorithm converges to the same performance as MF at low SNR and provides equivalent performance to ZF at high SNR.

[0054] By transmitting on the maximum eigenvector of the left pseudo-inverse of the composite multi-user channel matrix, the signal leakage plus noise ratio (SLNR) is focused on maximizing the desired SLNR for UEs 150a and 150b.

[0055] Digital precoding implicitly determines how the amplitude and phase of multiple generated / manipulated beams should be controlled. Using digital signal processing algorithms at the baseband of network node 200, different digital precoding modes differ from each other in the following ways: enhancing signal power in the direction of the desired (scheduled) UEs 150a and 150b; enhancing the overall signal power distribution in the geographic vicinity of the desired UEs 150a and 150b; suppressing or neutralizing interference in the direction of the desired (scheduled) UEs 150a and 150b; and suppressing or neutralizing interference in the geographic vicinity of the desired (scheduled) UEs 150a and 150b.

[0056] Analog beam manipulation (also known as analog beamforming) in this disclosure refers to applying appropriate phase shifts to an array of antenna elements constituting a baseband port to achieve directivity (spatial selectivity) of the signal transmitted from the elements, independent of digital precoding applied to the baseband port. The phase shift weights for each element are preferably constant-modulus to achieve maximum PA efficiency, such as coefficients similar to the Discrete Fourier Transform (DFT) or classical phased array coefficients. At each antenna element, the common baseband port output signal is multiplied by the corresponding phase shift weight of the element. In some examples of hybrid beamforming, multiple individually controlled phase shifts are applied to the same array of elements, for example, to independently manipulate the beam toward multiple users. Several non-limiting examples of analog beam manipulation modes will be disclosed below.

[0057] For random angle (RA) RF processing, the design of the simulated beamforming weights is accomplished using a set of random angles that are fed into the simulated beamforming vector. Angles are plotted from uniform distributions in the azimuth and elevation domains between [-60°, 60°] and [-15°, 15°], respectively. These correspond to the azimuth and elevation angles of the TRP 140.

[0058] According to the Selective Combination (SC) RF processing, the strongest path is selected for each UE 150a, 150b, and then the gain manipulation of the simulated beam is directed toward the strongest path by appropriately designing the simulated beamforming weights. For a given set of azimuth and elevation angles, the strongest path is selected by calculating the absolute value of the estimated propagation channel vector multiplied by the array manipulation response.

[0059] For the Aggregated Composite Channel Phase Extraction (AC-CPE) RF processing, weights are generated by extracting the phase of the conjugate transpose of the composite downlink channels from TRP 140 toward multiple UEs 150a and 150b. AC-CPE provides phase-only control at each RF chain by extracting the phase of the conjugate transpose of the aggregated downlink channels from network node 200 to each UE 150a and 150b. This is to align the phase of the channel elements and thus reap the large array gain provided by the numerous antennas at TRP 140.

[0060] Because analog beam manipulation explicitly defines how multiple analog beams are formed and manipulated, different analog beam manipulation modes differ from one another in terms of the signal power generated and manipulated toward the desired (scheduled) UEs 150a and 150b, and the overall signal power distribution within the geographic vicinity of the desired (scheduled) UEs 150a and 150b's location. To achieve this, different analog beam manipulation modes use analog RF front-end circuitry to generate and adapt the amplitude and phase of the analog beams (captured in the design of the beam manipulation weights). Different analog beam manipulation modes can also differ from one another in terms of beam coherence distance, user tracking consistency, and user tracking duration in multipath environments.

[0061] The different analog beam manipulation modes disclosed above are suitable for interfacing with any of the digital precoding modes disclosed above, and vice versa.

[0062] Assume the number of nodes at network node 200 (or TRP 140) is N. t The up / down conversion RF chains have N up / down conversions, and the simulated beamforming matrix (configured for power splitting, phase shifting, and power combining) has a dimension of N. t ×M, and is designed to generate up to N in the downlink. t One beam, targeting N t A discrete departure angle (AOD). The vast majority of analog beamformers require N values ​​to be designed via codebooks with a discrete Fourier transform (DFT) structure for a fixed beam. t ×M complex analog beamforming weights. The exact method for generating DFT entries can vary and depends on the specific analog beamforming mode under consideration. Specific examples of different analog beamforming modes are provided below. The adaptation of the analog beam occurs based on the long-term (second-order) time variation of the physical radio propagation channel conditions. This is in contrast to digital precoding, in which digitally assisted signal processing is used on an instantaneous basis to enable amplitude and phase manipulation of the transmission beam using the currently estimated physical radio propagation channel conditions. Thus, a system with dimensions M×N... t The digital beamforming matrix can be designed to increase the desired signal power of the served UEs 150a and 150b, or to assist in manipulating zeros in other directions (positions) toward interfering UEs 150c and 150d. Various methods exist for implementing digital beamforming, focusing on optimizing the trade-off between maximizing the desired signal or neutralizing interference (since both cannot be achieved due to conflicting design requirements). Specific examples of different digital precoding modes are provided below.

[0063] The approach of using digital precoding and analog beamforming modes that performs well for statically and / or randomly distributed UEs 150a and 150b may offer poorer performance or unfavorable performance / complexity trade-offs for mobile and / or closely spaced UEs 150a and 150b, because many of the multipath components in the physical radio propagation channel may exist over relatively large distances, increasing inter-UE channel correlation and generating higher interference power. If, at such a time, given knowledge of the link SNR, network node 200 is configured to make a decision about which combination of digital precoding and analog beamforming modes provides the optimal UE spectral efficiency, this can improve the SINR for each user, as well as throughput and overall system performance.

[0064] Figure 3 A block diagram of network node 200 is schematically illustrated. Network node 200 is configured to be operatively connected to a TRP 140 having a massive MIMO antenna array with cross-polarized elements connected via an interface to an integrated RF front-end circuit block. This is associated with modules 240:290 in network node 200, which are configured for hybrid beamforming and, more particularly, for multi-user transmission configurations that select a combination of specified digital precoding modes and analog beam manipulation modes. Parallel references are also provided. Figure 4 describe Figure 3 Network node 200.

[0065] Figure 4 This is a flowchart illustrating an embodiment of a method for transmitting MU-MIMO signals. The method is performed by network node 200. The method is advantageously provided as computer program 1220.

