PRE-CODING INFORMATION

Abstract

Method and apparatus for providing a channel status information report, comprising: receiving configuration information to configure a measurement window to form a compression matrix of a port selection codebook from a set of vector component codebooks, wherein the configuration information defines the size of the measurement window, which is common to all of the at least one layer to be reported; selecting a number of measurement window indices based on the configuration information to form the compression matrix from the set of vector component codebooks; and reassigning the selected indices, associated with vector components of the compression matrix, with respect to a reference vector component index, such that the reference vector component index is reassigned to a first measurement window index;Report channel status information including a pre-coding matrix indicator to a network; the pre-coding matrix indicator comprises compression matrix information after reassignment.

Description

Field This disclosure relates to methods, apparatus, and software products for communicating precoding information between communication devices. Background Communication sessions can be established between two or more communication devices, such as user devices or terminals, base stations / access points, and / or other nodes. A communication session can be provided, for example, by means of a communication network and one or more compatible communication devices. A communication device on one side of the network provides an access point to the system and is equipped with appropriate signal reception and transmission equipment to enable communication, for example, allowing other devices to access the communication system. Communication sessions can include, for example, data communication to carry communications such as voice, video, email, text messages, multimedia, and / or content data, and so on.Non-exhaustive examples of services provided include two-way or multi-way calls, data communication, multimedia services, and access to a data network system, such as the Internet. In a mobile or wireless communication system, at least part of a communication session between at least two devices occurs via a wireless or radio link. Examples of wireless systems include public land mobile networks (PLMNs), satellite-based communication systems, and various wireless local area networks (WLANs). A user can access the larger communication system through an appropriate communication device or terminal. A communication device is commonly referred to as user equipment (UE) or a user device. A communication device is equipped with signal receiving and transmitting apparatus to enable communication, such as access to a communications network or direct communication with other users.A communication device can access a carrier provided by a station in a radio access network, for example, a base station, and transmit and / or receive communications on the carrier. A feature of modern systems is the ability to operate in multiple paths, where a communication device can communicate through multiple routes. Multipath communication can be provided through an arrangement known as multiple-input / multiple-output (MIMO). The communication system and associated devices typically operate in accordance with a given standard or specification that establishes what the various entities associated with the system are permitted to do and how this should be achieved. The communication parameters and / or protocols to be used for the connection are also typically defined. An example of a communication system is UTRAN (3G radio). Other examples of communication systems include Long-Term Evolution (LTE) of the Universal Mobile Telecommunications System (UMTS) radio access technology and so-called fifth-generation (5G) or New Radio (NR) networks. 5G is being standardized by the Third Generation Partnership Project (3GPP). Successive versions of the standard are known as releases. Within 3GPP, 5G NR standardization work is underway to further improve aspects of MIMO channel state information (CSI). Compendium The subject matter of the independent claims is provided in respect of certain aspects. Additional aspects are defined in the dependent claims. Brief description of the figures Some aspects will be described in more detail below, solely as examples, with reference to the following examples and the accompanying figures, in which: Figure 1 illustrates an example of a system in which the invention can be put into practice; Figure 2 shows an example of a control device; Figure 3 is an example of frequency domain (FD) components that are outside the reporting window; Figures 4 and 5 are flowcharts according to certain examples; Figure 6 is a signaling flow diagram between two communication devices according to an example; Figures 7 and 8 illustrate measurement windows and reports configured according to two examples; Figures 9 and 10 are flowcharts according to additional examples; Figure 11 is another example of measurement and reporting windows; Figures 12 to 17 show examples of index mapping; Figure 18 shows another solution to solve the problem of being out of the window according to a modality; Figures 19 to 20 show methods, according to some modalities. Detailed description of the examples The following description provides an illustrative account of some possibilities for implementing the invention. Although the description may refer to one or more embodiments in various places, this does not necessarily mean that each reference is to the same embodiment, or that a particular feature applies only to a single embodiment. Individual features of different embodiments may also be combined to provide other embodiments. Wireless communication systems provide wireless communication to connected devices. Typically, an access point, such as a base station, is provided to enable this communication. The following will describe different scenarios using a 3GPP 5G radio access architecture with MIMO capability as an example. However, the available options are not necessarily limited to this architecture.Some examples of other possible systems are the Universal Mobile Telecommunications System (UMTS), the Radio Access Network (UTRAN or E-UTRAN), Long-Term Evolution (LTE), LTE-A (LTE Advanced), Wireless Local Area Network (WLAN or Wi-Fi), Worldwide Interoperability for Microwave Access (WiMAX), Bluetooth®, Personal Communications Services (PCS), ZigBee®, Wideband Code Division Multiple Access (WCDMA), systems using Ultra Wideband (UWB) technology, sensor networks, Mobile Ad-hoc Networks (MANET), Cellular Internet of Things (IoT), RAN and Internet Protocol Multimedia Subsystems (IMS), or any combination thereof. Figure 1 shows a wireless system 1 comprising a radio access system 2. A radio access system may comprise one or a plurality of access points, or base stations 12. A base station may provide one or more cells. An access point may comprise any node that can transmit / receive radio signals (e.g., a TRP, a 3GPP 5G base station such as gNB, eNB, a user device such as a UE, etc.). A communication device 10 is located within the service area of ​​radio access system 2, and device 10 can then communicate with access point 12. Communications 11 from device 10 to access point 12 are commonly referred to as the uplink (UL). Communications 13 from access point 12 to device 10 are commonly referred to as the downlink (DL). In the example, the downlink is shown schematically to comprise up to four beams per polarization in the spatial domain (SD). It should be noted that the broader communication system is only shown as cloud 1 and may comprise a number of elements that, however, are not shown for clarity. For example, a 5G-based system may consist of a user equipment (UE), a 5G radio access network (5GRAN) or a next-generation radio access network (NG-RAN), a 5G core network (5GC), one or more application functions (AF), and one or more data networks (DN). The 5G-RAN may comprise one or more gNodeBs (gNBs) or one or more gNodeB distributed unit functions (gNBs) that connect to one or more gNodeB centralized unit functions (gNBs). The 5GC may also comprise entities such as a network segment selection function (NSSF); a network exposure function; a network repository function (NRF); a policy control function (PCF); unified data management (UDM); an application function (AF); an authentication server function (AUSF); an access and mobility management function (AMF); a session management function (SMF); and so on. Device 10 can be any suitable communications device adapted for wireless communications. A wireless communications device can be provided by any device capable of sending and receiving radio signals. Non-exhaustive examples include a mobile station (MS) (e.g., a mobile device such as a mobile phone or what is known as a 'smartphone'), a computer equipped with a wireless interface card or other wireless interface installation (e.g., a USB dongle), a personal data assistant (PDA) or tablet equipped with wireless communication capabilities, machine-type communication (MTC) devices, Internet of Things (IoT) communication devices, or any combination thereof or similar. The device can be provided as part of another device.The device can receive signals via an air or radio interface using appropriate receiving equipment and can transmit signals using appropriate radio signal transmission equipment. Communications can occur via multiple paths. To enable MIMO communications, devices 10 and 12 are equipped with multi-antenna elements. These are schematically represented by antenna assemblies 14 and 15. A communications device such as access point 12 or user device 10 is provided with a data processing apparatus comprising at least one processor and at least one memory. Figure 2 shows an example of a data processing apparatus 50 comprising processor(s) 52, 53 and memory