TECHNICAL FIELD
[0001] The present invention relates to the field of wireless communications, and in particular to methods, user equipments and computer program products for providing channel state information, CSI, feedback from a user equipment in the form of one or more CSI reports in a wireless communication system.
BACKGROUND
[0002] In a wireless communications system, such as New Radio, also called 3GPP Fifth Generation wireless communications system or 5G for short, downlink (DL) and uplink (UL) signals convey data signals, control signals comprising DL control information (DCI) and/or uplink control information (UCI), and a number of reference signals (RSs) used for different purposes. A radio network node or a radio base station or a gNodeB (or gNB or gNB/TRP (Transmit Reception Point)) transmits data and DCI through the so-called physical downlink shared channel (PDSCH) and the physical downlink control channel (PDCCH), respectively.
[0003] A UE transmits data and UCI through the so-called physical uplink shared channel (PUSCH) and physical uplink control channel (PUCCH), respectively. Moreover, the DL or UL signal(s) of the gNB respectively the user equipment (UE or a radio device) may contain one or multiple types of RSs including a channel state information RS (CSI-RS), a demodulation RS (DM-RS), and a sounding RS (SRS). The CSI-RS (SRS) is transmitted over a DL (UL) system bandwidth part and used at the UE (gNB) for CSI acquisition. The DM-RS is transmitted only in a bandwidth part of the respective PDSCH/PUSCH and used by the UE/gNB for data demodulation.
[0004] One of many key feature of 5G is the use of multi-input multi-output (MIMO) transmission schemes to achieve high system throughput compared to previous generations of mobile systems. MIMO transmission generally demands the availability of accurate CSI used at the gNB for a signal precoding using a precoding matrix of the data and control information. The current third Generation Partnership Project Release 15 specification (3GPP Rel. 15) therefore provides a comprehensive framework for CSI reporting. The CSI is acquired in a first step at the UE based on received CSI-RS signals transmitted by the gNB. The UE determines in a second step based on the estimated channel matrix a precoding matrix from a predefined set of matrices called 'codebook'. The selected precoding matrix is reported in a third step in the form of a precoding matrix identifier (PMI) and rank identifier (RI) to the gNB.
3GPP Rel.-15 dual-stage precoding and CSI reporting
[0005] In the current Rel.-15 NR specification, there exist two types (Type-I and Type-II) for CSI reporting, where both types rely on a dual-stage (i.e., two components)
W_{1}W_{2} codebook. The first component, or the so-called first stage precoder,
W_{1}, is used to select a number of beam vectors and (if configured) the rotation oversampling factors from a Discrete Fourier Transform-based (DFT-based) matrix which is also called the spatial codebook. The spatial codebook comprises an DFT- or oversampled DFT matrix of dimension
N_{1}N_{2} ×
N_{1}O_{1}N_{2}O_{2}, where
O_{1} and
O_{2} denote the oversampling factors with respect to the first and second dimension of the codebook, respectively. The DFT vectors in the codebook are grouped into (
q_{1},
q_{2}), 0 ≤
q_{1} ≤
O_{1} - 1, 0 ≤
q_{2} ≤
O_{2} - 1 subgroups, where each subgroup contains
N_{1}N_{2} DFT vectors, and the parameters
q_{1} and
q_{2} are denoted as the rotation oversampling factors. The second component, or the so-called second stage precoder,
W_{2}, is used to combine the selected beam vectors.
[0006] Assuming a rank-
R transmission and a dual-polarized antenna array at the gNB with configuration (
N_{1},
N_{2}, 2), the Rel.-15 double-stage precoder disclosed in [1] for the
s-th subband and r-th transmission layer is given by
where the precoder matrix
W^{(r)}(s) has 2
N_{1}N_{2} rows corresponding to the number of antenna ports, and
S columns for the reporting subbands/PRBs. The matrix
W_{1} ∈
is the wideband first-stage precoder containing 2
U spatial beams for both polarizations, which are identical for all
S subbands, and
F_{A} is a diagonal matrix containing 2
U wideband amplitudes associated with the 2
U spatial beams, and
is the second-stage precoder containing 2
U subband (subband amplitude and phase) complex frequency-domain combining-coefficients associated with the 2
U spatial beams for the s-th subband.
[0007] For the 3GPP Rel.-15 dual-stage Type-II CSI reporting, the second stage precoder,
W_{2} is calculated on a subband basis such that the number of columns of
W_{2} =
depends on the number of configured subbands. Here, a subband refers to a group of adjacent physical resource blocks (PRBs). One major drawback of the Type-II CSI feedback is the large feedback overhead for reporting the combining coefficients on a subband basis. The feedback overhead increases approximately linearly with the number of subbands, and becomes considerably large for large numbers of subbands. To overcome the high feedback overhead of the Rel.-15 Type-II CSI reporting scheme, it has recently been decided in 3GPP RAN#81 [2] (3GPP radio access network (RAN) 3GPP RAN#81) to study feedback compression schemes for the second stage precoder
W_{2}. In several contributions [3]-[4], it has been demonstrated that the number of beam-combining coefficients in
W_{2} may be drastically reduced when transforming
W_{2} using a small set of DFT basis vectors into the delay domain. The corresponding three-stage precoder relies on a three-stage (i.e., three components)
codebook. The first component, represented by matrix
W_{1}, is identical to the Rel.-15 NR component, independent of the layer (
r), and contains a number of spatial domain (SD) basis vectors selected from a spatial codebook. The second component, represented by matrix
is layer-dependent and used to select a number of delay domain (DD) basis vectors from a Discrete Fourier Transform-based (DFT-based) matrix which is also called the delay codebook. The component, represented by matrix
is third layer-dependent and contains a number of combining coefficients that are used to combine the selected SD basis vectors and DD basis vectors from the spatial and delay codebooks, respectively.
[0008] Assuming a rank-
R transmission the three-component precoder matrix or CSI matrix for a configured 2
N_{1}N_{2} antenna/DL-RS ports and configured
N_{3} subbands is represented for the first polarization of the antenna ports and
r-th transmission layer as
and for the second polarization of the antenna ports and
r-th transmission layer as
where
b_{u} (
u = 0, ...,
U - 1) represents the
u-th SD basis vector selected from the spatial codebook,
(
d = 0,
..., D - 1) is the
d-th DD basis vector associated with the
r-th layer selected from the delay codebook,
is the complex delay-domain combining coefficient associated with the
u-th SD basis vector,
d-th DD basis vector and
p-th polarization,
U represents the number of configured SD basis vectors,
D represents the number of configured DD basis vectors, and
α^{(l,p)} is a normalizing scalar.
[0009] A major advantage of the three-component CSI reporting scheme in equation (2) is that the feedback overhead for reporting the combining coefficient of the precoder matrix or CSI matrix is no longer dependent on the number of configured frequency domain subbands (i.e., it is independent from the system bandwidth). Moreover, the feedback overhead and the performance of the precoder matrix or CSI matrix can be controlled by the gNB by configuring to the UE a maximum number of non-zero combining coefficients,
K, per layer, or all layers, that can be contained in the third component,
and are reported by the UE. As only the amplitude and phase information of non-zero combining coefficients are reported, an indicator such as a bitmap is required that indicates which of the
2UD coefficients per layer are selected and reported by the UE. According to [5], the selected non-zero coefficients of the r-th layer are indicated by a bitmap, where each bit in the bitmap is associated with a polarization index (
p ∈ {1,2}), an SD basis index (0 ≤
u ≤
U - 1) and DD basis index (0 ≤
d ≤
D - 1). A "1" in the bitmap indicates that the combining coefficient associated with the polarization index
p, SD basis index
u, and DD basis index
d is non-zero, selected and reported by the UE. A "0" in the bitmap indicates that the combining coefficient associated with the polarization index
p, SD basis index
u, and DD basis index
d is zero, and hence not reported by the UE.
[0010] According to [6], the strongest combining coefficient per layer is normalized to 1 and not reported. In order to indicate which of the
2UD coefficients of a layer is the strongest combining coefficient, a strongest coefficient indicator (SCI) is reported per layer by the UE.
[0011] According to [6], the non-zero combining coefficients
(contained in
are quantized as follows:
where the amplitude of the combining coefficient
is given by two amplitudes, the first and the second amplitudes denoted by
and
respectively. Here,
denotes the polarization reference amplitude defined for each polarization which is common for all amplitude values associated with a polarization
p (
p = 1,2). For the polarization index of the
U SD components associated with the SCI,
and not reported. The polarization reference amplitude associated with the other polarization
p' ≠
p is quantized with
a' bits. In addition, the amplitude
and the phase
of each combining coefficient
is quantized with
a bits and
b bits, respectively.
Configuration and reporting of the three-component CSI scheme
[0012] For the configuration of the precoder matrix or CSI matrix, a CSI report configuration may be signaled via higher layer (e.g., RRC) from the gNB to the UE, wherein the higher layer CSI report configuration may contain the following information [7]:
- A parameter U indicating a number of SD basis vectors to be selected by UE from the spatial codebook for the calculation of W_{1},
- A parameter D (or variants thereof) indicating a number of DD basis vectors to be selected by UE per layer from the delay codebook for the calculation of
- A parameter K (or variants thereof) indicating a maximum number of non-zero coefficients contained in matrix
per layer, or all layers, and used by the UE to combine the selected SD basis vectors and DD basis vectors, and - A parameter N_{3} indicating the number of frequency domain subbands of the CSI matrix and the dimension of the DD basis vectors in the delay codebook, and
- Additional parameter(s) for the configuration of the reporting of the DD basis vectors.
