Methods, systems, and computer readable media for message flood suppression during access node-gateway (an-gw) unavailability and after an-gw restoration

TECHNICAL FIELD

0001The subject matter described herein relates to reducing or preventing message flooding in mobile communications networks. More particularly, the subject matter described herein relates to suppressing message flooding between a policy and charging enforcement function (PCEF) or packet gateway (PGW) and a policy and charging rules function (PCRF) during AN-GW unavailability and after AN-GW restoration.

BACKGROUND

0002The AN-GW is the node in the mobile communications network through which user devices establish sessions with the network. In some instances, an AN-GW may handle hundreds or even thousands of user sessions. When an AN-GW goes down or becomes unavailable, there is currently no mechanism in the network for session-independent notification and handling of the AN-GW unavailability. As a result, when an AN-GW becomes unavailable, the PCEF must signal the PCRF individually for each session involving the unavailable AN-GW. Signaling the PCRF individually for each session involving the unavailable AN-GW is burdensome on the PCEF, the PCRF, and the intervening network. In addition, if the PCRF does not recognize the per-session AN-GW status information being conveyed by the PCEF, the PCRF may continue to send messages to the PCEF for the failed sessions, resulting in rejections from the PCEF. When the AN-GW is restored, per-session messaging is required to restore each session for which AN-GW unavailability was noticed. If the PCRF does not support the exchange of per-session AN-GW restoration information, multiple requests and responses may be sent between the PCEF and the PCRF with rejections from the PCEF. For example, when the PCEF does not support an AN-GW status attribute value pair (AVP), the PCEF drops all messaging towards the PCRF. The PCEF also rejects all messaging from the PCRF with UNABLE _TO _COMPLY (message code 5012). No AN-GW status information is shared.

0003Accordingly, there exists a need for methods, systems, and computer readable media for suppressing message flooding during AN-GW unavailability and after AN-GW restoration. <nplcit id="ncit0001" npl-type="s"><text>ZTE: "Error handling when the PCC rule is removed due to the S-GW restoration support", 3GPP DRAFT; 29212_CR1078R2_(REL-12)_C3-131607, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE; 650, ROUTE DES LUCIOLES; F-06921 SOPHIA-ANTIPOLIS CEDEX; FRANCE, vol. CT WG3, no. Porto, Portugal; 20131007 - 20131011</text></nplcit> describes access node-gateway communications between network nodes in a mobile communication network.

SUMMARY

0004The scope of the invention is set out in the appended claims. The subject matter described herein includes methods, systems, and computer readable media for suppressing message flooding during AN-GW unavailability and after restoration. According to one aspect, the subject matter described herein includes a system for suppressing message flooding during access node-Gateway (AN-GW) unavailability and after AN-GW restoration. The system includes a network node including at least one processor and a memory. The network node includes an interface for receiving an indication of unavailability of an AN-GW. An AN-GW status maintainer implemented by the at least one processor maintains, in the memory, a node-level AN-GW status indicator indicating availability or unavailability of the AN-GW and setting the node-level AN-GW status indicator to unavailable in response to receiving the indication of unavailability of the AN-GW. An AN-GW status checker implemented by the at least one processor checks status of the AN-GW using the node-level AN-GW status indicator and suppresses, based on the status of the AN-GW determined from the node-level AN-GW status indicator, messaging for plural sessions involving the AN-GW.

0005According to another aspect, the subject matter described herein includes a method for suppressing message flooding during AN-GW unavailability and after AN-GW restoration. The method includes, at a network node including at least one processor and a memory, receiving an indication of unavailability of an AN-GW. The method further includes maintaining, in the memory, a node-level AN-GW status indicator indicating availability or unavailability of the AN-GW and setting the node-level AN-GW status indicator to unavailable in response to receiving the indication of unavailability of the AN-GW. The method further includes checking status of the AN-GW using the node-level AN-GW status indicator. The method further includes suppressing, based on the status of the AN-GW determined from the node-level AN-GW status indicator, messaging for plural sessions involving the AN-GW.

0006The subject matter described herein can be implemented in software in combination with hardware and/or firmware. For example, the subject matter described herein can be implemented in software executed by a processor. In one exemplary implementation, the subject matter described herein can be implemented using a non-transitory computer readable medium having stored thereon computer executable instructions that when executed by the processor of a computer control the computer to perform steps. Exemplary computer readable media suitable for implementing the subject matter described herein include non-transitory computer-readable media, such as disk memory devices, chip memory devices, programmable logic devices, and application specific integrated circuits. In addition, a computer readable medium that implements the subject matter described herein may be located on a single device or computing platform or may be distributed across multiple devices or computing platforms.