or memories 51. Figure 2 further shows connections between the apparatus elements and an interface for connecting the data processing apparatus to other device components. The at least one memory may comprise at least one ROM and / or at least one RAM. The communications device may comprise other possible components for use in the software- and hardware-assisted execution of the tasks for which it is designed, including access control and communications with access systems and other communications devices, and implement the features described herein for device positioning. The at least one processor may be coupled to the at least one memory. The at least one processor may be configured to execute software code appropriate for implementing one or more of the following aspects. The software code may be stored in the at least one memory, for example, in the at least one ROM. The following describes certain aspects of measurements, configurations, and signaling for operations related to multipath or multibeam wireless transmissions using 5G terminology. In frequency-division duplex (FDD) systems, full uplink-downlink (UL-DL) reciprocity cannot be assumed due to the duplexing distance between the uplink (UL) and downlink (DL) channels. However, partial channel reciprocity can be assumed based on properties such as angles of departure (AoD), angles of arrival (AoA), and multipath propagation delays. These partial UL-DL reciprocity properties can be considered in the signaling between communication devices.For example, a gNB can estimate UL probe reference signals (SRS) to acquire delay-related information, such as frequency domain (FD) components, which may be the same as a UE selection made via the channel DL status information reference signal (CSI-RS). The gNB can then use the selected FD components to further precode CSI-RS resources formed by beams that already contain space domain (SD) beams. The CSI-RS ports must be configured to transmit multiple sets of FD components across the CSI-RS. It has been recognized that CSI MIMO feedback operation can be improved by exploiting partial uplink / downlink (UL / DL) reciprocity.The improvement in CSI measurement and reporting can be based on evaluating and, if necessary, specifying improvements to the port selection codebook (e.g., based on the existing 3GPP ReL15 / 16 Type II port selection codebook). In this improvement, angle and delay information are estimated in the gNB based on SRS using DL / UL angle and delay reciprocity, and the UE reports the remaining DL CSI. This has primarily focused on frequency-division duplexing (FDD) of frequency range 1 (FR1) to achieve a better balance between UE complexity, throughput, and reporting overhead. For example, the Type II port selection (PS) codebook was improved in 3GPP Rel-16 by introducing frequency-domain compression (FD) operations in the 3GPP Rel-15 Type II port selection codebook.This enhanced Type II (eType II) port selection (PS) codebook is described, for example, in section 5.2.2.2.6 of 3GPP TS 38.214 v16.3.0 September 2020. Enhanced Type II (eType II) Port Selection (PS) Code Books (CBs) of 3GPP Rel-16 comprises three codebook components: the port selection matrix (M4), the discrete Fourier transform (DFT)-based compression matrix (U^), and the combination coefficient matrix (W2). In one example, the beam grid matrix Wts, of size 2N1N2 x 2L, provides spatial domain components; a linear combination coefficient (LCC) matrix W2 is of size 2L x Mv; and the DFT-based compression matrix Wts provides delay information (each column represents the delay tap) and is of size N3 x Mv. The parameter N is the number of antenna ports in the horizontal domain, N2 is the number of antenna ports in the vertical domain, L is the number of orthogonal vectors / beams per polarization, and N3 is the number of PMI frequency sub-bands. M is the number of frequency domain (FD) components. The same codebook structure is adopted in the further enhanced 3GPP Rel-17 PS Type II (FeType II) codebook for partial UL-DL channel reciprocity in frequency-division duplex (FDD) operation. An overview of the differences in the notification mechanism for the compression matrix (Wf) in FeType II PS compared to the eType II PS codebook, and their potential consequences, is provided below. In 3GPP Rel-16, the network configures the number of DFT vectors (Mv) that form the compression matrix (Wf), which depends on the reported range (v), by setting a combination of parameters (paramCombination-r16). Mv components can be understood as denoting basis or component elements in the frequency domain (FD). More generally, this parameter comprises the appropriate vectors or components. In the accompanying figures, the components are defined by the X-axis to indicate the DFT transformation of the frequency domain. The components correspond to channel delays, so the domain can also be called the delay domain. A UE can then select Mv vectors of length N3 from a DFT codebook for each layer, where N3 is the number of sub-bands in the precoding matrix indicator (PMI), which is equal to the size of the DFT codebook. The selection is made separately for each layer, and therefore each 14^ is considered layer-specific.The UE selects these DFT vectors from the entire set of codebooks of N3 vectors if N3 < 19, while it is restricted to selecting vectors from a window of size N = 2MV if N3 > 19. In the windowing mechanism, a UE chooses the best position for the window, provided that component 0 is included in the window and the window position is common to all layers. In practice, the UE selects and reports the initial window component, Miniciai, as {—N + 1, -N + 2, ..., 0}. Note that a negative value of Miniciai means that the window encloses the set of components N3; that is, all candidate components are represented modulo N3. A design principle in the Wfen 3GPP Rel-16 report is that the strongest coefficient for each layer is found in the first column of the bitmap associated with the non-zero coefficients of VK2, corresponding to component 0. IV2 is reported using a bitmap of size K^MV. The bitmap indicates the position of the reported non-zero coefficients, and the amplitudes and phases of the coefficients are reported sequentially after the bitmap indication, where Kλ < P is the number of CSI-RS ports that a UE is configured to select. This is ensured by the UE configured to apply two cyclic offsets for each layer l: a first cyclic offset, -n^p, is applied modulo / V2 to all Mv components of IT}· for a given layer, so that the component of the strongest coefficient is moved to position 0.A second cyclic shift, - / / , which is related to the first, is applied modulo Mv to the columns of W2, i.e., to the bitmap of size 2L x Mvv, in the corresponding order in which the amplitude and phase coefficients are reported. Note that fj ∈ {0,1, -1} is the column index of the strongest coefficient of W2 before the shift, while is the index of the FD component of the strongest coefficient before the shift. After these cyclic shifts, a UE reports the non-zero Mv-1 components of Wf in relation to the 0 component, which is, after the shift, the component of the strongest coefficient.A UE also reports W2 after the shift, so the columns of W2 correspond to the Mv components in ascending order of their index: the first column of W2 corresponds to component 0 (the strongest coefficient component after the shift), the second column of IV2 corresponds to the first reported component of Wf, the third column of W2 corresponds to the second reported component of Wf, and so on. The gNB does not need to know fj because the first cyclic shift is applied to the Wf components before reporting. This principle exploits a property of the precoder vectors indicated by a Precoding Matrix Indicator (PMI), whose performance is unaffected by a cyclic shift of the components in Wf. According to 3GPP Rel-16, the position of the strongest coefficient in the W2 matrix is ​​reported for each layer because it serves as an amplitude reference for the other nonzero coefficients in IV2.The reported position may only include the row index because the gNB may assume the column is the first column. W2 is layer-specific. Therefore, the gNB receiving the report may expect the strongest coefficient to be in component 0, i.e., the first column of W2, and thus the Strongest Coefficient Indicator (SCI) only indicates the beam index, i.e., the row index of W2 corresponding to its location. In practice, the windowing mechanism used in 3GPP Rel-16 for W3 > 19 works such that the UE shifts the strongest coefficient component for each layer to position 0, then determines the best value of Miniciai, common to all layers, and then selects different Mvcomponents for each layer within a component window of length N starting from Miniciai.In this way, all the Mvcomponents of each layer, after the cyclic displacement, fall within the report window starting in Minicial. It can also be assumed that W{is common to the layers. In that case, the proposed indicators such as MiTliciaÍo or the position of the strongest coefficient can be common to the layers. For PS FeType II codebooks, the current agreement is that the Mv components, also known as frequency domain (FD) basis points, are limited within a single window / set of size N. At least for range 1, the FD basis points used for Wf quantization are limited within a single window / set of size N configured for the UE. It has been proposed that the FD basis points in the window must be consecutive from an orthogonal DFT array. An alternative proposal is that the FD basis points in the set can be consecutive or non-consecutive, and gNB freely selects them from an orthogonal DFT array. Other conditions, such as whether Wf can be turned on / off and / or the associated value of Mvy if this applies when Wf is turned off, are still open. A measurement window configured by gNB can restrict the selection of the Mv