3GPP RAN1 #94 Meeting 會議報告
Transcript of 3GPP RAN1 #94 Meeting 會議報告
1
會議報告(會議類別:其他)
出席第三代合作夥伴計畫無線存取網路
3GPP RAN1 #94 Meeting會議報告
出席人員:羅立中、李建民、包偉丞、陳志軒
派赴地區:瑞典/哥德堡
會議期間:107年 8月 20日至 107 年 8月 24日
報告日期:107年 9月 26日
2
摘 要
本團隊出席在瑞典/哥德堡舉辦的無線存取網路第 1 工作組
(Radio Access Network working group #1, RAN1) # 94會議,本次會議
由 3GPP 歐洲友人(The European Friends of 3GPP)舉辦,包含:Apple、
BlackBerry、Deutsche Telekom、Ericsson、Huawei、Intel、InterDigital、
KPN、Motorola、Nokia、Orange、Qualcomm、SIMAlliance、Sony、
Telefonica、Telenor、Thales、TIM、Vodafone等公司。此次會議共有
約 600人註冊,實際到場人數約為 500人。本團隊依規劃有 4位成員
出席會議,此行主要任務在於參與第五代行動通訊(5th Generation
mobile communication, 5G)之新無線電技術(New Radio Technology, NR)
的相關技術討論,關注 5G NR標準制定的動向及各家公司提案方向,
以利掌握下世代通訊系統的趨勢和系統模擬平台的建構。並與其他大
廠接觸以討論合作項目。
本次會議是關於 NR 第 16版(Release 16, R16)開始的第一次起始
會議,內容除了先前就已經開始的研究項目,像是 NR非正交多重接
取(Non-Orthogonal Multiple Access, NOMA)、基於新無線電技術之非
授權頻譜存取(NR-based Access to Unlicensed Spectrum, NR-U)、NR整
之合存取與後端網路(Integrated Access and Backhaul, IAB)、以及 NR
車聯網外,同時新增了 NR 遠距干擾管理 (Remote Interference
Management)和 NR 之超高可靠性與低延遲通訊(Ultra Reliable and
Low Latency Communication, URLLC)於實體層的技術加強兩個研究
項目,預計下次會期將會增加多重輸入多重輸出 (Multiple Input
Multiple Output, MIMO)增強的工作項目。這次 NR MIMO方面主要的
討論在於規格修正與澄清,以及一些用詞與錯字修正。相較於去年緊
湊的議程,今年議程進行可以說較為人性化。Rel-16新的研究和工作
項目也在此會期陸續展開,建議可以及早投入。
此外,本次會議亦針對 NR在第 15版(Release 15, R15)之標準規
範於 2017 年年底宣告完成後,對於其細節尚未定義清楚的部分繼續
討論,主要參與 MIMO、載波聚合(Carrier Aggregation, CA)以及部分
3
頻寬(Bandwidth Part, BWP)等議題的討論。本團隊除了持續追蹤相關
技術發展外,亦持續關注 NR相關新研究項目的技術研發方向。
縮寫與中英文對照表
英文全稱 英文縮寫 中文全稱
3rd Generation Partnership Project 3GPP 第三代合作夥伴計畫
5th Generation mobile communication 5G 第五代行動通訊
Augmented Reality AR 擴增實境
backhaul 後端網路
Bandwidth Part BWP 部分頻寬
beam 波束
Beam failure detection 波束故障偵測
Beam failure recovery BFR 波束故障恢復
Beam Management 波束管理
bundling 綑綁
Channel Quality Indicator CQI 通道品質指標
Carrier Aggregation CA 載波聚合
Channel State Information CSI 通道狀態訊息
Channel State Information-Reference
Signals CSI-RS 通道狀態訊息參考訊號
Channel Status Information
Interference Measurement CSI-IM 通道狀態訊息干擾量測
Code Block Group CBG 碼塊群組
Codebook 編碼簿
Codebook based transmission for UL 基於碼簿之上行傳輸
codeword 碼字
Common search space 公共搜索空間
Configured grant 配置授權
contention free 免競爭
Contention window size CWS 競爭時窗大小
Control Resource Set CORESET 控制資源集合
Cross-Link Interference CLI 跨連結干擾
CSI processing unit CPU 通道狀態訊息處理單元
DeModulation Reference Signal DMRS 解調參考訊號
4
Discovery Reference Signal DRS 探索參考訊號
Down Link DL 下行
Downlink control information DCI 下行控制訊息
Downlink Feedback Indicator DFI 下行回傳指標
enhanced Mobile BroadBand eMBB 增強型行動寬頻
Frequency Division Multiplexing FDM 分頻多工
Guard Period GP 保護時間
Hybrid Automatic Repeat reQuest HARQ 混合型自動重傳請求
identification ID 識別
Integrated Access and Backhaul IAB 整合存取與後端網路
Interlace 交織
Liaison Statement LS 聯絡說明
Licensed-Assisted Access to
unlicensed spectrum LAA
授權頻譜輔助之非授權頻譜
存取
link-level simulation LLS 鏈結層模擬
Listen-Before-Talk LBT 先聽後說
Maintenance 維護
Medium Access Control Control
Element MAC CE 媒體存取控制之控制元件
Millimeter band 毫米波頻帶
Multi-antenna scheme 多天線方案
Multiple Input Multiple Output MIMO 多重輸入多重輸出
Multiplexing 多工
new generation NB gNB 新世代基地台
New Radio Multiple Input Multiple
Output NR MIMO
新無線多重輸入多重輸出技
術
New Radio Technology NR 新無線電技術
Non-ACKnowledgement NACK 否定應答
non-codebook 非編碼簿
Non-codebook based transmission for
UL 非基於碼簿之上行傳輸
Non-Orthogonal Multiple Access NOMA 非正交多重存取
Non-Zero Power NZP 非零功率
NR-based Access to Unlicensed
Spectrum NR-U
基於新無線電技術之非授權
頻譜存取
Paging 呼叫
Peak-to-Average Power Ratio PAPR 峰均功率比
5
Physical Downlink Control CHannel PDCCH 實體下行控制通道
Physical Downlink Share CHannel PDSCH 實體下行共用通道
Physical Random Access CHannel PRACH 實體隨機存取通道
Physical Resource Block PRB 實體資源區塊
Physical Uplink Control Channel PUCCH 實體上行控制通道
Physical Uplink Share Channel PUSCH 實體上行共用通道
port 埠
Preamble 前置符元
Precoder 預編碼器
precoding matrix indicator PMI 預編碼陣列指示
Primary Secondary Cell PSCell 主要次細胞
Quasi-Colocated QCL 準同位
Radio Access Network working group
#1 RAN1 無線存取網路第 1工作組
Radio Access Network working group
#2 RAN2 無線存取網路第 2工作組
Radio Access Network working group
#3 RAN3 無線存取網路第 3工作組
Radio Resource Control RRC 無線電資源控制
Random Access Response RAR 隨機存取回應
rank 級別
Reference Signal RS 參考訊號
Release 14 R14 第 14版
Release 15 R15 第 15版
Release 16 R16 第 16版
Remaining Minimum System
Information RMSI 其餘最少所需系統資訊
Remote Interference Management RIM 遠端干擾管理
resource set 資源集合
search space 搜尋空間
Secondary Cell SCell 次細胞
Semi-Persistent Channel State
Information SP-CSI 半持續性通道狀態訊息
slot 槽
Sounding Reference Signal SRS 探測參考訊號
Sounding Reference Signal resource SRS resource 探測參考訊號資源
6
Sounding Reference Signal resource
set
SRS resource
set 探測參考訊號資源集合
SRS Resource Index SRI 探測參考訊號資源指標
StandAlone SA 獨立運作
SubCarrier spacing SCS 子載波間距
Synchronization Signal Block SSB 同步訊號區塊
Tracking Reference Signal TRS 追蹤參考信號
Transmission Configuration Indication TCI 傳送配置指示
Transmission scheme 傳送機制
transport block TB 傳輸區塊
Ultra Reliable and Low Latency
Communication URLLC 超高可靠性與低延遲通訊
Unlicensed band 非授權頻帶
Unlicensed carriers 非授權頻譜載波
Up Link UL 上行
Up Link grant UL grant 上行調度
Uplink control information UCI 上行控制信息
User Equipment UE 用戶設備
Virtual Reality VR 虛擬實境
Wi-Fi 無線區域網路
技術貢獻
本團隊目前主要鎖定 NR MIMO議題進行研究,目標是相關議題
的討論以及尋找其他可研究的題目,並與其他大廠接觸以討論合作
項目並參與技術協商及立場表達為此次與會主要目的。本次會議共
提出 1件技術貢獻。
本團隊目前主要鎖定 NR 的 BWP 存取技術、NR MIMO 議題進
行研究、非授權頻帶存取技術以及 URLLC 於實體層的技術加強等
相關議題進行研究,此次會議並參與上述議題的討論並提出兩篇相
關的技術貢獻論文。
會議解說
1. 多重輸入多重輸出(MIMO)
7
NR MIMO這次會期主要仍然是在規格的修訂與文字的澄清討論,
在 Multi-antenna scheme 方面,主要討論了先前的 TB 關閉功能以及
基於 codebook 和 non-codebook 的上行傳輸相關文字修訂。在通道狀
態訊息(CSI)獲取方面,主要討論了預編碼陣列指示(PMI)回報傳送機
制的一些無線電資源控制(RRC)開銷問題、下行控制訊息(DCI)啟動多
個識別 CSI回報問題、通道和干擾量測資源的準同位(QCL)澄清問題、
通道狀態訊息處理單元(CPU)佔用問題和半持續性 CSI 回報問題等。
在 beam management 方面,主要討論了 RAN2 關於波束故障恢
復(BFR)問題的回覆、非週期性通道狀態訊息參考訊號(CSI-RS)排程
低於用戶裝置(UE)能力時討論、跨載波排程的傳送配置指示(TCI)澄
清以及是否擴充 TCI狀態的討論。
此項目討論在毫米波頻帶的多天線技術以維持細胞和手機間的
連接提供更高的頻譜效率。會議中此項目的議程安排分為以下幾個討
論大綱:
Maintenance for Multi-antenna scheme
Maintenance for CSI acquisition
Maintenance for beam management
Maintenance for Reference signals and QCL
2. 載波聚合(CA)與部分頻寬(BWP)
此項目討論 CA 以及 BWP 技術於 NR R15 的未完成部分的細部
討論。
3. 新無線電技術(NR)超高可靠性與低延遲通訊(URLLC)實體層增強
之研究
此項目討論 URLLC於實體層的技術加強,預期進一步增強 R15
版本中 URLLC的技術。會議中此研究項目的議程安排分為以下幾個
討論大綱:
Layer 1 enhancements
8
UL inter UE Tx prioritization/multiplexing
Enhanced UL grant-free transmissions
4. 新無線電技術(NR)之遠端干擾管理技術研究
此項目討論遠端干擾管理技術,用以基地台偵測來自遠端的基地台干
擾,並進一步發展基地台間的協調機制,與干擾消除或通報等管理技
術。會議中此研究項目的議程安排分為以下幾個討論大綱:
Mechanisms for improving network robustness
Mechanisms for identifying strong gNB interferers
5. 基於新無線電技術之非授權頻譜存取(NR-U)
此項目討論 NR-U的模擬設定、架構設計、實體層通道設計和實
體層程序設計。會議中此研究項目的議程安排如下:
Study on NR-based Access to Unlicensed Spectrum
Simulation Methodology for NR-U operation
Frame structure for NR-U operation
Potential physical layer channel design
Potential physical layer procedures
Others
9
目 錄
摘 要 ........................................................................................................... 2
一、會議名稱........................................................................................... 10
二、參加會議目的及效益 ...................................................................... 10
