Methods and apparatus for processing error control messages in a wireless communication system

The present invention generally relates to wireless communication systems, and more specifically to the processing of error control messages in wireless communication systems that schedule link resources.

3GPP (The Third-Generation Partnership Project) has launched a program to develop advanced wireless communication system standards under the initiative known as "Long Term Evolution" or LTE. In the standardization discussion, it has been agreed that LTE systems will utilize the automatic repeat request (ARQ) mechanism in the wireless link control (RLC) protocol. Its specified ARQ protocol is a selective repeat protocol (used in RLC Acknowledged Mode). Protocol), which provides a means for the sending node to poll the receiving node for status, as well as a means for sending status reports from the receiving node to the sending node. Depending on the receipt of the status report, the sending node may resend any of the lost data, or may take other appropriate action. In response to polling, the receiver typically sends a status report. However, sending status reports may be prohibited under certain circumstances. For example, a status-prohibition timer started when the last status report was sent can block the status report for a period of time.

The developers of 3GPP also agreed to specify a timer that regulates the node's response to either polling or status reporting, along with a set of polling and status triggers. Examples of triggers and timers soon agreed to be included in the 3GPP LTE standard include: · Automatic status report on detection of missing protocol data unit (PDU); · Automatic polling in response to the transmission of the last PDU in the transmit buffer, which provides the transmitting node with confirmation that the burst of data has been completely received; Status to prevent nodes from sending status reports too frequently-prohibition timers, because excessively frequent status reports cause unnecessary retransmissions; -Polling to ensure that the polling that would have been missing was retransmitted-Retransit timer

Of course, additional triggers and timers may also be adopted in LTE. It can be expected that the RLC ARQ protocol for LTE will ultimately have many similarities to the wideband CDMA (W-CDMA) RLC protocol specified in 3GPP TS 25.322. .. In addition to Acknowledged Mode, the RLC protocol for LTE will include transparent mode as well as unacknowledged mode.

Traditional RLC In the ARQ scheme, RLC timers and status reports are generated in response to certain triggers. For example, if the receiving node receives the poll and the status prohibition timer is not running, the receiving node immediately generates a status report showing the current status of the receiver. A typical status report may include an identifier for the last protocol data unit (PDU) received and / or a negative response for one or more PDUs that were not successfully received. The status report is then provided to the media access control (MAC) layer for transmission to the sending node. (So-called one of ordinary skill in the art will understand that each wireless communication node typically includes a transmitter and a receiver. Further, the ARQ scheme may be implemented in both directions. Purpose of the present disclosure. For this reason, the term "transmitting node" generally refers to a node that sends one or more data PDUs to a "receiving node" in confirmation mode. In 3GPP technical terms, confirmation mode data refers to "confirmation mode." Sent from the "sender" of the RLC entity; the PDU is transmitted to the "peer entity" or the "receiver" of the RLC entity in confirmation mode. Given this usage, the receiving node has a status on the sending node. A PDU can be sent. Similarly, a transmitting node can receive a status PDU.)

In systems that utilize the status-prohibition timer, the receiving node typically activates the status-prohibition timer when the status report is transferred from the RLC layer to the MAC layer. Then, until the timer expires, no further status reports are allowed, even if one or more new triggers for the status report occur in the meantime. Such triggers can be other polls received from the transmitting node, or detection of missing PDUs. Thus, the status prohibition time ensures that subsequent status reports are delayed by at least the period specified by the status-prohibition timer.

<p> In LTE systems, the uplink (for the transmission of mobile base stations) is the resource to be scheduled, and scheduling is controlled by the serving base station (called advanced node B or eNodeB in LTE). As a result, at a given moment, the mobile station will not have immediate access to transmit resources. If the MAC layer at the mobile station receives a status report from the RLC layer when the uplink resource is not scheduled, the mobile station must first request the resource before it can send the status report to eNodeB. Since scheduling is controlled by eNodeB, the permission of uplink resources may be significantly delayed. For example, multi-user scheduling can delay resource allocation, or initial scheduling requests can be lost on transmission. As a result, the transmission of the status report can be delayed to the extent that the state of the receiver described by the status report becomes obsolete even before its transmission.</p>