[0013] The CSI report may contain at least a rank indicator (RI) indicating the selected number of layers of the CSI matrix, the number of selected number of non-zero combining coefficients across all layers,
K_{NZ}, and a PMI defining the three components of the CSI matrix, wherein the PMI contains at least the following information [7]:
- A spatial domain subset indicator (SD basis indicator) indicating the selected U SD basis vectors and (if configured) the selected oversampling rotation factors from the spatial codebook for the RI layers of the CSI matrix,
- A delay domain subset indicator (DD basis indicator) indicating per layer the selected DD basis vectors,
- A strongest coefficient indicator (SCI) per layer indicating the SD basis index (or the SD and DD basis indices) associated with the strongest combining coefficient (which is not reported),
- Amplitude and phase information associated with the K_{NZ,r} selected non-zero quantized delay domain combining coefficients per layer,
- A bitmap per layer indicating the SD basis indices and DD basis indices associated with the K_{NZ,r} non-zero coefficients per layer,
- A polarization specific reference amplitude per layer, and
- Possible additional parameter(s) associated with the DD basis subset indication.
UCI omission for 3GPP Rel.-15 CSIreporting
[0014] UCI omission [1] for PUSCH-based resource allocation and CSI reporting was introduced in 3GPP Rel-15. It allows a UE to drop some parts of one or more CSI report(s) in the case that the PUSCH resource allocation is not sufficient to carry the entire content of the CSI report(s). UCI omission may happen when the base station did not accurately allocate the PUSCH resources when scheduling the CSI report(s). For example, the base station may allocate resources for a rank-1 (RI=1) CSI report, but the UE determines a rank-2 transmission and reports a rank-2 (RI=2) CSI report of which size is larger than the size of the allocated PUSCH resources. In such a case, the UE has to drop a portion of the UCI content. In 3GPP Rel. 15 the dropping is achieved by decomposing the UCI payload associated with the CSI reports into smaller portions, the so-called priority levels, see Table 5.2.3-1 of [1], where priority level 0 has the highest priority, and
N_{REP} represents the total number of CSI reports configured to be carried on the PUSCH. Each priority level is associated with a part of a CSI report. The UE drops the CSI portions with lower priority such that the payload size of the CSI reports fits with the PUSCH resource allocation. Moreover, the CSI payload is portioned into two parts: CSI part 1 and CSI part 2. The CSI part 1 contains the RI and an indicator that indicates the size of CSI part 2. The size of CSI part 1 is fixed, whereas the size of CSI part 2 varies depending on the determined RI by the UE and some other factors. Since the gNB needs to know CSI part 1 in order to decode CSI part 2, UCI omission is only performed on CSI part 2.
[0015] The CSI part 2 is composed on 2
N_{REP} + 1 CSI portions. Here, 2
N_{REP} CSI portions, the so-called subband PMIs, contain the CSI content(s) associated with the even and odd subbands of the
N_{REP} CSI report(s). Moreover, each subband PMI is associated with a priority level (starting from index 1 to 2
N_{REP}). In addition, the first CSI portion which is associated with priority level index 0 contains information for all 2
N_{REP} subband PMIs (i.e., for the entire CSI reporting band). The motivation behind the Rel. 15 subband-based CSI decomposition and omission method is that in case of omission of a first subband PMI of CSI report
n, the gNB may use the CSI content of the reported second subband PMI of CSI report
n to estimate the CSI of the omitted first subband PMI by using an interpolation scheme. In this way, a severe degradation of the performance can be avoided as neighbored subbands are typically highly correlated.
SUMMARY
[0016] For the known three-component CSI reporting scheme, the 3GPP Rel. 15 UCI omission procedure cannot be reused, since subband-based PMI does not exists and a decomposition of the CSI part 2 into a number of subband PMIs is not possible. Consequently, new UCI omission rules are required.
[0017] Note that the three-component CSI reporting scheme, the CSI payload of a CSI report can be controlled by the UE by the number of non-zero coefficients to be reported. In case of UCI omission, the UE may simply reduce the number of non-zero coefficients to be reported for one or more of the CSI reports based on the available PUSCH resources. However, a reduction of the number of non-zero combining coefficients would require a recalculation of the combining coefficients, SD and DD basis vectors of the CSI matrices for the one or more CSI reports, occupying additional UE resources. Such additional UE resources may not be available at the UE. Therefore, the UCI omission scheme should not be require a recalculation of CSI matrices for one or more CSI reports.
[0018] For the three-component CSI reporting scheme, the size of the payload of the CSI reports is mainly determined by the bitmaps and the amplitude and phase information of the reported non-zero combining coefficients of the CSI reports.
[0019] In this invention, different segmentation schemes for the bitmaps and the amplitude and phase information of the reported non-zero combining coefficients of the CSI reports for the three-component CSI reporting scheme are proposed.
[0020] In one solution of this invention, the UCI omission scheme is based on dropping a portion of the amplitude and phase information of the non-zero combining coefficients of a CSI report.
[0021] In another solution of this invention, the UCI omission schemes is based on dropping a portion of the amplitude and/or phase information of the non-zero combining coefficients and a portion of the bitmaps that are associated with the dropped combining coefficients.
[0022] The present invention proposes a method performed by a user equipment, UE, for providing channel state information, CSI, feedback in the form of one or more CSI reports in a wireless communication system, the method comprising:
- receiving, from a network node, gNB, higher layer configuration(s) of one or more downlink reference signals, and one or more CSI report configuration(s) associated with the downlink reference signal configuration(s), and a radio signal via a MIMO channel, the radio signal including the downlink reference signal(s) according to the one or more downlink reference signal configurations,
- estimating, the downlink MIMO channel based on measurements on the received one or more downlink reference signals, the downlink reference signals provided over a configured number of frequency domain resources, time domain resources and one or more ports,
- determining, for each CSI report configuration, a precoding matrix based on an estimated channel matrix and two codebooks, the two codebooks including
- a spatial codebook comprising one or more spatial domain (SD) basis components of the precoder, and
- a delay codebook comprising one or more delay domain (DD) basis components of the precoder,
and one or more non-zero combining coefficients for complex combining of the one or more SD and DD basis vectors, and - reporting to the network node, the one or more CSI reports for the one or more CSI report configurations.
[0023] Each CSI report contains the selected precoding matrix in the form of a precoding matrix identifier, PMI, and a rank identifier, RI, indicating the transmission rank for the RI layers of the precoding matrix, and each CSI report comprises two parts: CSI part 1 and CSI part 2, where part 1 has a fixed payload size and comprises information indicating the size of the payload of part 2. Part 2 comprises at least the amplitude and phase information of the selected non-zero combining coefficients of the CSI report, and a portion, or the entirety, of part 2 is considered for omission from the CSI report.
[0024] According to one proposed aspect of the invention part 1 contains at least the information on the selected number of non-zero combining coefficients across all RI layers and an indication of the transmission rank for the RI layers of the selected precoding matrix.
[0025] It is also proposed that part 2 contains at least the following information for the RI layers of the selected precoding matrix:
- a spatial domain, SD, basis subset indicator, including, if configured, the rotation oversampling factors, indicating the selected SD basis vectors from the spatial codebook,
- one or more delay domain, DD, basis subset indicators indicating the selected DD basis vectors from the delay codebook,
- the phase and amplitude of the selected non-zero delay-domain combining coefficients,
- a strongest coefficient indicator, SCI, indicating the DD and SD vector associated with the strongest coefficient per layer,
- a polarization reference amplitude per layer,
- a bitmap for indicating the non-zero combining coefficients per layer, and
- possible additional parameter(s) associated with the DD basis subset indication.
[0026] One aspect of the invention teaches that part 2 of the
N_{REP} CSI reports can be segmented into
TN_{REP} CSI subgroups, where always
T CSI subgroups are associated with a single CSI report, where each CSI subgroup is associated with a priority (priority level).
[0027] Another aspect of the invention teaches that the part 2 of the
N_{REP} CSI reports can be segmented into
TN_{REP} + 1 CSI subgroups, wherein always
T CSI subgroups are associated with a single CSI report, and one CSI subgroup contains information associated with all
N_{REP} CSI reports, wherein each CSI subgroup is associated with a priority (priority level), and the CSI subgroup that contains information associated with all
N_{REP} CSI reports has the highest priority (priority level 0).
[0028] It is proposed that in case of omission, the UE may drop the CSI subgroups with lower priority until the payload size of the CSI reports fits with the resource allocation from the gNB. When omitting a CSI subgroup for a particular priority level, the UE may omit all the CSI content at that priority level.
[0029] The parameter
T may indicate the number of CSI subgroups per CSI report and be related to the granularity of the CSI content that is omitted from a CSI report, where a high value of
T indicates a high granularity and a low value of
T indicates a low granularity, and where each CSI report is associated only with two CSI subgroups when the parameter
T is given by the value of 2.
[0030] The parameter
T may depend on the information content contained in a CSI report, in which case:
- the parameter T may depend on the RI value indicated in the CSI report, wherein T = 2 if the RI value indicated in the CSI report is larger than a specific threshold value and T = 1 if the RI value in the CSI report is smaller than a specific threshold value, or
- the parameter T depend on the number of non-zero coefficients K_{NZ} indicated in the CSI report, wherein T = 2 if the number of non-zero coefficients indicated in the CSI report is larger than a specific threshold value, such as K_{NZ} > K̃_{NZ}, and T = 1 otherwise.