三、會議時間........................................................................................... 10
四、會議地點........................................................................................... 10
五、會議議程........................................................................................... 10
六、會議紀要........................................................................................... 12
七、心得與建議 ...................................................................................... 56
10
一、會議名稱
3GPP RAN1 #94 Meeting
二、參加會議目的及效益
參與在瑞典/哥德堡舉辦的 3GPP RAN1 # 94會議,本計畫團隊主要
關注的焦點是參與 5G新無線電技術中多天線技術、非授權頻帶存取技
術以及 URLLC 技術的討論及尋找可研究的題目,同時關注 5G NR 標
準制定的動向及各家公司提案方向,以利掌握下世代通訊系統的趨勢和
系統模擬平台的建構。
三、會議時間
Aug. 20, 2018 ~ Aug. 24, 2018
四、會議地點
瑞典/哥德堡, Gothia Towers
五、會議議程
RAN1 #94會議議程如下:
1 Opening of the Meeting
2 Approval of Agenda
3 Highlights from Ran Plenary
4 Approval of Minutes from Previous Meeting
5 Incoming Liaison Statements
6 E-UTRA
7 NR
7.1 Maintenance of Release 15 NR
7.1.1 Maintenance for Initial access and mobility
7.1.2 Maintenance for MIMO
7.1.3 Maintenance for Scheduling/HARQ Aspects
11
7.1.4 Maintenance for NR-LTE Co-Existence
7.1.5 Maintenance for UL Power Control
7.1.6 Maintenance for URLLC
7.1.7 NR Release 15 UE features
7.1.8 Other
7.2 NR in Release 16
7.2.1 Study on non-orthogonal multiple access for NR
7.2.2 Study on NR-based Access to Unlicensed Spectrum
7.2.3 Study on Integrated Access and Backhaul for NR
7.2.4 Study on NR V2X
7.2.5 Study on Remote Interference Management for NR
7.2.6 Study on Physical Layer Enhancements for NR URLLC
7.2.7 Study on Self Evaluation towards IMT-2020 Submission
8 Closing of the Meeting
會議進行之會議室與時間安排如下:
未來會議時間與地點如下:
Meeting Date Venue
RAN1#94-Bis 8 - 12 Oct 2018 Chengdu
RAN1#95 12 - 16 Nov 2018 Spokane
RAN1-AH-1901 21 - 25 Jan 2019 Taipei
RAN1#96 25 Feb - 1 Mar 2019 Athens
12
六、會議紀要
1. 會議摘要
本次會議由“3GPP的歐洲友人(The European Friends of 3GPP)”主
辦,成員公司有Apple, BlackBerry, Deutsche Telekom, Ericsson, Huawei,
Intel, InterDigital, KPN, Motorola, Nokia, Orange, Qualcomm, SIMAlliance,
Sony, Telefonica, Telenor, Thales, TIM, Vodafone等,共約近有500人參加。
本計畫團隊依規劃有4位成員出席參加RAN1 #94會議。
台灣派員參與此次會議之公司/機構有 CHTTL、ITRI、III、MediaTek、
ASUS等 5家。
2. NR MIMO技術之相關重要進展
5G NR 中主要特徵技術即是採用大規模 MIMO,並且將運行頻率推昇
到 millimeter band 等級。NR MIMO議題是基於 R14之垂直波束成形和
全維度MIMO運行在高頻段所遇到的問題作探討。
以下就 multi-antenna scheme 各個子議題進行討論:
Maintenance for Multi-antenna scheme
在上次會期決議了可以利用將(IMCS, rvid)設為(26,1),把兩個傳輸
區塊(TB)中的一個給關掉,不過在 TS38.214中並未將此功能描
述清楚,主要模糊地帶在於是否能夠利用將兩個 TB的(IMCS, rvid)
都設為(26,1)關掉兩個 TB,因此即使沒有資料要傳送,也能夠
啟動非週期性 SRS。最後決議是此功能只能用於 2個 codeword
下關掉其中一個 TB,若兩個 TB 的(IMCS, rvid)都設為(26,1),那
麼還是一個正常的配置,並不會關掉兩個 TB,詳細決議如下:
Codebook based transmission for UL
13
當 UE沒有被配置到上行 transmission scheme,則 UE不期待藉
由 DCI格式 0_1 進行上行排程,另外當 UE的上行排程是根據
DCI 格式 0_0,則 PUSCH 傳輸則是考慮只有一個天線埠的情
形。另一方面,由於 codebook based transmission for UL以及 non-
codebook based transmission for UL 除了使用 UL grant 相關的
DCI指示運作外,也可藉由 configured grant配置觸發,而此運
作並未描述在規格書裡。具體規格文字調整決議如下:
Agreement
Text proposal for 38.214:
>>>>>>>>>>>> Start text proposal 1a >>>>>>>>>>>>
6.1.1 Transmission schemes
Two transmission schemes are supported for PUSCH: codebook based
transmission and non-codebook based transmission. The UE is configured
with codebook based transmission when the higher layer parameter txConfig
in PUSCH-Config is set to 'codebook', the UE is configured non-codebook
based transmission when the higher layer parameter txConfig is set to
'nonCodebook'. If the higher layer parameter txConfig is not configured, the
UE is not expected to be scheduled by DCI format 0_1. the PUSCH
transmission is based on one PUSCH antenna port, which is triggered by
DCI format 0_0. If PUSCH is scheduled by DCI format 0_0, the PUSCH
transmission is based on a single antenna port.
>>>>>>>>>>>> End text proposal 1a >>>>>>>>>>>>
>>>>>>>>>>>> End text proposal 1b >>>>>>>>>>>>
6.1.1.1 Codebook based UL transmission
For codebook based transmission, PUSCH can be scheduled by DCI format
0_0, DCI format 0_1 or semi-statically configured to operate according to
Subclause 6.1.2.3. If this if PUSCH is scheduled by DCI format 0_1, the
UE determines its PUSCH transmission precoder based on SRI, TPMI and
the transmission rank from the DCI, given by DCI fields of SRS resource
indicator and Precoding information and number of layers in subclause
7.3.1.1.2 of [TS 38.212]., where the The TPMI is used to indicate the
precoder to be applied over the antenna ports {0…ν-1} and that corresponds
to the SRS resource selected by the SRI when multiple SRS resources are
configured, or if a single SRS resource is configured TPMI is used to
14
indicate the precoder to be applied over the antenna ports {0…ν-1} and that
corresponds to the SRS resource. The transmission precoder is selected from
the uplink codebook that has a number of antenna ports equal to higher layer
parameter nrofSRS-Ports in SRS-Config, as defined in Subclause 6.3.1.5 of
[4, TS 38.211]. When the UE is configured with the higher layer parameter
txConfig set to 'codebook', the UE is configured with at least one SRS
resource. The indicated SRI in slot n is associated with the most recent
transmission of SRS resource identified by the SRI, where the SRS resource
is prior to the PDCCH carrying the SRI before slot n.
>>>>>>>>>>>> End text proposal 1b >>>>>>>>>>>>
>>>>>>>>>>>> Start text proposal 1c >>>>>>>>>>>>
6.1.1.2 Non-Codebook based UL transmission
For non-codebook based transmission, PUSCH can be scheduled by DCI
format 0_0, DCI format 0_1 or semi-statically configured to operate
according to Subclause 6.1.2.3. the The UE can determine its PUSCH
precoder and transmission rank based on the wideband SRI when multiple
SRS resources are configured, where the SRI is given by the SRS resource
indicator in DCI according to subclause 7.3.1.1.2 of [5, 38.212], or the SRI is
given by srs-ResourceIndicator according to subclause 6.1.2.3. The UE shall
use one or multiple SRS resources for SRS transmission, where the number
of SRS resources which can be configured to the UE for simultaneously
transmission in the same RBs is a UE capability. Only one SRS port for each
SRS resource is configured. Only one SRS resource set can be configured
with higher layer parameter usage in SRS-ResourceSet set to 'nonCodebook'.
The maximum number of SRS resources that can be configured for non-
codebook based uplink transmission is 4. The indicated SRI in slot n is
associated with the most recent transmission of SRS resource identified by
the SRI, where the SRS resource is prior to the PDCCH carrying the SRI
before slot n.
>>>>>>>>>>>> End text proposal 1c >>>>>>>>>>>>
由於 codebook based transmission for UL除了使用 UL grant相關
DCI指示運作外,也可藉由 configured grant配置觸發,而此運
作並未描述在規格書裡。具體規格文字調整決議如下:
Agreement
15
Text proposal for TS38.211
6.3.1.5 Precoding
The block of vectors Tiyiy )(...)( )1()0( , 1,...,1,0layersymb Mi shall be
precoded according to
)(
)(
)(
)(
)1(
)0(
)(
)(
1
0
iy
iy
W
iz
iz
p
p
where 1,...,1,0 apsymb Mi ,
layersymb
apsymb MM . The set of antenna ports 1,...,0 pp
shall be determined according to the procedure in [6, TS 38.214].