<p> Disclosed herein are methods and devices for processing error control messages such as RLC ARQ control messages. An example method is to signal the media access controller (MAC) that a link resource is needed to send the data and to indicate that the link resource to send the data is scheduled. Includes receiving from the MAC and generating an error control message based on the current error control status after receiving the above indicator. The error control message is then forwarded to the MAC for transmission. Queuing and because the generation of error control messages is delayed until their transmission is scheduled.<u style="single">Obsolete</u>(stale) Sending control messages is avoided.</p><p> In another aspect of the invention, the MAC may notify the wireless link controller (RLC) when transmission of the RLC ARQ control message is initiated or acknowledged. In response to the notification, the RLC controller may activate or resume the error control timer. The timer includes, for example, a polling timer or a status prohibition timer.</p>

<figref num="1">It is a simplified figure of the communication system which concerns on one Embodiment of this invention.</figref><figref num="2">The hierarchy of several communication protocols that can be utilized in the system of Fig. 1 is shown.</figref><figref num="3">It shows the effect of scheduling delays on the formation of ARQ status reports.</figref><figref num="4">It shows the effect of scheduling delay on the ARQ polling timer.</figref><figref num="5">It shows the timing of generating the ARQ status report according to one embodiment.</figref><figref num="6">The polling timer processing according to one embodiment of the present invention is shown.</figref><figref num="7">It is a logical flow diagram which shows the method as an example of processing an error control message.</figref><figref num="8">It is a block diagram of the wireless device which concerns on one or more embodiments of this invention.</figref>

FIG. 1 is a simplified diagram of communication system 10, which includes a transmitting node 110 and a receiving node 120. As discussed above, each of the transmitting node 110 and the receiving node 120 has a complete transmitter / receiver; the terms "transmit" and "receive" describe a particular endpoint in data transmission where acknowledgments are made. Used to do. Therefore, the transmitting node 110 transmits one or more data units including the protocol data unit (PDU) to the receiving node 120, and also transmits one or more error control messages such as a poll request. (Error control messages such as polling may be contained in the same protocol data unit as traffic data, as discussed below.) And the receiving node 120 has one or more error control messages such as status reports. May request to resend the PDU that failed to send or receive.

In some embodiments of the present invention, the transmitting node 110 may be configured by LTE eNodeB and the receiving node 120 may be configured by LTE compliant mobile stations. In that case, the data unit depicted in Figure 1 is transmitted over the downlink by eNodeB, and one or more error control messages are transmitted over the uplink. However, those skilled in the art will appreciate that the ARQ scheme can be implemented for the detection of errors in the transmission of PDUs over the uplink, even in the opposite direction. In that case, the roles of the transmitting and receiving nodes are reversed between the eNodeB and the mobile station.

The techniques disclosed herein will be described with reference to LTE systems, but the invention is not limited to such systems. In fact, a practitioner with technical knowledge who has read the following description and viewed the accompanying drawings will describe here to a variety of wireless systems, especially systems that dynamically schedule transmit resources on uplinks and / or downlinks. You will understand that the techniques to be applied can be applied.

Each of the communication nodes in Figure 1 is configured to operate according to a particular communication protocol, such as the LTE protocol specified by 3GPP. Figure 2 shows several protocol hierarchies; at each of the communication nodes, analog and digital hardware, programmable processors configured with the appropriate software, or a combination thereof. Can be implemented using. In particular, the transmitting node 110, which may be an LTE mobile station, may use, for example, the protocol hierarchy in the protocol stack 210 that communicates with the corresponding protocol layer in the protocol stack 220 of the receiving node 120, which is LTE eNodeB.

Each of the protocol stacks 210 and 220 includes a physical layer, a data link layer, and a network layer. The data link layer is divided into two sub-layers, the wireless link control (RLC) layer and the media access control (MAC) layer. In this embodiment as an example, the network layer is divided into a control plane protocol (RRC) and a user plane protocol (IP).

In LTE systems, the physical layer uses Orthogonal Frequency Division Multiple Access (OFDMA) technology for downlinks and the closely related Single Carrier Frequency Division Multiplexing Access (SC-FDMA) for uplinks. In general, the physical layer provides data transmission over an air (wireless) interface, transport channel multiplexing and demultiplexing, transport channel mapping onto physical channels, physical channel modulation and demodulation, forward. It has functions such as direction error correction coding and decoding, frequency and time synchronization, transmission power control, and RF processing.