[0031] One proposed aspect of the invention teaches that the first CSI subgroup associated with priority level 0 contains CSI information of all
N_{REP} CSI reports, and wherein the joint CSI subgroup contains the information of at least one of the following parameters:
- the selected SD basis subset indicator including (if configured) the rotation oversampling factors,
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers,
- the polarization reference amplitude value(s) for the RI layers,
- the bitmap(s) for indicating the K_{NZ} non-zero combining coefficients for the RI layers, and
- possible additional parameter(s) associated with the DD basis subset indication.
[0032] It is also proposed that the CSI subgroup with highest priority of a CSI report contains at least the information of the following parameters:
- the selected SD basis subset indicator including (if configured) the rotation oversampling factors,
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers,
- the polarization reference amplitude value(s) for the RI layers, and
- the bitmap(s) for indicating the K_{NZ} non-zero combining coefficients for the RI layers, and
- possible additional parameter(s) associated with the DD basis subset indication.
[0033] It is proposed that each CSI subgroup, that contains information of a fraction of the combining coefficients, may contain the amplitude values, or phase and amplitude values, associated with a maximum of
x > 1combining coefficients of a CSI report, and that the remaining CSI subgroups with lower priority may contain the remaining amplitude values, or remaining phase and amplitude values, of the CSI report.
[0034] It is also proposed that each CSI subgroup that has the highest priority and is associated with a single CSI report may contain at least the bitmaps associated with all SD components of one or more DD basis vectors, or one or more DD basis vector indices, for the RI layers of the precoding matrix indicated in the CSI report, and that the CSI subgroup may contains the corresponding amplitude and/or phase information of the combining coefficients associated with the bitmaps.
[0035] Another aspect of the invention teaches that bitmaps and the information of the combining coefficients for the RI layers of a CSI report may be segmented into D segments, wherein each segment contains the sub-bitmap associated with all SD components of a single DD basis index for all RI layers of the precoding matrix, and where the CSI subgroup that has the highest priority and is associated with a single CSI report contains one or more of the
D segments.
[0036] Here it is proposed that each segment may contain the associated amplitude and/or phase information of the combining coefficients for the RI layers associated with the sub-bitmap.
[0037] A CSI subgroup that is associated with the highest priority of a CSI report may contain at least the bitmap(s) for indicating the non-zero combining coefficients for a first subset of the RI layers, and a first fraction of the amplitude and phase values of the selected non-zero delay-domain combining coefficients associated with the first subset of the RI layers of the precoding matrix indicated in the CSI report.
[0038] It is proposed that the CSI subgroup that contains CSI information associated with
N_{REP} CSI reports, may contain for one or more CSI reports, the SCI(s) for the RI layers, the fraction of the bitmaps and the amplitude and/or phase information of the combining coefficients associated with all SD components and DD basis vector index 0 for the RI layers of each CSI report.
[0039] When the bitmaps of all RI layers of size 2
UD ×
RI associated with a single CSI report are contained in a single CSI subgroup, it is proposed that the bitmap may be segmented into
RI segments, each of 2
UD bits, in an increasing order with respect to the layer numbers, and each segment is associated with all SD and DD basis indices of a layer.
[0040] When a CSI subgroup contains the bitmaps of all RI layers, it is proposed that the bitmap of size 2
UD ×
RI is segmented into
D segments, and each segment has a size of 2
U ×
RI and is associated with a single DD basis vector index. It is also proposed that each segment of size 2
U ×
RI is segmented into
RI segments, and each segment has a size of 2
U × 1 and is associated with a single DD basis vector index and a single layer index.
[0041] It is proposed that each CSI subgroup contains the amplitude and/or phase information of the combining coefficients associated with a portion of the bitmaps of the RI layers, or with a portion of the bitmaps of the D segments.
[0042] The bit-width of the CSI subgroup associated with a single CSI report and highest priority is fixed and given by
A +
B, where
A is the combined bit-width of all components that are contained in the CSI subgroup apart from the number of non-zero combining coefficients, and
B is the bit-width associated with the amplitude (
a) and phase information (
b) of a fraction of the combining coefficients
[0043] The invention also relates to a user equipment, UE, comprising a processor and a memory, the memory containing computer program code executable by the processor whereby the UE is operative to perform any one of the subject matter of the inventive method.
[0044] The invention also relates to a computer program product comprising computer program code, which, when executed by a processor, enables the processor to perform any one of the subject matter of inventive method.
[0045] The invention provides a method through which new UCI omission rules are implemented enabling a UE using the known three-component CSI reporting scheme to make use of an omission procedure without requiring a recalculation of the combining coefficients, SD and DD basis vectors of the CSI matrices for the one or more CSI reports.
BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Examples of embodiments and advantages of the embodiments herein are described in more detail with reference to attached drawings in which:
- Figure 1
- is a schematic and simplified illustration of a user equipment in communication with a radio base station,
- Figure 2
- is an illustration of scheme 1 showing an example CSI part 2 decomposition into a number of CSI subgroups. N_{REP} CSI subgroups with priority level (t - 1)N_{REP} to tN_{REP} - 1 (1 ≤ t ≤ T) are always associated with N_{REP} CSI reports,
- Figure 3
- is an illustration of scheme 1 showing a Decomposition of CSI part 2 into a number of CSI subgroups. T CSI subgroups with priority level (n - 1)T to nT - 1 (1 ≤ n ≤ N_{REP}) are always associated with a single CSI report,
- Figure 4
- is an illustration of scheme 2 showing an example of decomposition of CSI part 2 into a number of CSI subgroups. The first CSI subgroup with the highest priority contains information of CSI report 1 to N_{REP}. T CSI subgroups with priority level (n - 1)T + 1 to nT (1 ≤ n ≤ N_{REP}) are always associated with a single CSI report,
- Figure 5
- is an illustration of scheme 2 showing an example of decomposition of CSI part 2 into a number of CSI subgroups. The first CSI subgroup with the highest priority contains information of CSI report 1 to N_{REP}. N_{REP} CSI subgroups with priority level (t - 1)N_{REP} + 1 to tN_{REP} (1 ≤ t ≤ T) are always associated with N_{REP} CSI reports,
- Figure 6
- is an illustration of an example of CSI content contained in the first joint CSI subgroup with priority level 0 and CSI content of the CSI subgroup that has highest priority and is associated with a single CSI report n,
- Figure 7
- is an illustration of Example of CSI content contained in first joint CSI subgroup with priority level 0 and CSI content of the T CSI subgroups associated with CSI report n,
- Figure 8
- is an illustration of CSI content of T = 2 CSI subgroups with respect to the third method for decomposition scheme 1,
- Figure 9
- is an illustration of CSI content of first and second CSI subgroup associated for CSI report t, with segmentation into D segments, where D denotes the number of configured delay vectors per layer of the CSI matrix of CSI report t.
- Figure 10
- is an illustration of further segmentation of segment k into sub-segments, each sub-segment is associated with all SD components and a DD basis index of a single layer,
- Figure 11
- is an illustration of segmentation of segment k into k_{r,2U} sub-segments, each sub-segment is associated with all SD components and a DD basis index of a single layer,
- Figure 12
- is an illustration of segmentation of a bitmap contained in a CSI subgroup into RI segments, each of 2UD bits,
- Figure 13
- is an illustration of segmentation of bitmap of size 2UD × 1 into subsegments, each of size 2U × 1,
- Figure 14
- is an illustration of segmentation of bitmap of size 2UD × 1 into subsegments, each of size 2D × 1,
- Figure 15
- is an illustration of segmentation of bitmap of size 2UD × 1 into subsegments, each of size D × 1,
- Figure 16
- is an illustration of segmentation of bitmap into D segments, each of size 2U × RI bits,
- Figure 17
- is an illustration of segmentation of 2U × RI bits segment into sub-segments of 2RI bits each,
- Figure 18
- is an illustration of segmentation of 2U × RI bits segment into sub-segments of RI bits each,
- Figure 19
- is an illustration of segmentation of 2U × RI bits into sub-segments of 2U bits each,
- Figure 20
- is an illustration of amplitude information followed by the phase information of each combining coefficient,
- Figure 21
- is an illustration of amplitude information of all X combining coefficients followed by the phase information of all X combining coefficients,
- Figure 22
- is an illustration of phase information of all X combining coefficients followed by the amplitude information of all X combining coefficients, and
- Figure 23
- is a schematic and simplified illustration of a user equipment and a computer program product,
DETAILED DESCRIPTION
[0047] In the following is presented a detailed description of the exemplary embodiments in conjunction with the drawings, in several scenarios, to enable easier understanding of the solution(s) described herein.