For non-codebook-based transmission, the precoding matrix W equals the
identity matrix.
For codebook-based transmission, the precoding matrix W is given by
1W for single-layer transmission on a single antenna port, otherwise by
Tables 6.3.1.5-1 to 6.3.1.5-7 with the TPMI index obtained from the DCI
scheduling the uplink transmission or the higher layer parameters according
to the procedure in [6, TS 38.214].
<unrelated part omitted>
因為規格38.214裡的專有名詞關於 codebook選擇限制與38.331
裡的專有名詞不一致,所以修正規格 38.214的專有名詞,使之
與規格 38.331 裡的內容一致,實際規格書文字調整決議如下:
Agreement
Text proposal for TS38.214
6.1.1.1 Codebook based UL transmission
<unrelated part omitted>
For codebook based transmission, the UE determines its codebook subsets
based on TPMI and upon the reception of higher layer parameter
codebookSubset in PUSCH-Config which may be configured with '
fullyAndPartialAndNonCoherent ', or 'partialAndNonCoherent' , or
'nonCoherent' depending on the UE capability. The maximum transmission
rank may be configured by the higher parameter maxRank in PUSCH-
Config .
16
A UE reporting its UE capability of 'partialAndNonCoherent' transmission
shall not expect to be configured by codebookSubset with '
fullyAndPartialAndNonCoherent '.
A UE reporting its UE capability of ' nonCoherent ' transmission shall not
expect to be configured by codebookSubset with '
fullyAndPartialAndNonCoherent ' or with 'partialAndNonCoherent '.
A UE shall not expect to be configured with the higher layer parameter
codebookSubset set to 'partialAndNonCoherent' when higher layer parameter
nrofSRS-Ports in an SRS-ResourceSet with usage set to ' codebook ' indicates
that two SRS antenna ports are configured.
For codebook based transmission, the UE may be configured with a single
SRS-ResourceSet set to ' codebook ' and only one SRS resource can be
indicated based on the SRI from within the SRS resource set. The maximum
number of configured SRS resources for codebook based transmission is 2. If
aperiodic SRS is configured for a UE, the SRS request field in DCI triggers
the transmission of aperiodidc SRS resources.
When multiple SRS resources are configured by SRS-ResourceSet with
usage set to ' codebook ', the UE shall expect that higher layer parameters
nrofSRS-Ports in SRS-ResourceSet in SRS-ResourceSet shall be configured
with the same value for all these SRS resources.
Non-codebook based transmission for UL
因為規格 38.214 裡的專有名詞關於 SRS resource set 的用詞與
38.331裡的專有名詞不一致,所以調整規格 38.214的專有名詞,
使之與規格 38.331裡的內容一致,實際規格書文字調整決議如
下:
Agreement:
Agree this TP to 6.1.1.2 of 38.214:
----------Start of TP ----------------------------------------
For non-codebook based transmission, the UE does not expect to be
configured with both spatialRelationInfo for SRS resource and associatedCSI-
RS in SRS-ConfigSRS-ResourceSet for SRS resource set.
For non-codebook based transmission, the UE can be scheduled with DCI
format 0_1 when at least one SRS resource is configured in SRS-ResourceSet
with usage set to 'nonCodebook'.
17
--------------End of TP ---------------------------------
關於非週期 SRS 傳輸觸發,應該是指示觸發狀態與被觸發的
SRS resource set之間的配對關係。而此運作並未準確描述在規
格書裡。具體規格文字調整決議如下:
Agreement
Adopt the following TP for Subclause 6.1.1.2 of TS38.214:
----Start of TP ---------------------------------------
- If aperiodic SRS resource set is configured, the associated NZP-CSI-RS
is indicated via SRS request field in DCI format 0_1 and 1_1, where
AperiodicSRS-ResourceTrigger (indicating the association between
aperiodic SRS triggering state and SRS resource sets), triggered SRS
resource(s) srs-ResourceSetId, csi-RS (indicating the associated NZP-
CSI-RS-ResourceId) are higher layer configured in SRS-ResourceSet. A
UE is not expected to update the SRS precoding information if the gap
from the last symbol of the reception of the aperiodic NZP-CSI-RS
resource and the first symbol of the aperiodic SRS transmission is less
than 42 OFDM symbols.
--------------End of TP ------------------------------------
UE接收的 DCI時槽包含觸發非週期 SRS傳輸當下,且非週期
的 CSI-RS 與非週期 SRS 傳輸有相關聯時,此時排程中的 TCI
狀態不會被配置類別 D 的 QCL。而此運作並未描述在規格書
裡。具體規格文字調整決議如下:
Agreement
Adopt the following TP for Subclause 6.1.1.2 of TS38.214:
-----------Start of TP ------------------------------
- If the UE configured with aperiodic SRS associated with aperiodic NZP
CSI-RS resource, the presence of the associated CSI-RS is indicated by
the SRS request field if the value of the SRS request field is not '00' as
in Table 7.3.1.1.2-24 of [5, TS 38.212]. The CSI-RS is located in the
same slot as the SRS request field. If the UE configured with aperiodic
18
SRS associated with aperiodic NZP CSI-RS resource, any of the TCI
states configured in the scheduled CC shall not be configured with
'QCL-TypeD'.
---End of TP ---------------------------------------------------------------
由於 non-codebook based transmission for UL除了使用 UL grant
相關 DCI指示運作外,也可藉由 configured grant配置觸發,而
此運作並未描述在規格書裡。具體規格文字調整決議如下:
Agreement
Agree the TP to 6.1.1.2 of 38.214
---------------Start of TP --------------------------------------------
6.1.1.2 Non-Codebook based UL transmission
<unrelated part omitted>
The UE shall perform one-to-one mapping from the indicated SRI(s) to the
indicated DM-RS ports(s) given by DCI format 0_1 or
ConfiguredGrantConfig according to subclause 6.1.2.3 in increasing order.
<unrelated part omitted>
----------End of TP --------------------------------------------------------------
在規格中,UE 接收來 DCI 裡的 SRI 缺乏參照來源,於是文字
補充連結到 38.212 裡關於 SRI 的相關細節。另外 SRI 也可從
configured grant 相關設定來指派,而此運作並未描述在規格書
裡。根據以上兩點,實際規格書裡文字修改決議如下:
Agreement
Confirm the working assumption of TP to 38.214, Section 6.1.1.2:
------------------Start of TP for 38.214---------------------------------
6.1.1.2 Non-Codebook based UL transmission
For non-codebook based transmission, the UE can determine its PUSCH
precoder and transmission rank based on the wideband SRI given by SRS
resource indicator field from the DCI when multiple SRS resources are
configured, where the SRI is given by the SRS resource indicator in DCI
according to subclause 7.3.1.1.2 of [TS 38.212], or the SRI is given by srs-
19
ResourceIndicator according to subclause 6.1.2.3.
-----------------------End of TP ---------------------------------------
因為規格 38.214 裡的專有名詞關於 SRS resource set 的用詞與
38.331裡的專有名詞不一致,所以調整規格 38.214的專有名詞,
使之與規格 38.331裡的內容一致,另外修正錯別字。實際規格
書文字調整決議如下:
Agreement
Agree this TP to 6.2.1 of 38.214:
-----------------Start of TP ---------------------------------
6.2.1 UE sounding procedure
The UE can be configured with one or more Sounding Reference Symbol
(SRS) resource sets as configured by the higher layer parameter SRS-
ResourceSet. For each SRS resource set, a UE may be configured with 1K
SRS resources (higher laterlayer parameter SRS-Resource), where the
maximum value of K is indicated by [SRS_capability [13, 38.306]]. The SRS
resource set applicability is configured by the higher layer parameter SRS-
SetUseusage in SRS-ResourceSet. When the higher layer parameter
usageSRS-SetUse is set to 'BeamManagement', only one SRS resource in
each of multiple SRS sets can be transmitted at a given time instant. The SRS
resources in different SRS resource sets can be transmitted simultaneously.
