WO2017088920A1

WO2017088920A1 - Error recovery mechanisms at lower layers of mobile communication systems

Info

Publication number: WO2017088920A1
Application number: PCT/EP2015/077705
Authority: WO
Inventors: Frank Frederiksen; Troels Emil Kolding; Gilberto BERARDINELLI; Klaus Ingemann Pedersen; Saeed Reza KHOSRAVIRAD; Juho Mikko Oskari Pirskanen
Original assignee: Nokia Solutions And Networks Oy
Priority date: 2015-11-25
Filing date: 2015-11-25
Publication date: 2017-06-01

Abstract

Configuration data is acquired (S21) at a transmitting apparatus, which indicates a number of transmission channels to be used for processing data packets at a receiving apparatus. The number of transmission channels is higher than a number of soft decoding buffers available at the receiving apparatus, the soft decoding buffers being usable for storing data packets to be retransmitted. The configuration data is transmitted (S22) to the receiving apparatus via an air interface. The receiving apparatus receives (S11) the configuration data from the transmitting apparatus and controls (S12) processing of data packets received from the transmitting apparatus based on the configuration data.

Description

ERROR RECOVERY MECHANISMS AT LOWER LAYERS OF MOBILE

COMMUNICATION SYSTEMS

DESCRIPTION

BACKGROUND OF THE INVENTION

Field of the invention The present invention relates to error recovery mechanisms at lower layers of mobile communication systems. In particular, the present invention relates to HARQ operation at a lower layer of a protocol stack.

Related background Art

The following meanings for the abbreviations used in this specification apply:

2G 2^nd Generation

3G 3^rd Generation

5G 5^th Generation

AP Access Point

ARQ Automatic repeat request

BB Baseband

CAPEX Capital Expenditure

DCI Downlink Control Information

EDGE Enhanced Data Rates for GSM Evolution

GERAN GSM EDGE Radio Access Network

GRPS General Packet Radio Service

GSM Global System for Mobile communications

HARQ Hybrid ARQ

HSPA High Speed Packet Access

LTE Long Term Evolution

LTE-A Long Term Evolution - Advanced

OFDM Orthogonal Frequency Division Multiplexing

OPEX Operational Expenditure

PDCCH Physical Downlink Control Channel PUCCH Physical Uplink Control Channel

PDSCH Physical Downlink Shared Channel

PUSCH Physical Uplink Shared Channel

RLC Radio Link Control

RRC Radio Resource Control

RRH Remote Radio Head

SAW Stop-and-Wait

TCP Transmission Control Protocol

TTI Transmission Time Interval

UE User Equipment

UMTS Universal Mobile Telecommunication System

W-CDMA Wideband-Code Division Multiple Access

WiMAX Worldwide Interoperability for Micro-wave Access In order to optimize spectral efficiency of mobile communication systems there are mechanisms to address the need to recover from transmission errors over the wireless transmission channel and to allow a scheduling node to transmit with aggressive data rates across the channel. For example, such error recovery mechanisms comprise:

- forward error correction through channel coding for correcting single bit errors,

- hybrid automatic repeat request (HARQ), reusing already transmitted energy when processing retransmitted packets (also referred to as soft combining principle), and

- higher layer retransmission protocols, like RLC layer or TCP layer

retransmissions.

In current LTE implementation, there are two different HARQ implementations with respect to timing. In LTE downlink, asynchronous HARQ is used, which means that the transmission timing of a scheduled retransmission can vary in time with a minimum latency for processing, while LTE uplink uses the concept of synchronous HARQ, meaning that one HARQ SAW process is tied to the system timing and can only occur at predetermined time instants (e.g. in case of LTE UL

- for each 8 ms). One of the challenges in the current system design of LTE is that it is very much focused on pre-determined and standardized processing time limits in both ends of a communication link. For instance, the current LTE specification setup does not bring any benefits in case an eNB or UE is capable of doing faster processing. Correspondingly, there is no room for addressing needs for longer processing times, which may cause the system to stall. SUMMARY OF THE INVENTION

The present invention is concerned with enabling insertion of latency in a mobile communication system also in a centralized scenario while maintaining end-user throughput close to peak without significantly increasing a data buffer in a receiving apparatus.

This is achieved by the methods, apparatuses and computer program product as defined in the appended claims. According to an aspect of the present invention, a mixed HARQ and ARQ operation within a transmission pipeline towards a given UE is provided. In particular, the UE is made aware of two specific parameters related to the transmission: (a) number of HARQ buffers for soft combining, and (b) the maximum number of "addressable channels", i.e. channels that can be signaled for transmission, wherein the number of addressable channels is higher than the number of HARQ buffers for soft combining. In other words, only some addressable channels are protected with HARQ and allow for soft combining, while other addressable channels are not protected by HARQ. Although the description uses HARQ as a non-limiting example, any soft combining or other technique using previously received information is applicable to the invention.

According to a non-limiting embodiment of the invention, in an example with eight addressable channels and four HARQ processes, a fixed mapping is adopted such that transmission IDs 0-3 are protected by HARQ process ID 0-3, while transmission IDs 4-7 are not associated with any HARQ buffers. According to another embodiment of the invention, a more sophisticated option is adopted in which the UE is allowed to assign internal HARQ process numbers to received data. The UE assigns a data packet to a HARQ buffer when a reception fails, and hence the UE needs several pending retransmissions before running out of HARQ buffers. In this way, it is possible to fill the transmission pipeline with data packets without running out of HARQ buffers.