BRIEF DESCRIPTION OF THE DRAWINGS

0007The subject matter described herein will now be explained with reference to the accompanying drawings of which: <ul id="ul0001" list-style="none" compact="compact"><li><figref idref="f0001">Figure 1</figref> is a network diagram illustrating an exemplary operating environment for the subject matter described herein;</li><li><figref idref="f0002">Figures 2A</figref> and <figref idref="f0003">2B</figref> are a message flow diagram illustrating suppression of message flooding using a node-level AN-GW status indicator according to an aspect of the subject matter described herein;</li><li><figref idref="f0004">Figure 3</figref> is a block diagram illustrating an exemplary PCEF that utilizes a node-level AN-GW status indicator to suppress message flooding during AN-GW unavailability and after AN-GW restoration according to an aspect of the subject matter described herein;</li><li><figref idref="f0005">Figure 4A</figref> is a flow chart illustrating exemplary operations performed by the PCEF illustrated in <figref idref="f0004">Figure 3</figref> to suppress message flooding during AN-GW unavailability according to an aspect of the subject matter described herein;</li><li><figref idref="f0006">Figure 4B</figref> is a flow chart illustrating exemplary operations performed by the PCEF of <figref idref="f0004">Figure 3</figref> to suppress message flooding after AN-GW restoration according to an aspect of the subject matter described herein;</li><li><figref idref="f0007">Figure 5</figref> is a block diagram illustrating an exemplary Diameter routing agent (DRA) for using a node-level AN-GW status indicator to suppress message flooding during AN-GW unavailability and after AN-GW restoration according to an aspect of the subject matter described herein;</li><li><figref idref="f0008">Figure 6A</figref> is a flow chart illustrating exemplary operations performed by the DRA in <figref idref="f0007">Figure 5</figref> for suppressing message flooding during AN-GW unavailability according to an aspect of the subject matter described herein;</li><li><figref idref="f0009">Figure 6B</figref> is a flow chart illustrating exemplary operations performed by the DRA of <figref idref="f0007">Figure 5</figref> for suppressing message flooding after AN-GW restoration according to an aspect of the subject matter described herein;</li><li><figref idref="f0010">Figure 7</figref> is a block diagram illustrating an exemplary PCRF for suppressing message flooding during AN-GW unavailability and after AN-GW restoration according to an aspect of the subject matter described herein;</li><li><figref idref="f0011">Figure 8A</figref> is a flow chart illustrating exemplary operations performed by the PCRF in <figref idref="f0010">Figure 7</figref> for suppressing message flooding during AN-GW unavailability according to an aspect of the subject matter described herein; and</li><li><figref idref="f0012">Figure 8B</figref> is a flow chart illustrating exemplary operations performed by the PCRF in <figref idref="f0010">Figure 7</figref> for suppressing message flooding after AN-GW restoration according to an aspect of the subject matter described herein.</li></ul>

DETAILED DESCRIPTION

0008<figref idref="f0001">Figure 1</figref> is a network diagram illustrating an exemplary operating environment for the subject matter described herein. Referring to <figref idref="f0001">Figure 1</figref>, a plurality of user equipment (UEs) <b>100</b> establish sessions with a network through an AN-GW <b>102.</b> AN-GW <b>102</b> may handle hundreds or even thousands of user sessions at any given time. A PGW or PCEF <b>104</b> (hereinafter, PCEF <b>104</b>) functions as an intermediary between AN-GW <b>102</b> and a PCRF <b>106.</b> PCRF <b>106</b> performs policy and charging functions for user sessions with the network.

0009When AN-GW <b>102</b> becomes unavailable (e.g., due to operational failure or loss of network connectivity), messages from other network nodes, such as PCEF <b>104</b> and PCRF <b>106,</b> can flood the network because messaging is sent on a per-session basis and the only state information regarding AN-GW <b>102</b> maintained by PCRF <b>106</b> or PCEF <b>104</b> is per-session state. As a result, AN-GW unavailability is detected for one session, and that unavailability is not communicated to other sessions. Similarly, when an AN-GW is restored, the restoration is not communicated across all sessions associated with the AN-GW.

0010In <figref idref="f0001">Figure 1</figref>, a DRA <b>108</b> is responsible for routing Diameter messaging in the network. Because DRA <b>108</b> routes Diameter messages between PCEF <b>104</b> and PCRF <b>106,</b> DRA <b>108</b> can likewise become overloaded due to message flooding that occurs when AN-GW <b>102</b> fails.

