WO2018172542A1

WO2018172542A1 - Data buffer handling for dual connectivity

Info

Publication number: WO2018172542A1
Application number: PCT/EP2018/057519
Authority: WO
Inventors: Torsten DUDDA; Henning Wiemann; Jose Luis Pradas; Martin Skarve
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2017-03-24
Filing date: 2018-03-23
Publication date: 2018-09-27

Abstract

The invention relates to a method performed by a user equipment, UE (102), for managing or handling uplink, UL, data transmissions in a radio communications network, wherein the UE is served by a first radio base station (103) and a second radio base station (104) providing dual connectivity to the UE in the radio communications network, the method comprising preprocessing UL data prior to receiving a UL grant; and storing the preprocessed data in an UL data buffer (302); the invention further related to a corresponding UE (102).

Description

DATA BUFFER HANDLING FOR DUAL CONNECTIVITY TECHNICAL FIELD

[001] Disclosed are embodiments related to sending uplink (UL) data for a split bearer using dual connectivity (DC).

BACKGROUND

[002] The 3 GPP is working on a New Radio (NR) standard for 5G, building on

LTE/EUTRAN. For NR, the dual connectivity (DC) protocol architecture of a split bearer will be standardized, building on the protocol architecture used for LTE for the DC split bearer. In a DC split bearer arrangement, the UE maintains a packet data convergence protocol (PDCP) entity for the split bearer connected to multiple (e.g., two) radio link control (RLC) and medium access control (MAC) entities. These are each associated to a cell group, the master cell group and secondary cell group respectively. Transmission via the master cell group goes to the Master eNB (MeNB) (MgNB in NR or 5G terminology); transmission via the secondary cell group goes to the Secondary eNB (SeNB) (SgNB in NR or 5G terminology). The MeNB and SeNB maintain their own RLC and MAC entities associated to this single split bearer. A further node or function is involved, the packet processing function (PPF), which may be separate or collocated with the MeNB or SeNB, and the PPF terminates the PDCP protocol on the network side.

[003] When sending UL data for the split bearer in DC, the UE may transmit a PDCP protocol data unit (PDU) via either the MeNB or SeNB side. To steer this operation, for LTE, two configuration parameters have been standardized: split threshold and prioritized cell group. If the data in the PDCP UL buffer (i.e., UL queue) is below the split threshold, transmission of the data is only allowed via the prioritized link/cell group/eNB; if the data in PDCP is higher than the threshold, transmission is allowed via both links/cell groups/eNBs. The prioritized cell group can be configured to be either the MeNB or SeNB.

SUMMARY

[004] A problem with sending UL data for the split bearer in DC, is that current approaches result in unnecessary latency. Prior to transmitting data, the data must be processed into a PDCP PDU format. This processing is depending upon whether the MeNB or SeNB will be used for transmission. Current approaches therefore only begin processing after an UL grant has been received (and therefore after it is known whether the MeNB or SeNB will be used for transmission). This means that there is a delay between when the UE is authorized, via the UL grant, to transmit, and when that transmission takes place. This delay is at least as long as the processing time required to process the UL data into PDCP PDU format. Accordingly, improved approaches to sending UL data for split bearer in DC are needed.

[005] It should be understood that the problems and embodiments described herein are equally applicable to wireless access networks and user equipments (UEs) implementing access technologies and standards other than NR (e.g., 3G, 4B). NR (e.g., 5G NR) is used as an example technology for purposes of describing embodiments, but this disclosure is not so limited.

[006] In NR, ultra-low latency for transmission, and processing (i.e. preparing of a transmission), is a fundamental requirement. In particular, transmission of data in uplink in response to reception of an uplink grant should be done within a very short time (to be standardized). To make this ultra- low latency possible, data should be preprocessed for transmission even before the uplink grant is received. This preprocessing may comprise assigning PDCP and RLC sequence numbers (SNs) to PDCP service data units (SDUs). As indicated above, when data in the UL buffer (queue) exceeds the split threshold, it is uncertain via which link (i.e., RLC entity), the data will be transmitted. According to previous approaches, an RLC sequence number cannot be assigned beforehand. In contrast to such previous approaches, embodiments provide that sending UL data for split bearer in DC can be efficiently operated with the use of the split threshold. For example, embodiments allow for immediate transmission of the UL data upon reception of an uplink grant. This is possible because of prescheduling or preprocessing of the UL data. Embodiments achieve this efficiency by using further pre-scheduling thresholds and other methods as described herein.

[007] According to one aspect there is provided a method performed by a user equipment (UE) for managing or handling uplink (UL) data transmissions in a radio

communications network, wherein the UE is served by a first radio base station and a second radio base station providing dual connectivity to the UE in the radio communications network. The method includes the UE preprocessing data prior to receiving a UL grant; and storing the preprocessed data in an UL data buffer.

[008] In some embodiments, preprocessing the UL data comprises preprocessing at least a first portion of the data specific to the first radio base station. In some embodiments, a size of the first portion is less than a difference between a first preprocessing high threshold and a first preprocessing low threshold.

[009] In some embodiments, preprocessing the first portion of the data specific to the first radio base station comprises re-processing at least a subpart previously preprocessed specific to the second radio base station.

[0010] In some embodiments, wherein preprocessing the data further comprises preprocessing at least a second portion of the data specific to the second radio base station.

[0011] In some embodiments, a size of the second portion is less than a difference between a second preprocessing high threshold and a second preprocessing low threshold.

[0012] In some embodiments, the first portion and the second portion overlap.