[0066] Generally, network node 200 adapts its hybrid beamforming mode by dynamically selecting a multi-user transmission configuration from candidate multi-user transmission configurations based on user characteristics (such as the mobility of UEs 150a and 150b). The adaptation is based on a predetermined mapping rule. User characteristics are provided as input to a mapping function, which then provides the selected multi-user transmission configuration as output. Specifically, network node 200 is configured to execute steps S102, S104, and S106:

[0067] S102: Network node 200 obtains parameter values ​​of statistical data for jointly scheduled UEs 150a and 150b served by network node 200. The statistical data relates to the mobility of at least each UE in UEs 150a and 150b. The parameter values ​​can be provided in parameter value module 250 at network node 200.

[0068] S104: Network node 200 dynamically selects a multi-user transmission configuration from candidate multi-user transmission configurations. The selection of the multi-user transmission configuration can be implemented at network node 200 in configuration selector module 240. Each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beamforming mode selected from candidate analog beamforming modes. Parameters defining each candidate digital precoding mode can be provided at network node 200 in candidate digital precoding mode module 255. Parameters defining each candidate analog beamforming mode can be provided at network node 200 in candidate analog beamforming mode module 260. The multi-user transmission configuration is selected based on parameter values ​​and a configuration mapping from parameter values ​​to candidate multi-user transmission configurations. Pre-configured mappings can be implemented at network node 200 in mapper module 245. Each multi-user transmission configuration can define a hybrid beamforming mode, such as one defined by a digital precoding mode and an analog beamforming mode.

[0069] S106: Network node 200 transmits MU-MIMO signals to UEs 150a and 150b using the selected multi-user transmission configuration. The selected configuration can be provided in the selection configuration module 275 at network node 200.

[0070] As will be further disclosed below, network node 200 may further include candidate scheduling mode module 265 and / or candidate grouping mode module 270.

[0071] Given Figure 2 In the scenario presented, network node 200 can be like Figure 3 China and according to Figure 4 The method is configured with multiple candidate multi-user transmission configurations, wherein each candidate multi-user transmission configuration specifies at least a digital precoding mode and an analog beam manipulation mode selected from candidate digital precoding modes.

[0072] For the selected multi-user transmission configuration, known techniques for channel sounding, user packetization, precoding weight calculation, etc., can be used to perform MU-MIMO signal transmission toward UEs 150a and 150b. Assuming other beams are negligible, network node 200 can, for example, perform uplink sounding and estimate the uplink physical radio propagation channel by estimating a subset of relevant beams. This is consistent with the sparsity nature of millimeter-wave channels, as most often only a certain number of directions are prominent in the channel.

[0073] This method enables network node 200 to optimize beamforming signal transmission and reception. Multi-user transmission configurations can be dynamically selected and tuned based on a dynamic handover mechanism that tracks the distribution, mobility status, link SNR, and other operational network parameters of UEs 150a and 150b.

[0074] In conventional hybrid beamforming systems, precoding and scheduling methods are typically fixed, and dynamic switching between different beamforming modes is not feasible. The disclosed method enables dynamic selection between different hybrid beamforming modes, depending on attributes such as user equipment mobility, but optionally based on other factors, parameters, and conditions. Thus, the optimal multi-user transmission configuration can be selected for each multi-user scenario to produce best spectral efficiency performance.

[0075] Advantageously, the adaptive selection of digital precoding mode and analog beam manipulation mode can be combined with the adaptive selection of multi-user scheduling mode to enable the co-design of beamforming and scheduling.

[0076] Advantageously, the disclosed method improves the trade-off between digital processing and fronthaul reduction and potential performance loss for hybrid beamforming architectures with typical radio conditions at millimeter wave frequencies.

[0077] Advantageously, compared to the case where a fixed beamforming pattern is deployed at any given site, the disclosed method can significantly enhance performance (e.g., for a given level of processing complexity in network nodes, UE SINR, spectral efficiency, and cell throughput), improve network resource utilization and user experience, regardless of network configuration, load, and mobility levels.

[0078] Advantageously, the disclosed method enables the reduction of processing complexity while maintaining the desired performance, thereby reducing network device costs and power consumption by shifting a large amount of digital processing to the analog signal processing domain.

[0079] Advantageously, the disclosed method is applicable to cluster scenarios, such as UE 150a and 150b located in dense urban centers, various indoor environments (shopping malls, airports, cafes, and offices), and is also applicable to public transportation scenarios.

[0080] Advantageously, the disclosed method can provide insights and guidance for network deployment at an early stage by optimizing the network’s radio access components (e.g., selecting the most appropriate network equipment for a given or expected distribution of UE 150a, 150b at a particular site or area).

[0081] An embodiment involving the transmission of MU-MIMO signals, such as that performed by network node 200, will now be disclosed in further detail.

[0082] As network conditions and the UE population change, new multi-user transport configurations can be selected. Specifically, in some embodiments, the multi-user transport configuration is dynamically selectable for each scheduling instance of UEs 150a and 150b. The selection of the multi-user transport configuration can alternatively be performed over a long period, applied to a large number of subsequent scheduling instances. Updates can be performed periodically, with the update rate determined by the frequency of changes in the network environment (e.g., seconds, minutes, or hours). Alternatively, updates can be performed when changes in the network environment exceed a threshold.

[0083] In some embodiments, the mobility of UE 150a and 150b is defined by the respective user equipment movement trajectory and mobility status of UE 150a and 150b.

[0084] In some embodiments, the statistics are further related to at least one of the following: spatial separation between UEs 150a and 150b, and link quality of each UE in UEs 150a and 150b.

[0085] Parameter values ​​may be related to other factors, parameters, or conditions. In particular, in some embodiments, parameter values ​​are also parameter values ​​of network operation parameters. In some non-limiting examples, network operation parameters are related to at least one of the following: the current traffic load of network node 200, the current number of MU-MIMO layers available at network node 200. These characteristics can be estimated using known techniques, such as using reported reference signal received power (RSRP) dynamics, beam switching rates in beam management, and Doppler estimation based on channel estimation for (i): channel estimation correlation, (ii): Radio Resource Management (RRM) SINR reports or Channel State Information Reference Signal (CSI-RS) reports, (iii): the number of UEs 150a and 150b served, and (iv) scheduler occupancy information, etc.

[0086] Different aspects of the mapping based on the selection of the multi-user transmission configuration will now be disclosed. These aspects also provide different alternatives on how the mapper module 245 can be implemented. In some aspects, the multi-user transmission configuration is selected based on optimization criteria. In particular, in some embodiments, the mapping is configured based on optimization criteria that maximize the SINR and spectral efficiency of UEs 150a and 150b. In some aspects, the mapping is based on previous performance observations or simulations. In particular, in some embodiments, the mapping is determined by at least one of the following: previous performance observations of network node 200, simulations of operating network node 200.