[0048] As previously described, in 3GPP new radio system, a UCI omission procedure has been standardized in Rel. 15, which omission procedure cannot be reused, since subband-based PMI does not exists and a decomposition of the CSI part 2 into a number of subband PMIs is not possible. Consequently, new UCI omission rules are required
[0049] Figure 1 is a simplified illustration of a method performed by a user equipment, UE, and a radio base station, gNB, for providing channel state information, CSI, feedback in the form of one or more CSI reports in a wireless communication system A, the method comprising:
- receiving, from a network node, gNB, higher layer configuration(s) of one or more downlink reference signals, and one or more CSI report configuration(s) associated with the downlink reference signal configuration(s), and a radio signal via a MIMO channel, the radio signal including the downlink reference signal(s) according to the one or more downlink reference signal configurations,
- estimating, the downlink MIMO channel based on measurements on the received one or more downlink reference signals, the downlink reference signals provided over a configured number of frequency domain resources, time domain resources and one or more ports,
- determining, for each CSI report configuration, a precoding matrix based on an estimated channel matrix and two codebooks, the two codebooks including
- a spatial codebook comprising one or more spatial domain (SD) basis components of the precoder, and
- a delay codebook comprising one or more delay domain (DD) basis components of the precoder,
and one or more non-zero combining coefficients for complex combining of the one or more SD and DD basis vectors, and - reporting to the network node, the one or more CSI reports for the one or more CSI report configurations.
[0050] Each CSI report contains the selected precoding matrix in the form of a precoding matrix identifier, PMI, and a rank identifier, RI, indicating the transmission rank for the RI layers of the precoding matrix, and each CSI report comprises two parts: CSI part 1 and CSI part 2, where part 1 has a fixed payload size and comprises information indicating the size of the payload of part 2. Part 2 comprises at least the amplitude and phase information of the selected non-zero combining coefficients of the CSI report, and a portion, or the entirety, of part 2 is considered for omission from the CSI report.
[0051] In accordance to an embodiment, the UE is configured with
N_{REP} CSI reports to be carried on the PUSCH, wherein each CSI report may comprise two parts: CSI part 1 and CSI part 2, where part 1 has a fixed payload size and is used to indicate the size of the payload of part 2. The CSI part 1 may contain at least the information on the number of the combining coefficients across all layers and an indication of the transmission rank (RI) for the RI layers of the selected precoding matrix. The CSI part 2 of a CSI report may contain at least the following information for the RI layers of the selected CSI matrix for the configured antenna ports and subbands:
- a selected SD basis subset indicator including (if configured) the rotation oversampling factors,
- a selected DD basis subset indicator per layer,
- the phase and amplitude of the selected non-zero delay-domain combining coefficients per layer,
- a strongest coefficient indicator (SCI) per layer,
- a polarization reference amplitude per layer,
- a bitmap for indicating the non-zero combining coefficients per layer, and
- possible additional parameter(s) associated with the DD basis subset indication.
Decomposition for CSI part 2 - Scheme 1
[0052] In accordance with embodiments, in the first decomposition scheme (Scheme 1), the CSI part 2 of the
N_{REP} CSI reports may be segmented into
TN_{REP} CSI subgroups, wherein always T CSI subgroups are associated with a single CSI report. Moreover, each CSI subgroup is associated with a priority level, wherein the first subgroup has the highest priority level 0. The remaining
TN_{REP} - 1 CSI subgroups are associated with the lower priority levels 1 to
TN_{REP} - 1. The last CSI subgroup
TN_{REP} - 1 may be associated with the lowest priority level
TN_{REP} - 1.
[0053] Figure 2 illustrates a first example of Scheme 1, where always
N_{REP} CSI subgroups with priority level (
t - 1)
N_{REP} to
tN_{REP} - 1 (1 ≤
t ≤
T) and associated with
N_{REP} CSI reports are grouped together.
[0054] Figure 3 illustrates a second example of Scheme 1, where always
T CSI subgroups with priority level (
n - 1)
T to
nT - 1 (1 ≤
n ≤
N_{REP}) and associated with a single CSI report are grouped together
[0055] The parameter
T indicates the number of CSI subgroups per CSI report and is related to the granularity of the CSI content that is omitted from a CSI report. A high value of
T indicates a high granularity and a low value of
T indicates a low granularity. When the parameter
T is given by the value of 2, each CSI report is associated only with two CSI subgroups.
[0056] The parameter
T indicating the number of CSI subgroups per CSI report may also depend on the CSI report. In one example, the parameter
T may be dependent on the rank indicated in the CSI report. For instance,
T = 2, if the rank indicated in the CSI report is larger than one (RI>1) and
T = 1 if the rank indicated in the CSI report is one (RI=1). In another example, the parameter
T may be dependent on the number of non-zero coefficients
K_{NZ} indicated in the CSI report. For instance,
T = 2, if the number of non-zero coefficients indicated in the CSI report is larger than a specific threshold value, i.e.,
K_{NZ} > K̃_{NZ}, and
T = 1 otherwise.
[0057] In contrast to Rel. 15 CSI decomposition, where the first CSI subgroup contains information of all
N_{REP} CSI reports, each subgroup in the proposed decomposition contains the information associated only with a single CSI report.
[0058] In case of UCI omission, the UE drops the CSI subgroups with lower priority until the payload size of the CSI reports fits with the PUSCH resource allocation. When omitting a CSI subgroup for a particular priority level, the UE omits all the CSI content at that priority level.
Decomposition for CSI part 2 - Scheme 2
[0059] A drawback of the above CSI decomposition scheme 1 (shown in Fig. 2 and Fig. 3) is that when all CSI subgroups associated with a single CSI report are dropped, the complete CSI report is dropped. To avoid the complete dropping of the CSI content of a CSI report, the following embodiment proposes a CSI decomposition that allows the gNB to recalculate partly the CSI matrices for all
N_{REP} CSI subgroups even all CSI subgroups are dropped by the UE, except the first CSI subgroup with the highest priority (priority level 0).
[0060] In accordance with embodiments, the CSI part 2 of the
N_{REP} CSI reports may be segmented into
TN_{REP} + 1 CSI subgroups, wherein always
T CSI subgroups are associated with a CSI report. The CSI first subgroup contains information associated with all
N_{REP} CSI reports.
[0061] Each CSI subgroup is associated with a priority level, wherein the first subgroup has the highest priority level 0. The remaining
TN_{REP} CSI subgroups are associated with the lower priority levels 1 to
TN_{REP}, where the last CSI subgroup
TN_{REP} may be associated with the lowest priority level
TN_{REP}.
[0062] The parameter T indicates the number of CSI subgroups per CSI report and is related to the granularity of the CSI content that is omitted from a CSI report. A high value of
T indicates a high granularity and a low value of
T indicates a low granularity. When the parameter
T is given by the value of 2, each CSI report is associated only with two CSI subgroups.
[0063] The parameter
T indicating the number of CSI subgroups per CSI report may also depend on the CSI report. In one example, the parameter
T may be dependent on the rank indicated in the CSI report. For instance,
T = 2, if the rank indicated in the CSI report is larger than one (RI>1) and
T = 1 if the rank indicated in the CSI report is one (RI=1). In another example, the parameter
T may be dependent on the number of non-zero coefficients
K_{NZ} indicated in the CSI report. For instance,
T = 2, if the number of non-zero coefficients indicated in the CSI report is larger than a specific threshold value, i.e.,
K_{NZ} ≤
K̃_{NZ}, and
T = 1 otherwise.
[0064] In case of UCI omission, the UE drops the CSI subgroups with lower priority until the payload size of the CSI reports fits with the PUSCH resource allocation. When omitting a CSI subgroup for a particular priority level, the UE omits all the CSI content at that priority level.
[0065] Figure 4 illustrates a first example of Scheme 2, where always
T CSI subgroups with priority level (
n -
1)
T + 1 to
nT (1 ≤
n ≤
N_{REP}) and associated with a single CSI report are grouped together.
[0066] Figure 5 illustrates a second example of Scheme 2, where always
N_{REP} CSI subgroups with priority level (
t - 1)
N_{REP} + 1 to
tN_{REP} (1 ≤
t ≤
T) and associated with
N_{REP} CSI reports are grouped together.
Content of a CSI subgroup
[0067] In accordance with embodiments, when the first CSI subgroup (associated with priority level 0) contains CSI information of all
N_{REP} CSI reports, the joint CSI subgroup may contain the information of at least one of the following parameters:
- the selected SD basis subset indicator including (if configured) the rotation oversampling factors,
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers,
- the polarization reference amplitude value(s) for the RI layers,
- the bitmap(s) for indicating the K_{NZ} non-zero combining coefficients for the RI layers, and
- possible additional parameter(s) associated with the DD basis subset indication.
[0068] For the segmentation of the phase and amplitude values of the
N_{REP} CSI reports for Scheme 2, two partitioning approaches are proposed in the following.
[0069] In the first approach, a first portion of the phase and amplitude values of the selected non-zero delay-domain combining coefficients of the
N_{REP} CSI reports is contained in the first joint CSI subgroup that has the highest priority. An example of the CSI content of the first joint CSI subgroup and of the CSI subgroup that has the highest priority and is associated with a single CSI report with respect to the first approach is shown in Figure 6.
[0070] In the second approach, the first CSI subgroup that has the highest priority does not contain any phase and amplitude values of the selected non-zero delay-domain combining coefficients for the
N_{REP} CSI reports, and only the remaining CSI subgroups contain the information of the non-zero combining coefficients. An example of the CSI content of the first joint CSI subgroup and of the CSI subgroups associated with a single CSI report with respect to the second approach is shown in Figure 7.
[0071] In addition, depending on the CSI content of the first joint CSI subgroup with priority level 0, each CSI subgroup with the highest priority and associated with a single CSI report may contain the following parameters (when not already listed in the first joint CSI subgroup):
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers,
- the polarization reference amplitude value(s) for the RI layers,
- the bitmap(s) for indicating the K_{NZ} non-zero combining coefficients for the RI layers, and
- possible additional parameter(s) associated with the DD basis subset indication.