The media access control (MAC) layer generally provides unacknowledged transmission of service data units (SDUs) between peer MAC entities. The MAC feature chooses the appropriate transport format depending on the data rate for each transport channel, handles priorities between different user data flows (in base stations that support multiple users), and controls messages. It may include scheduling, as well as multiplexing and demultiplexing of higher layer PDUs. In LTE systems, resource scheduling is also performed by the MAC layer. In particular, uplink resources can be requested by the MAC layer for the mobile station and are allocated between the mobile stations by the corresponding MAC layer in the eNodeB.

The RLC layer establishes, releases and maintains RLC connections, segments and rejoins, concatenates, and retransmits (ARQ) error correction between variable-length upper layer PDUs with smaller RLC PDUs, upper layer PDUs Performs a variety of functions, including in-sequence delivery, duplicate detection, flow control, and other functions.

The RRC protocol deals with control signaling on wireless interfaces such as radio access bearer control signaling, measurement reporting, and handover signaling. The user plane portion of the network layer contains traditional functions performed by Layer 3 protocols such as the well-known Internet Protocol (IP).

The wireless link control (RLC) protocol hierarchy within the protocol stacks 210 and 220 includes an automatic repeat request (ARQ) mechanism. The RLC layer in the transmitting node 110 receives user data, segments it, and converts it into an RLC PDU. In some embodiments, the transmitted RLC PDU may include a field indicating whether the transmitted message is a data PDU or a control PDU. Some other fields correspond to polling fields and may contain bits indicating that the sending node 110 wants a status report from the receiving node. The RLC PDU may further include a "sequence number" field indicating the sequence number of the data PDU; this sequence number may be added for each new data PDU. Finally, the data field contains a segment of higher level data information. The Length Indicator and extended E fields may also be included in the RLC PDU.

Depending on the PDU with the polling bit P set to 1, the RLC layer of the receiving node may generate a status report indicating that the RLC PDU was received correctly. Positive or negative acknowledgments, or a combination of both, can be used. In LTE, status report PDUs include an acknowledgment sequence number (ACK_SN) field that indicates the smallest sequence of PDUs that were not received by the receiving node and no missing was detected. This status report PDU may also contain one or more negative response sequence numbers (NACK_SN) that identify the PDUs for which missing was detected by the receiving node. Therefore, when the transmitting node 110 receives the status report PDU, all PDUs up to the PDU corresponding to ACK_SN (not including itself) are received except for the PDU identified by one or more NACK_SN fields. It is judged.

As briefly mentioned above, dynamic scheduling of resources in some systems is between the generation of error control messages such as polling PDUs or status report PDUs discussed here and the actual transmission of those messages. It can cause delays. 3GPP In systems where transmit resources are permanently available, such as broadband CDMA systems, when RLC control messages (eg status report PDUs or polling PDUs) are generated, they typically, except for short processing delays. Will be sent immediately. So, for example, when a status report is sent, it generally represents an accurate "snapshot" of the state of the receiver. In LTE, in contrast, uplinks are tightly scheduled and mobile stations typically do not have any permanent resources. If the mobile station is not scheduled, the mobile station's MAC layer must first request an uplink resource before the status report PDU can be sent. The resulting delay can make the status report PDU outdated before it is actually sent to eNodeB.

This problem is shown in Figure 3, where events in eNodeB are shown along the upper horizon and events in mobile stations are shown along the lower horizon. The event flow in Fig. 3 starts by sending a poll request from eNodeB to the mobile station, as shown on the left in the figure. As discussed above, in traditional systems, mobile stations generate status reports immediately (except for processing delays). In LTE systems, as discussed above, this status report may indicate that one or more confirmation mode PDUs (or some of them) are missing or failed to process.

When the RLC layer generates a status report, the status report is forwarded to the MAC layer for delivery to eNodeB. However, in the drawn event flow, resources are not immediately available on the uplink when trying to send data. Therefore, the MAC layer requests the uplink resource from the eNodeB by transmitting the scheduling request. As discussed above, permitting uplink resources is subject to considerable delays. Such delays can occur simply because the eNodeB is servicing a large number of mobile stations, or because the eNodeB is allocating resources to higher priority requests at that time. In some cases, failure to receive a resource request by eNodeB causes a delay or exacerbates the delay, in which case the request must be retransmitted.