[0013] In some embodiments, the method also includes: receiving a UL grant from the first or the second radio base station; and in response to receiving the UL grant, transmitting a portion of the preprocessed data to the corresponding first or second radio base station (e.g., as a result of receiving the UL grant, the device transmits the data immediately - e.g., the data is transmitted in a TTI for which the uplink grant is valid).

[0014] In some embodiments, the UL grant is specific to the first radio base station, and the transmitted portion of data corresponds to a subset of the first portion. In some embodiments, the UL grant is specific to the second radio base station, and the transmitted portion of data corresponds to a subset of the second portion.

[0015] In some embodiments, the method further includes receiving a first UL grant from the first radio base station; receiving a second UL grant from the second radio base station; in response to receiving the first UL grant, transmitting a portion of the data in the UL data buffer that corresponds to a subset of the first portion to the first radio base station (e.g., this data is transmitted in the TTI for which the first UL grant is valid); and in response to receiving the second UL grant, transmitting a portion of the data in the UL data buffer that corresponds to a subset of the second portion to the second radio base station (e.g., this data is transmitted in the TTI for which the second UL grant is valid).

[0016] In some embodiments, the method further includes: determining that a current size of the UL data buffer is above or below a split threshold; in response to determining that the current size of the UL data buffer is below a split threshold, sending a scheduling request (SR) only to the first radio base station; in response to determining that the current size of the UL data buffer is above a split threshold, sending a SR to both of the first radio base station and the second radio base station.

[0017] In some embodiments, the method further includes: receiving one or more of the first preprocessing high threshold, the first preprocessing low threshold, the second

preprocessing high threshold, and the second preprocessing low threshold, from a node in the radio communications network.

[0018] In some embodiments, the received thresholds are updated based on current network conditions. In some embodiments, the current network conditions comprise at least one of: the UE UL throughput to first radio base station and/or to second radio base station, the ratio of throughput between the first radio base station and the second radio base station, the TTI length for this UE, the link quality (SINR) to first radio base station and/or to second radio base station.

[0019] In some embodiments, the UE receives the updated thresholds from the network and/or the UE performs the updating.

[0020] In some embodiments, preprocessing the data in the UL data buffer comprises processing at least a portion of the data in the UL data buffer to create one or more packet data convergence protocol (PDCP) service data units (SDUs). In some embodiments, creating one or more PDCP SDUs comprises one or more of the following steps: assigning a PDCP sequence number (SN) to each of the PDCP SDUs; building a PDCP header for each of the PDCP SDUs; encrypting each of the PDCP SDUs to create one or more PDCP protocol data units (PDUs); assigning radio link control (RLC) SN to each of the PDCP PDUs; and building an RLC header for each of the PDCP PDUs. [0021] In some embodiments, the method includes placing the data in the UL data buffer prior to preprocessing the data.

[0022] In another aspect, there is provided a UE that has a preprocessing module configured to preprocess UL data prior to receiving a UL grant; and a storing module configured to store the preprocessed data in an UL data buffer. The UE is further adapted to perform the method steps described above and herein.

BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings, which are incorporated herein and form part of the specification, illustrate various embodiments.

[0024] FIG. 1 illustrates a split bearer network arrangement according to some embodiments.

[0025] FIG. 2 is a flow chart illustrating a process according to some embodiments.

[0026] FIG. 3. illustrates an UL data queue (buffer).

[0027] FIGs. 4A, 4B and 4C illustrate different scenarios according to some

embodiments.

[0028] FIG. 5A illustrates an arrangement according to some embodiments.

[0029] FIG. 5B illustrates an arrangement according to some embodiments.

[0030] FIG. 6 is a block diagram of a UE according to some embodiments.

[0031] FIG. 7 is a diagram showing functional modules of a UE according to some embodiments.

DETAILED DESCRIPTION [0032] FIG. 1 illustrates a split bearer network arrangement according to some embodiments. As shown in FIG. 1, UE 102 is served by a first radio base station 103 (e.g., Master eNB or gNB, or MeNB) and a second radio base station 104 (e.g., Secondary eNB or gNB, or SeNB). Both MeNB 103 and SeNB 104 are coupled to a packet processing function (PPF) node 106. UE 102 further comprises a UL data buffer (i.e., a UL data queue). Generally, after data for uplink transmission is placed in the UL data buffer, UE 102 will transmit the data. As shown, UE 102 is utilizing a dual connectivity (DC) split bearer, with a split threshold. For a given Transmission Time Interval (TTI), if the UL data buffer is filled to a value less than the split threshold, then the UE 102 may only transmit the UL data to MeNB 104. And if the UL data buffer is filled to a value greater than or equal to the split threshold, then the UE 102 may use one or both of the MeNB 103 and SeNB 104 (e.g., some of the buffered data is transmitted to MeNB and another portion of the buffered data is transmitted to SeNB).

[0033] In order to be prepared for immediate transmission upon receiving a UL grant (in order to be prepared to transmit in the TTI for which the UL grant is valid), UE 102 preprocesses data in the UL data buffer. The preprocessing is dependent on the link ultimately selected (i.e., on whether the MeNB or SeNB is used for data transmission). The preprocessing may also be based on several threshold values, as presently described. For example, the split threshold as already been described above, may determines whether only MeNB or one or more of MeNB and SeNB will be used in uplink. Further preprocessing thresholds, such as the M-preProc-low threshold (first preprocessing low threshold, e.g. 0), M-preProc threshold (first preprocessing high threshold) both associated to the MeNB (first radio base station), and S-preProc-low threshold (second preprocessing low threshold, e.g. 0), and S-preProc threshold (second preprocessing high threshold), both associated to the SeNB (second radio base station). For purposes of an easy understanding, the UL data buffer will be described as having a bottom or first position starting at 0 (lowest address of the buffer). For data at the position indicated by the M-preProc-low threshold (taken to be 0 throughout this discussion), and up to the position indicated by the M-preProc threshold, UE 102 preprocesses such data for transmission to the MeNB 103. For data at the position indicated by the S-preProc-low threshold (e.g. taken to be equal to the split threshold, unless otherwise noted), and up to the position indicated by the S- preProc threshold, UE 102 preprocesses such data for transmission to the SeNB 104.