For aperiodic SRS at least one state of the DCI field is used to select at least
one out of the configured SRS resource set
…
--------------End of TP ------------------------------------------
在 PRB bundling方面,也修正一些名詞,詳細決議如下:
20
Maintenance for CSI acquisition
首先是非 PMI回報下,Precoder也是需要根據 rank標準化:
目前對於非 PMI 回報配置的最大資源數量與 RRC 參數
maxNrofNZP-CSI-RS-ResourcesPerConfig相關,最大可達 128,
而當網路配置非 PMI回報時,RRC需要配置大量的 CSI-RS資
源的 port指標,造成大量的傳輸資源需求,因此有公司建議應
該要降低這個數目,不過有公司認為目前不適宜更改 RRC 參
數,因此用折衷的方式,採用預設值的方式希望避免這種情況
發生,並且會在 RAN1的相關規格上加上不超過 64個NZP CSI-
RS的限制,不過並不會去修正相關的 RAN2 規格:
21
目前標準允許一個 DCI可以啟動多個回報,不過有可能發生兩
個 CSI 回報有著同樣的回報 ID 連結到不同的 resource set,在
這個情況下同樣 ID 的 CSI 回報如何對應到 UCI 的優先順序目
前仍然沒有定義,為了簡便起見,最後決議排除一個 DCI可以
啟動同樣 ID的 CSI回報的情形:
以下是為了澄清通道量測和干擾量測的資源是 QCL條件下,只
有在執行 CSI-IM時才成立,若干擾量測資源在 NZP CSI-RS上
執行,那麼所有位於用作干擾量測NZP CSI-RS資源組中的NZP
CSI-RS資源都會和通道量測的 NZP CSI-RS QCL:
22
以下修正是為了澄清當 CSI回報是 Type II時,resource set裡不
會配置超過一個 CSI-RS資源:
以下修正是為了澄清 UE 不會在同一個 slot 裡面收到多個非週
期性的 CSI 要求,且澄清了 UE 不會在一個 slot 裡面傳送多個
含有非週期性的 CSI回報:
由於目前關於通道狀態訊息處理單元(CPU)佔用定義問題,導
致在基於週期性和半持續性 CSI-RS 的非週期性 CSI 回報時,
會出現 CPU 會在啟動非週期性 CSI 回報前就已經被佔用的情
況,因此針對這種情境需要修正 CPU佔用的定義,詳細決議如
下:
以下決議是修正關於上行和下行的子載波間距(SCS)不同時,
23
CSI回報的 CSI-RS資源參考位置:
以下決議是為了澄清當兩個 CSI回報碰撞時,只有在使用不同
的實體層通道傳送時才會直接丟棄,否則將按照 UCI複用的步
驟決定優先性:
對於 15kHz的 SCS,(Z1,Z1’)的 CSI時間要求決議如下:
24
在 MAC CE啟動延遲方面,有公司認為考慮到 NR有基於 CBG
的 HARQ傳輸機制,因此需要額外考慮這個狀況在規格中定義,
不過最後的結論是目前規格已經夠清楚不需要額外修訂:
目前的規格允許啟動多個 SP-CSI 回報,而每個 SP-CSI 回報有
著不同的偏移,以下的規格修正是針對這個機制進行一些限制,
避免重新啟動同一個 ID的 SP-CSI回報,因為這將會複雜化 CSI
回報時間和 CPU佔用問題:
關於 SP-CSI啟動和反啟動的欄位對應方面,有公司提出可以利
用全 0的欄位設置來反啟動所有的 SP-CSI,避免一個一個反啟
動以節省開銷,不過其他公司認為目前不需要這種最佳化的額
外作法,並且也做了相關的規格澄清:
25
以下修正是為了與先前會議的決議一致:在 UCI中的 CQI欄位
的數量只從 rank限制決定:
Maintenance for beam management
這次會期在 BFR方面,首先是回覆了 RAN2 有關於在 UE傳送
PRACH之後監控 PDCCH以及 BFR成功的條件:
在上次的會期有做成工作假設,若未配置 BFR的 CORESET或
是 BFR 的 search space,那麼 contention free 的 BFR 將無法執
行,有公司嘗試修正在規格中的相關敘述,不過由於有公司認
為目前 38.331 已經有定義清楚了,因此建議將以下敘述刪除:
26
目前的規格將每個 BWP有兩個 beam failure detection資源和先
前的工作假設:若有兩個資源指標在 TCI狀態中,那麼有 QCL
Type-D的資源將用來作 beam failure detection的概念混在一起,
因此有了以下修正:
在 beam management 方面,首先,若排程非週期性 CSI-RS 低
於 UE能力回報的值,那麼這個非週期性的 CSI-RS資源的 QCL
假設是什麼尚未有相關決議,一個方式是允許基地台做這種排
程,不過僅限於 CSI獲取用途的非週期性 CSI-RS,另一種方式
是不允許基地台做這種排程,最後是採取前者的方式,不過
QCL的假設參照仍需要進一步討論:
關於跨載波排程時,當 DCI有 TCI時,那麼這個 TCI指的是排
程的載波的 PDSCH還是被排程載波的 PDSCH,最後決議是指
的是被排程載波的 PDSCH:
因為MAC CE重新配置 TCI狀態有延遲問題,導致在新的 TCI
狀態生效前的 PDCCH內所指示的 TCI狀態未明(如下圖所示),
因此以下決議決定在這個狀況下,TCI 狀態仍要根據被排程的
那個 slot的 TCI狀態狀態而定:
27
Scheduling PDCCH(TCI = 0)
PDSCH is QCLed with TCI state index 0:Whether TCI state index 0 is based on old candidate
TCI states or new reconfigured TCI states?
time
ACK for MAC CEMAC to reconfigure candidate TCI states
3ms
在先前的會議中有決議透過 DCI format 0_0 排程的 PUSCH 的
QCL資訊由 PUCCH獲取,不過由於透過 DCI format 0_0 排程
的 PUSCH的細胞上有可能沒配置 PUCCH,在這個情況下就無
法獲取 QCL 資訊,一個可能性是 QCL 資訊由另一個細胞的
PUCCH獲取,另一個可能性就是限制沒有配置 PUCCH的細胞
上,不能透過 DCI format 0_0排程的 PUSCH,最後決議是採用
者,詳細決議如下:
以下修正是由於 CSI-ResourceConfig 這個參數在 38.214 和
38.331的定義敘述有出入:
目前最多可以配置 64個 TCI狀態,同時也決定支援最高 64個
動態的波束指配,由於後續對於 TCI狀態配置的決議,導致 TRS
和其他的參考信號與通道無法共用 TCI狀態,因此決議要提高
可配置的 TCI狀態數量到 128個:
28
3. 載波聚合(CA)與部分頻寬(BWP)
首先,針對 RAN2 所送來 LS 進行討論。此 LS 包含兩個部分,第
一部分針對 SSB 以及相對應 RMSI 的 multiplexing 方式為 pattern 2 以
及 pattern 3(如下圖所示)時的量測行為有了以下的問題:
Issue #1: In case multiplexing pattern 2 or 3 is used for the initial DL
BWP, for an active DL BWP that overlaps with the initial DL BWP but
not with the SS/PBCH block associated to the initial DL BWP, is it
possible for the UE to:
o Question #1a: Perform BM based on the SS/PBCH block associated
to the initial DL BWP?
o Question #1b: Perform BFD based on the SS/PBCH block
associated to the initial DL BWP?
o Question #1c: Perform RLM based on the SS/PBCH block
associated to the initial DL BWP?
o Question #1d: Monitor the SS/PBCH block associated to the initial
DL BWP to receive broadcast SI / paging based on CORESET#0 and
searchspace#0? (assuming the same SCS is used in the active BWP
as that of initial BWP)
29
SS
/PB
CH
Blo
ck
CO
RE
SE
T
PD
SC
H
time
frequency
SS
/PB
CH
Blo
ck
CO
RE
SE
T
PD
SC
H
frequency
time
Pattern 2 Pattern 3
圖一:Multiplexing pattern 2及 pattern 3示意圖。
此部分議題牽扯到 UE頻寬能力以及射頻與基頻的調整行為,考慮
到 UE複雜度問題,此部分的回應主要以簡化 UE行為為基準。相關決
議如下:
Agreements:
Issue #1: In case multiplexing pattern 2 or 3 is used for the initial DL BWP, for
an active DL BWP that overlaps with the initial DL BWP but not with the
SS/PBCH block associated to the initial DL BWP, is it possible for the UE to:
o Question #1a: Perform BM based on the SS/PBCH block associated to the
initial DL BWP?
Answer: No, a UE is not expected to perform BM based on the
SS/PBCH block outside the active DL BWP. Note that the
scenario can be supported with appropriate CSI-RS
configuration.
o Question #1b: Perform BFD based on the SS/PBCH block associated to the
initial DL BWP?
Answer: No, a UE is not expected to perform BFD based on the
SS/PBCH block outside the active DL BWP. Note that the
30
scenario can be supported with appropriate CSI-RS
configuration.
o Question #1c: Perform RLM based on the SS/PBCH block associated to the
initial DL BWP?
Answer: No, a UE is not expected to perform RLM based on the
SS/PBCH block outside the active DL BWP. Note that the
scenario can be supported with appropriate CSI-RS
configuration.
o Question #1d: Monitor the SS/PBCH block associated to the initial DL BWP
to receive broadcast SI / paging based on CORESET#0 and searchspace#0?
(assuming the same SCS is used in the active BWP as that of initial BWP)
Answer: No, a UE is not expected to monitor the SS/PBCH
block outside the active DL BWP and associated to the initial DL
BWP to receive broadcast SI / paging based on CORESET#0 and
searchspace#0.
第二部份議題與使用 BWP傳送系統訊息相關細節有關,RAN2 希
望釐清當UE目前啟動的BWP其資源並未與初始BWP有任何重疊時,
UE在接收系統訊息時是否需要與接收 SSB的行為有關,細節如下:
Issue #2: For an active DL BWP that overlaps with neither the initial
DL BWP nor with the SS/PBCH block associated to the initial DL
BWP (from now on: "non-overlapping" active DL BWP), RAN2
understands that additional CORESET/search space(s) need to be
configured in the active DL BWP to provide SI broadcast and paging
to a connected UE. What is not clear is whether an SS/PBCH block
necessarily needs to be associated to the additional CORESET/search
space(s) and, specifically, if a UE needs to monitor such SS/PBCH
block.
o Question #2a: In a "non-overlapping" active DL BWP, does a UE
necessarily need to monitor an SS/PBCH block associated to the
additional CORESET/search space to be able to receive SI
broadcast? If yes, could this be the SS/PBCH block associated to the
initial DL BWP? And under which conditions, if any, could the UE
monitor it for this purpose?
o Question #2b: In a "non-overlapping" active DL BWP, does a UE
necessarily need to monitor an SS/PBCH block associated to the
additional CORESET/search space to be able to receive paging? If
31
yes, could this be the SS/PBCH block associated to the initial DL
BWP? And under which conditions, if any, could the UE monitor it
for this purpose?
由於目前每一 CORESET皆已定義相對應的傳送配置指示訊息,為
了使 UE 行為一致,即使接收系統訊息,UE 仍依循目前所規範的方式
運行即可,。相關決議如下:
Agreements:
Issue #2: For an active DL BWP that overlaps with neither the initial DL BWP
nor with the SS/PBCH block associated to the initial DL BWP (from now on:
"non-overlapping" active DL BWP), RAN2 understands that additional
CORESET/search space(s) need to be configured in the active DL BWP to
provide SI broadcast and paging to a connected UE. What is not clear is whether
an SS/PBCH block necessarily needs to be associated to the additional
CORESET/search space(s) and, specifically, if a UE needs to monitor such
SS/PBCH block.
o Question #2a: In a "non-overlapping" active DL BWP, does a UE
necessarily need to monitor an SS/PBCH block associated to the
additional CORESET/search space to be able to receive SI broadcast?
If yes, could this be the SS/PBCH block associated to the initial DL
BWP? And under which conditions, if any, could the UE monitor it for
this purpose?