Existing Solution for Handling AN-GW Failure

0011According to<nplcit id="ncit0002" npl-type="s"><text> Third Generation Partnership Project (3GPP) Technical Specification (TS) 29.212 v13.5.0 (2016-04</text></nplcit>), the disclosure of which is incorporated herein by reference in its entirety, an event trigger attribute value pair (AVP) may include an AN_GW_CHANGE code. This value is used in credit control authorization (CCA) and reauthorization request (RAR) commands by the PCRF to indicate that upon the change of the serving access node-gateway, policy and charging control (PCC) rules shall be requested. When used in a credit control request (CCR) command, the AN_GW_CHANGE code indicates that the PCEF generated the request because the serving access node-gateway changed. The address of the new serving access node-gateway is indicated in the AN-GW-Address AVP.

0012The AN-GW-Address AVP (AVP code 1050) is of type address and contains the control plane Internet protocol version 4 (IPv4) or Internet protocol version 6 (IPv6) addresses of the access node-gateway. An access node-gateway may be a signaling gateway (SGW) for 3GPP networks or an access gateway/evolved packet data gateway (AGW/ePDG) for non-3GPP networks.

0013The AN_GW_STATUS AVP (AVP code 211) is of type enumerated. It is sent from the PCEF to the PCRF to indicate the status of the AN-GW to the SGW. One status value that is defined is AN_GW_FAILED, which indicates that the AN-GW has failed and that the PCRF should refrain from sending policy decisions to the PCEF until the PCRF is notified that the AN-GW has recovered.

0014One problem with the AN_GW_STATUS AVP is that it is only a session based parameter. There is no node-level indicator that indicates that an AN-GW is down. Accordingly, if the AN_GW_STATUS AVP is supported, messaging must be exchanged between the PCEF and the PCRF for each session formerly handled by the unavailable AN-GW in order for the unavailable state of the AN-GW to be known for each session. This can result in hundreds or even thousands of messages exchanged between the PCRF and the PCEF. A similar problem occurs from the PCEF to the PCRF when the AN-GW recovers. That is, messages will need to be sent for each session for which failure has been notified when the AN-GW recovers. Such messaging is undesirable as it floods the network with unnecessary per-session signaling involving the unavailable or restored AN-GW.

Proposed Solution

0015According to an aspect of the subject matter described herein, an AN-GW status indicator may be set on PCEF <b>104,</b> DRA <b>108,</b> and/or PCRF <b>106</b> if any associated AN-GW becomes unavailable. The setting of the AN-GW status indicator may indicate that an AN-GW is unavailable. The AN-GW status indicator may be reset once all of the associated AN-GWs are restored or become available. The AN-GW status indicator may be a node-level indicator, meaning that once a node, such as PCEF <b>104,</b> DRA <b>108,</b> or PCRF <b>106,</b> is notified that the AN-GW status indicator has been set, that node will suppress session based signaling for all sessions involving the unavailable AN-GW.

0016In addition to maintaining the AN-GW status indicator at the PCEF <b>104,</b> DRA <b>108,</b> and PCRF <b>106,</b> AN-GW restore data may also be maintained at the PCEF <b>104,</b> DRA <b>108,</b> and PCRF <b>106.</b> The AN-GW restore data may include an AN_GW_DOWN_IP data structure which maintains the IP address of the AN-GWs which are down. The AN_GW_DOWN_IP data structure may be populated on PCEF <b>104,</b> DRA <b>108,</b> and PCRF <b>106</b> and identifies which AN-GWs are down. The AN-GW restore data may also include an AN_GW_DOWN_ACK parameter that is set on PCEF <b>104</b> and DRA <b>108</b> after receiving a successful response from PCRF <b>106</b> for the request carrying an AN-GW status AVP as AN_GW_FAIL (zero) for the corresponding AN-GW. The AN-GW restore data may also include an AN_GW_RESTORED_IP parameter. This parameter may store the restored AN-GW IP address on PCEF <b>104,</b> DRA <b>108,</b> and PCRF <b>106.</b>

0017In one aspect of the subject matter described herein, in-session messaging may be used to communicate the indication of AN-GW unavailability among PCEF <b>104,</b> PCRF <b>106,</b> and DRA <b>108.</b><figref idref="f0002">Figures 2A</figref> and <figref idref="f0003">2B</figref> illustrate a message flow diagram illustrating suppression of message flooding using an AN-GW status indicator according to an aspect of the subject matter described herein. Referring to <figref idref="f0002">Figure 2A</figref>, in line 1, session establishment signaling is exchanged between AN-GW <b>102</b> and PCEF <b>104.</b> The session establishment signaling may be exchanged for each session involving a UE (not shown in <figref idref="f0002">Figures 2A</figref> and <figref idref="f0003">2B</figref>) establishing a session with the network. In line 2, PCEF <b>104</b> sends a credit control request-initial (CCR-I) message with an AN-GW address AVP to DRA <b>108.</b> In line 3, DRA <b>108</b> stores the AN-GW IP address as part of session information for the session involving the CCR-I message. In line 4, DRA <b>108</b> sends a CCR-initial message to PCRF <b>106.</b> The CCR-initial message includes the AN-GW address AVP, which contains the IP address of AN-GW <b>102.</b> In line 5, PCRF <b>106</b> saves the AN-GW IP address as part of session information for the session involving the CCR-I message. In line 6, PCRF <b>106</b> sends a credit control answer-initial (CCA-I) message to DRA <b>108.</b> In line 7, DRA <b>108</b> sends a CCA-I message to PCEF <b>104.</b> In line 8, AN-GW <b>102</b> and PCEF <b>104</b> exchange additional session establishment messaging to establish the session. The process illustrated in lines 1-8 may occur for each session involving AN-GW <b>102.</b>