vectors of Wf and is common to the layers.Furthermore, its meaning is unclear and open to different interpretations. For example, the window could be used to limit the maximum separation between Mv components for each layer when Mv > 1. Note that this differs from the window in Rel-16 PS for JV3 > 19, where the IntS window, of size 2MV, is common to the layers, thus limiting the maximum separation between selected components across all layers after aligning the strongest coefficient components. Another possible interpretation is that the window configures a set of DFT vectors for W2 measurement; that is, a UE is configured to measure (at least) the DFT vectors within the window (but it can measure more) and select Mv components that span at most N components—that is, they fit within the configured window, after appropriate offsetting, if necessary.Note that a UE can calculate more than N components and choose the best Mv outside the configured measurement window comprising FD components from 0 to IV - 1, provided the selected components fit within the window after a cyclic shift common to all Mv components. In fact, two different Wf selections obtained from each other via a cyclic shift produce different W2 values, but can be reported as the same Wf without affecting precoder performance. Two more alternatives are possible with respect to the relationship between the parameters V and Mv. The candidate values ​​for Mv in Rel-17 are smaller than in Rel-16, with only 1, 2 (and possibly 4) considered. The window / set size of Rel-17 is also smaller compared to Rel-16, with the most likely candidate values ​​also being 1, 2, 3, and 4. At least for range 1, it has been proposed that the relationship between N and Mv is either N = Mv or N > Mv. The candidate value(s) of N are under investigation, but could be, for example, 2 or 4. One difference between the window / set mechanism in Rel-17 and that of Rel-16 is that the window / set choice is made by the gNB, rather than the UE, by exploiting the partial reciprocity of delays between the UL and DL channels in the FDD operation. However, the position of the strongest coefficient in each layer is determined by the UE because the fast-fading channel components are not reciprocal. Therefore, if the same design principle as Rel-16 is reused—meaning that a gNB expects the strongest coefficient in the first Wf column—it is possible for a UE to shift the Mv components outside the configured window. This out-of-window problem can occur in the design of the Rel-17 port selection (PS) codebook, for example, for Wf configurations with Mv > 1. To illustrate, consider an example with the parameter combination N = M = 2 where the window is formed by two consecutive Digital Fourier Transform (DFT) components, 0 and 1. Figure 3 shows the frequency domain (FD) components and the configured window of size N = 2 for two layers. For each layer, the vertical bars correspond to the Mv-2 FD components. The height of the bar illustrates the amplitude of the M / 2 combination coefficients for the port with the strongest coefficient in each layer. It should be noted that each layer consists of Ki < P ports selected by the UE from the total P CSI-RS ports, and therefore the size of the IV2 matrix is ​​Ki * Mv for each layer.It is also observed that the Mv components are common to all ports in a layer and, in this case, to all layers; that is, they are common to the layer. However, Figure 3 only illustrates the amplitudes of the Mv coefficients for the port where the strongest of the KiMr coefficients is located. For layer 1, the strongest coefficient is in component 1 of a port; therefore, a cyclic offset of -1, modulo N3 = 13, is applied to move the FD component from the strongest coefficient to component 0, reusing the same design principle as in Rel16 PS. The offset is cyclic because component 0 is shifted to -1, which corresponds to component -1 mod 13 = 12, while component 1 is shifted to component 0. For layer 2, the strongest coefficient is already in component 0; therefore, there is no need for an offset.After the offset in layer 1, one component moved outside the window configured by the gNB, as shown in one possible scenario for layer 1 (A). In one possible scenario for layer (B), both components are inside the window. Due to this difference, a UE needs to signal one of the two combinations to the gNB for each layer. Reusing flags defined in Rel-16 CB without modification is not possible in Rel-17. In fact, the indicator A15 (Common to the layer) defined in the windowing mechanism of the regular CBs and port selection Rel-16 is common to the layer and it is not possible to find a window position that is common to both layers in figure 7. The indicator (M^ specific to the layer) is also not suitable, because, if the window size N = 2 is replaced in the definition of i16il, a single code point for W¡ is obtained, as follows: ¿1,6, / e {o,1.....(^)-1) = (0}.(1) This does not allow distinguishing between the two combinations in (A) and (B). Therefore, by reusing the indicators for Wf defined in Rel-16 when Wf is limited within a window of size N, in the case of Figure 3 there are no reports of W^, but the gNB would not be able to distinguish between the two Wf for Layer 1 (A) and Layer 2 (B). The out-of-window problem illustrated in (A) of 3 could be avoided if the Rel-16 design principle that the strongest coefficient is in the first column of W2 is not followed. The component of the strongest coefficient could then be any of the Mv components. However, this approach would require cumbersome changes to the omission rules since the mapping of PMI uplink control information (UCI) in Part 2 of the CSI report would need to be changed.UCI omission is a mechanism for the UE to remove part of the PMI payload when the physical uplink shared channel (PUSCH) resource is insufficient to accommodate the full report. The omission rules for Type II Rel-16 CBs are designed to assign a higher priority value, Pri(l,i,f), to the first column of W2, so that W2 can be split into two parts assigned to different priority groups, making it less likely that the strongest coefficient for each layer will be omitted. If the strongest coefficient can be in any of the Mvcomponents, the probability of omitting the strongest coefficient increases, which can impact performance if an omission occurs. Examples of a reporting window for Wf and a measurement window for Wf are discussed in more detail below.The reporting window can be configured by a network entity, for example, by the gNB, and the UE can be configured accordingly. The size of the reporting window can be defined in different ways; examples of this are explained in detail below. Figure 4 shows a flowchart of an example for the operation of a device that receives and uses configuration information, for example, device 10 in Figure 1. The method provides a pre-coding matrix indicator for a channel status information report configured with a port selection codebook. This configuration information is used to configure a measurement window of a compression matrix from the port selection codebook. The configuration information defines the size of the measurement window, which is received in 100.The device can then select the number of components in the measurement window based on the configuration information in 102. A compression matrix reporting window is indicated in 104 based on a cyclical displacement of the measurement window, so that the strongest component in each layer is moved to a predefined component position. For example, the strongest component is moved to component index 0. Figure 7 shows a more detailed example of sending layer-specific indications of the initial position. Alternatively, the reporting window can be configured as an extension of the measurement window. An example of this is shown in Figure 8, where the extension is fixed without requiring a reporting window indication.In this case, the operation in block 104 of Figure 4 would involve using a configuration received from a reporting window that defines a measurement window extension such that the frequency domain component of the strongest coefficient in each layer is shifted to a predefined component position. For example, the strongest component is shifted to the component position at index 0. Figure 5 shows a flowchart of an example operation on a device provided in an access network, for example, access point 12 in Figure 1, to configure a device to report information related to multi-channel communications. In this method, the device sends configuration information to set up a measurement window for a compression matrix from a port selection codebook.The configuration information defines the number of coefficient components in the frequency domain of the measurement window. A channel status information report, including the precoding matrix indicator, is then received at 202. A report window is received from a port selection codebook based on a cyclical shift of the measurement window, such that the frequency domain component of the strongest coefficient in each layer is shifted to a predefined component, for example, the component with index 0. The device can then rebuild precoders based on the precoding matrix indicator at 204. Alternatively, instead of a cyclic shift, the report window is configured as an extension of the measurement window. An example of this is shown in Figure 8.More detailed examples of possible uses of grouped precoding are provided below. Figure 6 shows a signaling flow diagram according to an example between two communication devices, specifically between a UE 10 and a gNB 12, for configuring, measuring, and reporting the port selection codebook, where a specific window size is communicated to the UE. More specifically, the gNB 12 sends message 60 to the UE. This message includes a top-layer configuration with a window-side parameter N to restrict the selection of Mvvectors. In the