[0087] In some aspects, a lookup table (LUT) is used to implement the mapping, which specifies candidate multi-user transport configurations for different network and UE scenario characteristics (e.g., mobility states). Configuration parameters (e.g., the number of co-scheduled UEs 150a, 150b) may also be specified. In particular, in some embodiments, the mapping is provided as a LUT, wherein each entry in the lookup table specifies one of the candidate multi-user transport configurations in terms of at least a combination of a digital precoding mode and an analog beamforming mode.

[0088] To provide service to multiple UEs 150a, 150b, they need to be scheduled by network node 200, and the optimal scheduler needs to dynamically adapt to mmW propagation conditions, especially when considering spatial consistency characteristics. In some aspects, the multi-user transport configuration therefore also considers different scheduling modes. Specifically, in some embodiments, each candidate multi-user transport configuration in the candidate multi-user transport configuration further specifies the scheduling mode selected from the candidate scheduling modes. Parameters defining each candidate scheduling mode in the candidate scheduling mode module 265 can be provided at network node 200. In some non-limiting examples, each candidate scheduling mode in the candidate scheduling modes is associated with scheduling using one of the following: Round-Robin Scheduling (RRS), Proportional Fair Scheduler (PFS), and Maximum SINR Scheduling. Maximum SINR scheduling prioritizes UEs with the best channel quality, thus penalizing radio cell edge UEs. On the other hand, RRS scheduling selects and schedules UEs in a round-robin manner, thereby creating equal resource sharing regardless of propagation channel conditions. PFS scheduling provides a balance between maximum SINR scheduling and RRS by considering fairness criteria, while offering maximum effort to maximize the SINR of the UE (and thus maximize throughput). In some examples, when the link SNR is favorable (high) and when UEs 150a and 150b are relatively far apart, maximum SINR scheduling will optimize the spectral efficiency of each UE 150a and 150b due to interference neutralization and optimized analog beam weight calculation. PFS is preferred where UEs 150a and 150b are closely positioned and have high link SNR, because high inter-UE channel correlation will lead to high levels of interference that degrade achievable performance. In such scenarios, PFS will result in a broad improvement in performance. When the link SNR is low and UEs 150a and 150b are far apart, the relative difference between PFS and RRS is minimized, and therefore both will result in similar performance. PFS may be preferred when UEs 150a and 150b are closely positioned. Different scheduler modes can be used for different mobility speeds, taking into account factors such as the primary intervals of UEs 150a and 150b, link conditions, and network load. PF can be used to strike a balance between service and network performance optimization for each UE 150a and 150b. For low-mobility UEs 150a and 150b, performance optimization may become more important due to the higher level of correlation in the physical radio propagation channel introduced by spatial consistency, and therefore PF can be selected as the scheduling mode. For high-speed UEs 150a and 150b, since all UEs 150a and 150b may experience good link conditions over time, scheduling with PF tuned towards maximum SINR or even more towards SINR maximization can also be used.

[0089] As part of selecting a multi-user transmission configuration, the selection of a scheduling mode can be performed less frequently in time than the selection of a digital precoding mode or an analog beamforming mode. In this respect, the scheduling mode may not change on a per-slot basis, but rather on medium-term (such as physical time changes corresponding to multiple minutes) changes in UE group attributes, including mobility, density, separation, etc.

[0090] In some aspects, the multi-user transport configuration also considers different multi-user packet modes. Specifically, in some embodiments, each candidate multi-user transport configuration in the candidate multi-user transport configuration further specifies the multi-user packet mode selected from the candidate multi-user packet modes. Parameters defining each candidate multi-user packet mode in the candidate packet mode module 270 can be provided at network node 200. In some non-limiting examples, each candidate multi-user packet mode in the candidate multi-user packet mode relates to the multi-user packetization of UEs 150a and 150b in at least one of the following aspects: the criterion for how many UEs of UEs 150a and 150b to be co-scheduled, and the criterion for which UE of UEs 150a and 150b to be co-scheduled. More specifically, network node 200 can also adapt the user packetization strategy, i.e., the selection of co-scheduled MU-MIMO users, based on factors such as the mobility of UEs 150a and 150b, the separation of UEs 150a and 150b, and network load. Aspects that can be adapted to different multi-user packet modes include selecting the style (e.g., selecting any UE 150a, 150 if the minimum or usual separation is wide, and selecting a sufficiently separated subset of UE 150a, 150b if UE 150a, 150b are closely spaced in different clusters), the number of UE 150a, 150b to be co-scheduled (selecting more UE 150a, 150b if network load is high and / or the mobility of UE 150a, 150b is low, and consistently supporting a high number of tiers is expected), etc. Network node 200 can also apply packet criteria to co-schedule UE 150a, 150b with similar mobility attributes, similar SINR, etc., under some conditions, while not applying such constraints under other conditions.

[0091] Aspects of digital precoding patterns will now be disclosed. In some embodiments, each candidate digital precoding pattern involves controlling the overall radiation pattern defining a beamforming lobe in which MU-MIMO signals are transmitted toward UEs 150a and 150b by applying mode-related weights to MU-MIMO signals from different baseband ports using digital signal processing. According to the above non-limiting example, the candidate digital precoding patterns may differ from each other in terms of signal power level in the direction toward UEs 150a and 150b, overall signal power distribution in the geographic vicinity of UEs 150a and 150b, suppression or invalidation of interference in the direction toward UEs 150a and 150b, and suppression or invalidation of interference in the geographic vicinity of UEs 150a and 150b. According to the above non-limiting example, each candidate digital precoding pattern may be associated with digital precoding using one of the following: matched filter digital precoding, zero-forcing digital precoding, regularized zero-forcing digital precoding, and signal leakage plus noise ratio digital precoding.

[0092] Aspects of simulated beam manipulation modes will now be disclosed. In some embodiments, each candidate simulated beam manipulation mode involves controlling the directivity of the MU-MIMO signal by applying a mode-dependent phase shift at the radio frequency front-end circuitry to the antenna array constituting the baseband port. According to the above non-limiting examples, the candidate simulated beam manipulation modes may differ from each other in terms of the signal power level in the direction toward UE 150a, 150b and the overall signal power distribution in the geographic vicinity area of ​​UE 150a, 150b. According to the above non-limiting examples, each candidate simulated beam manipulation mode may be associated with simulated beam manipulation using a corresponding one of: random angle simulated beam manipulation, selected combination simulated beam manipulation, and aggregated composite channel phase extraction simulated beam manipulation.