0018In line 9, AN-GW <b>102</b> becomes unavailable, and PCEF <b>104</b> becomes aware of the unavailability of AN-GW <b>102.</b> In one example, PCEF <b>104</b> may detect the unavailability of AN-GW <b>102</b> by expiration of a Diameter watching timer or a session-based message timer. In line 10 of the message flow diagram, PCEF <b>104</b> sets the AN-GW status indicator to indicate that AN-GW <b>102</b> is unavailable. In line 10, the AN-GW status indicator is illustrated as AN_GW_Down. Also in line 10, PCEF <b>104</b> populates the AN_GW_DOWN_IP data structure with the IP address of the unavailable AN-GW. It should be noted that the AN-GW status indicator is a node-level parameter maintained by PCEF <b>104,</b> rather than a session-based parameter. PCEF <b>104</b> will also store the IP address of the unavailable AN-GW in the AN_GW_DOWN_IP data structure.

0019In line 11 of the message flow diagram, PCEF <b>104</b> sends a credit control request-update (CCR-U) message to DRA <b>108.</b> The CCR-U message includes an AN_GW_STATUS parameter indicating that the AN-GW is unavailable. In line 12 of the message flow diagram, DRA <b>108</b> sets the AN-GW status indicator indicating unavailability of AN-GW <b>102.</b> DRA <b>108</b> also populates the AN_GW_Down_IP data structure with the IP address of the impacted AN-GW <b>102.</b>

0020In line 13 of the message flow diagram, DRA <b>108</b> sends a CCR-U message to PCRF <b>106.</b> The CCR-U message includes an AN_GW_STATUS parameter indicating that the AN-GW <b>102</b> is unavailable. In line 14, PCRF <b>106</b> sets the AN-GW status indicator and populates the AN_GW_DOWN_IP data structure with the IP address of AN-GW <b>102.</b> In line 15 of the message flow diagram, PCRF <b>106</b> sends a credit control acknowledgement-update (CCA-U) message indicating successful setting of the AN-GW status indicator. In line 16 of the message flow diagram, once PCRF <b>106</b> has successfully set the AN-GW down acknowledgment parameter, PCRF <b>106</b> suppresses sending of session-based messaging involving the unavailable AN-GW <b>102.</b> For example, PCRF <b>106</b> may refrain from sending reauthorization requests (RARs) and credit control acknowledgements (CCAs) with rule installation and modification for sessions impacted by unavailability of AN-GW <b>102.</b> One exception may be rules for removing or releasing sessions.

0021In line 17, DRA <b>108</b> sets an AN-GW unavailability acknowledgement parameter for AN-GW <b>102.</b> In line 18, DRA <b>108</b> sends a CCA-U message to PCEF <b>104</b> indicating successful setting of the AN-GW unavailability acknowledgment parameter. In line 19 of the message flow diagram, once DRA <b>108</b> has set the AN-GW unavailability acknowledgment parameter, DRA <b>108</b> suppresses session based messaging for sessions impacted by the failure of AN-GW <b>102.</b> For example, DRA <b>108</b> may suppress sending of RARs for impacted sessions except for messages involving rule removal or session removal reports to PCEF <b>104.</b> DRA <b>108</b> may also send a reauthorization acknowledgement (RAA) with a failure or rejection code indicating that AN-GW <b>102</b> is down.

0022In line 20 of the message flow diagram, after receiving the CCA-U message from DRA <b>108,</b> in line 20, PCEF <b>104</b> sets the AN-GW unavailability acknowledgment parameter, and, in line 21, suppresses sending of CCRs for impacted sessions except for rule removal or session removal reports.

0023In line 22 of the message flow diagram, AN-GW <b>102</b> is restored or becomes available, and PCEF <b>104</b> becomes aware of the availability, including the restored IP address. In line 23, PCEF <b>104</b> removes the AN_GW_RESTORE_DATA record and sends, in the best case scenario (explained below), a single CCR-U message to communicate the AN-GW restoration data for all sessions for which failure notifications are sent. One example of such a message is illustrated in line 24 where PCEF <b>104</b> sends a CCR-U message to DRA <b>108.</b> The CCR-U message includes an AN_GW_CHANGE parameter indicating AN-GW restoration as well as the AN-GW-Address AVP.