[0034] Note that if the M-preProc threshold is greater than or equal to the S-preProc-low threshold (that may correspond to the split threshold), then there will be some overlap of data that UE 102 preprocesses for both MeNB 103 and SeNB 104. The amount of overlap will be equal to the difference between the M-preProc threshold and the S-preProc-low threshold (e.g., split threshold). On the other hand, if the M-preProc threshold is less than the S-preProc-low threshold (e.g., split threshold), there will not be any such overlap. [0035] An advantage to this preprocessing is that, when an UL grant is received, for example, an UL grant for MeNB 103 and another UL grant for SeNB 104, UE 102 is ready to send the preprocessed data. Therefore, it is advantageous for the difference between the M- preProc threshold and the M-preProc-low threshold to be greater than the maximum grant size on MeNB 103. Likewise, it is advantageous for the difference between the S-preProc threshold and the S-preProc-low threshold to be greater than the maximum grant size on SeNB 104. The reason for this is that otherwise it is possible that when a UL grant is received, not enough data will be ready (i.e. preprocessed) to fully utilize the UL grant resource. That is, either

transmitting the UL data using less than the allowed grant size, or waiting to process additional UL data, would both increase latency.

[0036] There may be more than one SeNB 104 that is serving UE 102.

[0037] FIG. 2 illustrates a process 200. Process 200 may begin with step 202, where UE

102 preprocesses data prior to receiving a UL grant. Process 200 may further include step 204, preprocessing a first chunk of data specific for a first radio base station, the first chunk having a size less than or equal to a first preprocess threshold. For example, this may involve

preprocessing data for MeNB 103, where the first chunk is equal to or less than the difference between M-preProc and M-preProc-low (e.g., 0). Process 200 may further include step 206, preprocessing a second chunk of data specific for a second radio base station, the second chunk having a size less than or equal to a second preprocess threshold. For example, this may involve preprocessing data for SeNB 103, where the second chunk is equal to or less than the difference S-preProc and S-preProc-low (e.g., split threshold). In step 208, the UE 102 stores the preprocessed data in an UL data buffer. In step 210, the UE 102 performs a comparison to determine whether the current size of the UL data buffer is below a split threshold. If so, only transmission to the first radio base station is permitted, and so the UE 102 sends in step 212 a scheduling request (SR) only to the first radio base station. Otherwise, transmission over both the first and second radio base stations is permitted, and so the UE 102 sends in step 214 a SR to both first and second radio base stations.

[0038] FIG. 3 illustrates an arrangement 300 showing an uplink data buffer 302, e.g. an uplink PDCP buffer with a PDCP split buffer threshold 304. As shown, the buffer 302 is filled up from the upper part (i.e. buffer position 0 is shown at the upper part, and the buffer position increases going downward). Data in a position "below" the split threshold 304 (i.e., data between buffer position 0 and the split threshold 304), can only be sent via the prioritized link (indicated here as "M", while the un-prioritized link is labeled "S"). Data in a position "above" the split threshold (e.g., data between the split threshold and the M-preProc threshold) can be sent via either M or S. The received uplink grant(s) determine where the data is sent.

[0039] Preprocessing data in the buffer includes, for example, taking steps to generate a

PDCP PDU that will include the data (or an encrypted version of the data) in the payload portion of the PDCP PDU. For example, preprocessing data in the buffer may include one or more of: assigning a PDCP SN for the PDCP PDU, generating the PDCP header for the PDCP PDU, encrypting the data, placing the encrypted data in a payload portion of the PDCP PDU.

Preprocessing may further include assigning an RLC SN to the PDCP PDU, assigning the RLC header, and parts of the MAC header, e.g. length field. In Dual Connectivity, the data routed over the MeNB (or master gNB) and the SeNB (or secondary gNB) will have independent RLC entities and, therefore, independent RLC SNs. That is, data must be processed specifically for either the MeNB (or master gNB) or the SeNB (or secondary gNB).

[0040] The UE should preprocess data ready for transmission over M up to a

preprocessing threshold 306 specific for M (shown as M-preProc threshold 306). This preprocessing threshold is assumed to be greater than or equal to the split threshold. Data above the split threshold, up to another threshold preprocessing threshold 308 (shown as S-preProc threshold 308), should be preprocessed for potential transmission on S. The other preprocessing threshold (S-preProc threshold) can be chosen as an offset to the split threshold and should be larger than the maximum grant size on S.

[0041] By using such preprocessing, it is ensured that once an uplink grant on M or S (or both) is received, the UE has preprocessed data for immediate transmission ready for all possible received grant sizes on M and S.

[0042] In one embodiment the UE can be configured with the preprocessing thresholds for M and S, respectively. That is, for example, the network (e.g., base station or core network node) may be configured to send to the UE a message (e.g. a Radio Resource Control (RRC) message or a non-access stratum (NAS) message) that includes information indicating the preprocessing thresholds.

[0043] In another embodiment, the UE preprocesses data according to these configured preprocessing thresholds or to UE implementation specific thresholds (thresholds determined by the UE e.g. based on throughput).