[0093] Therefore, some examples of candidate beamforming methods include MF, ZF, RZF and SLNR as different examples of digital precoding modes in the digital signal processing domain, and RA, SC and AC-CPE as different examples of analog beam manipulation modes in the analog signal processing domain.

[0094] In some respects, the selection of a multi-user transport configuration mode is based on a comparison between obtained parameters and one or more thresholds. Specifically, in some embodiments, a multi-user transport configuration is selected from candidate configurations based on a mapping, using a threshold for each type of parameter value, and comparing parameter values ​​with a set of thresholds. Examples of thresholds are provided below.

[0095] In some examples, for the high mobility of UEs 150a and 150b, a multi-user transmission configuration mode is selected to maximize signal power in the desired direction of UEs 150a and 150b, for example, using MF or RZF in combination with AC-CPE or SC. That is, in some embodiments, one type of parameter value is a mobility type parameter value, and when the mobility of each UE in UEs 150a and 150b is higher than a threshold of the mobility type parameter value, the multi-user transmission configuration is based on a mapping selected to maximize signal power in the direction toward UEs 150a and 150b.

[0096] In some examples, for the high mobility of UEs 150a and 150b, a multi-user transmission configuration mode can be selected to balance the desired power maximization and interference suppression, for example, using ZF, RZF, or SLNR in combination with AC-CPE. That is, in some embodiments, one type of parameter value is a mobility type parameter value, and when the mobility of each UE in UEs 150a and 150b is below a threshold of the mobility type parameter value, the multi-user transmission configuration is based on a mapping selected to balance maximizing signal power and suppressing interference in the direction toward UEs 150a and 150b.

[0097] In some examples, if the mobility state of a first number of UEs 150a, 150b, or a first portion of UEs 150a, 150b, is high, then network node 200 applies a multi-user transmission configuration defined by using MF as the digital precoding mode and SC as the analog beamforming mode, where a certain number can be an absolute number threshold, a relative portion threshold, or any such UEs 150a, 150b present. This multi-user transmission configuration is chosen because, in the case of high mobility, the physical radio propagation channel will be expected to decorrelate at a much faster rate (for both the desired and / or interfering UEs 150a, 150b), allowing for the opportunity to maximize the desired signal power, while mobility provides some implicit interference reduction due to the high probability of widely separated channels. The selection of the analog beamforming mode will also rely on identifying and adjusting the analog beamforming matrix such that it takes into account transmission in the direction of the strongest path of the desired UEs 150a, 150b, where the analog beam gain can be manipulated. This will produce the best performance achievable through hybrid beamforming.

[0098] Conversely, in some examples, if the mobility status of a second predetermined number or portion of UEs 150a and 150b is low, network node 200 applies a multi-user transmission configuration defined by using RZF as the digital precoding mode and AC-CPE as the analog beamforming mode. This is because low mobility may result in moderate channel correlation levels for both desired and interfering UEs 150a and 150b due to spatially consistent channels. This necessitates a more balanced approach between minimizing interference and maximizing desired signal strength by selecting RZF as the digital precoding mode and AC-CPE as the analog beamforming mode, using the acquired radio channel phases to provide the analog beamformer. Due to the low mobility of UEs 150a and 150b, these radio channel phases can be estimated with high accuracy to assist the analog beamformer.

[0099] While AC-CPE generally performs well and has the potential for use in a wide range of scenarios, one scenario in which it may not perform optimally is described below. When the mobility state of multiple UEs 150a, 150b (desired and / or interfering) is high (e.g., with millisecond coherence times), analog beamforming in the SC aspect will produce better performance than AC-CPE. In this case, to maximize spectral efficiency to each UE 150a, 150b, a multi-user transmission configuration is defined using MF as the digital precoding mode and SC as the analog beamforming mode. This is because, in high mobility scenarios, the physical radio propagation channel implicitly decorrelates to the desired and / or interfering UEs 150a, 150b at a much faster rate, allowing focus only on the opportunity to maximize the desired signal power. This is because mobility already provides an implicit interference reduction mechanism due to the high probability of widely separated channels over very short durations. This will maximize signal power in two phases; first, MF digital beamforming, and then SC analog beamforming.

[0100] In some respects, the goal of choosing a multi-user transport configuration is to maximize the SINR and spectral efficiency of UEs 150a and 150b while making cost trade-offs for the current scenario. One consideration is spatial distribution, and specifically the spatial separation of UEs 150a and 150b. While advanced approaches allow for zero manipulation for interference control, the effectiveness of zero manipulation decreases as co-scheduled UEs 150a and 150b become closer to each other in beam / angle space. Therefore, for closely spaced UEs 150a and 150b, a simpler beamforming alternative focused on maximizing desired signal power will offer nearly as good performance and is preferred due to lower processing complexity.

[0101] In some examples, the spatial separation metric S is defined to quantify the separation between UEs 150a, 150b. The spatial separation metric may be, for example, an average value or a predetermined percentile of the difference distribution of the main angles of arrival (AOA) of the signals of adjacent UEs 150a, 150b in azimuth and / or elevation. By comparing this metric with the threshold T, the selection of the digital precoding scheme can be based on such metric. In some examples, if S<T, MF digital precoding is applied; otherwise, digital precoding in terms of ZF, RZF or SLNR is used, while analog beam steering in terms of either SC or AC-CPE is used. The threshold or preference for selecting which digital precoding scheme to use may depend on a number of additional factors, such as the mobility patterns of UEs 150a, 150b that determine second-order variations in the propagation channels across multiple UEs 150a, 150b.

[0102] In mmW systems, it is advantageous to use the AOA-based spatial separation criterion to account for steering accuracy and separation limitations caused by, for example, main lobe width and channel estimation inaccuracies. Common millimeter-wave channels are dominated by specular reflection and line-of-sight (LOS) propagation. For the vast majority of measured mmW physical radio propagation channels, sparsity in the number of available multipath components is a key observable property, where only a limited number of dominant components are visible to UEs 150a, 150b and / or the network node 200. The mechanisms leading to denser multipath environments, diffuse scattering and diffraction, are generally not prominent at millimeter-wave frequencies. In any case, for dispersive physical radio propagation channels, separation does not necessarily have a geometric meaning, but may be quantified in terms of channel correlation over a set of small-scale fading mechanisms, for example. However, for the vast majority of measured physical radio propagation channels, sparsity in the number of available multipath components is a key observable property, where only a limited number of dominant components are visible to UEs 150a, 150b and / or the network node 200. The spatial separation threshold T used to select which digital precoding scheme to use may additionally be based on the antenna array size and the resulting beam resolution. Larger arrays can accommodate UEs 150a, 150b with closer separation (T can be reduced). It may also depend on the number of co-scheduled UEs 150a, 150b to serve the provided network load. If a small number of users are co-scheduled, the minimum separation T is lower.