According to an embodiment of the invention, a mixed HARQ and ARQ operation mode is adopted in the downlink direction of a radio access network, and an eNB functions as transmitter of data packets while a UE functions as receiver of the data packets.

According to another embodiment of the invention the mixed HARQ and ARQ operation mode is adopted in the uplink direction of the radio access network, and the UE functions as transmitter of data packets and while the eNB functions as receiver of the data packets.

In the following the invention will be described by way of embodiments thereof with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments described may be implemented in a radio system, such as in at least one of the following : Worldwide Interoperability for Micro-wave Access

(WiMAX), Global System for Mobile communications (GSM, 2G), GSM EDGE radio access Network (GERAN), General Packet Radio Service (GRPS), Universal Mobile Telecommunication System (UMTS, 3G) based on basic wideband-code division multiple access (W-CDMA), high-speed packet access (HSPA), Long Term

Evolution (LTE), LTE-Advanced, and/or 5G system.

Fig. 1 shows a schematic diagram illustrating a fronthaul and air interface propagation delay between a centralized BB site and a UE. Fig. 2 shows a schematic block diagram illustrating a configuration of control units in which examples of embodiments of the invention are implementable.

Fig. 3 shows flowcharts illustrating processes A and B which implement examples of embodiments of the invention.

Fig. 4 shows a diagram illustrating an HARQ scheme used in LTE and HSPA with a different number of HARQ buffers and different delays. Fig. 5 shows a diagram illustrating an HARQ scheme used in LTE and HSPA with a different number of HARQ buffers and different delays, and under the operation of long fronthaul delays which cause a gap of four TTIs.

Fig. 6 shows a diagram illustrating an HARQ scheme used in LTE and HSPA with eight HARQ processes, and under the operation of long fronthaul delays.

Fig. 7 shows a diagram illustrating a mixed HARQ-ARQ scheme under the operation of long fronthaul delays according to an embodiment of the invention. Fig. 8 shows a diagram illustrating a mixed HARQ-ARQ scheme with eight HARQ processes under the operation of long fronthaul delays according to an embodiment of the invention.

Fig. 9 shows a diagram illustrating the mixed HARQ-ARQ scheme according to the embodiment of the invention depicted in Fig. 7 with many HARQ

retransmissions.

DESCRIPTION OF THE EMBODIMENTS In today's mobile communication systems such as e.g. HSPA or LTE, typically a number of parallel stop-and-wait HARQ channels are defined to ensure continuous transmission on an air interface. For FDD LTE, a timing setup is such that there are eight HARQ SAW channels defined for each link direction, while for different configurations of LTE TDD, there is a different number of HARQ SAW channels defined. This is to accommodate for differences in signaling delays as well as availability of control channels when providing these associated control channels such as PDCCH and PUCCH for feedback for the data channels.

Further, ARQ is a technique used in case a receiver does not store any further information related to originally received data packets. In ARQ, a receiver simply informs a transmitter that a data packet failure occurred, and discard the received data packet. In any subsequent retransmission, the receiver evaluates newly received data without consideration of earlier information. In HARQ, the receiver stores the previously received information and does combination with new information to achieve a "combination gain" prior to decoding. By combining old and new data, it is possible to utilize transmission energy already invested also for retransmissions. The drawback of HARQ is that it requires more memory at the receiver while associated signaling becomes slightly more advanced.

However, the benefit from HARQ is in the order of 1-3 dB on top of gain from regular ARQ.

It is to be noted that the problems and principles outlined are applicable to FDD operation as well as TDD operation. As mentioned in the foregoing, current wireless network systems such as LTE are typically designed according to fixed and well-defined processing delays. For instance, FDD DL HARQ structure allows for 3 ms of processing time at a UE, such that the UE should start transmitting the A/N associated with the downlink data, in the 4^th TTI in the uplink (4 ms after the reception of the PDCCH for the assignment followed by the data in PDSCH). In order to allow for corresponding processing in a base station (eNB), this processing delay is also defined as 3 ms (after the TTI where the PUCCH is received). Correspondingly, for the uplink transmissions, there is a four TTI delay from the PDCCH with the UL assignment until the UE should be transmitting in the PUSCH. This delay is to allow for the UE to prepare for the data transmission. After the reception of the PUSCH, the eNB is allowed at maximum three TTIs (3 ms) for processing the received data until the UE should be instructed to make a retransmission or a new

transmission. If, by any chance, the eNB is suffering from a processing delay, it will not be able to provide the scheduling instruction for either downlink or uplink direction at the right time, so that there will be a missing transmit opportunity. In the downlink case, the result typically is a 1 ms delay or gap of transmission, while for the uplink which uses synchronized HARQ in LTE, the result is in the order of 8 ms delay or gap. Such a phenomenon is denoted HARQ stall. In deployments of mobile communication systems, operators may like to create centralized processing for multiple nodes. Such centralized processing brings significant benefits in terms of processing power and maintenance (e.g. both CAPEX and OPEX savings), but has the drawback that there is an additional signaling delay from a processing unit to a physical transmit antenna which is also denoted fronthaul delay. Depending on an operator infrastructure, there may be relatively large fronthaul delays, which can pose a challenge for the overall network performance. Such fronthaul delays typically are included into the processing delay budget, and potentially put extreme requirements on real processing to meet delay requirements and thereby support continuous transmission towards a single UE. The problem may be further exacerbated by tighter latency requirements. Fronthaul delays may in some implementations include pre-processing and post-processing delays to allow for conversion to suitable transmission format on the fronthaul link. Fig. 1 shows a schematic diagram illustrating a fronthaul and air interface propagation delay between a centralized BB site and a UE 1. In particular, the centralized BB site is connected via a fronthaul interface to a remote RF site, assuming a fronthaul latency of TFH seconds. In an embodiment, the BB site may comprise a remote control unit operatively coupled (e.g. via a wireless or wired network) to a remote radio head (RRH) located in the RF site. In an embodiment, at least some of the described processes may be performed by the remote control unit. In an embodiment, the execution of the processes may be shared among the RRH in the RF site and the control unit in the BB site. The air interface forming communication media between the RF site and the UE 1 is subject to a propagation delay denoted T_pro_P. It is to be noted that the schematic diagram of Fig. 1 is simplified in the sense that only one remote RF site is pictured, while in reality, the centralized BB site will of course have fronthaul connections towards a large number of RF sites. An apparatus comprising the centralized BB site and the remote RF site is denoted with reference sign "2". As a preliminary matter before exploring details of various implementations of embodiments of the present invention, reference is made to Fig. 2 for illustrating a simplified block diagram of various electronic devices that are suitable for use in practicing the embodiments of this invention.