example, the configuration comprises the element 'CodebookConfig-r17'. The UE can respond with a sound reference signal (SRS) 61, after which the gNB can determine a set of CSI-RS vector pairs in the spatial and frequency domains in 62.The gNB can exploit partial UL-DL reciprocity, for example, as explained earlier. Each CSI-RS port is then precoded in 63 via transmit (tx) antennas and frequency units with a pair of precode vectors. The precoded CSI-RS is communicated to the UE using message 64. The UE can now calculate, in 65, configured CSI report amounts, for example, in 'CSI-ReportConfig', including the PMI. The PMI reported by message 66 can include a per-layer index associated with the selected Mvvectors. The gNB can then combine, in 67, the received PMI with the CSI-RS precode vector pairs determined in 62 to obtain the precode for the downlink shared physical channel (PDSCH) and the demodulation reference signal (DMRS). The precoded data and the DMRS can then be signaled to the UE in 68.The following explains in more detail, by way of example, an enhanced codebook arrangement for Port Select Channel Status Information (PS CSI) signaling where UE reports a Precoding Matrix Indicator (PMI) in a Channel Status Information (CSI) report. This is configured with a Port Select codebook in which DFT-based Mvvectors are constrained to a configured measurement window of size N, and the strongest coefficient in each layer is expected to be located in the first column of the bitmap of reported non-zero W2 coefficients. The gNB can then reconstruct the precoders from the PMI. One or more PMI indices are associated with a reporting window for Wf, which can be different from the configured measurement window. This can be particularly relevant in the context of frequency domain (FD) compression operations.In certain applications, compression operations can be moved, at least partially or mostly, from the UE to the gNB. This improvement is based on the assumption of partial reciprocity of the array delays in the UL and DL channels and on flexibility in the use of frequency domain components. A first example is illustrated in Figure 7. Non-zero Mv-1 components, if reported, are indicated with respect to the reporting window. The reporting is based on a layer-specific reporting window of size N and initial point M^nicialle {-N + 1, -N + 2, ...,0}, for layer l = 1,v. One index can be reported per layer. The index can correspond to the cyclic shift required to move the strongest coefficient component to a predefined component, preferably 0. The measurement window configured by the gNB restricts the selection of the Mv vectors of Wfy, which are common to the layers.The report window can be a layer-specific, offset version of the measurement window common to all layers. The report window can be generated based on the configured measurement window by applying cyclic offsetting, such that the strongest coefficient component is moved to component 0, and then an index is reported to gNB indicating the offset for each layer. The Mvcomponents can be configured to be consecutive. The index associated with the report window can correspond to the strongest coefficient component within the configured measurement window.The UE can apply a cyclic shift of -index (“a cyclic shift to the left of the index”) modulo Mv to the columns of W2, before reporting, including the bitmap of the reported non-zero coefficients and the corresponding indices of the amplitudes and phases of these non-zero coefficients, so that the strongest coefficient is moved to the first column. The UE can be configured not to report Mv-1 non-zero components. A cyclic shift of a negative integer value, denoted above as -index, applied to a component x modulo y is a circular shift to the left, i.e., it subtracts the index from xy and is expressed by the operation: (x-index) mod y. Conversely, a cyclic shift of a positive integer value, index, is expressed by the operation: (x+index) mod y.A second example is based on a common report window for layers of size 2N - 1 and a fixed starting point M^initial = -N + 1, as shown in Figure 8. The report window can be obtained from the configured measurement window by extending the measurement window on one side with an additional TV - 1 consecutive components. As mentioned earlier, non-zero components, if reported, are indicated relative to the report window. The strongest component may already be in the position of the first component (component 0 in Figures 7 and 8). In such a case, the UE can selectively use a report window that corresponds to the measurement window. According to one example, a UE can be configured to determine if the strongest tap in any of the layers is on a component other than the first component and, in response, selectively apply changes to the report window from the measurement window.If necessary, the UE can selectively signal changes to the gNB. Figures 9, 10, and 11 show an example where the size and position of the measurement window are maintained. However, the UE takes a different action based on the stronger components, as will be explained in more detail later in its specification. In a more specific configuration, applicable, for example, to 3GPP Rel-17 port selection (PS), and where N = Mv = 2, the examples in Figures 7 and 8 allow one binary index per layer to be reported. In other possible configurations with N = Mv > 2 or N > Mv > 2 and the additional restriction that the Mv components must be consecutive, the example in Figure 7 can be more efficient in terms of signaling overhead compared to the second example.It should also be noted that, in these cases, the first example can also be described as reporting an index x, e (0,1, -1} for each layer l = 1, corresponding to the strongest coefficient component before cyclic shifting. This index, indexx, is equal to fj, and can therefore be reported as a column index of the strongest coefficient indicator. No further reporting of 14^ may be necessary if the components are set to be consecutive. A UE can apply a cyclic offset -index_x, = -fj to the columns of W2, including the bitmap of the reported nonzero coefficients and the arrays of amplitudes and phases of these nonzero coefficients, so that the strongest coefficient is moved to the first column. W2 is a Kr x Mv array of complex coefficients for each layer where only nonzero combination coefficients are reported. The components of W1, if reported, are reported with reference to the lowest index component, rather than the strongest coefficient component, so that the reported Mv-1 components are always guaranteed to be within the measurement window of size N configured by the gNB. In practice, the UE can apply a cyclic offset to the components of W1 equal to the index of the first component, -n^, instead of -n^p as would be the case in Rel-16.The gNB, when reconstructing the PMI, can apply an opposite cyclic shift, indexx, to the columns of the reconstructed W2 so that they correspond one-to-one to the components of Wf in ascending index order: the first column of W2, after the indexx, shift, corresponds to the component at index 0, the second column of Wf to the lowest index component among the reported Mr-1 components (i.e., the first reported component), and so on, until the last column of Wf corresponds to the last of the reported Mv-1 components. Equivalently, the gNB can obtain the components of Wf for layer l after applying the cyclic shift: (η^ / ι - n^dlcex,^mod N3. In this case, the gNB would apply the reconstructed W2 without cyclic displacements. The following describes further possible details for modifying the ¿15e ¿16i indices introduced in 3GPP Rel-16. In the first example, the Mvvectors of Wf, for the layer l = 1, can be identified by MRnicial ;y n3il M^nicialile(-N + lN + 2.....0}(2)n3,l — [n3,( । >n3,¡] where M¡nicial l is indicated by an index i1;5J with a value set at {04¿v-1} and n3; is indicated by a combinatorial index i1ÁÍ with a value set at ( (Nl\)¿^e¡o,i..... (3) (4) (5) This is illustrated in Figure 7. The mapping of M^mciaua¿i,5, / and the mapping of the non-zero indices in the report window for the i16 layer can reuse the description in Section 5.2.2.2.5 of TS 38.214 Re-16 with the following modifications: Miniciai is replaced by MRniciall from (2) and 2MV is replaced by N. More precisely: M^niciallseindica por ¿15 i, que se informe y se protege por ¿l,5,í — mr11inicial,l J^hiicial.l + ​​N Initial M, l 0 Initial,l θ (6) Only the non-zero indices and IntS are reported, where IntS = { (^MRinitial l+ ¿)modN3,i = 0.1, ...,N - 1}, where the indices f = 1, ...,Μυ- 1 are assigned so that n^ increases as f increases. See n^ = nínn3,l - (N3- N) n[fí<MRnicialil+ N-lηΪ,ί > initial,l + N~ 1 (7) then1C ( / V -1 - where C(x,y) is provided in Table 5.2.2.2.5 of 3GPP TS 38.214 Rel-16. To operate according to figure 7, the UE can be configured to apply a cyclic offset to the components of W¡, module N3: mod N3. The UE can apply a cyclic shift of -f to the columns of W2, modulo Mv: f = (f - f)modMv. The UE can report for f = 1, -1, after the shift to Γ. / N - 1usar l,,s' U, -1 bits. The UE can also report -M^nicial le {0,1, ...,N - 1}, for layer l by using [log2L / j bits. The UE reports the position of the strongest coefficient of layer l by i*¡, by using i) e {0,1, ...,Kr— l}[log2 / G1 bits. In the second example, the Mvvectors of Wf, for layer l = 1,..., v, are identified n3i¡de (3), indicated by a combinatorial index i16J with a set value The mapping of non-zero indices in the reports window for layer 16 can reuse the description in Sec. 5.2.2.2.5 of 3GPP TS 38.214 Rel-16 with the following modifications: Minitial is replaced by Minitial = N + 1 for all layers and 2Minitial is replaced by 2N - 1. More precisely: Only nonzero indices and IntS are reported, where IntS = { (—(V + 1 + i)modN3,i = 0,1, ...,N - 1}, where the indices f = 1, - 1 are assigned so that n^, increases as f increases. See .n3fí - -2N+ 