Answer: No, a UE does not necessarily need to monitor an
SS/PBCH block associated to the additional CORESET/search
space to be able to receive SI broadcast. The UE can be
configured with TCI states for the additional CORESET/search
space to enable SI broadcast reception.
o Question #2b: In a "non-overlapping" active DL BWP, does a UE
necessarily need to monitor an SS/PBCH block associated to the
additional CORESET/search space to be able to receive paging? If yes,
could this be the SS/PBCH block associated to the initial DL BWP?
And under which conditions, if any, could the UE monitor it for this
purpose?
Answer: No, a UE does not necessarily need to monitor an
SS/PBCH block associated to the additional CORESET/search
space to be able to receive paging. The UE can be configured
with TCI states for the additional CORESET/search space to
enable paging reception.
32
對於一 PSCell以及一 SCell,為了增加基地台對於資源配置的彈性,
其初始 DL BWP 的資源大小可以被重新配置,不再是固定為
CORESET#0的大小。為此,RAN1#92bis中對於 UE於 common search
space搜尋 DCI時的定義需要做相對應的調整,細節如下:
Agreements:
Modify the agreements in RAN1#92bis as follows.
o When monitoring for DCI in a BWP, the size of DCI format 0-0/1-0 is
given by
For format 0-0/1-0 (regardless of RNTI) in CSS, the size is given
by the initial DL BWPCORESET#0
For format 0-0/1-0 in USS, the size is given by the active BWP as
long as the DCI size budget is fulfilled
Otherwise, for format 0-0/1-0, the size is given by the
initial DL BWPCORESET#0
o For DCI format 1-0 in CSS with P-RNTI, SI-RNTI, RA-RNTI, C-
RNTI, CS-RNTI, or TC-RNTI:
the RB numbering for the scheduled PDSCH starts from the
lowest RB in the CORESET the DCI is received in
the maximum number of RBs possible to indicate in the DCI is
given by the size of the initial DL BWPCORESET#0.
o Payload sizes for 2-2 and 2-3 are padded (if needed) to match the size
of formats 0-0/1-0 as defined by the initial BWP in CSS
Modify the agreements in RAN1#91 as follows if Option #2 is supported in
PCell
o CORESET configured by RMSI is confined within the initial DL
BWPthe bandwidth of CORESET#0
For PCell, the initial DL BWP can be configured in SIB1 to be the same as or
different with the initial DL BWP as initially defined by CORESET#0
o The initial DL BWP configured in SIB1 includes the bandwidth of
CORESET#0
o If the initial DL BWP configured by SIB1 is different with the initial
DL BWP as initially defined by CORESET#0, the configuration of the
initial DL BWP configured by SIB1 is applicable after the initial access
33
For PSCell and SCell, the initial DL BWP of a cell can be configured to be the
same as or different with the initial DL BWP as defined by CORESET#0 of the
cell
o The initial DL BWP of a cell includes the bandwidth of CORESET#0
of the cell
Note: RAN1 assumes that the above agreements related to PCell, PSCell and
SCell have no RAN1 specification impact and will be captured in RAN2 specs
4. NR URLLC實體層增強之研究
以下針對各相關子議題說明:
實體層增強
首先,針對此研究項目的應用標的以及其模擬細節,依據其研究項
目的說明文件 RP-181477 以及會議的討論,相關的應用標的分別為運
輸產業、配電系統、工廠自動化以及 R15 已可支援的 AR/VR。相關決
議如下:
Agreements:
Select one or more representative use case(s) for the prioritized use cases in
the SID and/or the Rel-15 enabled use case for evaluation, which use case(s) to
evaluate is up to companies.
o Further discussion how/whether to capture them in the TR
o Further discussion other detailed simulation assumptions
The following table of representative use cases for selection for evaluation is an
example as the starting point for further discussion:
Use case
(Clause #)
Reliability
(%)
Latency
(ms)
# of UEs
(per cell)
Data packet size
and traffic model
Description
Transport
Industry
(22.186: 25)
[99.999] [5] (end to
end latency)
[30]
DL: [TBD] byte;
ftp model 3 with
arrival interval
[TBD] s
UL: [TBD] byte;
Periodic with
arrival interval
[TBD] s
Remote
driving
34
Power
distribution
(22.804:5.6.4
&5.6.6)
99.9999 5(end to
end latency)
8 [80] byte
ftp model 3 with
arrival interval
100ms
Power
distribution
grid fault and
outage
management
[99.999] 15(end to
end latency)
8 250 byte
Periodic and
deterministic with
arrival interval
0.833 ms
Differential
protection
Factory
automation
(22.804: 5.3.2)
99.9999 [2](end to
end latency)
[4, 40] 20 byte, 50 byte
Periodic and
deterministic
traffic model
Motion
control
Rel-15 enabled
use case (e.g.
AR/VR)
99.999 [1ms] (air
interface
delay)
1, 5, 10,
20
[32, 256] bytes
FTP model 2/3 or
periodic with
different arrival
rates
Companies
report the
combination
of the
requirement
o All the entries in the above table are subject to further discussion which
can be revisited in the next meeting
o Note: The details on above the requirements can refer to R1-1809337.
o Note: 3ms ~ 10ms CN delay for differential protection (i.e. power
distribution case 2) could be considered.
o Note: Rel-15 higher layer mechanisms for reliability may be applicable
for achieving the reliability requirement
o Note: The reliability and latency are as defined in 22.186.
o Note: For AR/VR, the requirement can refer to section 7.2.3 in TS
22.261.
o Note: FFS whether the packet size is based on application layer or
L2/L3. The packet size listed in the table needs to further discussed,
especially depending on the outcome of whether the packet size is
based on application layer or L2/L3
o Further discussion on how to map the requirements (e.g., reliability,
latency, etc.) to RAN-level requirements
Agreements:
35
Further discussion till next meeting regarding whether/how to evaluate the
number of users, the % of users, etc., satisfying reliability and latency
requirements.
Agreements:
Companies are encouraged to report the CDF of UE geometry
Further discussion whether/how to re-use the deployment and channel models
in the existing TRs (e.g. 38.802, 37.885 and 38.901)
其次,對於實體層的增強技術,可能的方向有強化 PDCCH的傳送
效率、強化 HARQ 回覆的效率以及通道狀態的量測與回報等。由於此
次會期為此議題的第一次討論,因此決議上並未有明確的技術方向,僅
決議朝上述幾個可能的技術進一步作評估,相關決議如下:
Agreements:
Further evaluate the potential PDCCH enhancements for NR Rel-16 URLLC.
Further evaluate PDCCH reliability
Further evaluate PDCCH blocking o Companies describe the resource utilization
Complexity should be considered
Latency of the enhancement(s) should be considered
Agreements:
Study further how to enable more than one PUCCH for HARQ-ACK
transmission within a slot.
Agreements:
Study further whether/how to enable enhanced reporting procedure/feedback for
HARQ-ACK.
Enhanced HARQ-ACK multiplexing on PUSCH and PUCCH
Finer indication for HARQ feedback timing, e.g. symbol-level, half-slot, etc. o Note: this may be related to more than one PUCCH for HARQ-ACK tx
within a slot
Other enablers are not precluded
Agreements:
Study the need for enhanced CSI reporting/measurement mechanisms. E.g.,
DMRS based CSI
A-CSI on PUCCH
Trigger by DL assignment
36
Enhanced CSI reporting mode
Other approaches are not precluded
以下列舉此議題相關的提案摘要,可作為後續收斂方向的判斷參考:
R1-1809163 Physical Layer Enhancements for NR URLLC NTT DOCOMO, INC.
TOPIC 1: On the limits of PDCCH BDs/CCEs
For example, for SCS = 15kHz, possible PDSCH scheduling occasion should be, e.g.,
per 2-symbol. Assuming that PDCCH schedules PDSCH, from the above tables, for
per 2-symbol PDSCH, each PDCCH monitoring occasion can have up to 6 PDCCH
candidates and 8 non-overlapped CCEs. Assuming the UE should be able to receive
DL assignment and UL grant at one time of PDCCH monitoring occasion, each
PDCCH (i.e., DL assignment or UL grant) can have up to 4 CCEs. This means that
for this particular case, the PDCCH aggregation level cannot be higher than AL = 4.
Therefore, following options can be considered to relax the obvious restrictions for
PDCCH monitoring specified in Rel.15.
Option 1: Specify higher numbers for the limits of BDs/CCEs.
Option 2: Support PDCCH-less PDSCH reception.
Option 3: Support nested search space structure.
Option 1 must be supported at least for the UE running both eMBB and URLLC
traffic;
Option 2 can be viewed as the enhancements or variations to the NR DL SPS
37
mechanism. Similar to the UL configured grant transmission, the transmission purely
rely on the RRC can be considered which is suitable for particular URLLC services
with periodic traffic profile deployed indoor;
Option 3 can reduce the CCE estimation efforts while may result in PDCCH
blocking. However, since the current search space structure is already non-nested, the
benefit would be restrictive.
Proposal 1:
Capture in the TR the essential need of solutions for the restrictive PDCCH
monitoring capability in Rel.15.
Possible solution 1: higher numbers for the limits of BDs/CCEs
Possible solution 2: PDCCH-less PDSCH reception
TOPIC 2: PDCCH repetition
Proposal 2:
Study following PDCCH repetitions and capture the benefits and advantages of
that in the conclusion of the TR for Rel.15 URLLC SI.
Opt. 1: PDCCH repetition using a single TRP
For example, a PDCCH is repeated within a CORESET using the
same/different AL(s)
Repetition uses same AL is almost equivalent to double the aggregation
level, using the different ALs can enable new aggregation levels. For
example, if a PDCCH is transmitted using AL=4 and AL=8, it is equivalent
to transmit a PDCCH using AL=12.
Opt. 2: PDCCH repetition using multiple TRPs
For example, a PDCCH is repeated across CORESETs using the same or
different AL(s)
DCI
Copied
Each repetition issend from each TRP
Time
Freq
TRP#1 TRP#2
TRP#3 TRP#4
Slot
38
Different CORESETs can be configured with different CCE-to-REG
mapping, different resource size, and/or different beams (QCLs).