0024The solution in <figref idref="f0002">Figures 2A</figref> and <figref idref="f0003">2B</figref> after AN-GW restoration can be contrasted with the existing solution. In the existing solution, CCR-Us will be sent individually for each session for which AN-GW failure has been notified, even after receiving notification from the PCRF that AN-GW restoration has been successfully communicated to the PCRF for some of the sessions. In the proposed solution, PCRF <b>104</b> sends CCR-Us to PCRF <b>106</b> (and hopefully only a single CCR-U) only until PCEF <b>104</b> receives confirmation that PCRF <b>106</b> has been successfully notified that the AN-GW has been restored for one of the sessions. Once PCEF <b>104</b> receives from PCRF <b>106</b> a CCA-U with success for the CCR-U communicating the restoration, PCEF <b>104</b> does not need to send additional messaging to PCRF <b>106</b> to notify to PCRF <b>106</b> for the remaining sessions involving the restored AN-GW. For example, if PCRF <b>106</b> receives a message indicating recovery of AN-GW for session #1, PCRF <b>106</b> interprets AN-GW restoration information from the message received for session #1 to apply to all sessions involving the restored AN-GW. In the best case scenario, PCEF <b>104</b> would send a single message indicating AN-GW restoration for session #1, apply the restoration information on PCEF <b>104</b> to all sessions by not sending such notifications. PCRF <b>106</b> would also apply the restoration information received for session #1 and update AN-GW status to restored for all sessions associated with the restored AN-GW. However, the subject matter described herein is not limited to suppressing all messaging associated with a restored AN-GW. For example, PCEF <b>104</b> may send AN-GW restoration information to PCRF <b>106</b> for each session involving the restored AN-GW until a response indicating successful updating of the AN-GW status is received from PCRF <b>106</b> for one of the sessions. When PCEF <b>104</b> receives such a response, PCEF <b>104</b> may suppress sending restoration notifications for any remaining sessions for which restoration has not been individually communicated to PCRF <b>106.</b>

0025Returning to <figref idref="f0003">Figure 2B</figref>, in line 25 of the message flow diagram, DRA <b>108</b> sets an AN_GW_RESTORED_IP parameter for the corresponding record for the AN-GW restore data. DRA <b>108</b> also generates and sends CCR-Us to notify PCRF <b>106</b> for all sessions for which DRA <b>108</b> sent AN-GW failure notification to PCRF <b>106</b> on its own. After all of the AN-GWs are available or restored, DRA <b>108</b> resets the AN-GW status indicator. In line 26, DRA <b>108</b> sends the CCR-U message to PCRF <b>106</b> as the message was received. The CCR-U message includes the AN_GW_Change attribute indicating restoration of AN-GW <b>102</b> and the IP address of the restored AN-GW. In line 27 of the message flow diagram, PCRF <b>106</b> removes the AN-GW restore data record for the corresponding restored AN-GW <b>102.</b> If all of the AN-GWs are restored, PCRF <b>106</b> resets the AN-GW status indicator.

0026Thus, using the messaging illustrated in <figref idref="f0002">Figures 2A</figref> and <figref idref="f0003">2B</figref>, much of the session based signaling, such as CCRs, RARs, and CCAs transmitted between PCEF <b>104</b> and PCRF <b>106</b> is suppressed. The use of a node-level AN-GW status indicator in each of PCEF <b>104,</b> DRA <b>108,</b> and PCRF <b>106</b> ensures the maximum message suppression. However, the subject matter described herein is not limited to using the AN-GW status indicator among each of PCEF <b>104,</b> DRA <b>108,</b> and PCRF <b>106.</b> Using the AN-GW status indicator at any one or more of these nodes is intended to be within the scope of the subject matter described herein.

0027As stated above, one of the nodes that may be modified to support the AN-GW status indicator is PCEF <b>104.</b><figref idref="f0004">Figure 3</figref> is a block diagram illustrating an exemplary PCEF that utilizes a node-level AN-GW status indicator according to an aspect of the subject matter described herein. Referring to <figref idref="f0004">Figure 3</figref>, PCEF <b>104</b> includes a processor <b>300</b> and a memory <b>302.</b> Memory <b>302</b> includes an AN-GW status indicator <b>304</b> that is a node-level indicator that indicates unavailability of one or more AN-GWs. PCEF <b>104</b> further includes an AN-GW status maintainer <b>305</b> that maintains status indicator in memory <b>302.</b> PCRF <b>106</b> also includes an AN-GW status checker <b>306</b> that checks the status of the AN-GW using AN-GW status indicator <b>304.</b> PCEF <b>104</b> further includes an AN-GW interface <b>308</b> that communicates with one or more AN-GWs and a PCRF interface <b>310</b> that communicates with one or more PCRFs.