1) Afí > o' (9) then i1;6;í= ~2-np,Mv- f}, where C(x,y) is provided in Table 5.2.2.2.5-4 of 3GPP TS 38.214 Rel-16. To operate according to Figure 8, the UE can be configured to apply a cyclic offset of -n^) to the components of Wf, modulo N3: = (n^ - n^[ mod N3. The UE can apply a cyclic shift of to the columns of W2, modulo Mv: f = (f - Mv). The UE reports for f = 1, after the shift by using (2N — 2\] log2 M_ i) bits, and the UE reports the position of the strongest coefficient of layer l by means of i,, with i, e {0,1, - 1} by using [log2j bits For a special case of N = Mv = 2, both of the previous examples require only 1 bit of signaling per layer. In particular, for the first example, only {0,1} is reported, and i1,β,i = 0 is not reported. On the other hand, for the second example, only i1;6i = {0,1} is reported. In the case N = Mv > 2, however, the first example requires fewer bits than the second example. In particular, for the first example, only {0, 1, -1} is reported, and i1; 6; i = 0 is not reported. Therefore, [log2Mv] bits are needed per layer. On the other hand, the second example quickly becomes inefficient since it requires 3 bits for N = Mv = 3, 5 bits for N = Mv = 4, etc., while the first example requires 2 bits in both cases. Furthermore, if the gNB configures the selected Mvcomponents to be consecutive within a measurement window of size N > Mv, the first example can still be used by specifying the same signaling as for the case N = Mv, i.e., by reporting only i15>ze {0,1, -1}. Such a configuration will have a positive effect on reducing UL overhead since the UE does not need to return any additional information about Wf apart from this specific layer index. This configuration refers to the Mvcomponents of Wf that are configured to be consecutive within the window of size N. Figures 9, 10, and 11 illustrate an aspect where no changes are made to the size and position of a report window relative to the measurement window configuration. Instead, the strongest component reported can be indicated in a way that places it in a different position than the first column. In the flowchart in Figure 9, a device receives configuration information at 300 to configure a measurement window for a compression array from the port selection codebook, with the configuration information defining the size of the measurement window. The device then selects the number of coefficient components in the measurement window at 302 based on the configuration information. When reporting, the device indicates at 304 the row and column index of the position of the strongest coefficient components in a non-zero combination coefficient array for each layer and applies a cyclic shift to the columns of the non-zero combination coefficient array for each layer so that the strongest coefficient is shifted to the first column. The flowchart in Figure 10 relates to the operation of a network device receiving the report. The device sends configuration information on port 400 to configure a measurement window for a compression matrix from a port selection codebook, defining the number of components in the measurement window. The CSI report, which includes the PMI, is then received on port 402. The report specifies a row and column index for the position of the strongest coefficient in a non-zero combination coefficient matrix for each layer, and a cyclic shift is applied to the columns of the non-zero combination coefficient matrix for each layer so that the first column shifts to the specified column index. The UE can report Wf with respect to the first component instead of the strongest coefficient component. A layer-specific M¡n¡ciai is not required because there are no components outside the window. The UE can still apply the offset - / / with respect to the strongest coefficient to W2 and then report the value of this offset as part of the strongest coefficient indicator. Wf is not aligned with W2, but since gNB knows the f¡*, it can correct the offset in Wf (see Figures 12-14) or undo the offset in W2 (see Figures 15-17). Either way, gNB can align Wf and W2 due to the additional reporting of the strongest coefficient column index. The layer-specific Wf can be reported for each layer and can comprise Mv-1 components (component 0 is always present and does not need to be reported). The UE can be configured to apply a cyclic offset of -n^ to the components of Wf (no N3 modulo needed): -n^). The UE can apply a cyclic offset of -f!* to the columns of W2, modulo Mv: f = ( / - y / Jmod Mv). The UE reports for f = 1, - 1, after the shift when using i / N-13og2\MV- i) bits. The UE reports the position of the strongest coefficient of layer l by [i / , f¡*] with i / e {0,1, ...,KA- 1} and / / e {0,1, - 1} when using [log2(XiAfr)] bits (or [log2^il + {log2 bits). Figures 12–17 show index mappings for N values ​​2, 3, and 4. Alignment of the W2 and Wf indices can be achieved by shifting W2 indices (examples in Figures 15–17) or by shifting Wf indices (examples in Figures 12–14). The flowchart in Figure 10 relates to shifting gNB in ​​W2. As noted earlier, in the two previous examples, N = Mv = 2 can also be described more simply by reporting an index x, {0, 1, -1} for each layer l = 1, corresponding to the strongest coefficient component. Another possible description is to use tables to map the index value (either i15 or i16) to the index of the FD component n, as shown in Figure 15, which shows the mapping of the index report Wfal to the component index for N = Mv = 2. It can be observed that = 0 by construction denotes component 0. Therefore, the table shows the value of component 1, obtained in the gNB by taking the non-zero value after applying the cyclic shift: (n^ ^index x,·)^mod - (j _ index x,) mod N3, for n^, {0, 1}. In other words, the first row of the table corresponds to layer 2 (B) of figure 7, while the second row corresponds to layer 1 (A) of figure 7.N3 denotes the number of PMI sub-bands and is equal to 13 in Figure 7. Figure 16 shows an example of mapping the Wfal index report component index [η^,η^] for N = Mv= 3. Figure 17 shows an example of mapping the Wfal index report component index [r / 1) n(2) n(3)l nara N _ M — 4. n31, n31, n3। pdrd in — iviv — 4*. For N = Mv = 3 and N = Mv = 4, Figure 16 and Figure 17, respectively, show the values ​​of the Mv components obtained in the gNB after the displacement: ( / - mdicex,) mod N3, for f E {0,1, ...Mv} and when reordering in ascending order. A communication device, for example, a user device, may have the means to report a PMI in a CSI report configured with a port selection codebook in which the DFT-based Mvvectors of W are restricted to a configured measurement window of size N, and the strongest coefficient in each layer is expected to be located in the first column of the bitmap of non-zero coefficients reported from W2. A network device, for example, a gNB, may have the means to reconstruct the precoders from the PMI. One or more PMI indices may be associated with a reporting window for 14^, which may be different from the configured measurement window. The reporting window may also be common to all layers.The Mv components can be configured to be consecutive, and the index associated with the reporting window corresponds to the strongest coefficient component within the configured measurement window. The UE can apply a cyclic -index shift to the W2 columns before reporting, including the bitmap of reported non-zero coefficients and the arrays of the amplitudes and phases of these non-zero coefficients, so that the strongest coefficient moves to the first column. In this case, non-zero Mv-1 components may not be reported. Let's look at another possible solution to the out-of-window problem with respect to Figures 18-20. In practice, in Rel-16, the layer-specific Wf components are shifted accordingly with a modulo- / V3 operation. In the ongoing work on Rel-17 to specify the PS FeType II codebook, it was agreed that the Mv components (also known as frequency domain or FD bases) are limited within a single window of size N and starting point Minic = 0. However, in Rel-17, due to the window / set size restriction N configured by the gNB, the SCI must be modified by adding information related to the FD component of the strongest coefficient, i.e., or / }*. The need for this modification can be illustrated, for example, in the case N = Mv = 2, for which Wf = [0,1] is common to the layers and is not reported. The measurement window for 1 / 14 restricts the maximum separation between the selected components. Note that the codebook indices associated with the vector components of the compression matrix Wf are given by ηψ* G {0,1, ...,N - 1}, with f = 0,1, — 1, and are indexed by f, which indicates the corresponding column of the non-zero combination coefficient matrix, W2. In this description, we also refer to n3n as FD (frequency domain) component indices and f as the column index of the non-zero combination coefficient matrix. Figure 3 illustrates an example of this window configuration, in which the strongest coefficient of layers 1 and 2 is in two different components. Because Wf is common to the layers and is not reported, the first component, n® = 0, may not correspond to the strongest coefficient component for some layer. Therefore, both [i / *, / / ] coordinates of the strongest coefficient must be reported for each layer l in order to locate the strongest coefficient, where i¡ G {0,1, ...,Kr-1} indicates the row of the non-zero combination coefficient matrix (corresponding to one of the selected CSI-RS ports) and f¡* e {0,1,..., Mv- 1} indicates the column of the non-zero combination coefficient matrix.Note that in this case, when N = Mv, the index of the FD component of the strongest coefficient, n3fl\, and the column index of the strongest coefficient in W2, take the same values, i.e., n / 1. Therefore, / / can be reported separately with [log2 IV] = [log2Mv] bits, or jointly with i*¡, by using [log2( / <1Mv)l bits. In the current agreement regarding the relationship between N and Mv, there is also the possibility that N > Mv, in which case Wf requires reporting and can be common to the layers or layer-specific. To minimize overhead and following the Rel-16 design, we assume that in this case Wf is reported using a combinatorial coefficient by using (N - 1 YOg2\MV-1) bits, so the 0 component is always included in Wf and is not reported. Because only Mv-1 components of FD are reported, it is necessary to specify a reference component, which determines the modulo- / V shift operation applied to the Wf components before reporting. For the port selection codebook Rel-17, to report the Mv components of Wf when using bits, it is proposed to reassign the components with respect to the component with the strongest coefficient using a modulo-N operation and report only the non-zero Mv1 components after the reassignment. If no reassignment is applied, log2 N Mvbits to report all Mvcomponents, which would generate a greater feedback overhead. In one mode, the FD bases are shifted with respect to the strongest coefficient basis, n^, so that nf* = -n^(3f)^(mod N), after the shift. This operation is similar to what is done in Rel-16, except for the modulo-N operation, which in Rel-16 is modulo-V3, because there is no gNB-defined measurement window for it in Rel-16. The formula also assumes that the FD bases nf* (i.e., the codebook indices of the vector components of the compression matrix W!) are common to the layers; that is, a single set is reported for all layers. However, the operation is also applicable in the layer-specific case of W! In the gNB, using the reported FD bases in the precoder reconstruction would not be correct, if / / > 0, because of the modulo-N operation that reassigns the FD bases to the left of , i.e., nf\inside the window, like n3Mv~fl\... ,n^ív~1->. This (iϊ modulo-N) operation can be undone in the gNB by reassigning with respect to n3, so that n3 = (íf) (f*} n3 + n3j mod N, after reassignment. However, this would require reporting n3¡e (0,1,... ,N - 1} when using [log2AI] bits. Instead, the proposal allows for proper deallocation by reporting / / e {0,1, ...,MV- 1}, which is one of the two SCI indices, [¿ / / / ]. In the proposal, the gNB reassigns / deallocates the reported FD bases with respect to n3Mv~fi> as = (n3^ - n3Mv~fl^mod N, so that = 0 after deallocation. The proposed solution is an alternative to reassigning the Mvcomponents with respect to the first base, n^, so that = {n'p — n®), after the shift. This alternative does not require a modulo-N operation. For the proposed solution, the result of the deallocation in the gNB is the same as if the reassignment in the UE were performed with respect to the first base. In Rel-16, where Wf is layer-specific, the reference component is that of the strongest coefficient for a layer l, so the component indices are reassigned with respect to n3l} as - n^^mod N3. This choice of offset for Wf in Rel-16 allows for reduced overhead in both the Wf report and the SCI report because only the first coordinate of the pair [i / / / ] that identifies the position of the strongest coefficient for layer l needs to be reported. For this latter overhead saving to be possible in Rel-16, a second offset must be applied to the column indices of IV2, which are reassigned with respect to f¡* as f = ( / - f¡*) mod Mv, so that the columns of W2 correspond to those of Wf after the offset. In Rel-17, however, it is necessary to report both coordinates of the SCI even if Wf is reported (for N > Mv) and the strongest coefficient is shifted to = 0. Therefore, if Wf is reported, a different offset from Rel-16 can be specified for the Wfy components, and the offset may not be applied to the W72 columns. There are two possible components that can be used as a reference for Wf: the component with the strongest coefficient, as in Rel-16, or the 'first' selected component, that is, the component with the lowest index value in the window: {0,1 ...,N - 1}. Depending on whether Wf is common to layers or specific to layers, the following mapping options are available for Wf. 1. Wf = [n¡0),..., common to the layers a. The reference is the strongest coefficient component for layer 1, n^\ The FD components are reassigned with respect to as n^ = (n^~ (f*}\ (f*\ nf J mod N, so that 1= 0 after reassignment. b. The reference is the 'first' component, n^. The FD components are reassigned with respect to an¡0) as = (n^ - n¡0)), so that n¡0)= 0 after reassignment. 2. Wf = [n^,..., layer specific, for l = 1,..., v a. The reference is the strongest coefficient component for the layer / , . The FD components are reassigned with respect to as = (n^ — n^'f^mod N, so that = 0 after reassignment. b. The reference is the 'first' component, n^. The FD components are reassigned with respect to an® as = (n^ - n®), so that n® = 0 after reassignment. If no offset is specified for 14γ, a UE can understand that Wf = [0,n^\ ...,n3Mv-1)], in the common case to layers, for example, i.e., it can assume that the first component, = 0 is fixed and only select and report Mv- 1 non-zero components in the window of size N. Regarding the shift applied to the IF2 columns, in practice there are two possible options: do not shift at all or shift the index f with respect to f = (f - mod Mv, so that the index of the strongest coefficient is f¡* = 0 after reassignment. For all the above alternatives for mapping the indexes of Wf and W2, reporting the pair [i / , / / ] as the SCI for layer l is sufficient to locate the strongest coefficient of W2 and ensure the correct reconstruction of Wf. Reporting the values ​​as an alternative also works, but requires more feedback bits if N > Mv because e ⊆ {0,1, ...,N - 1}, while fi ⊆ {0,1,.....Mv- 1}. Let us consider an example for cases 1.a and 1.b above, with N = 4, Mv = 2, and Wf = [1, 2] common to the layers. Suppose that, for layer 1, IV2 = [^] and the strongest coefficient is c at position [? / , / TJ = [0, 1]; for layer 2, IV2 = [^] and the strongest coefficient is f at position [i2, / 2*] = [1, 0]. This example is illustrated in Figure 18. A star subscript denotes the strongest component. On the left, the UE and gNB operations are shown in case 1.a, where the strongest coefficient for each layer is shifted to the 0 component of FD and the shift applied to Wf follows that of layer 1. Case 1.b is shown on the right, where the strongest coefficient is not shifted and the shift applied to Wf follows that of the selected component of the lowest index. íf*') ÍH In case 1.a, the components of Wf are reassigned with respect to n = 2 as n = -2^mod 4, f = 0,1, so that Wf = [0,3] is the reported FD basis set. The index f is reassigned with respect to n = 1 as f = (n - 1) mod 2, for layer 1, and with respect to n = 0, i.e., no shift is applied, for layer 2. The reported combining coefficient matrices are W2 = n = n, ... In gNB, using Wf=[0,3] in the precoder reconstruction would not be correct because the selected components on the left, den^1-1, i.e., have been reassigned within the window by the modulo operation N, such as ...,n3Mv~^·. Therefore, gNB needs to reassign the Mv components of Wf with respect to , if / j* > 0, such that η^ν~^ = G after unassignment. If / / = 0, no reassignment is necessary because the modulo operation did not reassign any components of Wf. An equivalent way of describing deallocation in the gNB is by reassigning the components , , í(Mv-f j)mod Mv) (f) í (f) (CMv-fflmodMjX , .. , , of Wf with respect to n3, as n3= \n3-n3J)modN, so that n((Mv / jmodMp) _ θ, |disassignment. In our example, Wf is reassigned with respect to = 3 as - 3) mod 4, f = 0,1. The gNB also deals off f as f = (f - (Mv- f,*)) mod Mv= (f + f,*) mod Mv, for l = 1,2 so that 14^ = [0,1] and W2= [bd]'[fh] are used in the reconstruction^θI precoder.It is observed that the UE reassigns = [1,2] to [^3^,7131)] = [3,0], but when reporting these components, the f indices are assigned in such a way that n3 increases with f, so the gNB receives [ti®, n^] = [0,3].. This is shown in figure 18. In case 1.b, the components of Wf are reassigned with respect to n = 1 as (n^-1 - 1), f = 0,1, so that Wf = [0,1] is the reported FD basis set. No mapping is applied to the index f; therefore, the combination coefficient matrices 14 / ? are reported without shifting the strongest coefficient to the FD component 0, as H dl'[f S. In the gNB, no deallocation is required, and the quantities reported for Wf and W2 are used in the precoder reconstruction. This example illustrates that, regardless of the solution adopted for the representation of Wfy M / 2, the SCI can be reported as the index pair [í;*, / / ] for l = 1, Table 1 and Table 2 summarize the mapping options analyzed above for the 14^ components and the W2 column index in the case of Wf common to layers and layer-specific. Note that the mapping and deallocation of 14^ only applies when N > Mv, while for N = Mv, 14^ is common to layers and is set via configuration. EU N > Mv(Wf informed) N>MV choice of displacement for Wf and W2 map of n^ / n^ (W^) map of f (W2) SCI / Wf layer common to the layers The strongest coefficient shifts to H = o = (n^ — mod N f = (f - fr) mod Mv [ii.fi] The strongest coefficient >4° = (n / > - n<»>) - [ii.fi] The strongest coefficient does not shift. Wf specific to the layer. The strongest coefficient shifts to / ; = on( / ) 3,1 — )) mod N f = ( / - / / ) mod Mv [iiJi] The strongest coefficient does not shift ηω = íncn _ n(o)\ 3,1 V L3,l -3,1 J - lílJl] Table 1. Summary of UE mapping options for Wfy components with column index 14 / . In all cases, the SCI can be reported as [ / ,, / / ] for layer I. gNB N > Mv (Wf informed) N>MV shift choice for Wf and W2 Deallocation of n^ / n^ (Wf) Deallocation of f(W2) Wf common to the layers The strongest coefficient shifts to f! = 0 , f> (n3^ ~ n3Mv ))mod N, > 0 / / = 0 / = ( / + / / ) mod Mv The strongest