Therefore, extra benefits/gains can be expected compared to Opt.1.
Proposal 3:
Study also data repetitions over multi-TRPs/panels.
TOPIC 3: UCI enhancement
Proposal 4:
Study the following features to realize multiple HARQ-ACK transmissions within a
slot:
More than two PUCCHs within a slot
More than one long PUCCHs within a slot
Study how to construct HARQ-ACK codebook between eMBB and URLLC.
TOPIC 4: CSI feedback enhancements
Proposal 5:
Study the necessity of aperiodic CSI on PUCCH.
Realize multiplexing of UCI and DMRS within a symbol.
TOPIC 5: PUCCH repetition enhancement
Proposal 6:
Support PUCCH repetition with any collision with PUCCH/PUSCH
Study PUCCH repetition with multi-TRPs (CM: no soft combine, just selection)
Evaluate the performance gain of PUCCH repetition with multi-TRPs
Study how to indicate PUCCH resources regarding PDCCH/PDSCH with multi-
TRPs
PDSCH#1 PDSCH#2 PUCCH
DCI#1 DCI#2
K1K1
the same slot for
HARQ-ACK feedback
HARQ-ACK feedback for
PDSCH #1 and PDSCH #2
DCI#2 is the last DCI
PUCCH#1
PUSCH#1
PUCCH#1 PUCCH#1 PUCCH#1
Slot n n+1 …
39
Study PUCCH format enhancement and UCI coding procedure enhancement,
if necessary
TOPIC 6: PUSCH enhancements
Proposal 7:
Study mini-slot repetitions as the promising candidates for URLLC enhancements
and capture the benefits and advantages of them in the TR.
PUSCH repetitions shorter than one repetition per slot (e.g., repetitions within
a slot).
PUSCH repetitions with multiple-TRPs.
Enabling different QCLs/precoders/beams across repetitions.
R1-1809400 Layer 1 URLLC Enhancements Motorola Mobility, Lenovo
TOPIC1: PDCCH repetition
Multiple PDCCHs can be transmitted in
(a) different CORESETs in the same TTI (via multiple TRP/beam/panel having
different QCL relationships between the downlink reference signals (RS) in a
RS Set (TCI-State) and the PDCCH DMRS ports) providing spatial and
frequency domain diversity gains.
(b) different TTIs (if the corresponding PDSCH is repeated in multiple TTIs)
providing time domain diversity gain as specified for LTE Rel-15 operation
(HRLLC).
Observation 1: Linking PDCCH candidates of multiple PDCCH transmissions
could increase the blocking probability and limit the scheduling flexibility.
Proposal 1: For PDSCH scheduling, multiple PDCCH transmissions in (a) different
CORESETs with different QCL assumptions and/or (b) different TTIs (similar to
A channel/signal
Combined/Selected
Each repetition is send to each TRP
Time
Freq
40
LTE-HRLLC) should be further studied to improve PDCCH detection reliability.
TOPIC 2: PDSCH relibility
To increase PDSCH reliability, diversity transmission of PDSCH can be transmitted
via
(a) multiple TRP/beam/panel in different time-frequency resources. Each PDSCH
can be scheduled by its own PDCCH.
(b) multiple TRP/beam/panel of same PDSCH layer(s) (on same time-frequency
resources). The PDSCH can be scheduled by a single PDCCH or multiple
PDCCH (as discussed in section 1).
(c) multiple time-domain repetitions (already supported for slot-level TTI), with
mini-slot TTI to reduce the latency.
Proposal 2: For PDSCH reliability improvement, diversity transmission of PDSCH
in (a) same/different time-frequency resources from multiple TRP/beam/panel with
different QCL assumptions, and/or (b) different TTIs (similar to LTE-HRLLC)
should be further studied.
R1-1809457 Layer 1 enhancements Qualcomm Incorporated
TOPIC 1: PHY-Layer Differentiation for eMBB and URLLC
Proposal 1: Adopt a PHY-layer signalling for differentiating the channels with different
traffic types in Rel. 16 URLLC.
TOPIC 2: Compect DCI
Pros: The main observation made was that reducing the DCI size even by about
10bits can only bring a negligible performance gain (0.4~0.6dB). On the other hand,
reducing the DCI size by a larger factor may limit the scheduling flexibility.
Cons: Reducing the DCI size could be to use the new DCI format as a service
indicator.
Proposal 3: Designing a compact DCI format can be considered for Rel. 16 URLLC.
CM: Carrier indicator, Rate-matching indicator, Waveform indicator
TOPIC 3: PDCCH repetition
If the URLLC requirements are not met even assuming the largest AL value, PDCCH
repetition can be considered as one solution.
41
The PDCCH repetition can be envisioned in two ways:
(1) the same DCI is sent multiple times, but the UE is not required to combine
different copies for decoding,
(2) the same DCI is sent multiple times (PDCCH can be repeated over multiple
(re-)transmissions in the same HARQ instance), but at each decoding attempt,
the UE is expected to combine PDCCHs that might have been sent in different
dimensions for larger processing gain.
Proposal 4: If the PDCCH reliability cannot be satisfied even with the largest
defined AL, the gains from PDCCH repetition can be studied for NR URLLC.
The starting point could be to evaluate PDCCH repetition (in CC, time, and
TRP domain with increased diversity order) that does not require combining
across different PDCCH occasions by the UE.
TOPIC 4: Enhanced HARQ-ACK reporting capability per slot
Proposal 5: To reduce the URLLC latency and improve its PUCCH coverage, multiple
HARQ-ACK reporting per slot can be supported.
TOPIC 5: A-CSI triggering by DL grant
Observation 1: The benefit of A-CSI triggering via a DL grant is two-fold: (1) There is
no need to spend resources on sending an UL grant to only trigger the A-CSI report, and
(2) the up-to-date CSI report can be made available at the gNB to make proper
scheduling decisions for TB re-transmissions.
Proposal 6: For Rel. 16 URLLC, allow for triggering A-CSI reporting via the downlink
grant.
UL inter UE Tx prioritization/multiplexing
此議題討論 UE間傳輸信號的配置方式,主要考慮不同 UE間分別
傳送不同應用情景的封包,例如 eMBB以及 URLLC,而這些信號又在
同一時間彼此碰撞,在此狀況下傳送 URLLC相關封包的 UE應該優先
保護,以保證其應用情景的需求。
Agreements:
RAN1 to study the potential enhancements for UL inter UE Tx
prioritization/multiplexing
42
o Performance study of the enhanced UL inter UE Tx
prioritization/multiplexing mechanisms using Re-15 mechanisms as
the performance benchmark
The use cases and scenarios adopted in L1 enhancements for
URLLC are considered for the evaluation of UL inter UE Tx
prioritization/multiplexing
Other factors to be considered such as overhead, capability,
etc.
o Study the UE UL cancelation mechanisms, including at least the
following aspects
The potential mechanisms may include UE UL
cancelation/pausing indication, UL continuation indication,
UL re-scheduling indication
Physical channel/signal used for the UL cancelation indication
UE Processing timeline for the UL cancelation indication
UE monitoring behaviours for the UL cancelation indication
UE PDCCH monitoring capability, if the UL cancelation
indication is by PDCCH
Methods to ensure the reliability of the indication for UE UL
cancelation
o Study the UL power control enhancements
o Study other enhancements for the multiplexing between a grant-based
UL transmission from a UE and a grant-free UL transmission from
another UE
Enhanced UL grant-free transmissions
此議題為針對 R15 已定義的 UL configured grant 傳輸技術的進一
步加強,以更符合 URLLC的應用需求。此技術於 R15的缺點主要有持
續性的傳輸碰撞、接收端的誤檢測、以及無法保證的重複傳輸次數。上
述問題皆會影響到 URLLC的傳輸品質需求。但由於此次會期為此議題
的第一次討論,因此決議上並未有明確的技術方向,僅針對目前 R15之
configured grant 的資源組態以及觸發方式(type 1 或 type 2)進行評估的
決議。其餘部分也是需要進一步評估以釐清上述問題是否有討論的必要
性,相關決議如下:
Agreements:
43
Study further whether/how multiple active configured grants for a BWP of a
serving cell.
o Identify potential specification impacts and options for both type 1 and
type 2
At least Activation/deactivation mechanism for Type2
E.g., whether each configuration is
activated/deactivated or multiple configurations are
activated/deactivated
Study how to support repetitions with multiple configurations
for a BWP of a serving cell
FFS HARQ process ID determination for both type 1 and type
2
FFS other specification impacts for both type 1 and type 2
o Study the performance impacts
Agreements:
Study further whether/how on ensuring K repetitions.
Study further on PUSCH repetitions within a slot for configured grant.
以下列舉此議題相關的提案與技術討論摘要,可作為後續收斂方向
的判斷參考:
R1-1808100 Enhanced UL configured grant transmissions Huawei, HiSilicon
TOPIC 1: Mini-slot-based repetitions within a slot
44
……
(c) The 1st TO is not available due to packet arrival (d) No TO is available due to packet arrival
……slot
(a) Two TOs with each consisting 7 symbols (b) One TO consisting 14 symbols
…
……
(e) The 1st TO is not available due to symbol
direction confliction
(f) No TO is available due to symbol
direction confliction
slot
2nd
TO with 7 OS1st TO with 7OS Only one TO with 14 OS
slot
…slot
√
D F F F U U U U U U U U U U
slot
…D F F F U U U U U U U U U U
slot
√
Observation 1: Mini-slot-based repetitions can provide more opportunities within a
slot to deliver a packet timely upon its arrival, and hence is a key approach to meet
the stringent latency requirement of URLLC services with periodic or a-periodic
traffic model in Rel.16.
Proposal 1: For both Type 1 and Type 2 PUSCH transmissions with a configured
grant, more than one mini-slot-based repetition within a slot should be supported for
Rel.16.
Observation 2: Explicit HARQ-ACK feedback can facilitate early-termination of the
repetitions and also trigger the flush of the HARQ buffer timely for the delivery of
new packets, thus can improve both latency and reliability performance for URLLC.
Proposal 5: For both Type 1 and Type 2 PUSCH transmission with a configured grant,
explicit HARQ-ACK feedback during or after K repetitions should be supported f
Rel.16.