[0103] In Figure 5 , the inter-user correlation coefficient as a function of the angular separation between two UEs 150a, 150b is illustrated. It can be observed that when UEs 150a, 150b are separated by 10° in azimuth, there is a moderate level of correlation. Due to spatial consistency increasing the level of the overall correlation coefficient, the limit to zero correlation is more than 50°. Figure 5Results for the TRP 140 equipped with an antenna array of 256 elements in a dual-polarized uniform planar array configured in a 16×16 element configuration are shown. When the two UEs 150a and 150b are relatively closely spaced (i.e., 2° angular separation), the inter-UE correlation is 0.73 regardless of sector distance. When UEs 150a and 150b are 10° apart, the inter-UE correlation drops to 0.25, and the limit toward zero correlation exceeds 50°. This is because even after a 50° relative spatial separation in azimuth angle, some of the spatially consistent multipath components remain with both UEs 150a and 150b. This has an impact on the spectral efficiency performance of the embodiments disclosed herein.

[0104] The threshold T can also depend on the achievable channel estimation quality. Typically, interference suppression schemes require high-quality channel estimation to accurately manipulate zeros. Channel estimation quality under poor link conditions can be improved by allocating additional training resources, but this comes at the cost of reduced SE due to reduced data resources. Therefore, for the low link quality of co-scheduled UEs 150a and 150b, the threshold T can be higher.

[0105] The threshold T can also depend on user mobility / movement speed. To maintain up-to-date Channel State Information (CSI), channel estimates need to be updated at the rate of channel coherence time. For UEs 150a and 150b with high mobility / Doppler, this means that data spectral efficiency will decrease due to the allocation of resources used for channel estimation / training. Therefore, for higher Doppler user groups or co-scheduled sets of UEs 150a and 150b, the threshold T can be higher.

[0106] The threshold T can also depend on the number of UEs 150a and 150b to be processed. For a larger number of co-scheduled UEs 150a and 150b, the complexity of multi-user transmission is higher, and the threshold T can therefore be higher to ensure sufficient performance improvement.

[0107] Generally, when UE mobility is high, the focus of analog beamforming and digital precoding is maximizing signal power in the direction of the desired (scheduled) UEs 150a and 150b. In such a scenario, interference suppression is not preferred because high mobility decorrelates both the desired and interfering UE channels. For this reason, a multi-user transmission configuration defined using SC as the analog beamforming mode and MF as the digital precoding mode can be selected. This will not be changed if UEs 150a and 150b are closely separated from each other while having high mobility.

[0108] For the lower mobility of UEs 150a and 150b, the desired and interfering radio channels contain significant spatial correlations due to their spatially consistent nature. Here, achieving a balance between maximizing desired power and suppressing interference using both analog and digital signal processing domains becomes more important for performance optimization. Consequently, a multi-user transmission configuration can be chosen that uses AC-CPE as the analog beamforming mode and RZF as the digital precoding mode. This will not be changed if UEs 150a and 150b are spatially far apart.

[0109] If the link SNR is high, it indicates that the network is operating under interference-limited conditions, where network node 200 focuses on interference suppression. Therefore, a multi-user transmission configuration defined using SC as the analog beamforming mode and RZF as the digital precoding mode can be selected. Conversely, if the link SNR is low, it indicates that the system is operating in a noise-limited region, where power maximization is prioritized, and thus a multi-user transmission configuration defined using AC-CPE as the analog beamforming mode and MF as the digital precoding mode can be selected.

[0110] If the network load is high, as long as RZF can be used as the digital precoding mode to perform interference suppression in the digital signal processing domain, then using the analog beamforming mode of RA will produce similar performance to other analog beamforming modes. Conversely, if the network load is low, interference suppression remains the criterion for performance optimization, and therefore RZF is retained as the digital precoding mode.

[0111] The following will disclose aspects of multi-user transmission configurations regarding the estimation of per-UE spectral efficiency. For efficient scheduling, the scheduler requires an estimate of the current system capacity to balance the risks of latency and network saturation. In some examples, network node 200 obtains a cell capacity estimate for a cell in which hybrid beamforming is applied by: (i) estimating the unconstrained UE spectral efficiency and / or cell throughput; (ii) estimating the spectral efficiency loss due to the hybrid beamforming used (i.e., the selected multi-user transmission configuration); and (iii) adjusting the unconstrained spectral efficiency estimate using the loss estimate.

[0112] Estimating unconstrained UE spectral efficiency and / or cell throughput can be performed using known techniques, such as Shannon capacity class expressions. To estimate the spectral efficiency loss due to the hybrid beamforming employed, a spectral efficiency expression for a given digital precoding mode (MF, ZF, RZF, or SLNR) can be derived using tools from multivariate statistics and Shannon theory. Then, using hybrid beamforming, another expression for spectral efficiency is derived for a given type of digital precoding and analog beamforming mode (SC, AC-CPE, or RA). Subsequently, for any given combination of analog beamforming and digital precoding, the difference between the resulting spectral efficiencies from fully digital precoding and hybrid beamforming is calculated and analyzed. This difference in spectral efficiency is then used to characterize the loss imposed by hybrid beamforming (any combination) relative to fully digital beamforming. The adjusted estimate can be used to more dynamically derive per-UE or total spectral efficiency limits, enabling scheduler optimization over a medium time horizon. The scheduler can then use the obtained spectral efficiency estimate to ensure that the scheduled load remains at a robust percentage of the achievable spectral efficiency, for example, no more than 70-80%.