[0044] In one particular embodiment, the UE should at least preprocess all data below the split threshold, intended for the prioritized link direction. In other words, the difference between S-preProc threshold and S-preProc-low threshold is zero (i.e., no preprocessing for S).

[0045] As shown in FIG. 3, there is an overlap area, where data is preprocessed twice: once for M and again for S. That is, data 310 that is preprocessed for M and data 312 that is preprocessed for S share some data in common. This overlap depends on M-preProc threshold: if M-preProc threshold is less than S-preProc-low threshold (e.g., split threshold), there is no overlap; otherwise, the overlap size is the difference between M-preProc threshold and S- preProc-low threshold. The UE in this case (where there is an overlap area), can keep duplicate data ready for potential transmission, where such duplicate data is differently preprocessed for M and S, respectively.

[0046] Embodiments are described below that describe the procedure for preprocessing in more detail. In particular, when preprocessing is done for a given link and how the UE acts upon receiving uplink grants.

[0047] In FIGs. 4A, 4B and 4C illustrate three example scenarios 411, 412, and 413. In each scenario, the UL data buffer 302 is filled with data exceeding the split threshold 304.

Therefore, some of the data is preprocessed for transmission on M (i.e. the data between buffer position 0 and threshold 306), some of the data is preprocessed for transmission on S (i.e., the data above threshold 304). For example, the UE may send a scheduling request (SR) to one or both of the M and S links. In case 1, only an UL grant for the M link is received. In cases 2 and 3, UL grants for both the M and S links are received. As shown, uplink data arrives in the UL data buffer 302, and is not yet preprocessed (gray, squares), at transmission time interval (TTI) n-x. The data will be preprocessed according to the preprocessing thresholds described above (see diagonal cross-marks labeled "Preprocess for M", horizontal cross-marks labeled "Preprocess for S").

[0048] For scenario 411, at TTI n, when a grant for M is received, buffer 302 includes some data that is ready for transmission on M, having been preprocessed already. That is, because the data is preprocessed it can be sent immediately, without the need for further processing into an appropriate format. As shown in FIG. 4A, the M grant received indicates that the amount of UL data the UE can send as a result of receiving the M grant is less than the amount of data that has been preprocessed for M. Accordingly, as a result of receiving the M grant, the UE transmits the data that exists between buffer position 0 and the position indicated by reference number 415. This transmitted data is then removed from the buffer, and as shown in FIG. 4A, at TTI n+1 the data that was preprocessed for M but not transmitted is repositioned in the buffer such that the data starts at buffer position 0 and ends at the position indicated by reference 416, and the data that was preprocessed for S but not transmitted is discarded.

[0049] Note that the overlap previously mentioned is such that, at TTI n+1, all remaining data has already been preprocessed. Depending on threshold sizes and the sizes of any UL grants, as well as the UL data buffer, that is not guaranteed to be the case.

[0050] For scenario 412 (FIG. 4B), at TTI n, an UL grant for M and an UL grant for S are received. Since the data above the split threshold 304 was already preprocessed for S at TTI n, it can also be sent immediately. It is noteworthy that, in this scenario, the remaining data at TTI n+1 has a longer queuing time (lower SN) than the data sent on S at TTI n (which has higher SN). This is not a problem, since the PDCP receiver applies reordering, i.e. waits for the remaining data before delivering the data in sequence to higher layers. Since the remaining data in this case was also preprocessed for M, it does not need further preprocessing.

[0051] For scenario 413 (FIG. 4C), at TTI n, an UL grant for M and an UL grant for S are received as in case 2. However, in case 3, the UL grant size for S is less than the size of data processed for S. The remaining data at TTI n+1 is lower than the split threshold and thus can only be sent via M. The preprocessed data above the data sent to S was already preprocessed for M and S, however needs reprocessing anyway, since the preprocessed data for M above the split threshold was preprocessed with the assumption that no S grant is received, and further the data preprocessed for S above the data sent on the grant for S needs preprocessing since it is part of the remaining data and can only be sent via M. It is assumed here that the re-preprocessing can be done until the TTI n+1. That is, the re-preprocessing can start at TTI n after the UL grants are received. The re-preprocessing in this case implies among other things that the RLC SN needs to be updated to account for the fact that the data is to be transmitted on the M link. Thus, the SN needs to be assigned accordingly. Depending on how much larger the split threshold is chosen compared to the M grant size, this data would remain anyway longer in the queue so that multiple TTIs could be spend for preprocessing.

[0052] A further case is illustrated in FIG. 5A, where preprocessing of data is assumed to take 2 TTIs. As shown, the split threshold and the M-preProc threshold are configured larger than a multiple of the maximum grant size, or as multiples of what can be transmitted within one TTI. To allow for 2 TTI preprocessing, the thresholds should be set to more than 2 TTI (as shown in FIG. 5A, the thresholds are set to 3 times the maximum grant size). The same applies to the S-preProc threshold, for which the offset should also be larger than multiple TTIs. This is advantageous so that preprocessed data is also available at a potential S-grant (not shown).

[0053] In FIG. 5A, the preprocessed data ready for transmission is shown with diagonal striped lines, while the data currently being preprocessed and thus not ready for transmission yet is shown as horizontal or vertical striped lines.

[0054] Block based preprocessing

[0055] Due to the reordering functionality of PDCP in the receiver, PDCP PDUs are OK to be received out of sequence in the receiver. The out of sequence delivery would start the reordering timer, and the receiver waits to deliver subsequent PDUs until the outstanding PDUs are received (within reordering time).