- Both group common DCI and UE-specific DCI can be considered for the
delivery of HARQ-ACK indication.
- NACK can be assumed if no ACK or UL grant for retransmission scheduling is
received when a grant-free transmission timer expires; a grant-free
retransmission can be performed by a UE upon NACK.
Proposal 6: For both Type 1 and Type 2 PUSCH transmission with a configured grant
in Rel.16, to guarantee a low-latency and yet reliable transmission, repetitions
45
should be continued in time as long as the repetition number hasn’t reached K and
no early ACK or UL grant is received.
R1-1809459 Enhanced SPS and grant-free transmissions Qualcomm Incorporated
Potential open issues with the configured-grant uplink operation are listed below:
Persistent-collisions: If the number of UEs in the system is very large (e.g., mMTC),
and/or the number of resources required for each UE in the system is high in order to
meet the reliability requirements (e.g., URLLC) then it may not be possible to allocate
orthogonal resources to each UE within the available resource pool.
In such cases, multiple UEs may be assigned the same periodicity, offset and symbol
allocation within the slot. This could result in persistent collisions if such UEs become
active at the same time. While Rel. 15 allows for frequency hopping to address this
problem, the effectiveness of such a solution may also be limited if the number of RBs
available is small, which makes it difficult for UEs with narrowband RF chain.
Mis-detection handling: As per the current mechanism defined in Rel. 15, the UE
cannot distinguish between the following two scenarios:
1) a TB transmitted using the configured-grant uplink mechanism has been
successfully received by gNB.
2) the gNB did not detect the presence of the transmission (DMRS blind detection
failure).
This is because in both cases the UE may not receive an explicit ack feedback. Thus,
even if detection failed at the gNB, the UE currently moves on to the next TB after the
expiration of the ConfiguredGrantTimer. This could significantly impact the reliability
of the grant-free mechanism.
Ensuring K repetitions: There are some cases where the repetition mechanism does
not ensure that K repetitions would occur:
Some of the K consecutive slots may be downlink slots, or some symbols within
the symbol allocation in the configured grant may later be determined as
downlink symbols based on the slot configuration. In this case, the repetition is
omitted on such slots.
According to Rel. 15, the repetitions shall be terminated at the last transmission
occasion among the K repetitions within the period P.
TOPIC 1: Time hopping to mitigate persistent collisions in UL configured-granted
transmissions
46
Figure 2: Overloaded allocation with period 3, persistent collision possible, for SPS
without repetition.
Figure 3: Time-hopping allocation to mitigate persistent collision problem for SPS
without repetition
Proposal 1: Consider time-hopping based resource allocation to improve the
latency-reliability tradeoff for configured grant uplink operation.
TOPIC 2: Misdetection handling
Current mechanism defined in Rel. 15 the UE cannot distinguish between the following
two scenarios:
A TB transmitted using the configured-grant uplink mechanism has been
successfully received by gNB.
The gNB did not detect the presence of the transmission (DMRS detection failure).
SOULTION: (2 options)
Option 1: Design more reliable uplink transmissions to reduce the mis-detection
probability at the gNB.
Observation 1: Allowing SR associated with grant-free data transmission may be
a promising scheme to resolve uplink collision, and hence reducing the mis-
47
detection probability at the gNB. Proposal 2: Study mechanisms to reduce
collision probability for PUSCH with configured grant.
Option 2: Use explicit (downlink) ack/nack feedback to help the UE distinguish
between mis-detection and Ack.
Observation 2: An explicit HARQ-Ack for PUSCH with configured grant may
require large downlink overhead.
Proposal 3: For some URLLC use cases, the UE can be expected to transmit an
ACK once the SPS activation/de-activation DCI is detected.
TOPIC 3: Ensuring K repetitions
Observation 3: As supported in NR Rel-15, the reliability of grant-free uplink
transmission may be ensured by grant-based retransmissions.
5. NR之遠端干擾管理技術研究
此研究項目關注於遠端基地台所造成的干擾現象,干擾的來源可能
來自數十甚至數百公里遠的基地台,其傳送電波經過大氣折射與傳導,
造成基地台接收上行資料時,同時也受到來自遠端基地台的遠端干擾,
遠端干擾的現象可能持續數分鐘甚至數小時,但由於此干擾來至遠端,
甚至可能不只一個基地台產生干擾或受到干擾,不免影響到服務品質,
對整個網路也呈現不穩定的狀態;由於是網路中的上下行傳輸的自我干
擾情況,增長上下行轉換的保護時間,是避免干擾現象的方案之一,但
保護時間的長短也相對影響可利用資源的數量,因此,適當的保護時間
長度設計,是此研究項目的目標之一。
此會期的一項決議便是針對保護時間的研究,假設系統網路對上下
行的傳輸設定有共同的認知,而各基地台再依此共同認知,動態地調整
其所需之上下行轉換保留時間,以避免遠端干擾的情勢,此轉換保留時
間亦進一步用於干擾偵測機制之所需。決議如下:
Agreements:
48
To include the following in the TR: As shown in Figure 1, it is assumed in the RIM
study that the whole network with synchronized macro cells has a common
understanding on a DL transmission boundary (denotes as the 1st reference point)
which indicates the ending boundary of the DL transmission, and an UL reception
boundary (denotes as the 2nd reference point) which denotes the starting boundary of
the first allowed UL reception within a DL-UL transmission periodicity.
o The boundary may be considered for RS design
o The 1st reference point locates before the 2nd reference point.
D D D S U
10 2 214
GP U U UDDDDDD D D D U U U U U U U U U U U U U
14
slot
D
1414Common understanding
among the whole network
D D D S U
10 2 214
U U U U U U U U U U U U U U
14
slot
1414Frame Structure
Configuration for gNB1
XXXDDDD D D D U UX X
1st Reference Point
(maximum DL transmission boundary)
2nd Reference Point
(maximum UL transmission boundary)
1st Reference Point 2nd Reference Point
D D D S U
10 2 214
U U U U U U U U U U U U U U
14
slot
1414Frame Structure
Configuration for gNB2
XDDDDDD D D D U UX X
1st Reference Point 2nd Reference Point
DL-UL switching period
Figure 1. Illustration of DL and UL transmission boundaries within a DL-UL transmission
periodicity
遠端干擾之現象,可能是單方面造成之現象,也可能是雙方的基地
台皆可感受到或產生干擾,因此在設計相關管理技術,須考量到此兩種
現象。決議如下:
Agreements:
In terms of the IoT (interference over thermal) increase between two sets of gNBs
causing remote interference to each other, two scenarios should be considered for NR-
RIM,
1. Scenario #1: IoT increases are detectable by one or more gNBs in both sets,
2. Scenario #2: IoT increase is detectable by one or more gNBs in only one set.
本會期中,提出了干擾偵測與管理的基本架構,執行步驟與程序可
49
進一步討論,受到影響與產生影響的基地台皆須針對遠端干擾,提出適
當的應變機制或干擾消除方案。基本的遠端干擾管理架構可歸納幾項重
點:當基地台感受的遠端干擾影響時,可傳送對應的參考訊號,而遠端
的基地台應適時地偵測此參考訊號,已確認是否產生遠端干擾行為,當
偵測到此參考訊號時,產生影響的基地台變執行干擾消除技術,並通知
受到影響的基地台已進行適當處置,當遠端干擾消除時,受到影響的基
地台可停止參考訊號的傳送。此架構中的執行細節又可有多項技術方案,
例如透過後端網路訊號或再由另一個參考訊號,告知受到影響的基地台
停止傳送參考訊號,利用參考訊號得知訊號的來源基地台,或協調建議
的干擾消除方案等,皆可進一步設計。決議如下:
Agreements:
Framework-1, Framework-2.1, Framework-2.2 below are used as starting point for further
study, using Framework-0 as basis for comparison.
Note:
- Not all the steps need to be included when making use of a given framework.
- Mechanisms for improving network robustness at both victim and aggressor side can be
studied under the NR-RIM frameworks.
A victim cell may take actions applying remote mitigation scheme. This detail is FFS
- An aggressor may also be a victim (and vice versa) at least for Scenario #1
0. Framework-0
Workflow of Framework-0
Step 0: Atmospheric ducting phenomenon happens and the remote interference appears
Step 1:
• Victim experiences “sloping” like IoT increase and start RS transmission
• Aggressor starts monitoring RS as configured by OAM
Step 2: Upon reception of RS, Aggressor reports the detected RS to OAM
Step 3: OAM sends remote interference mitigation scheme to Aggressor
50
Step 4: Aggressor applies remote interference mitigation scheme
Step 5: OAM stops RS monitoring and restores original config. at aggressor side and stop RS
transmission at victim side.
1. Framework-1
Workflow of Framework-1
Step 0: Atmospheric ducting phenomenon happens and the remote interference appears
Step 1:
• Victim experiences “sloping” like IoT increase and start RS transmission/monitoring
- This RS marked as RS-1 is used to assist aggressor(s) to recognize that they are
causing remote interference to the victim and to detect/deduce how many UL
resources of the victim are impacted by the aggressors.
• Aggressor starts monitoring RS as configured by OAM or when it experiences remote
interference with “sloping” IoT increase.
Step 2: Upon reception of RS-1, Aggressor starts remote interference mitigation solutions such
as muting some DL transmission symbols and transmits RS to inform victim that the
atmospheric ducting phenomenon still exist
• This RS marked as RS-2 is used to assist the victim to decide whether the atmospheric
ducting phenomenon still exist.
- It does not preclude the possibility of using RS-2 for other purposes, pending on
further study.
Step 3: Victim continues RS-1 transmission while receiving RS-2. Upon “disappearance” of
RS-2, victim stops RS transmission
Step 4: Aggressor continue remote interference mitigation while receiving RS-1. Upon
“disappearance” of RS-1, Aggressor restores original configuration when “disappearance” of
RS-1.
Note: Although RS-1 and RS-2 carry different functionalities, it might be beneficial to achieve
a common design for RS-1 and RS-2.
2. Framework-2.1
51
Workflow of Framework-2.1
Step 0: Atmospheric ducting phenomenon happens and the remote interference appears
Step 1:
• Victim experiences “sloping” like IoT increase and start RS transmission
- A set of gNBs might use the same RS, which may carry the set ID.