Fig. 2 shows a control unit 10 which may be part of and/or used by the UE 1, and includes processing resources (e.g. processing circuitry) 11, memory resources (e.g. memory circuitry) 12 that may store a program, and interfaces (e.g.

interface circuitry) 13 which may comprise a suitable radio frequency (RF) transceiver (not shown) coupled to one or more antennas (not shown) for bidirectional wireless communications over one or more wireless links 15 e.g. with the RF site or the apparatus 2. The processing resources 11, memory resources 12 and interfaces 13 are connected via a bus 14. According to an embodiment of the invention, the processing resources 11, memory resources 12 and interfaces 13 implement receiving means and controlling means.

In addition, Fig. 2 shows a control unit 20 which may be part of and/or used by the RF site and/or the BB site, and includes processing resources (e.g.

processing circuitry) 21, memory resources (e.g. memory circuitry) 22 that may store a program, and interfaces (e.g. interface circuitry) 23 which may comprise a suitable radio frequency (RF) transceiver (not shown) coupled to one or more antennas (not shown) for bidirectional wireless communications over one or more wireless links 15 e.g. with the UE 1. The processing resources 21, memory resources 22 and interfaces 23 are connected via a bus 24. According to an embodiment of the invention, the processing resources 21, memory resources 22 and interfaces 23 implement acquisition means and transmitting means.

The programs stored in the memory resources 12, 22 are assumed to include program instructions that, when executed by the associated processing resources 11, 21, enable the electronic device to operate in accordance with the exemplary embodiments of this invention, as detailed below. Inherent in the processing resources 11, 21 is a clock to enable synchronism among the various apparatus for transmissions and receptions within the appropriate time intervals and slots required, as the scheduling grants and the granted resources/subframes are time dependent. The transceivers include both transmitter and receiver, and inherent in each is a modulator/demodulator commonly known as a modem.

In general, the exemplary embodiments of this invention may be implemented by computer software stored in the memory resources 12, 22 and executable by the processing resources 11, 21, or by hardware, or by a combination of software and/or firmware and hardware in any or all of the devices shown.

In general, the various embodiments of the UE 1 can include, but are not limited to, mobile stations, cellular telephones, personal digital assistants (PDAs) having wireless communication capabilities, portable computers having wireless communication capabilities, image capture devices such as digital cameras having wireless communication capabilities, gaming devices having wireless communication capabilities, music storage and playback appliances having wireless communication capabilities, Internet appliances permitting wireless Internet access and browsing, as well as portable units or terminals that incorporate combinations of such functions.

The memory resources 12, 22 may be of any type suitable to the local technical environment and may be implemented using any suitable data storage technology, such as semiconductor-based memory devices, magnetic memory devices and systems, optical memory devices and systems, fixed memory and removable memory. The processing resources 11, 21 may be of any type suitable to the local technical environment, and may include one or more of general purpose computers, special purpose computers, microprocessors, digital signal processors (DSPs) and processors based on a multi-core processor architecture, as non-limiting examples.

Further, as used in this application, the term "circuitry" refers to all of the following :

(a) hardware-only circuit implementations (such as implementations in only analog and/or digital circuitry) and

(b) to combinations of circuits and software (and/or firmware), such as (as applicable): (i) to a combination of processor(s) or (ii) to portions of

processor(s)/software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions) and

(c) to circuits, such as a microprocessor(s) or a portion of a microprocessor(s), that require software or firmware for operation, even if the software or firmware is not physically present.

This definition of "circuitry" applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term "circuitry" would also cover an implementation of merely a processor (or multiple processors) or portion of a processor and its (or their) accompanying software and/or firmware. The term "circuitry" would also cover, for example and if applicable to the particular claim element, a baseband integrated circuit or applications processor integrated circuit for a mobile phone or a similar integrated circuit in server, a cellular network device, or other network device.

Now reference is made to Fig. 3 showing flowcharts illustrating processes A and B which implement examples of embodiments of the invention.

Process A is carried out by a transmitting apparatus transmitting data packets to a receiving apparatus via an air interface. Process B is carried out by the receiving apparatus. In a downlink direction, the transmitting apparatus may comprise the RF site and/or the BB site illustrated in Fig. 1, and the receiving apparatus may comprise the UE 1 illustrated in Fig. l. In an uplink direction, the transmitting apparatus may comprise the UE 1 illustrated in Fig. l, and the receiving apparatus may comprise the RF site and/or the BB site illustrated in Fig. 1.