coefficient does not shift - - Wf specific to the layer The strongest coefficient shifts to / , = 0 (7) ί(η3^ — h 01710 N, fi > 0 n31 = ) ml fl = 0 / = ( / + / / ) mod Mv The strongest coefficient does not shift - - Table 2. Summary of gNB deallocation options for Wf components and W2 column index. In all cases, the SCI can be reported as [i / / / ] for the / layer. Figure 19 shows a method for the reporting device, such as for a user device, that follows the principles described above, for example, in Figure 18. The method can be for providing a report of channel status information configured with a port selection codebook. The method can comprise, in step 1900, receiving configuration information to configure a measurement window of a compression matrix from the port selection codebook, where the configuration information defines the size of the measurement window. In one mode, the size of the measurement window N is greater than the number Mv of codebook indices that the device is allowed to select to define the compression matrix Wf. In step 1902, the UE selects a number of codebook indices from the measurement window based on the configuration information to form the port selection codebook W¡ compression matrix, where the codebook indices are associated with vector components of the compression matrix. In step 1904, the UE indicates the row and column index of the position of the strongest coefficient in a W2 matrix of nonzero combining coefficients for the layers. This step may involve reassigning the columns of the W2 matrix of nonzero combining coefficients for the layers so that the strongest coefficient is reassigned to the first column. In step 1906, the UE reassigns the codebook indices, so that the codebook index corresponding to the column index of the first layer's predefined coefficient is reassigned to the first index of the measurement window. This reassignment can also be called cyclic shifting. In one mode, the predefined coefficient is the strongest coefficient. In another mode, the predefined coefficient is the first coefficient. In another mode, this step may involve reporting all codebook indices of the port selection codebook compression array, except the first, after subtracting the codebook index corresponding to the column index of the first layer's strongest coefficient from all codebook indices of the compression array and applying a modulo operation of the measurement window size Λ / .In one mode, the measurement window codebook indices and compression matrix are DFT vectors. Figure 20 shows a method for the device to receive the report, such as a gNB, following the principles described above, for example, in Figure 18. In step 2000, the gNB sends configuration information to configure a measurement window of a compression array Wf from a port selection codebook, where the configuration information defines the number N of components in the measurement window. In one mode, the size of the measurement window N is greater than the number Mv of codebook indices that the UE is allowed to select to define the compression array Wf. In step 2002, the gNB receives a channel status information report that includes a precoding matrix indicator (PM), where row and column indices are indicated for the positions of the strongest coefficients in layers of an I44 matrix of non-zero blending coefficients, and the PMI comprises codebook indices of an I44 compression matrix. In one mode, these indices have been reassigned on the transmitter side as described above, e.g., with the strongest coefficient shifting to the first index. In step 2004, gNB deals the codebook indices of compression array 144 with respect to one of the codebook indices whose position depends on the column index of the first-layer predefined coefficient and the number of components in the compression array. In one mode, the predefined coefficient is the strongest coefficient. This occurs when the column index of the first-layer strongest coefficient is found to be non-zero. This step 2004 may involve understanding that the deallocation of codebook indices from the compression array comprises subtracting one of the codebook indices from all the codebook indices of the compression array 144 and applying a modulo operation of size N of the measurement window. In one modality, the function that identifies the position of a codebook index is obtained by subtracting the non-zero column index of the strongest coefficient of the first layer from the number Mv of components of the compression array Wf. In one mode, the non-zero combination coefficient matrix 144 and the codebook compression matrix 144 are aligned based on the indicated column indices of the strongest coefficients. In step 2006, the gNB builds pre-coders based on the received channel status information report. Although in some instances the description has assumed that Wfe is common to the layers, the solutions are also applicable to layer-specific Wfe. In one modality, the port selection codebook compression matrix for each layer is reassigned with respect to the corresponding codebook index to the column index of the strongest coefficient of the respective layer. It should be noted that, while the foregoing describes illustrative modalities, several variations and modifications can be made to the described solution without departing from the scope of the present invention. Different features from different modalities can be combined. Therefore, the embodiments may vary within the scope of the appended claims. In general, some embodiments may be implemented in special-purpose hardware or circuitry, software, logic, or any combination thereof. For example, some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software that can be executed by a controller, microprocessor, or other computing device, although the embodiments are not limited to this.While various modalities may be illustrated and described as block diagrams, flowcharts, or using some other pictorial representation, it is understood that these blocks, devices, systems, techniques, or methods described herein may be implemented in, as non-exhaustive examples, hardware, software, firmware, special-purpose circuits or logic, general-purpose hardware or controller, or other computer devices, or some combination thereof. The procedures can be implemented using computer software stored in memory and executable by at least one data processor of the entities involved, or by hardware, or by a combination of software and hardware. Furthermore, it should be noted that any of the above procedures may represent program steps, or interconnected logic circuits, blocks, and functions, or a combination of program steps and logic circuits, blocks, and functions. The software may be stored on physical media such as memory chips or memory blocks implemented within the processor, magnetic media such as hard drives or floppy disks, and optical media such as, for example, DVDs and their data variants, CDs. Memory can be of any type suitable for the local technical environment and can be implemented using any appropriate data storage technology, such as semiconductor-based memory devices, magnetic memory devices and systems, optical memory devices and systems, fixed memory, and removable memory. Data processors can be of any type suitable for the local technical environment and may include one or more general-purpose computers, special-purpose computers, microprocessors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), gate-level circuits, and processors based on multi-core processor architectures, among other examples. Alternatively or additionally, some methods can be implemented using circuitry. The circuitry can be configured to perform one or more of the functions and / or procedures of the method described above. This circuitry can be provided in the network entity and / or the communications device and / or a server and / or a device. As used in this application, the term "circuitry" may refer to one or more of the following: (a) hardware-only circuit implementations (such as analog-only and / or digital-only circuit implementations); (b) combinations of hardware and software circuits, such as: (i) a combination of analog and / or digital hardware circuitry with software / firmware and (ii) any part of the hardware processor(s) with software (including digital signal processors), the software and memories that work together to enable the communications device and / or device and / or server and / or network entity to perform the various functions described above; and (c) hardware circuit(s) and / or processor(s), such as a microprocessor(s) or a part of one or more microprocessors, that require software (e.g., firmware) for operation, but the software may not be present when it is not required for operation. This definition of circuitry applies to all uses of this term in this application, including any claim. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or a portion of a hardware circuit or processor and its accompanying software and / or firmware. The term circuitry also covers, for example, an integrated device. It should be noted that although modalities have been described in relation to certain architectures, similar principles can be applied to other systems. Therefore, although certain modalities were described above by way of example with reference to certain example architectures for wireless networks, technologies, standards, and protocols, the features described herein can be applied to any other suitable form of systems, architectures, and devices other than those illustrated and described in detail in the preceding examples. It is also noted that different combinations of different modalities are possible. It is also indicated herein that, while the above describes example modalities, various variations and modifications can be made to the described solution without departing from the spirit and scope of the present invention.