In either case, the uplink resource authorization in the event flow depicted in Figure 3 is ultimately received by the mobile station after a significant scheduling delay. The duration of this delay is shown in Figure 3 from the RLC's point of view, from the moment the error control message (status report) is forwarded to the MAC by the RLC until the MAC receives the permission of the uplink resource. .. During this period, some additional RLC PDUs, PDU1 and PDU2, are received by the mobile station. Therefore, as shown on the right in the figure, the status report is already obsolete by the time the status report is sent to eNodeB. Since the status report does not reflect the reception of PUD1 and PDU2, eNodeB is provided with inaccurate information as the current status of the mobile station receiver. This causes wasted downlink resources in the retransmission of PDU1 and PDU2, and potentially further delays in queuing data.

Scheduling delays can also cause problems with the operation of timers associated with error control processes. For example, if the RLC timer that controls the transmission of the status report, such as the status-prohibit timer, is activated when the PDU carrying the status report is supplied to the MAC layer, the timer may end too early. It can happen. In the worst case, the RLC layer can supply several RLC status reports to the MAC layer, all of which are queued for transmission. If these status reports contain a negative response for the same RLC PDU, the peer entity will resend the same PDU over and over again.

The same problem can occur with timers that control polling. If the polling timer is activated when the polling-carrying PDU is supplied to the MAC layer, it is possible that the polling timer elapses too quickly due to the delay in sending the polling-carrying PDU. This can result in unnecessary polling being queued and sent to the receiver. This is shown in the event flow in Figure 4. Similar to Figure 3, events at eNodeB are shown along the upper horizon, and events at mobile stations are shown along the lower horizon. On the left side of the figure, the polling control message POLL1 is generated by the RLC layer of eNodeB. The polling control message is immediately forwarded to the MAC layer for transmission to the mobile station, activating a polling timer that establishes the minimum delay before other polling requests are generated. However, the POLL1 message is not actually sent to the mobile station until after a considerable delay, which can cause a scheduling delay from the backup send queue.

In some cases, the delay in sending a poll request can be extended until after the poll timer expires, as shown in FIG. When the poll timer expires, the RLC layer, unaware that the previous poll was not sent, generates a second poll request POLL2. This second poll request is forwarded to the MAC and the polling timer is restarted.

Finally, the first poll request (POLL1) is sent to the mobile station. In the scenario shown, the uplink delay is not significant and the status report is responded quickly. Soon, a second poll request (POLL2 sent by eNodeB and received by the mobile station. Status-prohibition timer prohibits the mobile station from generating and sending other status reports, but a second poll request. Is obviously unnecessary and a waste of system resources.

Figures 3 and 4 show only some of the timing issues of RLC error control that can result from scheduling and queuing delays in wireless systems with scheduled resources. Another problem is that status reports or polling can be delayed because higher priority data is already queued. Consider a mobile terminal with multiple bearers or "logical channels" with different priorities (receive node from the downlink point of view, transmit node from the uplink point of view). The mobile terminal has data in the transmit buffer for the highest priority bearer and also receives scheduling permission from eNodeB, leaving bandwidth for transmission of the PDU associated with the lower priority bearer. It is assumed that it has not been done. Further assume that a trigger has occurred to send a status report or poll request on a low priority bearer. Status reporting or polling is significantly delayed because the mobile terminal does not have the resources to send its PDU. As in the scenario discussed above, this status report is in some cases obsolete before it is sent. In extreme cases, a number of status reports or polls are queued before low priority bearer resources are available, causing unnecessary retransmissions.

The solution to some of the problems mentioned above is the traditional ARQ process described above, where the resource for sending the RLC PDU contains the contents of the error control message (eg, the RLC PDU that carries the ARQ control information). The modification is to generate only when it can be supplied. In some embodiments, this can be achieved by providing an additional or modified interface between the RLC layer and the MAC layer. This additional interface allows the RLC layer to require the MAC layer to send RLC PDUs that carry control information. In addition, this additional interface allows the MAC layer to notify the RLC layer when link resources are available for transmission of the RLC PDU layer carrying control information. By this method, the RLC layer can substantially delay the generation of control information up to the resource, so that the error control information finally delivered contains the updated information of the state of the RLC layer. ..