[0072] When UE is configured to decompose the CSI payload of the
N_{REP} CSI reports with respect to decomposition scheme 1, and the UE drops up to
T-1 CSI subgroups which are associated with a single CSI report, the gNB should still be able to recalculate a part of the CSI matrix for the RI layers based on the remaining non-dropped CSI subgroup(s) associated with that CSI report. In this way, a certain minimum performance may be guaranteed even if a large part of the CSI content of a CSI report is dropped. In order to recalculate a part of the CSI matrix from a CSI report, the gNB needs to know at least some of parameters such as the selected SD and DD basis subset indicator(s), the SCIs, the bitmap(s) and the polarization reference amplitude value(s) for the RI layers, the window parameter
M_{init} from the CSI report. This information must be contained for each CSI report in the CSI subgroup that has the highest priority.
[0073] In accordance with embodiments, when the CSI payload containing the CSI information of
N_{REP} CSI reports is decomposed into a number of CSI subgroups, and each CSI subgroup contains CSI information that is associated only with a single CSI report, the CSI subgroup with highest priority of a CSI report may contain at least the information of the following parameters:
- the selected SD basis subset indicator including (if configured) the rotation oversampling factors,
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers,
- the polarization reference amplitude value(s) for the RI layers,
- the bitmap(s) for indicating the K_{NZ} non-zero combining coefficients for the RI layers,
- possible additional parameter(s) associated with the DD basis subset indication.
[0074] For the partitioning of the amplitude and phase values to the CSI subgroups several methods are proposed in the following: In a first method, the CSI subgroup with highest priority per CSI report does not contain any information of the selected non-zero combining coefficients, and only the remaining
T - 1 CSI subgroups with lower priority contain the amplitude and phase values of the selected non-zero combining coefficients of the CSI report. In a second method, the CSI subgroup with highest priority per CSI report may contain information of a first fraction of the amplitude values of the selected non-zero delay-domain combining coefficients, and the remaining
T - 1 CSI subgroups with lower priority contain the remaining fraction of amplitude values and all phase values of the CSI report. In a third method, the CSI subgroup with the highest priority per CSI report may contain information of a first fraction of the amplitude and phase values of the selected non-zero delay-domain combining coefficients, and the remaining
T - 1 CSI subgroups with lower priority contain the remaining fraction of amplitude and phase values of the CSI report.
[0075] Figure 8 shows the CSI content of two CSI subgroups associated with a single CSI report and the corresponding CSI content for each CSI subgroup when partitioning the amplitude and phase values with respect to the third method.
[0076] In the case of UCI omission and the CSI decomposition illustrated in Figure 2, the UE drops firstly the CSI subgroups that contain CSI information related to the non-zero combining coefficients. The remaining CSI subgroup(s) still contain the CSI content that can be used to recalculate partly the CSI matrix indicated in a CSI report. Only in the case that the UE drops all
T CSI subgroups associated with a CSI report, the full CSI report is dropped.
[0077] In the case of UCI omission and the CSI decomposition illustrated in Figure 3, the UE drops firstly the CSI subgroups that contain CSI information related to the non-zero delay-domain combining coefficients of the CSI report with lowest priority. In the case that the UE drops all
T CSI subgroups associated with the CSI report with lowest priority, the full CSI report is dropped.
[0078] In accordance with embodiments, each CSI subgroup that contains information of a fraction of the combining coefficients may contain the amplitude values, or phase and amplitude values associated with a maximum of
x > 1 combining coefficients of a CSI report. The remaining CSI subgroups with lower priority may contain the remaining amplitude values, or remaining phase and amplitude values of the CSI report.
[0079] For example, when
T = 2 and
x = 2, the phase and amplitude values of the combining coefficients are segmented into two CSI subgroups, where the first CSI subgroup contains the phase and amplitude values associated with
combining coefficients and the second CSI subgroup contains the phase and amplitude associated with the remaining
combining coefficients of a CSI report.
Segmentation of bitmaps and combining coefficients
[0080] The payload of the CSI part 2 is mainly determined by the bitmaps and the phase and amplitude information of the non-zero combining coefficients.
[0081] In the proposed scheme 1 the bitmaps for one or more of the RI layers of a CSI report are contained in the first CSI subgroup with the highest priority. Therefore, the payload size of this CSI subgroup can be higher than the payload size of other CSI subgroups. Due to the higher payload size, the UE may drop in some cases the CSI subgroups containing the bitmaps, and hence the full CSI report(s), when the UCI omission rate is high, i.e., when the UE has to drop a large part of the CSI part 2 content.
[0082] Similar, in the proposed scheme 2 all bitmaps for the RI layers of the
N_{Rep} CSI reports may be contained in the first joint CSI subgroup, or they may be contained per CSI report in the first CSI subgroup that has the highest priority. The payload size of this/these CSI subgroup(s) can be high, and the UE may drop in some cases the CSI subgroup(s) containing the bitmaps for the RI layers, when the UCI omission rate is high.
[0083] The following embodiments propose different schemes that reduce the payload size of the CSI subgroup(s) containing the bitmaps, and hence the probability of dropping this/these CSI subgroup(s), by partitioning the bitmaps and the phase and amplitude information of the combining coefficients to different CSI subgroups.
Segmentation with respect to DD basis subset
[0084] The first segmentation scheme partitions the bitmaps and the combining coefficients with respect to the number of DD basis vector indices of a CSI report.
[0085] When the UE drops a CSI subgroup, the CSI content of the remaining CSI subgroup(s) with higher priority should allow the gNB to partly reconstruct the CSI matrix for the RI layers indicated in the CSI report. To do this, the gNB requires the knowledge of the SD and DD basis indices associated with the strongest combining coefficient per layer. This information can be obtained from the bitmaps and the SCIs for the RI layers. In order to interpret the SCI(s) in a correct way, the first joint CSI subgroup or the first CSI subgroup associated with a single CSI report shall contain at least the fraction of the bitmaps and the information of the combining coefficients associated with the DD basis vector index of the SCI for the RI layers.
[0086] In accordance to an embodiment, each CSI subgroup that has the highest priority and is associated with a single CSI report may contain at least the bitmaps associated with all SD components of one or more DD basis vectors (or one or more DD basis vector indices) for the RI layers of the CSI matrix indicated in the CSI report. In addition, the CSI subgroup may contain the corresponding amplitude and/or phase information of the combining coefficients for the RI layers of the CSI matrix.
[0087] The bitmaps and the corresponding amplitude and/or phase information of the combining coefficients for the RI layers of the CSI subgroup may be segmented into
D' segments, where each segment contains the bitmaps and the corresponding amplitude and/or phase information of the combining coefficients for one or more of the RI layers associated with a single DD basis vector (or a DD basis vector index). Here,
D' ≤
D, and
D denotes the configured number of DD basis vectors to the UE for each of the RI layers of the CSI matrix.
[0088] In the case that
D' <
D for the CSI subgroup, the CSI subgroups with lower priority and/or the joint CSI subgroup (if present) may contain the remaining segments associated with the remaining DD basis vectors (DD basis vector indices).
[0089] In accordance to an embodiment, each CSI subgroup that has the highest priority and is associated with a single CSI report may contain the fraction of the bitmaps and the information of the combining coefficients associated with all SD components of the DD basis vector index which is associated with the SCI(s) for the RI layers.
Examples for segmentation
[0090] In accordance to an embodiment, the bitmaps and amplitude and/or phase information contained in a CSI subgroup is segmented in one or more segments, wherein each segment may contain a part of the bitmap (sub-bitmap) for each of the RI layers and the associated amplitude and/or phase information of the combining coefficients associated with the sub-bitmap.
[0091] For an example configuration of
T = 2, the CSI content of two CSI subgroups associated with a single CSI report is shown in Figure 9. As observed from the figure, the sub-bitmap and phase and amplitude information of the combining coefficients associated with all SD components and a DD basis vector index for the RI layers are always packed together into a single segment.
[0092] Note that the last segment of a CSI subgroup may contain only a fraction of a sub-bitmap or a fraction of the amplitude and/or phase information associated with that sub-bitmap. In such a case, the remaining fraction of the sub-bitmap or the remaining fraction of the amplitude and/or phase information associated with that sub-bitmap that is not contained in the CSI subgroup may be contained in a CSI subgroup of the same CSI report with lower priority.
[0093] In accordance to an embodiment, the bitmaps and amplitude and/or phase information contained in a CSI subgroup are segmented in one or more segments, wherein each segment may contain a part of the bitmap (sub-bitmap) for each of the RI layers and the associated amplitude and/or phase information of the combining coefficients associated with the sub-bitmap.
[0094] In one example, the k-th segment is further segmented into RI sub-segments, wherein each sub-segment contains the sub-bitmap associated with all SD components and a DD basis vector index of a single layer and the combining coefficients associated with the sub-bitmap, see Figure 10.
[0095] In one instance, each sub-segment
k_{r} associated with
r-th layer may be further segmented into
k_{r,2U} sub-segments, as shown in Figure 11. Each sub-segment
k_{r,u} contains a single-bit from the sub-bitmap and the corresponding amplitude and phase information of the combining coefficient. Note that the bit
k_{r,u} is immediately followed by the sub-segment
k_{r,u+1}, when
k_{r,u} is '0'. This is due to the absence of the amplitude and phase information associated with bit
k_{r,u}.