[0113] Figure 6 The first mobility scenario is illustrated schematically, in which two UEs 150a and 150b are moving along their respective trajectories in the horizontal xy plane, as illustrated by the corresponding arrow lines, and are served by a network node 200 (represented by its TRP 140) equipped with a hybrid beamformer. The 3GPP TR 38.901 NLOS UMi propagation model is used to calculate the propagation channel impulse responses from TRP 140 to UEs 150a and 150b, respectively. The moving speeds of both UEs 150a and 150b are fixed at 0.83 m / s, and the distance between each realized movement of the channel is fixed at 0.1 m. This is consistent with the 3GPP TR38.901 NLOS UMi recommendation. Figure 7 This demonstrates different multi-user transmission configurations and for purely digital beamforming. Figure 6 The per-UE spectral efficiency of the scenario described is presented. The best achievable spectral efficiency is achieved by utilizing a combination of ZF and AC-CPE, which converges to a relatively small performance difference at low SNR and results in greater variation compared to a fully digital beamforming mechanism at high SNR. The superior digital beamforming technique is RZF, which consumes significant computational resources compared to the low-complexity alternative of hybrid ZF and AC-CPE. Figure 7 The results presented show the gain based on the average per UE. This would mean that the overall gain for L UEs 150a and 150b across the system would be Figure 7The overall gain shown is an average of L times. Furthermore, the depicted gain is shown per 1Hz bandwidth, and therefore in bits per second per Hz. This will naturally be multiplied by the provided bandwidth to produce a rate characterized in bits per second.

[0114] Figure 8 A second mobility scenario is illustrated schematically, in which two UEs 150a and 150b are served by a network node 200 (represented by its TRP 140) equipped with a hybrid beamformer. Mobility in the second mobility scenario is higher than that in the first mobility scenario. UE 150a is moving at 10 m / s (36 km / h) in the horizontal xy plane, while UE 150b is stationary in the horizontal xy plane. The overall trajectory of UE 150a is indicated by a dashed line. UE 150a follows its trajectory in the clockwise direction indicated by the arrow. All remaining system parameters are related to... Figure 6 The parameters are the same. Figure 9 This illustrates different multi-user transport configurations with MF digital precoding. Figure 8 The per-UE spectral efficiency of the scenario described in the text. For example, in... Figure 9 As can be seen, especially for medium and large SNR, SC outperforms AC-CPE in terms of analog beam manipulation methods.

[0115] In summary, the embodiments disclosed herein enable the selection of a suitable hybrid beamforming mode based on statistical data of co-scheduled UEs 150a and 150b, as well as other optional factors, parameters, or conditions. The selection of the digital precoding mode can be independent of the selection of the analog beamforming mode, and combinations of digital precoding and analog beamforming modes can be selected to maximize desired signal power and / or minimize multi-user interference. Specific combinations of digital precoding and analog beamforming modes can be selected to ensure the maximization of relevant quantities across all UE conditions (high and low mobility, high and low UE 150a and 150b density, closely separated and distant UEs 150a and 150b, etc.). For closely spaced UEs 150a and 150b, a combination of MF or RZF with AC-CPE or SC can be applied to maximize received power in both the analog and digital signal processing domains. For spatially separated UEs 150a and 150b, a combination of ZF, RZF, or SLNR with AC-CPE can be selected to provide optimal spectral efficiency as a result of effective interference suppression. The appropriate scheduling mode (in terms of choosing between RRS, PFS, or maximum SINR scheduling) and / or MU-MIMO user packet mode (in terms of the number of UEs 150a and 150b simultaneously co-scheduled, and the selection criteria for UEs 150a and 150b) can be selected based on the feasible number of MU-MIMO layers, UE location and / or distribution, mobility, and / or predictability. Table 1 provides an overview of how multi-user transmission configurations can be selected as a function of UE spatial separation, UE mobility, UE link quality, and network load, based on the embodiments, aspects, and examples disclosed herein.

[0116]

[0117]

[0118]

[0119] Table 1: Overview of hybrid beamforming modes, scheduling modes, and multi-user packet modes as functions of UE spatial separation, UE mobility, UE link quality, and network load.

[0120] Figure 10 The components of the network node 200 according to an embodiment are schematically illustrated in terms of multiple functional units. The processing circuitry 210 is provided using any combination of one or more of the following: capable of executing a computer program product 1210 stored (e.g., in the form of storage medium 230). Figure 12The software instructions in the (in the middle) can be suitable for a central processing unit (CPU), multiprocessor, microcontroller, digital signal processor (DSP), etc. The processing circuitry 210 can be further provided as at least one application-specific integrated circuit (ASIC) or field-programmable gate array (FPGA).

[0121] Specifically, processing circuitry 210 is configured to cause network node 200 to perform a set of operations or steps as disclosed above. For example, storage medium 230 may store the set of operations, and processing circuitry 210 may be configured to retrieve the set of operations from storage medium 230 to cause network node 200 to perform the set of operations. The set of operations may be provided as a set of executable instructions.

[0122] Therefore, processing circuitry 210 is thus arranged to perform the methods disclosed herein. Storage medium 230 may also include a persistent storage device, which may be any single or combination of magnetic storage, optical storage, solid-state storage, or even remotely mounted storage. Network node 200 may further include a communication interface 220, which is configured at least to communicate with other entities, functions, nodes, and devices of communication network 100, as well as entities, functions, nodes, and devices (such as user equipment 150a:150d) served by communication network 100. In this way, communication interface 220 may include one or more transmitters and receivers, which include analog and digital components. Processing circuitry 210 controls the general operation of network node 200, for example, by sending data and control signals to communication interface 220 and storage medium 230, by receiving data and reports from communication interface 220, and by retrieving data and instructions from storage medium 230. Other components and related functionalities of network node 200 are omitted to avoid obscuring the concepts presented herein.

[0123] Figure 11 The components of the network node 200 according to the embodiment are illustrated schematically in terms of multiple functional modules. Figure 11 The network node 200 includes multiple functional modules: an acquisition module 210a configured to execute step S102; a selection module 210b configured to execute step S104; and a transmission module 210c configured to execute step S106. Figure 11The network node 200 may further include a plurality of optional functional modules, such as represented by functional module 210d, which is configured to perform any further steps included in the embodiments disclosed herein. Generally, each functional module 210a:210d may be implemented in hardware only in one embodiment and in software by means of software in another embodiment, i.e., the latter embodiment has computer program instructions stored on storage medium 230 that, when executed on processing circuitry, cause the network node 200 to perform the above-described combined... Figure 11 Regarding the corresponding steps mentioned, it should also be noted that even if the modules correspond to part of a computer program, they do not need to be separate modules within it; however, the way they are implemented in software depends on the programming language used. Preferably, one or more or all functional modules 210a:210d can be implemented by processing circuitry 210 (possibly cooperating with communication interface 220 and / or storage medium 230). Thus, processing circuitry 210 can be configured to retrieve instructions such as those provided by functional modules 210a:210d from storage medium 230 and be configured to execute these instructions, thereby performing any steps disclosed herein.