In step S21 of process A, configuration data is acquired, which indicate a number of transmission channels (also referred to as transmission opportunities or transmission identifications (IDs)) to be used for processing data packets (to be used for signaling for transmission) at the receiving apparatus. The number of transmission channels is higher than a number of soft decoding buffers available at the receiving apparatus, which are usable for storing data packets to be retransmitted by the transmitting apparatus. In step S22, the configuration data is transmitted to the receiving apparatus via the air interface. 6. Then, process A ends.

In step Sl l of process B, the configuration data is received from the transmitting apparatus via the air interface.

In step S12, processing of data packets received from the transmitting apparatus is controlled based on the configuration data. According to an embodiment of the invention, the soft decoding buffers comprise buffers to be used in a hybrid automatic repeat request (HARQ) operation mode.

According to an embodiment of the invention, the configuration data is transmitted to the receiving apparatus in control signaling along with payload data. In some embodiments of the invention the configuration data is

transmitted as a separate control message as part of the downlink control information (DCI), while the data part of the message is sent as part of the downlink shared channel. In an embodiment of the invention, the initial configuration of the downlink control information (DCI) content is configured and defined through radio resource control (RRC) messages.

According to an embodiment of the invention, the receiving apparatus provides the soft decoding buffers of a number lower than the number of transmission channels indicated in the configuration data.

According to an embodiment of the invention, the configuration data further indicates a number of the soft decoding buffers, and the receiving apparatus provides the soft decoding buffers of a number equal to the number indicated in the configuration data.

According to an embodiment of the invention, the receiving apparatus allocates a transmission channel to a soft decoding buffer for storing a data packet transmitted from the transmitting apparatus the detection of which has failed at the receiving apparatus.

According to an embodiment of the invention, the receiving apparatus sends an acknowledgment/negative acknowledgment message to the transmitting apparatus, which indicates whether a soft decoding buffer is used for storing an already received data packet.

According to an embodiment of the invention, the transmitting apparatus, e.g. a scheduling node of the BB site illustrated in Fig. 1, determines the number of transmission channels based on expected delays in transmission of data packets to the receiving apparatus.

According to an embodiment of the invention, the transmitting apparatus, e.g. the scheduling node, schedules retransmissions of data packets to the receiving apparatus based on an estimated state of storing data packets in the soft decoding buffers at the receiving apparatus.

According to an embodiment of the invention, the transmitting apparatus, e.g. the scheduling node, determines that the receiving apparatus has not stored a data packet to be retransmitted in the soft decoding buffers, and uses a robust coding scheme for retransmission of the data packet.

According to an embodiment of the invention, the transmitting apparatus, e.g. the scheduling node, determines that the soft decoding buffers each are occupied by data packets, and causes the RF site to transmit further data packets to the receiving apparatus anyway.

In an embodiment, the transmitting apparatus does not send the soft buffer identifier (ID) to the receiving apparatus along with the payload data. Instead an identifier of transmission channel (transmission ID) currently used for this payload data is transmitted to the receiver device along with the payload data. In some embodiments, the transmission ID is indicated in another channel such as the downlink control channel, while the actual data is carried on the downlink shared channel. In principle, these two IDs may be similar for the case where the receiver has sufficient amount of soft buffers, but in case the receiver is not having sufficient amount of buffers, these may vary. By receiving the

transmission ID, the receiver has the information needed to make a mapping from "transmission process ID" to the soft buffers being used - retransmissions would be scheduled to the same "transmission process ID", no matter whether it is an ARQ retransmission or a HARQ retransmission. In an embodiment, the resources for indicating the transmission ID from the transmitter to the receiver may utilize the resources currently used for the transmission of the soft buffer ID (e.g. HARQ ID). Thus, no extra resources are needed.

As described above, a framework is introduced which adopts a mixture of HARQ and ARQ operation modes within a transmission pipeline towards a receiving apparatus, e.g. a UE. The UE is made aware of a maximum number of transmission channels as "addressable channels", i.e. the channels that can be signaled for transmission. The UE may be made aware also of a number of HARQ buffers for soft combining. The number of addressable channels is higher than the number of HARQ buffers for soft combining. The number of addressable channels defines the maximum number of channels that can be used to transmit separate data to the receiver in parallel without the transmitter knowing the outcome of the reception (successful or unsuccessful).

This approach allows avoiding HARQ stall situations, as will be discussed in the following implementation examples of the present invention. For describing operation of an implementation example of the present invention, an example system with 4 HARQ SAW channels and a possibility to signal 8 channels for transmission is assumed. In this implementation example, the UE as receiving apparatus has four buffers for storing "soft data" for retransmissions, while it also has a "decoding buffer" for any new data arriving. The principle set out in the following generally is applicable to any system aiming at providing continuous flow of data towards a UE, even when long fronthaul delays are experienced.

Fig. 4 illustrates operation of a HARQ scheme used in LTE and HSPA with short fronthaul delay without adopting a mixture of HARQ and ARQ operation modes according to embodiments of the present invention. Scheduling is performed within a frame. Downlink direction of packet data transmission from an eNB as transmitting apparatus to a UE as receiving apparatus is considered.

The upper row in Fig. 4 indicates packet numbers of data packets of a data packet stream, the row in the middle packet numbers of the data packets for transmission in a transmission pipeline, and the lower row an HARQ process number. E∑l denotes a transmission failure of a data packet, and EQ denotes a successful retransmission of a data packet.