[0096] In accordance to an embodiment, the bitmaps and the information of the combining coefficients for the RI layers of a CSI report are segmented into
D segments, where each segment contains the bitmap (sub-bitmap) associated with all SD components of a single DD basis index for all RI layers. In addition, each segment may contain the associated amplitude and/or phase information of the combining coefficients for the RI layers associated with the sub-bitmap. The CSI subgroup that has the highest priority and is associated with a single CSI report may contain one or more of the
D segments.
[0097] Each CSI subgroup with the highest priority may contain in addition a set of parameters required from the CSI report such that the gNB may be able to recalculate the CSI matrix based on the fraction of the bitmaps for the RI layers and the amplitude and/or phase information of the combining coefficients contained in the CSI subgroup.
[0098] In accordance to an embodiment, when the CSI payload containing the CSI information of
N_{REP} CSI reports is decomposed into a number of CSI subgroups, and each CSI subgroup contains CSI information that is associated only with a single CSI report, and in addition, when the UE is configured to perform a segmentation of the bitmaps and amplitude and/or phase information of the combining coefficients to the CSI subgroups with respect to DD basis vector indices, each CSI subgroup with the highest priority may contain in addition at least the following parameters:
- the selected SD basis subset indicator including the rotation oversampling factors,
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers, and
- the polarization reference amplitude value(s) for the RI layers, and
- possible additional parameter(s) associated with the DD basis subset indication.
[0099] In accordance to an embodiment, when the CSI payload containing the CSI information of
N_{REP} CSI reports is decomposed into a number of CSI subgroups, and the first CSI subgroup contains CSI information that is associated to
N_{REP} CSI reports, and in addition, when the UE is configured to perform a segmentation of the bitmaps and amplitude and/or phase information of the combining coefficients to the CSI subgroups with respect to DD basis vector indices, the joint CSI subgroup may contain at least one of the following parameters:
- the selected SD basis subset indicator including (if configured) the rotation oversampling factors,
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers, and
- the polarization reference amplitude value(s) for the RI layers,
- possible additional parameter(s) associated with the DD basis subset indication.
[0100] The remaining CSI subgroups contain at least the CSI content associated with the bitmaps for the RI layers and the phase and amplitude information of the combining coefficients. Each CSI subgroup with the highest priority and associated with a single CSI report may contain at least a fraction of the bitmaps for the RI layers and the phase and amplitude information of a fraction of the
K_{NZ} combining coefficients.
[0101] Depending on the structure of the joint CSI subgroup with the highest priority (priority level 0), each CSI subgroup with the highest priority associated with a single CSI report may also contain the following parameters (when not already listed in the joint CSI subgroup):
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers, and
- the polarization reference amplitude value(s) for the RI layers,
- possible additional parameter(s) associated with the DD basis subset indication.
[0102] When the UE is enforced to perform a cyclic shift operation on the selected combining coefficients and the selected DD basis vectors per layer with respect to the DD basis vector index that is associated with the SCI, after the cyclic shift operation only the part of the bitmap which is associated with all SD basis vectors and the first DD basis vector (DD basis vector index 0) is required for each of the RI layers to identify the SD and DD basis indices associated with the strongest combining coefficient. In order to interpret the SCI(s) in a correct way, the first CSI subgroup should contain the fraction of the bitmap and the information of the combining coefficients associated with the DD basis vector index 0.
[0103] Note that when the SCI is given by a ┌
log_{2}(2
U)┐) bit indicator, the SCI may not be indicated in the sub-bitmap associated with DD basis vector index 0, and hence the sub-bitmap per layer may have a size of 2
U - 1 × 1 instead of 2
U × 1.
[0104] In accordance to an embodiment, when the UE is configured for a CSI report to perform a cyclic shift operation on the selected combining coefficients and the selected DD basis vectors per layer with respect to the DD basis vector index that is associated with the SCI, the CSI subgroup with the highest priority associated with a CSI report contains the sub-bitmap which is associated with DD basis vector index 0 and the corresponding amplitude and/or phase information of the combining coefficients associated with that sub-bitmap for the RI layers. For example, the first segment of a CSI subgroup with highest priority may be associated with DD basis vector index 0.
[0105] When the UE is not enforced to perform a cyclic shift on the selected combining coefficients and the selected DD basis vectors per layer with respect to the DD basis vector index that is associated with the SCI, the complete bitmap for all selected SD and DD basis vectors per layer may be required by the gNB to identify the SD and DD basis vector indices associated with the strongest combining coefficient. Therefore, the above segmentation of bitmaps and combining coefficients with respect to DD basis vector indices may not be possible.
[0106] In accordance to an embodiment, when UE is not enforced to perform a cyclic shift operation on the selected combining coefficients and the selected DD basis vectors per layer with respect to the DD basis vector index that is associated with the SCI for a CSI report, the CSI subgroup of the CSI report with highest priority may contain at least the bitmaps for the RI layers and all or a fraction of the phase and/or amplitude information of the combining coefficients of the CSI matrix of the CSI report. In the case that the CSI subgroup contains only a fraction of the phase and/or amplitude information of the combining coefficients, the remaining CSI subgroups of the CSI report with lower priority may contain the remaining fraction of the phase and/or amplitude information of the combining coefficients of the CSI matrix.
[0107] In accordance to an embodiment, when UE is not enforced to perform a cyclic shift operation on the selected combining coefficients and the selected DD basis vectors per layer with respect to the DD basis vector index that is associated with the SCI for a CSI report, and when the CSI payload containing the CSI information of
N_{REP} CSI reports is decomposed into a number of CSI subgroups that comprise a joint CSI subgroup which contains CSI information that is associated to
N_{REP} CSI reports, the joint CSI subgroup may contain the bitmaps for the RI layers of that CSI report.
Segmentation with respect to layers
[0108] The second segmentation scheme partitions for a CSI report the bitmaps and the combining coefficients with respect to the RI layers of the CSI matrix indicated in the CSI report.
[0109] In accordance with embodiments, a CSI subgroup that is associated with highest priority of a CSI report may contain at least the bitmap(s) for indicating the non-zero combining coefficients for a first subset of the RI layers, and a first fraction of the amplitude and phase values of the selected non-zero delay-domain combining coefficients associated with the first subset of the RI layers of the CSI matrix indicated in the CSI report.
[0110] The remaining CSI subgroups with lower priority associated with the same CSI report may contain at least the bitmap(s) for the remaining (second subset) of the RI layers and the remaining fraction of amplitude and phase values associated with the second subset of the RI layers of that CSI report.
[0111] When the UE is configured to decompose the CSI payload with respect to above-described scheme 1, the CSI subgroup with highest priority per CSI report may contain in addition the following parameters:
- the selected SD basis subset indicator including (if configured) the rotation oversampling factors,
- the selected DD basis subset indicator(s) for a first subset of the RI layers,
- the SCI(s) for a first subset of the RI layers,
- the polarization reference amplitude value(s) for a first subset of the RI layers,
- the bitmap(s) for indicating the non-zero combining coefficients for a first subset of the RI layers,
- possible additional parameter(s) associated with the DD basis subset indication.
[0112] When the UE is configured to decompose the CSI payload with respect to above-described scheme 2, the CSI subgroup with highest priority per CSI report of a CSI report may also contain in addition the following parameters (when not already listed in the joint CSI subgroup):
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers,
- the polarization reference amplitude value(s) for the RI layers,
- possible additional parameter(s) associated with the DD basis subset indication.
Wideband CSI information
[0113] In 3GPP Release 15, the CSI part 2 decomposition contains in the first CSI portion with priority level 0 the so-called wideband amplitudes of the
N_{REP} CSI reports. Based on these wideband amplitudes in the first CSI portion, the gNB can reconstruct per CSI report a wideband CSI matrix for the configured subbands even in the case that the UE has dropped all except the first CSI portion with priority level 0. A similar wideband CSI matrix can be defined for the two codebook-based CSI reporting scheme when applying the second CSI decomposition scheme described above. The gNB may derive a wideband CSI matrix for a CSI report when it has knowledge on the amplitude information of the combining coefficients and the fraction of the bitmap associated with the "strongest" DD basis vector index for each of the RI layers. In many cases, the "strongest" DD basis vector index for per layer corresponds to the DD basis vector index that is associated with the SCI. Therefore, based on the SCI, the bitmap associated with the DD basis vector index which is associated with the SCI and amplitude information of the combining coefficients, the gNB may reconstruct a wideband CSI matrix as proposed in the following embodiments.
[0114] In accordance to an embodiment, when the CSI payload containing the CSI information of
N_{REP} CSI reports is decomposed into a number of CSI subgroups, and the first CSI subgroup contains CSI information that is associated to
N_{REP} CSI reports, the joint CSI subgroup may contain the SCI(s) for the RI layers, the fraction of the bitmap and the amplitude and/or phase information of the combining coefficients associated with all SD components of the DD basis vector index which is associated with the SCI(s) for the RI layer(s) of one or more CSI reports, Based on the information contained in the joint CSI subgroup, the gNB is able to recalculate a wideband CSI matrix for each of the one or more CSI reports.
[0115] Similar to the discussion of the cyclic shift operation above, when the UE is configured for a CSI report to perform a cyclic shift operation on the selected combining coefficients and the selected DD basis vectors per layer with respect to the DD basis vector that is associated with the SCI, then the DD basis vector index 0 is associated with the SCI.