On this interface, the RLC layer of a communication node (eg, a mobile station) may first report to the MAC layer the need to send RLC ARQ control messages, such as status reports. After notifying the MAC layer to the RLC layer that the required resources are available (or soon available), the RLC layer generates relevant control information (eg status report information). Package the control information into an RLC PDU and submit the RLC PDU to the MAC layer for transmission to the remote node. Embodiments of the invention include the case where the RLC PDU carries either an RLC status report or polling, but can be applied to other error control messages as well.

FIG. 5 shows an event flow diagram similar to FIGS. 3 and 4, and shows the operation of the system according to some embodiments of the present invention. As in the case in FIG. 3, on the left side of the figure, the polling request is transmitted by eNodeB and received by the mobile station. In response to this trigger, the mobile station's RLC layer determines that an uplink resource will be needed to send the error control message, which in this case is the status report. Therefore, as shown in FIG. 5, the RLC layer signals to the MAC layer that an uplink resource is required. The MAC layer responds by requesting uplink resources from the scheduler if the resources have not yet been scheduled.

So-called one of ordinary skill in the art may or may not specifically indicate that the RLC-to-MAC layer signaling that an uplink resource is required indicates that the resource is required to send an error control message. You will understand the good things. Thus, in some embodiments, the signaling may simply indicate that the RLC PDU is pending and that the uplink resource should be scheduled if it is not yet available. In other embodiments, signaling that specifically indicates that the resource is needed for the control message can be beneficial.

At any event, substantial resource permission is obtained only after a significant delay, as shown in Figure 5. During this delay, multiple confirmation mode RLC PDUs, PDU3 and PDU4, are received at the mobile station during the scheduling delay. However, in this scenario, the status report is not generated and is not queued in the MAC layer as waiting for transmission. Instead, in Figure 5, the RCL layer is in the status report data until after the MAC layer notifies the RLC that the uplink resource request has been granted (ie, the uplink resource is available). Delay generation. Therefore, the status report has the data at that time (including the status of PDU3 and PDU4) when it is transferred to the MAC layer and transmitted to eNodeB. Although there are still processing and queuing delays, these delays are minimal compared to the situation depicted in Figure 3.

Some embodiments of the present invention utilize similar techniques for starting and resuming timers for error control processing such as RLC ARQ polling and status-prohibition timers discussed above. In these embodiments, the activation of the error control timer is by the MAC layer (to the RLC layer) that the RLC PDU carrying the error control information has been or will be transmitted to the remote node. Can be triggered by a notification.

In FIG. 6, an operation as an example of such an embodiment is depicted. As in the case of the event flow depicted in Figure 4, this event flow triggers a poll request on the eNodeB. The RLC PDU carrying the poll request is forwarded to the MAC for transmission to the mobile station. However, in this case, the polling timer does not start immediately. Instead, the polling timer does not start until after the MAC layer notifies the RLC layer that a polling request has been sent. As depicted in Figure 6, it can occur after significant scheduling / queuing delays. Since the activation of the polling timer was delayed until or before or after the polling request was actually sent, the timer does not expire before the status report is received from the mobile station. Therefore, unnecessary retransmission of the poll request is avoided.

In some embodiments, the MAC layer may be configured to notify the RLC layer that transmission of a PDU carrying RLC ARQ control information has begun or is about to begin. In other embodiments, the MAC layer may instead notify the RLC layer that an acknowledgment has been made on the MAC HARQ layer for the transmission of the PDU carrying the RLC ARQ control information. Those skilled in the art will appreciate that the techniques described above in terms of polling timers can also be applied to other error control timers, such as status-prohibited timers.

FIG. 7 is a logical flow diagram showing, for example, a method as an example for processing error control messages that can be executed by, for example, one or more RLC and MAC controllers described above. In the illustrated method, the techniques for the invention disclosed herein apply to both the generation of error control messages and the activation of error control timers. Those skilled in the art will understand that, of course, many embodiments of the present invention can be applied to both error control processes, while some can only be applied to one, not both. Let's do it.

For any event, the method as an example of FIG. 7 begins at block 710 with signaling to the MAC controller at the communication node that a link resource is needed to send the error control message. As discussed above, this can be triggered by any of a number of different events. For example, receiving a poll request at the receiving node will generally trigger status reporting processing. Then, in this case, the signaling for the link resource in block 710 is for the resource for sending the status report. Another possible trigger event is the expiration of the error control timer. For example, the expiration of the polling timer could trigger a new polling request, in which case the signaling in block 710 would require resources to trigger a new polling request.