[0056] By preprocessing of blocks of data and determining the block transmission direction already before the grant is received, these blocks of PDCP PDUs might be received out of sequence at the PDCP. This is however not an issue given the reordering process, i.e. this out of sequence delivery to PDCP would just become obvious as jitter to higher layers above PDCP . Therefore, allowing for preprocessing at the transmitter is traded for jitter introduced at the receiver. This should however not be considered a problem since it would become visible only at uplink queueing higher than the split threshold, for which anyway a considerable latency would be expected from the queuing time itself, being higher than the introduced jitter.

[0057] Dynamic changing of split threshold

[0058] In scenarios where throughput on M might be subject to sudden changes, e.g. for a fading channel, or for moving UEs, e.g. to the coverage edge, it is useful to be able to dynamically adapt the split threshold, so that transmission on S are enabled. In this case e.g. grants on M might be much lower than expected grants on S.

[0059] In one embodiment, the split threshold may be configurable dynamically, e.g. by

PDCP control signaling. A PDCP control PDU may be used for this purpose.

[0060] Starting threshold for S-preprocessing

[0061] So far, it was assumed that the starting point for the S-preprocessing is the split threshold. In case that dynamic changes of the split threshold are to be expected, it is however useful to decouple the starting point of the S-preprocessing from the split threshold. In one embodiment, the S-preprocessing-threshold-low (S-preProc-low) is defined as the starting point, and may be configurable by the network, or determined by the UE, e.g. based on measurements or estimations of coming grants.

[0062] This is useful, for example, if the split threshold needs to be suddenly adapted to a lower value, e.g. due to very low throughput on M, e.g. at a coverage edge. In this situation only very low M grant sizes can be expected and it is useful to send data on S. Therefore, the split threshold is set to a very low value. To allow immediately after this adaptation, transmission on S, data needs already to be preprocessed. Therefore, the S-preProc-low value should already before the split-threshold reconfiguration be set to a very low value, as illustrated in FIG. 5B.

[0063] Further embodiment - UE implementation based thresholds

[0064] In one embodiment the UE applies the preprocessing and may set at least one of split threshold, S-preProc threshold and M-preProc threshold according to measured, averaged over time window, or expected throughput or UL allocation size, measured, averaged over time window, or expected radio quality, or according to capabilities, or according to current or expected UL queueing buffer sizes. For example: if the UE expects a certain throughput on the secondary link, and/or data is buffered larger than the split threshold, the UE may for the next TTI (e.g., of 1ms) preprocess at least an amount of data equal or above the certain throughput, so that this amount of data can be send immediately (in the TTI for which an uplink grant is valid) if the uplink grant is received for this link. As a specific example, if the UE expects a throughput of lOOMbit/s on the secondary link, and data is buffered larger than the split threshold, the UE should for the next TTI (e.g., of 1ms) preprocess at least lOOMbit/s = lOOKbit/lms = 12.5kB of data so that this amount of data can be send immediately if an uplink grant is received for this link.

[0065] FIG. 6 is a block diagram of UE 102 according to some embodiments. As shown in FIG. 6, UE 102 may comprise: a data processing apparatus (DP A) 602, which may include one or more processors (P) 655 (e.g., a general purpose microprocessor and/or one or more other processors, such as an application specific integrated circuit (ASIC), field-programmable gate arrays (FPGAs), and the like); a transmitter 605 and a receiver 604 coupled to an antenna 622 for enabling UE 102 to transmit data to and receive data from an AN node (e.g., base station); and local storage unit (a.k.a., "data storage system") 608, which may include one or more nonvolatile storage devices and/or one or more volatile storage devices (e.g., random access memory (RAM)). In embodiments where UE 102 includes a general purpose microprocessor, a computer program product (CPP) 641 may be provided. CPP 641 includes a computer readable medium (CRM) 642 storing a computer program (CP) 643 comprising computer readable instructions (CRI) 644. CRM 642 may be a non-transitory computer readable medium, such as, but not limited, to magnetic media (e.g., a hard disk), optical media, memory devices (e.g., random access memory), and the like. In some embodiments, the CRI 644 of computer program 643 is configured such that when executed by data processing apparatus 602, the CRI causes UE 102 to perform steps described above (e.g., steps described above with reference to the flow charts). In other embodiments, UE 102 may be configured to perform steps described herein without the need for code. That is, for example, data processing apparatus 602 may consist merely of one or more ASICs. Hence, the features of the embodiments described herein may be implemented in hardware and/or software.

[0066] FIG. 7 is a diagram showing functional modules of UE 102 according to some embodiments. As shown in FIG. 7, UE 102 includes a preprocessing module 704 for preprocessing UL data prior to receiving a UL grant, a transmitting module 702 for employing a transmitter to transmit preprocessed data, a storing module 708 for storing data in a queue; and a receiving module 706 for employing a receiver to receive information (e.g., UL grants) transmitted by an AN node.

[0067] Further considerations:

[0068] For LTE dual connectivity, LTE DC, uplink split for the split bearer was introduced in Rel-13. This way, the UE is enabled to split its uplink PDCP queue among the two logical channels (RLCs) associated with the different cell groups (MAC entities) of MeNB and SeNB, and thus increase its uplink data rate. Hence, the UE performs "routing" of PDCP PDUs to the respective cell groups.