[0124] Network node 200 can be provided as a standalone device or as part of at least one other device. For example, network node 200 can be provided in a node of a radio access network or in a node of a core network. Alternatively, the functionality of network node 200 can be distributed among at least two devices or nodes. These at least two nodes or devices can either be part of the same network segment (such as a radio access network or a core network) or can be distributed among at least two such network segments. Generally, instructions requiring real-time execution can be executed in devices or nodes that are operationally closer to the cell than instructions that do not require real-time execution.

[0125] Therefore, the first portion of the instructions executed by network node 200 can be executed in a first device, and the second portion of the instructions executed by network node 200 can be executed in a second device; the embodiments disclosed herein are not limited to any particular number of devices on which the instructions executed by network node 200 can be executed. Therefore, the method according to the embodiments disclosed herein is suitable for execution by network node 200 residing in a cloud computing environment. Therefore, although in Figure 10 The document describes a single processing circuit 210, but processing circuits 210 can be distributed across multiple devices or nodes. This also applies to... Figure 11 Functional modules 210a: 210d and Figure 12 Computer program 1220.

[0126] Figure 12An example of a computer program product 1210 including a computer-readable storage medium 1230 is shown. On this computer-readable storage medium 1230, a computer program 1220 may be stored, which can cause processing circuitry 210 and entities and means operatively coupled thereto (such as a communication interface 220 and storage medium 230) to perform methods according to the embodiments described herein. Therefore, computer program 1220 and / or computer program product 1210 can provide components for performing any of the steps disclosed herein.

[0127] exist Figure 12 In the example, computer program product 1210 is illustrated as an optical disc, such as a CD (Compact Disc), DVD (Digital Versatile Disc), or Blu-ray Disc. Computer program product 1210 can also be embodied as a memory, such as random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or electrically erasable programmable read-only memory (EEPROM), and more particularly as a non-volatile storage medium of a device in an external memory such as USB (Universal Serial Bus) memory or flash memory (such as compact flash memory). Therefore, although computer program 1220 is schematically shown herein as a track on the depicted optical disc, computer program 1220 can be stored in any manner suitable for computer program product 1210.

[0128] The inventive concept has been primarily described above with reference to several embodiments. However, as will be readily apparent to those skilled in the art, other embodiments besides those disclosed above are also possible within the scope of the inventive concept as defined by the appended claims.

Claims

1. A network node (200) for transmitting multi-user multiple-input multiple-output (MU-MIMO) signals, the network node (200) comprising processing circuitry (210) configured to cause the network node (200) to: Obtain parameter values ​​of statistical data for user equipments (150a, 150b) jointly scheduled by the network node (200), the statistical data being at least related to the mobility of each user equipment (150a, 150b), wherein, One type of parameter value is a mobility-type parameter value; A multi-user transmission configuration is dynamically selected from candidate multi-user transmission configurations, wherein each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beam manipulation mode selected from candidate analog beam manipulation modes, and wherein the multi-user transmission configuration is selected based on parameter values ​​and a configuration mapping of the parameter values ​​to the candidate multi-user transmission configurations, wherein when the mobility of each user equipment (150a, 150b) is higher than a threshold value of the mobility type parameter value, the multi-user transmission configuration is selected based on the mapping to maximize signal power in the direction toward the user equipment (150a, 150b), and when the mobility of each user equipment (150a, 150b) is lower than the threshold value of the mobility type parameter value, the multi-user transmission configuration is selected based on the mapping to balance maximizing signal power and interference suppression in the direction toward the user equipment (150a, 150b); and The MU-MIMO signal is transmitted toward the user equipment (150a, 150b) using the selected multi-user transmission configuration.

2. The network node (200) according to claim 1, wherein, Each of the candidate multi-user transport configurations further specifies the scheduling mode selected from the candidate scheduling modes.

3. The network node (200) according to claim 2, wherein, Each of the candidate scheduling modes is associated with a scheduling mode using one of the following: round-robin scheduling, proportional fair scheduler, maximum signal-to-noise plus interference ratio scheduling.

4. The network node (200) according to claim 2 or 3, wherein, Selecting the scheduling mode as part of selecting the multi-user transmission configuration is performed less frequently in time compared to selecting the digital precoding mode and selecting the analog beam manipulation mode.

5. The network node (200) according to any one of claims 1 to 3, wherein, Each of the candidate multi-user transport configurations further specifies the multi-user packet mode selected from the candidate multi-user packet modes.

6. The network node (200) according to claim 5, wherein, Each of the candidate multi-user grouping modes relates to the multi-user grouping of the user equipment (150a, 150b) in at least one of the following aspects: a criterion for how many of the user equipment (150a, 150b) should be jointly scheduled, and a criterion for which of the user equipment (150a, 150b) should be jointly scheduled.

7. The network node (200) according to any one of claims 1 to 3, wherein, The mobility of the user equipment (150a, 150b) is defined by the corresponding user equipment movement trajectory and the mobility state of the user equipment (150a, 150b).

8. The network node (200) according to any one of claims 1 to 3, wherein, The statistics are further related to at least one of the following: spatial separation between the user equipments (150a, 150b), and link quality of each user equipment (150a, 150b).

9. The network node (200) according to any one of claims 1 to 3, wherein, The parameter values ​​mentioned are also the parameter values ​​for network operation parameters.

10. The network node (200) according to claim 9, wherein, The network operation parameters are related to at least one of the following: the current service load of the network node (200) and the current number of MU-MIMO layers available at the network node (200).

11. The network node (200) according to any one of claims 1 to 3, wherein, The mapping is configured based on an optimization criterion that maximizes the signal-to-noise plus interference ratio and spectral efficiency of the user equipment (150a, 150b).

12. The network node (200) according to any one of claims 1 to 3, wherein, The mapping is determined from at least one of the following: previous performance observations of the network node (200), and simulations of operating the network node (200).

13. The network node (200) according to any one of claims 1 to 3, wherein, The mapping is provided as a lookup table, wherein each entry in the lookup table specifies one of the candidate multi-user transmission configurations in terms of a combination of at least one digital precoding mode and one analog beam manipulation mode.