As shown in Fig. 4, through normal processing, the eNB is able to fill all the available HARQ buffers in the UE, and there is no need for introducing additional signaling. In particular, as can be seen in Fig. 4, two initial transmissions fail, the ones for data packets 1 and 2 of the data packet stream. It is also seen that the retransmission (with normal HARQ) for packet number 1 fails, and for packet number 2 succeeds. Following this, another retransmission for the packet number 1 is performed, and in the second retransmission attempt (third transmission attempt) it succeeds. In the lower row of Fig. 4 the HARQ process number used for each transmission is indicated.

Now, considering the case of long fronthaul delays, there is a situation as shown in Fig. 5, where the system experiences a long fronthaul delay of 4 TTIs marked by § in Fig. 5. Fig. 5 illustrates the HARQ scheme used in LTE and HSPA with long fronthaul delay without adopting a mixture of HARQ and ARQ operation modes according to embodiments of the present invention.

The upper row in Fig. 5 indicates packet numbers of data packets of a data packet stream, the row in the middle packet numbers of the data packets for transmission in a transmission pipeline, and the lower row an HARQ process number. denotes a transmission failure of a data packet, and EE denotes a successful retransmission of a data packet.

In the situation illustrated in Fig. 5, an HARQ stall is caused, where the eNB is not able to fill the transmission pipeline towards the UE due to lack of signaling space and due to lack of HARQ buffers to catch the data transmitted towards the UE. The problem is that at data packet #4 of the upper row, the ACK/NACK signal from the UE may have arrived at the radio head, but not at the central node, so the central processing node cannot do any scheduling at this time. Only when the information is available at the central node, it can be used for scheduling, and even after that, the scheduling decision needs to travel to the radio head and over the air interface. Hence, the "black-out" illustrated in Fig. 5 occurs. Fig. 6 illustrates a similar situation as depicted in Fig. 5. The first row from the top indicates the time, the second row data packet numbers in a transmission process at the eNB, the third row the data packet numbers in a transmission process at an RRH, the fourth row transmission IDs in an

acknowledgment/negative acknowledgment process at the UE, and the sixth row the transmission IDs in an acknowledgment/negative acknowledgment process at the eNB. The eNB may be seen here as the central node and RRH as the remote radio head (access point).

Fig. 6 shows an HARQ scheme used in LTE and HSPA with long fronthaul delay without adopting a mixture of HARQ and ARQ operation modes according to embodiments of the present invention, in which eight HARQ processes (and hence eight TX processes) are assumed.

According to Fig. 6, at time #0 the eNB decides to transmit DL data to the UE.

At time #3, the DL data arrives at the RRH from eNB with a 3ms fronthaul delay, and at the UE, i.e. no transmission or propagation delay is assumed in between RRH and UE. At time #7, the UE finishes processing of 4ms which includes time for receiving a data packet, and sends e.g. NACK, and the RRH receives the NACK from the UE.

At time #10, the eNB acquired the NACK from the RRH due to the 3ms fronthaul delay. At time #14, the eNB has processed the NACK (processing time of 4ms, including time for receiving the signal) and decides to retransmit DL data to the UE. Due to the fronthaul delays the eNB is not able to schedule any new

transmissions or retransmissions for HARQ processes since no knowledge exists relating to the success or failure of the transmission. This is due to the fact that the number of soft decoding buffers and transmission channels is the same, in this case eight. In the example shown in Fig. 6, the eNB is not able to schedule new data until time 14, meaning that there are six "empty" TTIs, and a loss of UE throughput of 6/14 = ~43%.

Fig. 7 shows an implementation example in which a mixture of HARQ and ARQ operation modes according to an embodiment of the present invention is applied. The upper row in Fig. 7 indicates packet numbers of data packets of a data packet stream, the row in the middle packet numbers of the data packets for transmission in a transmission pipeline, and the lower row an HARQ process number. ED denotes a transmission failure of a data packet, and 1-1-1 denotes a successful retransmission of a data packet.

In the implementation example of the invention shown in Fig. 7, the eNB (or a scheduling node) can indicate up to eight transmission IDs for data packets, while the UE maintains e.g. four HARQ buffers, i.e., assigns a data packet to an HARQ buffer when transmission of the data packet from the eNB to the UE fails. This means that it is possible for the eNB to transmit up to 8 parallel SAW channels, but only 4 of these will be under HARQ protection. In this way, it is possible to fill the transmission pipeline with data without running out of HARQ buffers. Only if there are several pending retransmissions the UE will run out of HARQ buffers.

The UE may in an embodiment do the book-keeping of which HARQ buffers are associated to which transmission IDs locally without following any rules from eNB. In prior art the eNB transmits an indication of the HARQ buffer ID to the UE, but this may not be needed in some embodiments of the invention. Assuming the eNB is the sender, the eNB may track the amount of negative acknowledgements (NACKs) from the UE and if the number of current NACKs exceeds the number of HARQ buffers available at the UE, the eNB knows that now it needs to transmit the current data packet again without soft combining possibility. In a more advanced embodiment, the UE may be able to transmit more than just the simple ACK/NACK in the uplink and may be able to indicate whether soft buffers were used or not for this particular packet corresponding to the ACK/NACK. The indication may be comprised in the ACK/NACK, or the indication may define the type of ACK/NACK used. In one embodiment, the time delay from the transmission of the packet to the reception of the ACK/NACK indicates whether soft buffers were used or not for this particular packet. E.g. a time delay of x TTIs denotes yes and a time delay of something else denotes soft buffers were not used. Similarly, some predefined coding (e.g. predefined MCS or codeword) used for the ACK/NACK signaling may indicate that soft buffer was used while another coding may indicate that soft buffers were not used. This way, the normal ACK/NACK need not be amended or any new bits need not be added to the ACK/NACK.