[0069] In LTE, according to 3GPP TS 36.323 (e.g. version 13.6.0), the scheme works as follows:

For split bearers, when indicating the data available for transmission to a MAC entity for BSR triggering and Buffer Size calculation, the UE shall:

- if ul-DataSplitThreshold is configured and the data available for transmission is larger than or equal to ul-DataSplitThreshold:

- indicate the data available for transmission to both the MAC entity configured for SCG and the MAC entity configured for MCG;

- else:

- if ul-DataSplitDRB-ViaSCG is set to TRUE by upper layer [3]:

- indicate the data available for transmission to the MAC entity configured for SCG only;

- if ul-DataSplitThreshold is configured, indicate the data available for transmission as 0 to the MAC entity configured for MCG;

- else:

- indicate the data available for transmission to the MAC entity configured for MCG only;

- if ul-DataSplitThreshold is configured, indicate the data available for transmission as 0 to the MAC entity configured for SCG. For split bearers, routing is performed in the transmitting PDCP entity, and reordering is performed in the receiving PDCP entity. When requested by lower layers to submit PDCP PDUs, the transmitting PDCP entity shall:

- submit the PDCP PDUs to either the associated AM RLC entity configured for SCG or the associated AM RLC entity configured for MCG, whichever the PDUs were requested by;

- else:

- if ul-DataSplitDRB-ViaSCG is set to TRUE by upper layers [3]:

- if the PDUs were requested by the associated lower layers configured for SCG:

- submit the PDCP PDUs to the associated AM RLC entity configured for SCG;

- else:

- if the PDUs were requested by the associated lower layers configured for MCG:

- submit the PDCP PDUs to the associated AM RLC entity configured for MCG.

[0070] Buffered data below the split threshold is only indicated as available for transmission for BSR reporting to the prioritized cell group, while when buffered data is above the split threshold, data is indicated to both cell groups. A transmission to lower layers is only done upon request from lower layers, i.e. at uplink grant reception. Otherwise the data remains on PDCP layer. At the time of the request, if the buffered data is below the threshold, data is only submitted to the prioritized cell group; if buffered data is above the threshold, data can be submitted to either MCG or SCG cell group, i.e. to the requesting cell group.

[0071] The following observations may be drawn:

• In LTE, the uplink split is configurable threshold based.

• Data below the threshold is reported as available and transmitted only via a

configurable prioritized cell group.

• If data is above the threshold, all data is reported to both cell groups, and

transmission of data above the threshold is possible via both cell groups. [0072] Alternative schemes to the split threshold based approach may be envisaged with respect to whether the UE maintains one PDCP buffer, from which data is pulled (as defined above), or whether the UE maintains two PDCP buffers (one for MeNB and one for SeNB), where incoming data is split into these two buffers based e.g. on a configured ratio. The latter approach would have the disadvantage of assigning PDCP data prematurely, leading to inefficiencies, in case the configured ratio does not correspond to the actual ratio of throughputs to MeNB and SeNB.

[0073] For the NR uplink split and PDCP design, the solution for LTE may be used as a baseline. For NR, one may consider the following further requirements as compared to LTE, for UL split design:

• NR link rates may be significantly different on MCG and SCG and may vary stronger. This applies both to NR-NR DC and NR-LTE DC, where MCG and SCG may correspond to links with different number of carriers, different carrier frequencies, different numerologies and TTI lengths. The UL split design should efficiently consider these varying throughputs on the links.

• Low latency: in NR a general design target is low latency, i.e. minimizing the processing and transmission time over L2, especially the best case latency when no queuing effects are present.

• Latency and processing difference among links: depending on numerology on NR link, or in case of LTE and NR integration, on difference in general latency and processing time, as well as backhaul connection, it is preferable to prioritize the faster link in case of no buffering instead of splitting data unnecessarily.

• Processing times in NR are significantly shorter, e.g. the time between grant reception and transmission. Therefore, preprocessing of data must be considered. Considering the example of single connectivity, at arrival of data in the uplink PDCP buffer, PDCP SN assignment, encryption, RLC SN assignment, as well as PDCP and RLC header construction can be done before an uplink grant becomes available. Also in NR UL split, preprocessing should be enabled as much as possible. [0074] Thus, UL split design in NR may consider: link rate differences, latency and processing time differences, as well as the generally significantly shorter latency and processing time in NR, i.e. between grant reception and transmission.

[0075] The split threshold base solution, can fulfil above-mentioned requirements for

NR:

• Data above the split threshold is split according to the received grants, i.e. according to the actual link data rate.

• Data below the split threshold is sent via a configurable prioritized link, which can be chosen to the be fastest link. This way, low latency and jitter is ensured.

• Preprocessing of data below and above the split threshold is possible, as further explained below.

[0076] Data below the split threshold may only be sent via the prioritized cell group.

Therefore, all PDCP data below the split threshold can be preprocessed readily before an uplink grant is received with the correct assumption that later transmission is done on the prioritized cell group. I.e. PDCP SN is assigned, encryption is done, as well as RLC SN of RLC associated with the prioritized cell group is done. When the uplink grant arrives, the data can be in-sequence transmitted via low layers. At the receiving side, PDCP would need to apply reordering only due to HARQ and RLC out of sequence deliveries within this cell group. This would lead to some jitter same as in single connectivity.

[0077] Data above the split threshold may be sent via one of both cell groups, depending on which cell group the uplink grant is available first. This way, all received grants can be utilized. By splitting the data, and due to the uncertainty about which grant arrives first, data may be transmitted and eventually received out of order at the PDCP receiver. The PDCP receiver applies reordering so that this out of order delivery will eventually be visible as jitter to higher layers, where the jitter relates to the skew time between the transmitted blocks among both the links. Preprocessing of data above the split threshold is also possible, so that it ready for transmission when a grant is available on MCG or SCG. [0078] Data above the split threshold can also be sent via the prioritized link, in case the split threshold is smaller than the maximum grant size on the prioritized link. This can be avoided by configuring higher split thresholds.

[0079] The UE could preprocess chunks of data according to expected grant sizes in subsequent TTIs.