0028<figref idref="f0005">Figure 4A</figref> is a flow chart illustrating exemplary steps performed by PCEF <b>104</b> using the AN-GW status indicator to suppress message flooding during AN-GW unavailability and after restoration. Referring to <figref idref="f0005">Figure 4A</figref>, in step <b>400,</b> PCEF <b>104</b> detects AN-GW unavailability. In one example, PCEF <b>104</b> may detect that the AN-GW is unavailable because the AN-GW fails to respond to a heartbeat or other message, such as an in-session message. In step <b>402,</b> PCEF <b>104</b> sets the node-level AN-GW status indicator in its memory. In step <b>404,</b> PCEF <b>104</b> informs PCRF <b>106</b> of the AN-GW unavailability. If session-based messaging is used to communicate the unavailability status to PCRF <b>106,</b> PCEF may transmit a CCR-U message to PCRF <b>106</b> indicating unavailability of the AN-GW and the IP address of the unavailable AN-GW. Until a CCA-U message is received from PCRF <b>106</b> indicating success for the CCR-U, PCEF <b>104</b> may keep sending CCR-Us with the AN_GW_STATUS AVP as AN_GW_FAILED for sessions triggering the CCR-U.

0029In step <b>406,</b> PCEF <b>104</b> receives an acknowledgement from PCRF <b>106.</b> In this example, the acknowledgement is a CCA-U message indicating successful setting of the AN-GW status indicator to unavailable by PCRF <b>106.</b> In step <b>408,</b> after receiving the acknowledgement, PCEF <b>104</b> suppresses sending of session-based messaging for sessions handled by the failed AN-GW. Suppressing session-based messages may include suppressing or stopping the sending of CCRs to report AN-GW failure for impacted sessions of the unavailable AN-GW. PCEF <b>104</b> may trigger CCR-Us with rule removals or session release reports along with AN_GW_FAILED status as per existing PCEF functionality. If PCEF <b>104</b> receives any CCA-Us or RARs containing anything but rule removals or session release, PCEF <b>104</b> may reject the specific operations but honor rule removals or session release as per existing PCEF functionality.

0030<figref idref="f0006">Figure 4B</figref> illustrates operations performed by PCEF <b>104</b> in response to detecting AN-GW restoration. Referring to <figref idref="f0006">Figure 4B</figref>, in step <b>410,</b> PCEF <b>104</b> detects AN-GW restoration. In step <b>412,</b> PCEF <b>104</b> removes the unavailable AN-GW IP address from its memory. In step <b>413,</b> PCEF <b>104</b> determines whether all of the failed AN-GWs are available. If all of the failed AN-GWs are not available, control returns to steps <b>410</b> and <b>412</b> where PCEF <b>104</b> waits for more restorations. In step <b>413,</b> if all AN-GWs are determined to be available, control proceeds to step <b>414</b> where the AN-GW status indicator is reset. In step <b>416,</b> PCEF <b>104</b> informs PCRF <b>106</b> of the AN-GW availability. In one example, the PCEF <b>104</b> may send a CCR-U message with an event trigger AVP set to AN_GW_Change and includes AN_GW_Address AVP to notify PCRF <b>106</b> that the AN-GW has been restored for all sessions for which failure was notified. In step <b>418,</b> PCEF <b>104</b> resumes normal communications with PCRF <b>106.</b> Resuming normal communications may include forwarding session based messaging between AN-GW <b>102</b> and PCRF <b>106.</b>

0031As indicated above, another node that may support message flooding suppression during AN-GW unavailability and after restoration is a Diameter routing agent. <figref idref="f0007">Figure 5</figref> is a block diagram of a Diameter routing agent with message flood suppression capability. Referring to <figref idref="f0007">Figure 5</figref>, DRA <b>108</b> includes a plurality of message processors <b>500, 502, 504,</b> and <b>506.</b> Message processors <b>500, 502, 504,</b> and <b>506</b> may each comprise a printed circuit board or blade in a computing platform. Each message processor <b>500, 502, 504,</b> and <b>506</b> includes a processor <b>508</b> and a memory <b>510.</b> In the illustrated example, message processors <b>500</b> and <b>502</b> each include a Diameter connection layer (DCL) <b>512</b> and a Diameter routing layer (DRL) <b>514.</b> DCL <b>512</b> may perform Diameter connection layer processing operations with external nodes. Such operations may include establishing and maintaining Diameter level connections with the external nodes. DRL <b>514</b> routes Diameter signaling messages based on Diameter routing information in the messages. Message processors <b>504</b> and <b>506</b> may include one or more Diameter applications <b>516.</b> Examples of Diameter applications that may be implemented in DRA <b>108</b> include database related applications, such as home subscriber server (HSS) address resolution, equipment identity register (EIR), or other applications.