• Aggressor starts monitoring RS as configured by OAM or when it experiences remote
interference with “sloping” IoT increase.
Step 2: Upon reception of RS, Aggressor informs the set of victim gNB(s) the reception of RS
through backhaul and apply interference mitigation scheme
- Message exchange in Step 2 could include other information, pending on further
study.
Step 3: Upon “disappearance” of RS, Aggressor informs the set of Victim gNB(s) the
“disappearance” of RS through backhaul and restore original configuration.
Step 4: Victim stop RS transmission upon the reception of the “disappearance of RS” info
through backhaul
3. Framework-2.2
Workflow of Framework-2.2
Step 0: Atmospheric ducting phenomenon happens and the remote interference appears
Step 1:
• Victim experiences “sloping” like IoT increase and start RS transmission
- A set of gNBs might use the same RS, which may carry the set ID.
• Aggressor starts monitoring RS as configured by OAM or when it experiences remote
interference with “sloping” IoT increase.
Step 2: Upon reception of RS, Aggressor informs the set of victim gNB(s) the reception of RS
through backhaul
52
Step 3: Upon reception of the “reception of RS” info received in the backhaul, victim sends
info to assist RIM coordination
Step 4: Aggressor applies remote interference mitigation scheme
Step 5: Upon “disappearance” of RS, Aggressor informs Victim the “disappearance” of RS
through backhaul.
Step 6: Victim stop RS transmission upon the reception of the “disappearance of RS” info
through backhaul
後端網路訊息與布建環境之意見,則進一步通知 RAN3,以確認討
論架構的可行性。決議如下:
Agreements:
Inform RAN3 that three frameworks are used as in RAN1 as a starting point for further
study. Following information will also be included is the LS.
o The distance between gNB aggressor and gNB victim can be up to 300 km.
o Action to RAN3: to provide feedback regarding feasibility of the frameworks
Draft LS in R1-1809875, which is approved and final LS in R1-1809987
以研究項目將進行鏈結層模擬,以確認參考訊號於討論架構中的效
能。決議如下:
Agreements:
For RIM SI, the evaluation is to be carried out via link-level simulation to evaluate
the performance of the reference signals in the NR-RIM frameworks.
鏈結層模擬所需之設定,也進一步說明。決議如下:
Agreements:
For simulation evaluation of reference signals in the NR-RIM frameworks
- Following Descriptions of the RS should be provided
o RS sequence
o Length of RS sequence
o Time/frequency pattern of RS
Time pattern (number of symbols)
Frequency pattern
- Following analytical metrics of the RS should be provided
o The complexity of reference signal detection at gNB
o Overhead
o Impact on UEs
53
o Others
- Simulation
o Simulation parameters
SCS: 30 kHz (mandatory) / 15 KHz (optional)
Simulation bandwidth: 20 MHz
gNB MIMO configuration: 1T1R (mandatory)/1T2R(optional)
Frequency offset: 0 Hz
FFT size: to be provided
Length of detection window Lsymbol: to be provided
Channel model:
- Option1: AWGN with random complex phase
- Option2: TDL-E (K-factor = [22] dB, DS = [30] ns, Doppler [0] Hz)
- FFS: whether one of the two options or both options are mandatory.
Delay of received RS: When multiple RSs arrive in the detection window,
the arrival time of the i-th RS respect to the start of the detection window,
△i , is uniformly distributed within [-Lsymbol, Lsymbol], where Lsymbol is the
length of UL symbol based on the numerology of RS.
Power of received RS:
- Option1: Pi of multiple RSs have a power offset with respect to the
reference power P0, where the power offset is randomly selected from
[-0.5dB, 0.5dB].
- Use option1 as starting point for evaluation, FFS other option(s), e.g.,
different power offset ranges.
o Simulation cases and related metrics
Case 1: Single RS + AWGN (mandatory)
- Metric: the minimum SNR where detection probability of [90%] and a
false alarm requirement of [1%]
- FFS: successful detection time, e.g., one-shot.
Case 2: Multiple RS + AWGN (mandatory)
- Number of total RSs arrived within one detection window: FFS
Number of base sequences arrived within the detection window:
FFS
- Metric: FFS.
6. 基於新無線電技術之非授權頻譜存取(NR-U)
54
針對 configured grants 之增強
關於配置授權傳輸的資源配置,NR的設計是連續的資源配置,在
NR-U的情境下,更彈性的資源配置是有幫助的。另外,型態一和型態
二的配置授權傳輸各有優點,同存於 NR-U的情境中是有好處的。因此
在此次會議的決議如下:
Agreement:
Allowing consecutive configured grant resources in time without any gaps in between
the resources and non-consecutive configured grant resources (not necessarily
periodic) with gaps in between the resources is beneficial and should be considered
for NR in unlicensed spectrum
Conclusion:
There is no necessity to exclude Type-1 or Type-2 configured grant mechanism for
operation of NR in unlicensed spectrum.
在 NR 的設計中,所配置的資源隱含著 HARQ 識別,但考慮到傳
輸的不確定性,UE在 PUSCH上傳送時,把 HARQ識別一起在 PUSCH
傳送。此次會議的決議如下:
Agreement:
UE selects the HARQ process ID from an RRC configured set of HARQ IDs for NR-
unlicensed configured grant transmission.
關於配置授權傳輸,gNB會回傳對應的 DFI給 UE,但考慮到 gNB
執行LBT可能會失敗,所以支援DFI包含之前的配置授權傳輸的HARQ
是有幫助的。此次會議的決議如下:
Agreement:
It is identified to be beneficial to support DFI to include pending HARQ ACK
feedback for prior configured grant transmissions from the same UE.
FFS: DFI to include HARQ ACK feedback for scheduled UL transmissions
using HARQ IDs configured for NR-unlicensed configured grant transmission.
55
配置授權傳輸,可以由一樣的配置授權資源來重新傳送,也可以透
過上行授權所排程的資源來重新傳送。此次會議的決議如下:
Agreement:
Retransmission via same configured grant resource is supported for a HARQ
process that was initially transmitted via configured grant resource.
Retransmission via resource scheduled by UL grant is supported for a HARQ
process that was initially transmitted via configured grant resource.
在配置授權傳輸下,UE可以自動的啟動重傳,當以下的條件符合
時
甲、 接收到 gNB的 DFI為 NACK時
乙、 繼續研讀的部分:在一段區間內,沒接收到 gNB所回傳的 DFI,
但考慮到有可能是 gNB 執行 LBT 沒成功,所以細節的設計需要
繼續研讀。
此次會議的決議如下:
Agreement:
UE may autonomously initiate retransmission for a HARQ process that was initially
transmitted via configured grant mechanism for NR-unlicensed when one of the
following conditions is met:
Reception of NACK feedback via DFI for the corresponding HARQ process
FFS: No reception of feedback from gNB upon the timer expiration.
o To introduce a new timer or reuse configuredGrantTimer.
配置授權傳輸的資源分配,可能是多個 UE共用,也可能和上行授
權所排程的資源共用。因此有效率的 UE多工和碰撞避免在配置授權的
傳輸之間或是配置授權的傳輸和上行授權傳輸之間是有幫助的。此次會
議的決議如下:
Agreement:
It is identified to be beneficial to consider UE multiplexing and collision avoidance
56
mechanisms between configured grant transmissions and between configured grant
and scheduled grant transmissions.
FFS: detailed mechanism.
配置授權傳輸的資源如果和配置給 NR-U DRS傳送的資源重疊時,
配置授權傳輸不允許傳送,NR-U DRS有較高的優先權。此次會議的決
議如下:
Agreement:
NR-unlicensed configured grant transmission is not allowed during the time when it
overlaps with occasions configured for potential NR-U DRS of the serving cell
irrespective of the configured time domain resource for configured grant transmission.
七、心得與建議
心得
在 R15 的 NR MIMO 已經在上半年已經結束,這次會期在 NR
MIMO 部分則是加強規格描述的精準度以及釐清與其他議題的
相關問題。各公司可能是為了各自 5G 商品時程的考量,在議
題的討論偏向保守,以盡可能不變更規格為主,即使有更動規
格的考量則是需要跟各自的商品實現部門確認才能進一步討論。
許多最佳化或是不影響大致運作的功能都被擱置,預計會在下
一次會議開展的工作項目再次討論,下一次 NR MIMO的工作
項目將會把討論初期的多收發節點、多面板傳輸等功能加入討
論。
URLLC及NR-U對於R15的 configured grant都有增強的討論,
提案方向與討論雷同。此外,部分概念也導入 LTE R15 URLLC
的方式,後續若有興趣可一併研究。
RIM 於本會期第一次討論,但此現象主要產生在低頻段運作,
由於是承接 LTE時代既有的問題,因此可猜想,電信營運商應
已有一套自身在運作的遠端干擾管理與偵測的機制,應是期望
57
在 R16將之標準化,因此可發現其討論以電信營運商為主,基
地台偵測與回報之架構已有明確步驟,對於 UE 的影響則尚未
明朗。
建議
由於 NR-U議題也包含MIMO相關議題,部分代表也開始關注
NR-U。對於未來 R16的設計,建議可從 NR-U與 NR MIMO一
起考慮在未來版本的布局。
NR R15 Maintenance相關議程有時會討論到同時處理 eMBB及
URLLC兩種資料相關的控制信號或程序的碰撞問題,由於目前
的狀態是無法再增加新的功能,使得處理問題的手段上略為受
限。考量到系統的複雜度,往後在問題的思考上要再嚴謹並考
慮多種面向的影響再做設計。
RIM的設計上雖說是以基地台為主,但可觀察其參考訊號的設
計與偵測方法,以及如何協調保護時間的設定等,進而避免跨
連結干擾的現象;更進一步來說 RIM的相關設計,可能進一步
成為討論跨連結干擾技術的討論基礎,得多注意對 UE 的影響
與新設計。
經過這幾次會議的觀察,主席的風格將會是由某家公司領導的
線下討論為主,在統整各公司的意見後才會在線上做決議,由
於線下會議時間不定且可能同時發生,建議單一議題應不只單
一代表參與。