[0080] If the UE also gives responsibility of this data to lower layers before the grant are received, the data needs to be transmitted also eventually via the lower layer the preprocessing was done for. This is not a problem if the grant is eventually received within the PDCP reordering time. This way, a potential delay because the grant was not received when/as large as expected, is eventually only visible as jitter to higher layers.

[0081] If the UE keeps all data on PDCP (e.g. assigns RLC SNs, but by this does not affect RLC transmitter state) and does only give responsibility of the data to lower layers once the grant is received, the before mentioned jitter can be mostly avoided. When the queue changes in between TTIs, e.g. due to received grant on one of the links, a re-preprocessing may to be done (e.g. discarding previous tentative SN assignments and assigning new SNs), so that at a next TTI preprocessed data is again available for the potentially received uplink grants.

[0082] It may thus be proposed to re-use LTE-UL-split threshold based solution in NR.

[0083] Furthermore, in LTE, the split threshold is configured based on RRC

reconfiguration. Since in NR however faster throughput changes are expected due to potential blocking and fading dips at higher carrier frequencies, it appears useful to make the split threshold in NR more dynamically configurable. This way, the UL transmission direction can be quickly adapted towards the link with the highest throughput.

[0084] In may be further proposed to make the split threshold configurable by PDCP control signaling (which by the network could be sent via the link with the highest throughput to the UE).

[0085] It may be further proposed to make NR UL split threshold configurable by RRC and dynamically adaptable by PDCP control signalling. [0086] The same reasoning applies to the PDCP parameter of prioritized link direction

(in LTE: ul-DataSplitDRB-ViaSCG). For example, in implementations, where the split threshold is not used, the uplink link direction should be quickly adaptable to the link with highest throughput, e.g. in case the other link suffers from throughput degradation, e.g. due to fading.

[0087] It may be further proposed to make the prioritized UL transmission direction

(similar as ul-DataSplitDRB-ViaSCG), configurable by RRC and dynamically adaptable by PDCP control signaling.

Claims

A method performed by a user equipment, UE (102), for managing or handling uplink, UL, data transmissions in a radio communications network, wherein the UE (102) is served by a first radio base station (103) and a second radio base station (104) providing dual connectivity to the UE (102) in the radio communications network, the method comprising:

• preprocessing (202) UL data prior to receiving a UL grant; and

• storing (208) the preprocessed data in an UL data buffer (302).

The method of claim 1 , wherein preprocessing the UL data comprises preprocessing (204) at least a first portion of the data specific to the first radio base station (103).

The method of claim 2, wherein a size of the first portion is less than a difference between a first preprocessing high threshold and a first preprocessing low threshold.

The method of any of claims 1-3, wherein preprocessing the first portion of the data specific to the first radio base station (103) comprises re-processing at least a subpart previously preprocessed specific to the second radio base station (104).

The method of any of claims 1-4, wherein preprocessing the data further comprises preprocessing (206) at least a second portion of the data specific to the second radio base station (104).

The method of claim 5, wherein a size of the second portion is less than a difference between a second preprocessing high threshold and a second preprocessing low threshold.

The method of any of claims 5-6, wherein the first portion and the second portion overlap.

The method of any of claims 1-7, further comprising:

• receiving a UL grant from the first or the second radio base station (104); and • in response to receiving the UL grant, transmitting a portion of the preprocessed data to the corresponding first or second radio base station (104).

9. The method of claim 8, wherein the portion of the preprocessed data to the corresponding first or second radio base station (104) is transmitted immediately in response to receiving the UL grant, e.g. in a TTI for which the uplink grant is valid.

10. The method of any of claims 8-9, wherein the UL grant is specific to the first radio base station (103), and the transmitted portion of data corresponds to a subset of the first portion.

11. The method of claim 8-9, wherein the UL grant is specific to the second radio base station (104), and the transmitted portion of data corresponds to a subset of the second portion.

12. The method of any of claims 1-7, further comprising:

• receiving a first UL grant from the first radio base station (103);

• receiving a second UL grant from the second radio base station (104);

• in response to receiving the first UL grant, transmitting a portion of the data in the UL data buffer (302) that corresponds to a subset of the first portion to the first radio base station (103); and

• in response to receiving the second UL grant, transmitting a portion of the data in the UL data buffer (302) that corresponds to a subset of the second portion to the second radio base station (104).

13. The method of claim 12, wherein the portion of the data in the UL data buffer (302) that corresponds to the subset of the first portion is transmitted in the TTI for which the first UL grant is valid and/or wherein the portion of the data in the UL data buffer (302) that corresponds to the subset of the second portion to the second radio base station (104) is transmitted in the TTI for which the second UL grant is valid.

14. The method of any of claims 1-13, further comprising:

• determining that a current size of the UL data buffer (302) is above or below a split threshold (304);

• in response to determining that the current size of the UL data buffer (302) is below a split threshold (304), sending a scheduling request, SR only to the first radio base station (103);

• in response to determining that the current size of the UL data buffer (302) is above a split threshold (304), sending a SR to both the first radio base station (103) and the second radio base station (104).

15. The method of claim 3, further comprising:

• receiving one or more of the first preprocessing high threshold, the first preprocessing low threshold, the second preprocessing high threshold, and the second preprocessing low threshold, from a node in the radio communications network.

16. The method of claim 15, wherein the received thresholds are updated based on current network conditions.

17. The method of claim 16, wherein the current network conditions comprise at least one of:

• the UE UL throughput to first radio base station (103) and/or to second radio base station (104),

• the ratio of throughput between the first radio base station (103) and the second radio base station (104),

• the TTI length for this UE (102), and

• the link quality (SINR) to first radio base station (103) and/or to second radio base station (104).