The association between transmission ID and HARQ buffer at the UE could be implemented in a number of ways. Possible options may include e.g. make a fixed mapping such that transmission ID 0-3 are protected by HARQ process ID 0-3, while transmission IDs 4-7 are not associated with any HARQ buffers. This may provide for simplicity, but may not be the optimum in terms of buffer utilization.

In another embodiment, the UE may assign soft buffers autonomously based on failed decoding attempts. Such an approach may complicate the detection process in the eNB correspondingly, as the eNB may need to keep track of which transmission IDs are protected by HARQ and which are not stored in HARQ buffers in the UE.

Let us take a look at this via a non-limiting example. The eNB starts transmission towards a UE with an aggressive MCS selection, and the first 4 receptions fail, and the UE will store soft data for these 4 transmission processes (indicated with transmission process ID 0, 1, 2, 3 to the UE). Now, the eNB continues

transmission towards the same UE with too aggressive MCS selection, and next 4 receptions also fail. As the UE does not have soft buffers for these processes, the UE will not store any soft values.

Then, the eNB starts retransmitting (as it now has the NACK from process #0), and UE will be combining data for this HARQ process (even when there is no direct indication from eNB to the UE of which HARQ process ID - the UE knows this by receiving the transmission ID (transmission process ID)).

At process #4, the UE does not have any information stored, so here the UE needs to use a "direct decoding buffer" to estimate the data packet received. In other words, the UE does not have any soft buffer for this data packet. Basically, the UE maintains an association between pending non-successful receptions stored in the soft buffers. Whenever a transmission process ID fails, this is coupled to the soft buffer for later processing when the same transmission process ID is scheduled from the eNB. It may be also worth noting that the timing for the HARQ roundtrip time (RTT) may not be fixed, as the eNB would always have the possibility of postponing the retransmissions. However, the HARQ processing times in the UE side may be constant (that is, a delay from packet is received until HARQ ACK/NACK is transmitted may be fixed). Thus, the embodiments are applicable also to asynchronous HARQ.

Fig. 8 illustrates fronthaul and processing delays of the implementation example shown in Fig. 7. The first row from the top indicates the time, the second row data packet numbers in a transmission process at the eNB, the third row the data packet numbers in a transmission process at an RRH, the fourth row transmission IDs in an acknowledgment/negative acknowledgment process at the UE, and the sixth row the transmission IDs in an acknowledgment/negative acknowledgment process at the eNB. Again, the eNB may be considered as the cloud (BB site in Fig. 1) and the RRH as the access point (RF site in Fig. 1). According to Fig. 8, the eNB (or a scheduling node) has indicated sixteen transmission IDs for data packets in configuration data sent to the UE, and the UE maintains eight HARQ buffers. The UE may then assign a data packet (e.g. data packet #0 in Fig. 8) to an HARQ buffer when transmission of the data packet #0 from the eNB to the UE fails. The UE may, in an embodiment, autonomously decide to which HARQ buffer this data packet is assigned. The UE may receive a transmission ID along with the data packet so as to help the UE to associate future data packets with the same transmission ID to the same HARQ buffer for soft decoding purposes.

In one embodiment, the RRH is "dumb" in the sense that it simply transmits information received from the eNB ("central unit" possibly managing multiple RRHs). In other words, the RRH may not make modifications to the information received from the eNB. In another embodiment, the RRH may modify the content received from the eNB, such as the transmission ID.

Further, in at least one embodiment of the invention, the eNB is not required to know which HARQ buffer ID the UE has assigned for a transmission ID. The eNB only may be required to know whether a data packet is protected by HARQ or not. Hence, there may not be any direct connection or association between transmission ID and HARQ buffer ID.

It is further noted here that, in Fig. 8, the rows of Ack/Nack signaling show UE understanding of time. Hence, a data packet received in time (or TTI) 0 has an Ack/Nack transmission in time (or TTI) 4.

According to Fig. 8, at time #0 the eNB decides to transmit DL data to the UE.

At time #3, the DL data arrives at the RRH from eNB due to the 3ms fronthaul delay, and at the UE, i.e. no transmission or propagation delay is assumed in between RRH and UE.

At time #7, the UE finishes processing of the data packet. This processing is assumed to take e.g. 4ms, including time to receive and process the data packet, as well as preparing for the transmission of the ACK/NACK signal. Then the UE sends e.g. NACK, and the RRH receives the NACK from the UE.

At time #10, the eNB acquired the NACK from the RRH due to the 3ms fronthaul delay.

At time #14, the eNB has processed the NACK (processing time of 4ms) and decides to retransmit the data packet to the UE. In the implementation example of Fig. 8, only one HARQ buffer is shown occupied for simplicity of representation, and the other transmission IDs can be used for processing data packets #1 to #15. In this example where eight soft decoding (HARQ) buffers are used, seven more data packets out of these 15 data packets can be protected under the HARQ (in case of failed transmissions). When the transmission is successful, the corresponding transmission ID (and possibly the soft decoding buffer used for this transmission) may be made available for a new transmission.

In this way, it is possible to fill the transmission pipeline with data packets. As can be seen when comparing Fig. 8 to prior art Fig. 6, there are no "empty" TTIs in Fig. 8. The eNB may schedule new transmissions #8-#15 to the UE even when it has no knowledge relating to the success or failure of the previous

transmissions #0-#7. This is due to the fact that the eNB may exceed the number of soft decoding buffers.