[0116] In accordance to an embodiment, when the CSI payload containing the CSI information of
N_{REP} CSI reports is decomposed into a number of CSI subgroups, and the first CSI subgroup contains CSI information that is associated to
N_{REP} CSI reports, the joint CSI subgroup may contain for one or more CSI reports, the SCI(s) for the RI layers, the fraction of the bitmap and the amplitude and/or phase information of the combining coefficients associated with all SD components and DD basis vector index 0 for the RI layers per CSI report.
[0117] Moreover, as discussed above, when the UE is not enforced to perform a cyclic shift operation for a CSI report on the selected combining coefficients, the complete bitmap for all selected SD and DD basis vector indices per layer and at least the amplitude and/or phase information of the combining coefficients associated with the DD basis vector index which is associated with the SCI(s) for the RI layer(s) is required to calculate the wideband matrix of that CSI report.
[0118] In accordance to an embodiment, when the CSI payload containing the CSI information of
N_{REP} CSI reports is decomposed into a number of CSI subgroups, and the first CSI subgroup contains CSI information that is associated to
N_{REP} CSI reports, and when the UE is not enforced to perform a cyclic shift operation on the selected combining coefficients for a CSI report, the joint CSI subgroup may contain the complete bitmap for all selected SD and DD basis vector indices per layer and at least the amplitude and/or phase information of the combining coefficients associated with the DD basis vector index which is associated with the SCI(s) for the RI layer(s) of that CSI report.
[0119] The size of the joint CSI subgroup depends on the number of non-zero combining coefficients associated with the DD basis vector index of the SCI(s) of the one or more CSI reports. The gNB may therefore even after decoding CSI part 1 of the CSI payload not know the payload size of the joint CSI subgroup. In the following, a method that fixes the payload size of the joint CSI subgroup is proposed.
[0120] In accordance to an embodiment, the joint CSI subgroup contains at least the SCI(s) for the RI layers, the fraction of the bitmap and the amplitude and/or phase information of the combining coefficients associated with all SD components and the DD basis vector index of the SCI for the RI layers, for the
N CSI reports that have the highest priority.
[0121] The value for the parameter
N may be either higher-layer configured by the gNB, or it is a priori known by the UE (e.g., fixed by specification), or it is determined by the UE.
[0122] For example, the UE may determine the value of the parameter
N (i.e., the number of CSI reports that have the highest priority) such that the payload size of the joint CSI subgroup is fixed.
Reading/packing rules when bitmaps of RI layers are contained in a single CSI subgroup
[0123] As explained before, the CSI subgroup that has the highest priority and is associated with a single CSI report may contain the bitmaps of all RI layers and only a fraction of the phase and/or amplitude information of the combining coefficients.
[0124] The following embodiment presents a decomposition scheme for the phase and/or amplitude information of the combining coefficients with respect to a number of CSI subgroups. The decomposition of the phase and/or amplitude information of the combining coefficients may depend on the ordering of the bit-sequences of the associated bitmap for the RI layers of a CSI report.
[0125] In accordance with embodiments, when the bitmaps of all RI layers of size 2
UD ×
RI associated with a single CSI report are contained in a single CSI subgroup, the bitmap is segmented into
RI segments, each of 2
UD bits, in an increasing order with respect to the layer numbers. Each segment is associated with all SD and DD basis indices of a layer, see Figure 12. In case of UCI omission, the UE drops first the CSI content associated with one or more of the higher layer(s), and then the CSI content associated with one or more of the lower layer(s). For example, for RI=4 and T = 2, the UE may first drop the CSI content associated with layer 3 and 4 (CSI subgroup with lower priority), and then drop the CSI content associated with layer 1 and 2 (CSI subgroup with high priority) of a CSI report.
[0126] The bit-sequence in each segment may be ordered with respect to one of the following two schemes: In a first ordering scheme (scheme 1), the bits in each segment of size 2
UD × 1 are ordered in a way that the first 2
U bits are associated with all 2
U SD components of the first DD basis index, followed by 2
U bits that are associated with all 2
U SD components of the second DD basis index, and so on, see Figure 13. In a second ordering scheme (scheme 2), the bits in each segment of size 2
UD × 1 are ordered in a way that the first
D bits are associated with all
D DD basis indices of the first SD basis index, followed by
D bits that are associated with all
D DD basis indices of the second SD basis index, and so on, see Figure 14 or Figure 15.
[0127] When the bitmaps are ordered as shown in Figure 13 or Figure 14 or Figure 15, the fraction of phase and/or amplitude information of the combining coefficients contained in a CSI subgroup is always associated with only a subset of the RI layers of a CSI report. Such an ordering/packing may result in a significant performance loss compared to dropping only a part of the phase and/or amplitude information of the combining coefficients of all RI layers. The following embodiment presents an alternative ordering/segmentation of the bitmaps and combining coefficients that avoids a dropping of the combining coefficients of one or more layers of a CSI report.
[0128] In accordance with embodiments, when a CSI subgroup contains the bitmaps of all RI layers, the bitmap of size 2
UD ×
RI is segmented into
D segments, and each segment has a size of 2
U ×
RI and is associated with a single DD basis vector index, see Figure 16. In case of UCI omission, the UE drops first the CSI content associated with one or more DD basis indices of all RI layers, and then the CSI content associated with the remaining DD basis indices of all RI layers. For example, for
D = 4, the UE may first drop the CSI content associated DD basis index 3 and 4, and then the CSI content associated with DD basis index 1 and 2 of a CSI report.
[0129] The bit-sequence in each segment may be ordered with respect to one of the following two schemes: In a first ordering scheme (scheme 1), the bits in each segment of size 2
U ×
RI are ordered in a way that the first
RI bits are associated with the first SD basis vector index of all
RI layers, followed by the next
RI bits associated with the second SD basis index of all
RI layers, and so on, see Figure 17 and Figure 18. In a second ordering scheme (scheme 2), the bits in each segment of size 2
U ×
RI are ordered in a way that the first
2U bits are associated with all
2U SD components of a DD basis index associated with the first layer, followed by the 2
U bits associated with all 2
U SD components of a DD basis index associated with the second layer, and so on, see Figure 19.
Packing/Ordering rules for the combining coefficients
[0130] As explained in an embodiment, a CSI subgroup associated with a CSI report may contain the amplitude and phase information of a fraction of the combining coefficients, where different ordering schemes for the amplitude and phase values may be applied.
[0131] In accordance with an embodiment, the amplitude and phase information of the combining coefficients in a CSI subgroup may be ordered by one of the following schemes. In scheme 1, for each combining coefficient in CSI subgroup, the amplitude information (
a bits) is followed by the phase information (
b bits), see Figure 20, where X denotes the number of combining coefficients contained in the CSI subgroup. In scheme 2, the amplitude information of
X combining coefficients (
Xa bits) is followed by the phase information of
X combining coefficients (
Xb bits), see Figure 21. In scheme 3, the phase information of
X combining coefficients (
Xb bits) is followed by the amplitude information of
X combining coefficients (
Xa bits), see Figure 22.
Bit-width of the subgroups
[0132] In accordance with embodiments, the bit-width of the first CSI subgroup associated with a single CSI report and highest priority may be fixed and given by
A +
B, where A is the combined bit-width of all components that are contained in the first subgroup apart from the number of non-zero combining coefficients, and
B is the bit-width associated with the amplitude (
a) and phase information (
b) of a fraction of the combining coefficients
[0133] For example, for
T = 2, when the first subgroup contains only the bitmaps of all layers and a fraction of the combining coefficients
then the bit-widths of the first CSI subgroup and second CSI subgroup 2 are given by
and
respectively.
[0134] In one example,
x =
T.
[0135] In order to perform the methods steps and operations of the UE previously described, there is also provided a UE 30, as shown in Figure 23, which UE 30 comprises a processor 31 or processing circuit or a processing module or a processor or means; a receiver circuit or receiver module 34; a transmitter circuit or transmitter module 35; a memory module 32; a transceiver circuit or transceiver module 33 which may include the transmitter circuit 35 and the receiver circuit 34. The UE 30 further comprises an antenna system 36 which includes antenna circuitry for transmitting and receiving signals to/from at least the UE 30.
[0136] The invention also relates to a computer program product 37, which in the figure is stored in the memory module 32, which computer program product 37 comprises computer program code 38, which, when executed by the processor 31, enables the processor 31 to perform any one of the subject matter of the inventive method.
References
[0137]
- [1] 3GPP TS 38.214 V15.3.0: "3GPP; TSG RAN; NR; Physical layer procedures for data (Rel. 15).", Sept 2018.
- [2] Samsung "Revised WID: Enhancements on MIMO for NR", RP-182067, 3GPP RAN#81, Gold Coast, Australia, Sept. 10th - 13th, 2018.
- [3] R1-1806124, Fraunhofer IIS, Fraunhofer HHI, Enhancements on Type-II CSI reporting scheme, RAN1#93, Busan, South Korea, May 21st - May 25th, 2018.
- [4] R1- 1811088, Fraunhofer IIS, Fraunhofer HHI, Enhancements on Type-II CSI reporting scheme, RAN1#94-Bis, Chengdu, China, Oct. 8th - Oct. 12th, 2018.
- [5] Chairman's Notes, RAN1#96, Athens, Greece, Feb. 25th - Mar. 1st, 2019.