At any event, at block 720, the MAC controller receives an indicator that the link resource is available. As discussed in detail above, this can occur immediately after a request for a resource, or after a significant scheduling delay. In either case, an error control message is generated in block 730 in response to the indicator that the link resource is available. Therefore, the content of the error control message is generated based on the status at that time, and the content does not become "stale" by any scheduling delay. At block 740, the error control message is forwarded to the MAC controller for transmission. Since the message generation of block 730 and the message transfer of block 740 are delayed until after the resources are available, the transmission delay after transfer to the MAC controller is minimized.

At block 750, the MAC controller receives a notification that the transmission of the error control message has started or completed. In response to this notification, an appropriate error control timer, such as a polling timer or a status-prohibition timer, is activated, as shown in block 760.

The RLC and MAC procedures described herein can be implemented by the RLC and MAC controllers that implement the RLC and MAC layers of the protocol stacks 210 and 220 described above, respectively. Those skilled in the art may implement these procedures by modifying conventional RLC and MAC controllers, as described above by one or more programmable processors, hardware circuits, or a combination thereof. You will understand that it is okay.

Further, the methods disclosed herein can be implemented at the ends of either or both of the wireless links such as the LTE mobile station or eNodeB described above. Thus, FIG. 8 shows the general features of a wireless communication device according to one or more embodiments of the present invention; the illustrated wireless device 800 is, in various embodiments, a mobile terminal (cellular telephone, It can include wireless PDA (personal digital assistant), wireless PC (personal computer), machine-to-machine (machine-to-machine) devices, etc.), base stations, repeaters, or other nodes that terminate wireless links.

The wireless device 800 of FIG. 8 has a wireless transmitter / receiver 810 capable of operating to communicate with the remote transmitter / receiver via the antenna 815 on one or more wireless links. In some embodiments, the radio transmitter / receiver 810 is configured to transmit and receive signals formatted according to a standard, such as one of the radio standards published by 3GPP. For example, radio transmitter / receiver 810 may be configured to transmit and / or receive OFDMA and SC-FDMA signals according to LTE standards.

The wireless device 800 further has a media access control function 820, a wireless link control function 830, and other processing 840. The general functionality of the MAC and RLC features is discussed above; these features may be implemented on any of the various combinations of programmable processors configured with analog and digital hardware and software. Those skilled in the art will appreciate that these and other functions required for the operation of the wireless device 800 can be implemented using one or more programmable processors. In many embodiments, the MAC820 and RLC830 functions are implemented on a single microprocessor or purpose-built integrated circuit configured with software that performs the various RLC and MAC functions described herein, eg, FIG. Implemented as a protocol stack such as Protocol Stack 210.

For example, the RLC controller 830 can be implemented with a microprocessor programmed with software that defines the RLC layer, and the media access controller states that the RLC layer requires link resources to transmit data. Signals to, receives an indication from the media access controller that the link resource for sending the data has been scheduled, and an error control message based on the error control status for the current RLC layer in response to the indication. Is configured to generate. The error control message may include, but is not limited to, a polling request or status report. In some embodiments, the RLC layer may be further configured to receive a notification that transmission of an error control message has begun and to activate an error control timer in response to the notification. The error control timer can include, but is not limited to, a polling timer or a status-prohibited timer.

Similarly, all or part of the MAC controller 820 may be implemented on the same microprocessor programmed with software defining the MAC layer, or on one or more other microprocessors. The MAC layer is configured to receive a signal from the RLC layer indicating that a link resource is required for data transmission and request the link resource as needed. The MAC layer is further configured to notify the RLC layer after receiving permission for the resource. In some configurations, the MAC layer is used when an error control message (provided by the RLC layer to the MAC layer) is transmitted, or in some embodiments, when an error control transmission is imminent. Further configured to notify the RLC layer.

Of course, the content of the present disclosure may be carried out in a manner different from that specifically exemplified herein, without departing from the essential features of the present invention. The present embodiment is considered in all respects as an explanation rather than a limitation, and all modifications within the meaning and equality of the appended claims are intended to be included herein. ..