Claims

1. An apparatus comprising: at least one processor and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured, with the at least one processor, to cause the apparatus to: receive configuration information to configure a measurement window to form a compression matrix of a port selection codebook from a set of vector component codebooks, wherein the configuration information defines the size of the measurement window, which is common to all of the at least one layer to be reported;Select a number of measurement window indices based on configuration information to form the compression matrix from the set of vector component codebooks, where the selected indices are associated with vector components of the compression matrix and are common to all of the at least one layer to be reported; and reassign the selected indices associated with the vector components with respect to a reference vector component index, such that the reference vector component index is reassigned to a first measurement window index; report channel status information that includes a precoding matrix indicator to a network, the precoding matrix indicator comprising compression matrix information after reassignment.

2. The apparatus of claim 1, wherein the configuration information further configures the number of vector components of the compression matrix.

3. The apparatus of claim 2, wherein the number of vector components of the compression matrix is ​​less than the number of indices in the measurement window.

4. The apparatus of any of claims 1 to 3, wherein the index of the reference vector component is the lowest of the selected indices and the first index of the measuring window is 0.

5. The apparatus of any of claims 1 to 4, wherein the reassignment comprises subtracting the index of the reference component from all the compression matrix indices.

6. The apparatus of any of claims 1 to 5, wherein the precoding matrix indicator indicates all compression matrix indices except the first one, after reassignment.

7. The apparatus of any of claims 1 to 6, wherein the at least one memory and the computer program code are configured, with the at least one processor, to cause the apparatus to: determine a position of the strongest coefficient in a non-zero combination coefficient array for each of the at least one informed layer; and indicate the row and column index of the position of the strongest coefficient in the non-zero combination coefficient array for each of the at least one informed layer.

8. The apparatus of any of claims 1 to 7, wherein the vector components of the codebook set are vector components of DFT.

9. The apparatus of any of claims 1 to 8, wherein the first index of the measuring window coincides with the first index of the set of vector component codebooks.

10. The apparatus of any of claims 1 to 8, wherein the first index of the measuring window is offset from the first index of the set of vector component codebooks.

11. The apparatus of any of claims 1 to 10, wherein the index of the reference vector component is offset with respect to the first index of the measuring window.

12. The apparatus of any of claims 1 to 11, wherein the index of the reference vector component is offset with respect to an index of the strongest coefficient of the measuring window.

13. The apparatus of any of claims 1 to 12, wherein the at least one layer to be reported comprises a plurality of layers.

14. An apparatus comprising: at least one processor and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured, with the at least one processor, to cause the apparatus to: send configuration information to configure a measurement window to form a compression matrix of a port selection codebook from a set of vector component codebooks, wherein the configuration information defines the number of measurement window components, which is common to all of the at least one layer to be reported;receive a channel status information report that includes a precoding matrix indicator comprising information from a compression matrix of a port selection codebook after reassignment, wherein, on reassignment, selected measurement window indices have been reassigned with respect to the index of a reference vector component, such that the index of the reference vector component has been reassigned to a first measurement window index, wherein the selected indices are associated with vector components of the compression matrix and are common to all of the at least one reported layer; and rebuild precoders based on the precoding matrix indicator.

15. The apparatus of claim 14, wherein the configuration information further configures a number of vector components of the compression matrix.

16. The apparatus of any of claims 14 to 15, wherein the precoding matrix indicator further comprises row and column index information indicating the position of the strongest coefficient in a non-zero combination coefficient matrix for each of the at least one informed layer.

17. A method comprising: receiving configuration information to configure a measurement window to form a compression matrix of a port selection codebook from a set of vector component codebooks, wherein the configuration information defines the size of the measurement window, which is common to all of the at least one layer to be reported; selecting a number of indices of the measurement window based on the configuration information to form the compression matrix from the set of vector component codebooks, wherein the selected indices are associated with vector components of the compression matrix and are common to all of the at least one layer to be reported;and reassign the selected indices associated with the vector components with respect to a reference vector component index, such that the reference vector component index is reassigned to a first index of the measurement window; report channel status information including a precoding matrix indicator to a network, the precoding matrix indicator comprising compression matrix information after reassignment.

18. The method of claim 17, wherein the configuration information further configures the number of vector components of the compression matrix.

19. The method of claim 18, wherein the number of vector components of the compression matrix is ​​less than the number of indices in the measurement window.

20. The method of any of claims 17 to 19, wherein the index of the reference vector component is the lowest of the selected indices and the first index of the measurement window is 0.

21. The method of any of claims 17 to 20, wherein the reassignment comprises subtracting the index of the reference component from all the compression matrix indices.

22. The method of any of claims 17 to 21, wherein the precoding matrix indicator indicates all compression matrix indices except the first one, after reassignment.

23. The method of any of claims 17 to 22, further comprising: determining a position of the strongest coefficient in a non-zero combination coefficient array for each of the at least one informed layer; and indicating the row and column index of the position of the strongest coefficient in the non-zero combination coefficient array for each of the at least one informed layer.

24. The method of any of claims 17 to 23, wherein the vector components of the codebook set are vector components of DFT.

25. The method of any of claims 17 to 24, wherein the first index of the measurement window coincides with the first index of the set of vector component codebooks.

26. The method of any of claims 17 to 24, wherein the first index of the measurement window is offset from the first index of the set of vector component codebooks.

27. The method of any of claims 17 to 26, wherein the index of the reference vector component is offset with respect to the first index of the measurement window.

28. The method of any of claims 17 to 27, wherein the index of the reference vector component is offset with respect to an index of the strongest coefficient of the measurement window.

29. The method of any of claims 17 to 28, wherein the at least one layer to be reported comprises a plurality of layers.

30. A method comprising: sending configuration information to configure a measurement window to form a compression matrix of a port selection codebook from a set of vector component codebooks, wherein the configuration information defines the number of measurement window components, which is common to all of the at least one layer to be reported;to receive a channel status information report that includes a precoding matrix indicator comprising information from a compression matrix of a port selection codebook after reassignment, wherein, in the reassignment, the 34 selected measurement window indices have been reassigned with respect to the index of a reference vector component, such that the index of the reference vector component has been reassigned with respect to a first measurement window index, wherein the selected indices are associated with vector components of the compression matrix and are common to all of the at least one reported layer; and to rebuild precoders based on the precoding matrix indicator.

31. The method of claim 30, wherein the configuration information further configures a number of vector components of the compression matrix.

32. The method of any of claims 30 to 31, wherein the precoding matrix indicator further comprises row and column index information indicating the position of the strongest coefficient in a non-zero combination coefficient matrix for each of the at least one informed layer.

33. A computer program product incorporated in a computer-readable distribution medium and comprising program instructions that, when loaded into an apparatus, execute the method according to any of claims 17 to 32.

34. A computer program product comprising program instructions that, when loaded into an apparatus, execute the method according to any of claims 17 to 32.

35. An apparatus comprising means for carrying out the method according to any of claims 17 to 32.

Images