- [6] R1-1902304, Samsung, Summary of CSI enhancement for MU-MIMO, RAN1#96, Feb. 25th - Mar. 1st, 2019
- [7] R1-1905629, Samsung, Feature lead summary for MU-CSI - revision on selected issues, Xi'an, China, Apr. 12th - 16th, 2019.
1. A method performed by a user equipment, UE, for providing channel state information, CSI, feedback in the form of one or more CSI reports in a wireless communication system, the method comprising:
- receiving, from a network node, gNB, higher layer configuration(s) of one or more downlink reference signals, and one or more CSI report configuration(s) associated with the downlink reference signal configuration(s), and a radio signal via a MIMO channel, the radio signal including the downlink reference signal(s) according to the one or more downlink reference signal configuration(s),
- estimating, the downlink MIMO channel based on measurements on the received one or more downlink reference signals, the downlink reference signals provided over a configured number of frequency domain resources, time domain resources and one or more ports,
- determining, for each CSI report configuration, a precoding matrix based on an estimated channel matrix and two codebooks, the two codebooks including
- a spatial codebook comprising one or more spatial domain (SD) basis components of the precoder, and
- a delay codebook comprising one or more delay domain (DD) basis components of the precoder,
and one or more non-zero combining coefficients for complex combining of the one or more SD and DD basis vectors, and
- reporting, to the network node, the one or more CSI reports for the one or more CSI report configurations,
wherein each CSI report contains the selected precoding matrix in the form of a precoding matrix identifier, PMI, and a rank identifier, RI, indicating the transmission rank for the RI layers of the precoding matrix, and wherein each CSI report comprises two parts: CSI part 1 and CSI part 2, wherein part 1 has a fixed payload size and comprises information indicating the size of the payload of part 2, wherein part 2 comprises at least the amplitude and phase information of the selected non-zero combining coefficients of the CSI report, and wherein a portion, or the entirety, of part 2 is considered for omission from the CSI report.
2. The method of claim 1, wherein part 1 contains at least the information on the selected number of non-zero combining coefficients across all RI layers and an indication of the transmission rank for the RI layers of the selected precoding matrix.
3. The method of claim 2, wherein part 2 contains at least the following information for the RI layers of the selected precoding matrix:
- a spatial domain, SD, basis subset indicator, including, if configured, the rotation oversampling factors, indicating the selected SD basis vectors from the spatial codebook,
- one or more delay domain, DD, basis subset indicators indicating the selected DD basis vectors from the delay codebook,
- the phase and amplitude of the selected non-zero delay-domain combining coefficients,
- a strongest coefficient indicator, SCI, indicating the DD and SD vector associated with the strongest coefficient per layer,
- a polarization reference amplitude per layer,
- a bitmap for indicating the non-zero combining coefficients per layer, and
- possible additional parameter(s) associated with the DD basis subset indication.
4. The method of any preceding claim, wherein part 2 of the N_{REP} CSI reports is segmented into TN_{REP} CSI subgroups, wherein always T CSI subgroups are associated with a single CSI report, wherein each CSI subgroup is associated with a priority (priority level).
5. The method of any one of claim 1 to 3, wherein the part 2 of the N_{REP} CSI reports is segmented into TN_{REP} + 1 CSI subgroups, wherein always T CSI subgroups are associated with a single CSI report, and one CSI subgroup contains information associated with all N_{REP} CSI reports, wherein each CSI subgroup is associated with a priority (priority level), and the CSI subgroup that contains information associated with all N_{REP} CSI reports has the highest priority (priority level 0).
6. The method of any preceding claim, wherein, in case of omission, the UE drops the CSI subgroups with lower priority until the payload size of the CSI reports fits with the resource allocation from the gNB.
7. The method of claim 6, wherein, when omitting a CSI subgroup for a particular priority level, the UE omits all the CSI content at that priority level.
8. The method of any one of claims 4 to 7, wherein the parameter T indicates the number of CSI subgroups per CSI report and is related to the granularity of the CSI content that is omitted from a CSI report, wherein a high value of T indicates a high granularity and a low value of T indicates a low granularity, and wherein each CSI report is associated only with two CSI subgroups when the parameter T is given by the value of 2.
9. The method of claim 8, wherein the parameter T depends on the information content contained in a CSI report.
10. The method of claim 9, wherein the parameter T depends on the RI value indicated in the CSI report, wherein T = 2 if the RI value indicated in the CSI report is larger than a specific threshold value and T = 1 if the RI value indicated in the CSI report is smaller than a specific threshold value.
11. The method of claim 9, wherein the parameter T depends on the number of non-zero coefficients K_{NZ} indicated in the CSI report, wherein T = 2 if the number of non-zero coefficients indicated in the CSI report is larger than a specific threshold value, such as K_{NZ} > K̃_{NZ}, and T = 1 otherwise.
12. The method of any one of claims 5 to 9, wherein the first CSI subgroup associated with priority level 0 contains CSI information of all
N_{REP} CSI reports, and wherein the joint CSI subgroup contains the information of at least one of the following parameters:
- the selected SD basis subset indicator including, if configured, the rotation oversampling factors,
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers,
- the polarization reference amplitude value(s) for the RI layers,
- the bitmap(s) for indicating the K_{NZ} non-zero combining coefficients for the RI layers, and
- possible additional parameter(s) associated with the DD basis subset indication.
13. The method of any one of claims 4 or 6 to 11, wherein the CSI subgroup with highest priority of a CSI report contains at least the information of the following parameters:
- the selected SD basis subset indicator including, if configured, the rotation oversampling factors,
- the selected DD basis subset indicator(s) for the RI layers,
- the SCI(s) for the RI layers,
- the polarization reference amplitude value(s) for the RI layers, and
- the bitmap(s) for indicating the K_{NZ} non-zero combining coefficients for the RI layers, and
- possible additional parameter(s) associated with the DD basis subset indication.
14. The method of any one of claims 4 to 13, wherein each CSI subgroup, that contains information of a fraction of the combining coefficients, contains the amplitude values, or phase and amplitude values, associated with a maximum of
x > 1 combining coefficients of a CSI report, and wherein the remaining CSI subgroups with lower priority contain the remaining amplitude values, or remaining phase and amplitude values, of the CSI report.
15. The method of any one of claims 4 to 14, wherein each CSI subgroup that has the highest priority and is associated with a single CSI report contains at least the bitmaps associated with all SD components of one or more DD basis vectors, or one or more DD basis vector indices, for the RI layers of the precoding matrix indicated in the CSI report, and wherein the CSI subgroup contains the corresponding amplitude and/or phase information of the combining coefficients associated with the bitmaps.
16. The method of any one of claims 4 to 14, wherein bitmaps and the information of the combining coefficients for the RI layers of a CSI report are segmented into D segments, wherein each segment contains the sub-bitmap associated with all SD components of a single DD basis index for all RI layers of the precoding matrix, and wherein the CSI subgroup that has the highest priority and is associated with a single CSI report contains one or more of the D segments.
17. The method of claim 16, wherein each segment contains the associated amplitude and/or phase information of the combining coefficients for the RI layers associated with the sub-bitmap.
18. The method of any one of claims 4 to 17, wherein a CSI subgroup that is associated with the highest priority of a CSI report contains at least the bitmap(s) for indicating the non-zero combining coefficients for a first subset of the RI layers, and a first fraction of the amplitude and phase values of the selected non-zero delay-domain combining coefficients associated with the first subset of the RI layers of the precoding matrix indicated in the CSI report.
19. The method of any one of claims 5 to 18, wherein the CSI subgroup that contains CSI information associated with N_{REP} CSI reports, contains for one or more CSI reports, the SCI(s) for the RI layers, the fraction of the bitmaps and the amplitude and/or phase information of the combining coefficients associated with all SD components and DD basis vector index 0 for the RI layers of each CSI report.
20. The method of any one of claims 4 to 19, wherein, when the bitmaps of all RI layers of size 2UD × RI associated with a single CSI report are contained in a single CSI subgroup, the bitmap is segmented into RI segments, each of 2UD bits, in an increasing order with respect to the layer numbers, and each segment is associated with all SD and DD basis indices of a layer.
21. The method of any one of claims 4 to 20, wherein, when a CSI subgroup contains the bitmaps of all RI layers, the bitmap of size 2UD × RI is segmented into D segments, and each segment has a size of 2U × RI and is associated with a single DD basis vector index.
22. The method of claim 21, wherein each segment of size 2U × RI is segmented into RI segments, and each segment has a size of 2U × 1 and is associated with a single DD basis vector index and a single layer index.
23. The method of claim 20 to 22, wherein each CSI subgroup contains the amplitude and/or phase information of the combining coefficients associated with a portion of the bitmaps of the RI layers, or with a portion of the bitmaps of the D segments.
24. The method of any one of claims 4 to 21, wherein the bit-width of the CSI subgroup associated with a single CSI report and highest priority is fixed and given by
A +
B, where
A is the combined bit-width of all components that are contained in the CSI subgroup apart from the number of non-zero combining coefficients, and
B is the bit-width associated with the amplitude (
a) and phase information (
b) of a fraction of the combining coefficients
25. A user equipment, UE, comprising a processor and a memory, the memory containing computer program code executable by the processor whereby the UE is operative to perform any one of the subject matter of claims 1 to 24.
26. A computer program product comprising computer program code, which, when executed by a processor, enables the processor to perform any one of the subject matter of claims 1 to 24.