14. The network node (200) according to any one of claims 1 to 3, wherein, Based on the mapping, the multi-user transmission configuration is selected from the candidate multi-user transmission configurations based on a comparison between the parameter values ​​and a threshold set, with a threshold for each type of parameter value.

15. The network node (200) according to any one of claims 1 to 3, wherein, Each of the candidate digital precoding modes involves controlling the overall radiation pattern defining a beamforming lobe by applying mode-related weights to the MU-MIMO signals from different baseband ports using digital signal processing, in which the MU-MIMO signals are transmitted toward the user equipment (150a, 150b).

16. The network node (200) according to any one of claims 1 to 3, wherein, The candidate digital precoding modes differ from each other in terms of signal power level in the direction toward the user equipment (150a, 150b), overall signal power distribution in the geographic vicinity of the user equipment (150a, 150b), suppression or invalidation of interference in the direction toward the user equipment (150a, 150b), and suppression or invalidation of interference in the geographic vicinity of the user equipment (150a, 150b).

17. The network node (200) according to any one of claims 1 to 3, wherein, Each of the candidate digital precoding modes is associated with digital precoding using one of the following: matched filter digital precoding, zero-forcing digital precoding, regularized zero-forcing digital precoding, and signal-to-leakage-plus-noise ratio digital precoding.

18. The network node (200) according to any one of claims 1 to 3, wherein, Each of the candidate analog beam manipulation modes involves controlling the directivity of the MU-MIMO signal by applying a mode-dependent phase shift at the radio frequency front-end circuitry to the antenna element group constituting the baseband port.

19. The network node (200) according to any one of claims 1 to 3, wherein, The candidate simulated beam manipulation modes differ from each other in terms of signal power level in the direction toward the user equipment (150a, 150b) and overall signal power distribution in the geographic vicinity of the user equipment (150a, 150b).

20. The network node (200) according to any one of claims 1 to 3, wherein, Each of the candidate simulated beam manipulation modes is associated with simulated beam manipulation using one of the following: random angle simulated beam manipulation, selected combination simulated beam manipulation, and aggregated composite channel phase extraction simulated beam manipulation.

21. The network node (200) according to any one of claims 1 to 3, wherein, The multi-user transmission configuration is defined by a hybrid beamforming mode consisting of a digital precoding mode and an analog beam manipulation mode.

22. The network node (200) according to any one of claims 1 to 3, wherein, The multi-user transport configuration is dynamically selectable for each scheduling instance of the user equipment (150a, 150b).

23. A method for transmitting multi-user multiple-input multiple-output (MU-MIMO) signals, the method being performed by a network node (200), the method comprising: (S102) Obtain parameter values ​​of statistics of user equipment (150a, 150b) jointly scheduled by the network node (200), the statistics being at least related to the mobility of each user equipment (150a, 150b), wherein one type of parameter value is a mobility type parameter value; (S104) Dynamically select a multi-user transmission configuration from candidate multi-user transmission configurations, wherein each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beam manipulation mode selected from candidate analog beam manipulation modes, and wherein the multi-user transmission configuration is selected based on the parameter values ​​and a configuration mapping of the parameter values ​​to the candidate multi-user transmission configurations, wherein when the mobility of each user equipment (150a, 150b) is higher than a threshold value of the mobility type parameter value, the multi-user transmission configuration is selected based on the mapping to maximize signal power in the direction toward the user equipment (150a, 150b), and when the mobility of each user equipment (150a, 150b) is lower than the threshold value of the mobility type parameter value, the multi-user transmission configuration is selected based on the mapping to balance maximizing signal power in the direction toward the user equipment (150a, 150b) and interference suppression; and The MU-MIMO signal is transmitted (S106) toward the user equipment (150a, 150b) using the selected multi-user transmission configuration.

24. A network node (200) for transmitting multi-user multiple-input multiple-output (MU-MIMO) signals, the network node (200) comprising: The acquisition module (210a) is configured to acquire parameter values ​​of statistics of user equipment (150a, 150b) jointly scheduled by the network node (200), the statistics being at least related to the mobility of each user equipment (150a, 150b), wherein one type of parameter value is a mobility type parameter value; The selection module (210b) is configured to dynamically select a multi-user transmission configuration from candidate multi-user transmission configurations, wherein each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beam manipulation mode selected from candidate analog beam manipulation modes, and wherein the multi-user transmission configuration is selected based on the parameter values ​​and a configuration mapping of the parameter values ​​to the candidate multi-user transmission configurations, wherein when the mobility of each user equipment (150a, 150b) is higher than a threshold value of the mobility type parameter value, the multi-user transmission configuration is selected based on the mapping to maximize signal power in the direction toward the user equipment (150a, 150b), and when the mobility of each user equipment (150a, 150b) is lower than the threshold value of the mobility type parameter value, the multi-user transmission configuration is selected based on the mapping to balance maximizing signal power and interference suppression in the direction toward the user equipment (150a, 150b); and The transmission module (210c) is configured to transmit the MU-MIMO signal toward the user equipment (150a, 150b) using a selected multi-user transmission configuration.

25. A computer program product comprising computer code for transmitting multi-user multiple-input multiple-output (MU-MIMO) signals, the computer code causing the network node (200) to: (S102) Obtain parameter values ​​of statistical data for user equipments (150a, 150b) jointly scheduled by the network node (200), the statistical data being at least related to the mobility of each user equipment (150a, 150b), wherein, One type of parameter value is a mobility-type parameter value; (S104) Dynamically select a multi-user transmission configuration from candidate multi-user transmission configurations, wherein each candidate multi-user transmission configuration specifies at least a digital precoding mode selected from candidate digital precoding modes and an analog beam manipulation mode selected from candidate analog beam manipulation modes, and wherein the multi-user transmission configuration is selected based on the parameter values ​​and a configuration mapping of the parameter values ​​to the candidate multi-user transmission configurations, wherein when the mobility of each user equipment (150a, 150b) is higher than a threshold value of the mobility type parameter value, the multi-user transmission configuration is selected based on the mapping to maximize signal power in the direction toward the user equipment (150a, 150b), and when the mobility of each user equipment (150a, 150b) is lower than the threshold value of the mobility type parameter value, the multi-user transmission configuration is selected based on the mapping to balance maximizing signal power in the direction toward the user equipment (150a, 150b) and interference suppression; and The MU-MIMO signal is transmitted (S106) toward the user equipment (150a, 150b) using the selected multi-user transmission configuration.

26. A computer-readable storage medium storing instructions, wherein, When executed by a processor, the instructions implement the method of claim 23.

Images