18. The method of claim 15, wherein the UE (102) receives the updated thresholds from the network and/or the UE (102) performs the updating.

19. The method of any of claims 1-18, wherein preprocessing the data in the UL data buffer (302) comprises processing at least a portion of the data in the UL data buffer (302) to create one or more packet data convergence protocol, PDCP, service data units, SDUs.

20. The method of claim 19, wherein creating one or more PDCP SDUs comprises one or more of the following steps:

• assigning a PDCP sequence number, SN to each of the PDCP SDUs;

• building a PDCP header for each of the PDCP SDUs;

• encrypting each of the PDCP SDUs to create one or more PDCP protocol data units, PDUs;

• assigning radio link control, RLC, SN to each of the PDCP PDUs; and

• building an RLC header for each of the PDCP PDUs.

21. The method of claim 1 , further comprising, prior to preprocessing the data, placing the data in the UL data buffer (302).

22. A user equipment, UE (102) adapted to manage or handle uplink, UL data transmissions in a radio communications network, wherein the UE (102) is served by a first radio base station (103) and a second radio base station (104) providing dual connectivity to the UE (102) in the radio communications network, the UE (102) comprising:

• a preprocessing module (704) configured to preprocess UL data prior to receiving a UL grant; and

• a storing module (708) configured to store the preprocessed data in an UL data buffer (302).

23. The UE (102) of claim 22, wherein the preprocessing module is further configured to

preprocess data by preprocessing at least a first portion of the data specific to the first radio base station (103).

24. The UE (102) of claim 23, wherein a size of the first portion is less than a difference between a first preprocessing high threshold and a first preprocessing low threshold.

25. The UE (102) of any of claims 23-24, wherein the preprocessing module (704) is further configured to preprocess the first portion of the data specific to the first radio base station (103) by re-processing at least a subpart previously preprocessed specific to the second radio base station (104).

26. The UE (102) of any of claims 22-25, wherein the preprocessing module (704) is further configured to preprocess the data by preprocessing at least a second portion of the data specific to the second radio base station (104).

27. The UE (102) of claim 26, wherein a size of the second portion is less than a difference between a second preprocessing high threshold and a second preprocessing low threshold.

28. The UE (102) of any of claims 26-27, wherein the first portion and the second portion overlap.

29. The UE (102) of any of claims 22-28, further comprising:

• a receiving module (706) configured to receive a UL grant from the first or the second radio base station (104); and

• a transmitting module (702) configured such that, in response to receiving the UL grant, the transmitting module (702) employs a transmitter to transmit a portion of the data in the UL data buffer (302) to the corresponding first or second radio base station (104), e.g. in the TTI for which the received UL grant is valid.

30. The UE (102) of claim 29, wherein the UL grant is specific to the first radio base station (103), and the transmitted portion of data corresponds to a subset of the first portion.

31. The UE (102) of claim 29, wherein the UL grant is specific to the second radio base station (104), and the transmitted portion of data corresponds to a subset of the second portion.

The UE (102) of any of claims 22-27, further comprising

the receiving module (706) further configured to receive a first UL grant from the first radio base station (103);

the receiving module (706) further configured to receive a second UL grant from the second radio base station (104);

the transmitting module (702) is configured such that, in response to receiving the first UL grant, the transmitting module (702) employs a transmitter to transmit a portion of the data in the UL data buffer (302) that corresponds to a subset of the first portion to the first radio base station (103); and

the transmitting module (702) is configured such that, in response to receiving the second UL grant, the transmitting module (702) employs a transmitter to transmit a portion of the data in the UL data buffer (302) that corresponds to a subset of the second portion to the second radio base station (104).

33. The UE (102) of any of claims 29-32, further comprising:

• the transmitting module (702) further configured to determine that a current size of the UL data buffer (302) is above or below a split threshold (304);

• the transmitting module (702) further configured to, in response to determining that the current size of the UL data buffer (302) is below a split threshold (304), send a scheduling request, SR, only to the first radio base station (103);

• the transmitting module (702) further configured to, in response to determining that the current size of the UL data buffer (302) is above a split threshold (304), send a SR to both of the first radio base station (103) and the second radio base station (104).

34. The UE (102) of any of claims 24-33, further comprising: • the receiving module (706) further configured to receive one or more of the first preprocessing high threshold, the first preprocessing low threshold, the second preprocessing high threshold, and the second preprocessing low threshold, from a node in the radio communications network.

35. The UE (102) of claim 34, wherein the received thresholds are updated based on current network conditions.

36. The UE (102) of any of claims 22-35, wherein the preprocessing module (704) is further configured to preprocess the data in the UL data buffer (302) by processing at least a portion of the data in the UL data buffer (302) to create one or more packet data convergence protocol, PDCP service data units, SDUs.

37. The UE (102) of claim 36, wherein the preprocessing module (704) is further configured to create one or more PDCP SDUs by one or more of the following steps:

• assigning a PDCP sequence number, SN, to each of the PDCP SDUs;

• building a PDCP header for each of the PDCP SDUs;

• assigning radio link control, RLC, SN to each of the PDCP PDUs; and

• building an RLC header for each of the PDCP PDUs.

38. A user equipment, UE (102) configured to manage or handle uplink, UL data transmissions in a radio communications network, wherein the UE (102) is served by a first radio base station (103) and a second radio base station (104) providing dual connectivity to the UE (102) in the radio communications network, the UE (102) comprising a processor and a computer readable medium, wherein the processor is configured to perform the steps of:

• preprocessing UL data prior to receiving a UL grant; and

• storing the preprocessed data in an UL data buffer (302).