Fig. 9 illustrates the implementation example of Figs. 7 and 8 in a situation with many packet errors. There is a burst of packet errors extending over the number of HARQ buffers available (it is assumed that four HARQ buffers are available), and the UE runs out of HARQ buffers. As shown in Fig. 9, data packet #5 does not have any available HARQ buffer to assign for later HARQ combining and, hence, the received soft data here is discarded (marked with "x"). When the eNB (or the scheduling node) performs retransmission of data packet #5 (illustrated by the UE performs regular decoding corresponding to layer 1 ARQ instead of HARQ. The scheduling node is aware of the UE discarding data packet #5 e.g. due to the sequence of negative acknowledgements received from the UE, and can take this fact into account when scheduling data packet #5 once more (i.e. at a later time). Knowing that data packet #5 has not been received in a proper way allows the eNB (or the scheduling node) to schedule the data with more robustness, for instance using higher power, more robust MCS, or more physical resources, i.e., "better" channel coding.

In terms of buffer management, the eNB may or may not be aware of how the UE arranges its memory for the different buffers for addressing HARQ

retransmissions.

In case the eNB is not able to have information on or assess which buffers are protected by HARQ and which are only protected by ARQ, uncertainty on the preferred retransmission scheme arises. In an embodiment of the invention this is addressed through autonomous detection of which packets are protected by HARQ buffers and which are not at the eNB through monitoring of the uplink ACK-NACK status messages from the UE. In another embodiment of the invention, there are rules as to which HARQ processes are protected by soft buffers, and which processes are just protected by ARQ operations, that is, the UE does not store any soft values for a selected set of buffers. For example, process IDs (or transmission IDs) 0-3 are protected by soft buffers, while process IDs (transmission IDs) 4-7 are not in any way protected by HARQ and thereby these are just subject to ARQ operation.

In yet another embodiment of the invention, the UE is allowed additional freedom in terms of autonomously assigning memory to different HARQ buffers. For this case, the UE may be able to provide HARQ buffers for all pending retransmissions (for cases with low data rate, but with a continuous transmission ongoing) by more intelligent memory handling. For such cases, the eNB also has to monitor and keep track of the used memory for HARQ operations, such that it has an assessment of how many processes are protected by HARQ, and which are only protected through ARQ. As indicated above, the present invention is not limited to the downlink di rection . In the uplink direction, there may be a case that the RF site and/or BB site, i .e. network side, is not sending DL ACK/NACK messages fast enough, and the UE only has l imited amount of Tx buffers to ma intai n data packets for tra nsmission to the network side. In LTE, there is a possibility to toggle "New Data indication" in UL grant to control whether the UE sends HARQ re-tra nsmission or new data packets. However, this means that the UE has to keep "old" data packets i n L2 buffer, leading to an i ncrease of L2 memory. According to a n embodiment of the invention, i n long latency deployments, there may be more uplink "cha nnels" than HARQ Tx buffers in the UE, and when the UE transmits a data packet to the channel that it will not receive HARQ

ACK/NACK for it may trash the data packet immediately and reuse that part of memory immediately. Thus, the network side may send UL grants first to n "channels" supporting HARQ re-transmission, and if processes are sti ll requiring feedback and the network side does not wish to trash the data in those processes yet, the network side may send gra nts to extra cha nnels constantly where the UE trashes the data packet i mmediately without possi bility of HARQ feedback. In practice, the above processing comprises a partial de-activation of HARQ in the uplink direction.

Some benefits from this approach include the fol lowing . In case of UE memory reduction, but static eNB latency requirements, the UE would be able to use less soft buffer memory for its HARQ implementation. In case of increased eNB latency requirements for long fronthaul delays, it is possible to "stretch" the UE memory such that the eNB can continue to have a transmission towards the UE in all TTIs, even if it may experience longer frontha ul delays compared to current assumption of maximum Vi-l ms for fronthaul delay. With this approach, it is possible to sacrifice link performance for "filling the pipeline", and the reduced link performance (for some of the transmission processes) can in pri nciple be predicted a nd compensated in the scheduling algorithm .

It is to be understood that the above description is illustrative of the invention and is not to be construed as li miting the invention. Various modifications and applications may occur to those skilled in the art without departing from the true spirit and scope of the invention as defined by the appended claims.

Claims

CLAIMS:

1. A method comprising :

acquiring configuration data indicating a number of transmission channels to be used for processing data packets at a receiving apparatus, that is higher than a number of soft decoding buffers available at the receiving apparatus, the soft decoding buffers being usable for storing data packets to be retransmitted; and

transmitting the configuration data to the receiving apparatus via an air interface.

2. The method of claim 1, wherein the configuration data further indicates a number of the soft decoding buffers.

3. The method of claim 1 or 2, comprising :

determining the number of transmission channels based on expected delays in a transmission process of data packets to the receiving apparatus.

4. The method of any one of claims 1 to 3, comprising :

modifying a downlink control information format to include information on the number of transmission channels.

5. The method of any one of claims 1 to 4, comprising :

scheduling retransmissions of data packets to the receiving apparatus based on an estimated state of storing data packets in the soft decoding buffers at the receiving apparatus.

6. The method of any one of claims 1 to 5, comprising :

determining that the receiving apparatus has not stored a data packet to be retransmitted in the soft decoding buffers; and

using a more robust coding scheme for retransmission of the data packet than used for retransmissions enjoying soft decoding.

7. The method of any one of claims 1 to 6, wherein the configuration data including an identifier of a current transmission channel is transmitted to the receiving apparatus along with payload data.

8. The method of any one of claims 1 to 7, comprising :

determining that the soft decoding buffers each are occupied by data packets; and

transmitting further data packets to the receiving apparatus.