0032Message processors <b>500</b> and <b>502</b> each include AN-GW status maintainer <b>305</b> that maintains AN-GW status indicator <b>304</b> (not shown in <figref idref="f0007">Figure 5</figref>) in memory of message processors <b>500</b> and <b>502.</b> Message processors <b>500</b> and <b>502</b> also include AN-GW status checker <b>306</b> that checks the status of one or more AN-GWs using the node-level AN-GW status indicator maintained in memory <b>510</b> and suppresses messaging during AN-GW unavailability and after restoration.

0033<figref idref="f0008">Figure 6A</figref> is a flow chart illustrating exemplary steps performed by DRA <b>108</b> in suppressing messaging using the node-level AN-GW status indicator. Referring to <figref idref="f0008">Figure 6A</figref>, in step <b>600,</b> DRA <b>108</b> receives from PCEF <b>104</b> an indication of AN-GW unavailability. In one example, the message from PCEF <b>104</b> may be a CCR-U message with a status indicating failure or unavailability of the AN-GW. In step <b>602,</b> DRA <b>108</b> sets the AN-GW status indicator in memory and stores the IP address of the unavailable AN-GW. In step <b>604,</b> DRA <b>108</b> forwards a message communicating the AN-GW unavailability to PCRF <b>106.</b> In one example, the message may be a CCR-U message. Until a CCA-U message is received from the PCRF indicating success of the CCR-U, DRA <b>108</b> may continue to forward CCR-Us of impacted sessions to PCRF <b>106.</b>

0034In step <b>606,</b> DRA <b>108</b> receives an acknowledgement from PCRF <b>106</b> confirming successful setting of the AN-GW status indicator by PCRF <b>106.</b> In step <b>608,</b> DRA <b>108</b> sets the AN_GW_DOWN_ACK that acknowledges the successful setting of the AN-GW status indicator by PCRF <b>106.</b> If the AN-GW is down and AN_GW_DOWN_ACK is set, in step <b>610,</b> DRA <b>108</b> suppresses messaging for sessions involving the unavailable AN-GW. Suppressing messages may include, if the DRA receives a RAR message from PCRF <b>106</b> for any session associated with the impacted AN-GW, then DRA <b>108</b> may not forward the message to PCEF <b>104</b> and may respond to PCRF <b>106</b> with a failure/rejection message. Suppressing the messaging may also include, if DRA <b>108</b> receives a CCR-U for the impacted session, then DRA <b>108</b> may respond with a failure/rejection. DRA <b>108</b> may continue to forward messaging associated with rule removal or session release between PCRF <b>106</b> and PCEF <b>104</b> as per existing AN-GW restoration functionality.

0035<figref idref="f0009">Figure 6B</figref> illustrates exemplary steps performed by DRA <b>108</b> after receiving an indication that a previously unavailable AN-GW is now available. Referring to <figref idref="f0009">Figure 6B</figref>, in step <b>612,</b> DRA <b>108</b> receives an indication of restoration of AN-GW <b>102.</b> In one example, the indication of restoration may be received in a CCR-U message from PCEF <b>104</b> with an AN-GW IP address indicating that AN-GW <b>102</b> is available. In step <b>614,</b> DRA <b>108</b> sets the AN-GW IP status indicator to restored and stores the IP address of the restored AN-GW. In step <b>616,</b> DRA <b>108</b> communicates the IP address and the indication of AN-GW availability to PCRF <b>106.</b> The communication may be effected using the CCR-U message with an event trigger AVP set to AN_GW_Change and includes the AN-GW-Address AVP to communicate the restored or new AN-GW IP address to PCRF <b>106.</b> If the AN-GW-Restored-IP variable is set, in step <b>618,</b> DRA <b>108</b> sends the new or restored AN-GW IP address to PCRF <b>106</b> for each session for which failure was noticed by DRA <b>108.</b> In step <b>620,</b> DRA <b>108</b> determines whether all of the unavailable or failed AN-GWs have been restored. If the all of the unavailable AN-GWs have not been restored, control returns to step <b>612</b> where DRA <b>108</b> continues to wait for information regarding restoration of any of the unavailable AN-GWs. If all of the AN-GWs have been restored, control proceeds to step <b>622</b> where the DRA <b>108</b> resets the AN-GW status indicator.