9. The method of any one of claims 1 to 8, wherein

the method is, at least partly, executed by a transmitting apparatus, and/or

the method is, at least partly, executed by a scheduling node.

10. A method for use by a receiving apparatus, the method comprising :

receiving, via an air interface, configuration data from a transmitting apparatus, the configuration data indicating a number of transmission channels to be used for processing data packets at the receiving apparatus, that is higher than a number of soft decoding buffers available at the receiving apparatus, the soft decoding being usable for storing data packets to be retransmitted; and controlling processing of data packets received from the transmitting apparatus based on the configuration data.

11. The method of claim 10, comprising :

providing the soft decoding buffers of a number lower than the number of transmission channels indicated in the configuration data; and/or

wherein the configuration data further indicates a number of the soft decoding buffers, and the method comprises:

providing the soft decoding buffers of a number equal to the number indicated in the configuration data.

12. The method of claim 10 or 11, comprising :

autonomously allocating a transmission channel to a certain soft decoding buffer for storing a data packet transmitted via the transmission channel from the transmitting apparatus, wherein detection of the transmitted data packet has failed at the receiving apparatus.

13. The method of any one of claims 10 to 12, comprising:

sending an acknowledgment/negative acknowledgment message to the transmitting apparatus, which indicates whether a soft decoding buffer is used for storing an already received data packet.

14. The method of claim 13, wherein the acknowledgment/negative

acknowledgment message to be sent to the transmitting apparatus is different when the soft decoding buffer is used for the data packet compared to a case when the soft decoding buffer is not used for the data packet.

15. The method of any one of claims 1 to 14, wherein

the receiving apparatus comprises a user equipment, and

the transmitting apparatus comprises an eNodeB or an access node, or the receiving apparatus comprises an eNodeB or an access node, and the transmitting apparatus comprises a user equipment, and/or the scheduling node comprises a central node managing the transmitting apparatus, and/or

the transmission channels comprise transmission opportunities and/or transmission identifications, and/or

the soft decoding buffers comprise buffers to be used in a hybrid automatic repeat request (HARQ) operation mode.

16. A computer program product including a program for a processing device, comprising software code portions for performing the steps of any one of claims 1 to 15 when the program is run on the processing device.

17. The computer program product according to claim 16, wherein the computer program product comprises a computer-readable medium on which the software code portions are stored.

18. The computer program product according to claim 16, wherein the program is directly loadable into an internal memory of the processing device.

19. An apparatus comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to perform :

20. The apparatus of claim 19, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform :

indicating in the configuration data a number of soft decoding channels to be used at the receiving apparatus; and/or

determining the number of transmission channels based on expected delays in a transmission process.

21. The apparatus of claim 19 or 20, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform :

22. The apparatus of any one of claims 19 to 21, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform :

23. The apparatus of any one of claims 19 to 22, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform :

24. The apparatus of any one of claims 19 to 23, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform :

transmitting the configuration data including an identifier of a current transmission channel to the receiving apparatus along with payload data.

25. The apparatus of any one of claims 19 to 24, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform :

determining that the soft coding buffers each are occupied by data packets; and

transmitting further data packets to the receiving apparatus.

26. The apparatus of any one of claims 19 to 25, wherein

the apparatus comprises a transmitting apparatus, or

the apparatus comprises a transmitting apparatus and a scheduling node.

27. A receiving apparatus comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to perform :

receiving, via an air interface, configuration data from a transmitting apparatus, the configuration data indicating a number of transmission channels to be used for processing data packets at the receiving apparatus, that is higher than a number of soft decoding buffers available at the receiving apparatus, the soft decoding buffers being usable for storing data packets to be retransmitted; and controlling processing of data packets received from the transmitting apparatus based on the configuration data.

28. The receiving apparatus of claim 27, and the at least one memory and the computer program code are configured to, with the at least one processor, cause the receiving apparatus to perform :

wherein the configuration data further indicates a number of the soft decoding buffers, and the at least one memory and the computer program code are configured to, with the at least one processor, cause the receiving apparatus to perform :

providing soft decoding buffers of a number equal to the number indicated in the configuration data.

29. The receiving apparatus of claim 27 or 28, the at least one memory and the computer program code configured to, with the at least one processor, cause the receiving apparatus to perform :

30. The receiving apparatus of any one of claims 27 to 29, the at least one memory and the computer program code configured to, with the at least one processor, cause the receiving apparatus to perform :

31. The receiving apparatus of claim 30, the at least one memory and the computer program code configured to, with the at least one processor, cause the receiving apparatus to perform :

sending a different acknowledgment/negative acknowledgment message to the transmitting apparatus when the soft decoding buffer is used for the data packet compared to a case when the soft decoding buffer is not used for the data packet.

32. An apparatus comprising :

acquisition means for acquiring configuration data indicating a number of transmission channels to be used for processing data packets at a receiving apparatus, that is higher than a number of soft decoding buffers available at the receiving apparatus, the soft decoding buffers being usable for storing data packets to be retransmitted; and

transmission means for transmitting the configuration data to the receiving apparatus via an air interface.

33. A receiving apparatus comprising :

reception means for receiving configuration data from a transmitting apparatus via an air interface, the configuration data indicating a number of transmission channels to be used for processing data packets at the receiving apparatus, that is higher than a number of soft decoding buffers available at the receiving apparatus, the soft decoding buffers being usable for storing data packets to be retransmitted; and

control means for controlling processing of data packets received from the transmitting apparatus based on the configuration data.