0036According to another aspect of the subject matter described herein, PCRF <b>106</b> may also suppress per-session messaging using an AN-GW status indicator as described above. <figref idref="f0010">Figure 7</figref> is a block diagram illustrating PCRF <b>106</b> that uses the AN-GW status indicator to suppress session-based messaging during AN-GW availability and after AN-GW restoration according to an aspect of the subject matter described herein. Referring to <figref idref="f0010">Figure 7</figref>, PCRF <b>106</b> includes a processor <b>700,</b> a memory <b>702,</b> and a policy database <b>704.</b> Policy database <b>704</b> may store network access and charging policies for user elements. When a new session is requested, PCRF <b>106</b> may access policy database <b>704</b> to determine the network access and usage policies applied to the session. PCRF <b>106</b> also includes a PCEF/DRA interface <b>706</b> for communicating with PCEF <b>104</b> and/or DRA <b>108.</b> PCRF <b>106</b> may also include other interfaces <b>708</b> for interfacing with other nodes, such as deep packet inspection nodes.

0037In the illustrated example, PCRF <b>106</b> also includes AN-GW status maintainer <b>305</b> that maintains the AN-GW status indicator and related information in memory <b>702.</b> PCRF <b>106</b> also includes AN-GW status checker <b>306</b> that checks the status of the AN-GW status indicator and suppresses per-session messaging during AN-GW unavailability and after restoration.

0038<figref idref="f0011">Figure 8A</figref> illustrates exemplary steps performed by PCRF <b>106</b> using the AN-GW status indicator to suppress messaging involving an unavailable AN-GW. Referring to <figref idref="f0011">Figure 8A</figref>, in step <b>800,</b> PCRF <b>106</b> stores the AN-GW IP address received in session based messaging. The messaging may be CCR messages and context may be maintained for each session with the contexts includes the AN-GW IP address. In step <b>802,</b> PCRF <b>106</b> receives an indication of AN-GW unavailability. The indication may be received in a CCR-U message with an AN-GW status AVP indicating that the AN-GW has failed. In response to receiving the message, in step <b>804,</b> PCRF <b>106</b> sets the AN-GW status indicator and stores the IP address of the impacted AN-GW in memory of PCRF <b>106.</b>

0039In step <b>806,</b> PCRF <b>106</b> continues to process requests and send responses to PCEF <b>104.</b> However, if any of the requests or responses correspond to the unavailable AN-GW, in step <b>808,</b> PCRF <b>106</b> suppresses such signaling. The procedure for suppressing such messages may be as follows. If the AN-GW status indicator is set and the IP address associated with the indicator matches the IP address in-session-based messaging received from PCEF <b>104,</b> RAR messages are not sent to PCEF <b>104.</b> In addition, CCA-U messages are not sent to PCEF <b>104</b> with rule installations or modifications, as such messages would result in failure.

0040<figref idref="f0012">Figure 8B</figref> illustrates exemplary steps performed by PCRF <b>106</b> in response to AN-GW restoration. Referring to <figref idref="f0012">Figure 8B</figref>, in step <b>810,</b> PCRF <b>106</b> receives an indication of AN-GW restoration. The indication may be received in CCR-U message from PCEF <b>104</b> indicating that the AN-GW is available with a restored AN-GW IP address. In step <b>812,</b> in response to receiving the indication of restoration, PCRF <b>106</b> removes the AN-GW restore data record for the corresponding restored AN-GW IP address record. In step <b>814,</b> PCRF <b>106</b> determines whether all AN-GWs that have active sessions for which PCRF <b>106</b> supplied policy rules have been restored. If all AN-GWs have not been restored, control returns to step <b>810</b> where PCRF <b>106</b> waits for more restoration messaging. If all AN-GWs have been restored, control proceeds to step <b>816</b> where PCRF <b>106</b> resets the AN-GW status indicator. In step <b>818,</b> PCRF <b>106</b> ceases suppression of session-based messaging involving the restored AN-GW.

Using Network Management Messaging in Diameter to Communicate AN-GW Unavailability

0041In the examples described above with regard to <figref idref="f0004 f0005 f0006 f0007 f0008 f0009 f0010 f0011 f0012">Figures 3 through 8B</figref>, in-session messaging is used to communicate AN-GW unavailability information among PCEF <b>104,</b> DRA <b>108,</b> and PCRF <b>106.</b> In an alternate implementation, Diameter network management messaging or other messaging may be used to communicate this information. For example, when AN-GW <b>102</b> is unavailable, instead of relying on an in-session CCR message, a new Diameter network management message may indicate that the AN-GW is down along with the IP address of the unavailable AN-GW. When AN-GW <b>102</b> becomes available, instead relying on an in-session CCR message, a new Diameter network management message may indicate AN-GW <b>102</b> is restored along with the restored AN-GW IP address. The Diameter network management message that is used may be a new message, i.e., one not currently specified in the Diameter standards, or a currently supported message, such as echo request-response, or other Diameter network management message.