Method for implementing an internet protocol (IP) charging and rating middleware platform and gateway system

[0001] patent application Ser. No. 10/307,335 entitled Improved method for implementing an Open Charging (OC) middleware platform and gateway system.

[0002] Not Applicable

[0003] Not Applicable

[0004] With the evolution and migration of Internet and related informational services to mobile and/or wireless handsets, devices and so on, telecommunications network operators are constantly seeking enhanced ways of rating for their respective data services. The prior art demonstrates considerable limitations, weaknesses and infirmities in this regard.

[0005] Consider U.S. Patent Application 20010055291 by Schweitzer, entitled System, method and computer program product for charging for competitive IP-over-wireless service, which details a means for charging Internet Protocol (IP) traffic but no substantial means for differentiating the traffic therein. Similarly, European Patent Application (EP) 1026853 by Yamaguchi et al., entitled Charging Method for Information Communication Network, teaches of art directed at merely counting the number of packets exchanged and nothing again of rating the different types of packets therein according to their pre-established utility and/or value.

[0006] Additionally, art which does address the problem of classifying IP traffic remains insufficient or lacks the sophistication of that of present. As with U.S. Patent Application 20020152321 by Le et al., entitled Method and apparatus for classifying IP data, makes reference to classifying said IP packets based only on IP header fieldsbasically Layer two (2) of the OSI stack (said stack has seven (7) layers). Our method disclosed goes well beyond Layer two (2) and involves classifying packets based on information obtained from Layer two (2) through to Layer seven (7). (For example, Multi-Media Service (MMS) classification performed by the disclosed invention depends on information obtained from layer six (6) and layer seven (7)). Our classification methodology incorporates correlation of data within the seven (7) layers to classify a packet. And similarly, U.S. Patent Application 20020103925 by Sheth et al., entitled Generic programmable internet protocol classification technique for a broadband engine, concentrates on IP classification specifically at the IPv4 header (i.e. layer 2/3). The art thereof is primarily directed at classification for more or less quality of service (QoS) reasons (indeed, the same can be said of the art identified former) and providing differentiated services. Our art is directed primarily at classification for the purpose of packet and application level data type identification(s).

[0007] U.S. Patent Application

[0008] 20020152321 October, 2002 Le et al. 709/238

[0009] U.S. Patent Application

[0010] 20020103925 August, 2002 Sheth et al. 709/236

[0011] U.S. Patent Application

[0012] 20010055291 December, 2001 Schweitzer 370/337

[0013] Foreign Patent Document(s)

[0014] 1026853 August, 2000 EPO.

[0015] The present invention relates generally to wireless communications and gateway services; and more specifically, to a method for implementing an Internet Protocol (IP) charging and rating middleware platform and gateway system.

[0016] The Internet Protocol (IP) charging and rating middleware platform and gateway system disclosed herewith is intended to equip, in this instance, telecommunications network operators with the ability to rate and bill IP traffic (such as File Transfer Protocol (FTP) and HyperText Transfer Protocol (HTTP)) based upon any number of informational variables, including volume, quality of service, source address, destination address, and/or time of day. The art also accommodates differentiated billing based upon service type, including Wireless Application Protocol (WAP), Multimedia Message Service (MMS), and other services which utilize the Internet Protocol (IP) as the transport protocol.

[0017] The implementation of the architecture preferentially resides at the access gateway point, between the telecommunication provider's and/or wireless operator's IP network (as for instance, GGSN (Gateway GPRS (General Packet Radio Service) Support Node) based or PDSN (Packet Data Service Node) based and the Internet/Intranet. The IP charging and rating middleware platform and gateway system effectively allows for the inspection of IP traffic at definable points within the packet information.

[0018] For the purposes of simplicity and elucidation, the architecture of the Internet Protocol (IP) charging and rating middleware platform and gateway system may be divided functionally among the IP Classification Engine (the core of the art seeking protection of Letters Patent) and the residual art, the IP charging controllers.

[0019] Inherent to the technology and methodology of the IP Classification Engine (IPCE) remain the rating rules. Such rules are loaded into the IPCE, whereupon a series of system level triggers are then armed to detect the occurrence of certain events as per the existing rating criteria. The arming of triggers and generation of event sets, providing specific charging and/or informational events form the functional foundation of the IPCE.

[0020] Triggers may be armed on a real time basis through the provisioning interface which link to a set of underlying Application Programming Interface's (API's) (in this instance), thereby providing application control for non-limiting instantiations of protocol state, destination URL or address, Time of Day, volume message content or type, and text string, among others. Indeed, those skilled in the art shall recognize that a variety of object oriented application programming interfaces will serve the purpose of notification without affecting the intent and scope of the present invention. Triggers may also be armed on a protocol-specific level, including among others, HTTP, POP3, SMTP, FTP, MMS, and WAP.

[0021] The IPCE also provides the logic for the blocking of data wherever said subscriber has insufficient funds.

[0022] The IP charging controllers in this instance represent much of the residual, generic components of the Internet Protocol (IP) charging and rating middleware platform and gateway system as performing balance queries (which decrements the usage values on each subscriber's temporary token account store, after applying rating specific charges on the data flow information, in real time) and other intermediating functions.

[0023] FIG. 1 illustrates a typical, non-limiting embodiment of the system level architecture employed in the disclosure of present;

[0024] FIG. 2A and FIG. 2B illustrate the Internet Protocol (IP) rating and classification techniques, methods and advances pertinent to the Internet Protocol (IP) charging and rating middleware platform and gateway system disclosed herein.

[0025] With reference to FIG. 1, wherever a mobile subscriber seeks to access the Internet 44 and/or related informational services via 45, 46, 47 (WAP Gateway, NMSC or Multi-Media Server, respectively), the GGSN 10 sends via the RADIUS protocol 11, an authorization request to the RADIUS Proxy Server 260 of the Internet Protocol (IP) charging and rating middleware platform and gateway system 200. Technicians skilled in the art will recognize that mobile subscriber's may wish to access content at their wireless handset and/or other similar wireless device other than those delimited prior without diluting the intent and scope of the invention of present. Further still, those practitioners will also recognize that a variety of protocols, including Authentication, Authorization, and Accounting (AAA) protocols, will satisfy the implementation of said architecture without affecting the intent and scope of the present invention. Fundamentally, the Internet Protocol (IP) charging and rating middleware platform and gateway system 200 resides between the GGSN 10 and the Master RADIUS server 20; in a lay manner, the GGSN 10 simply presupposes the gateway system 200 is the Master RADIUS server 20, and vice versa.

[0026] Continuing with reference to FIG. 1, the RADIUS Proxy Server 260 checks the master data to confirm that the account in question is active, via LDAP or RADIUS. Practitioners skilled in the art shall recognize that a variety of protocols will satisfy the implementation of said architecture without affecting the intent and scope of the present invention. Where the returned response is negative (not shown) an authorization reject 11 is returned to the GGSN.

[0027] The gateway system 200, makes a request to an Open Charging (OC) middleware platform and gateway system 250 as detailed in patent application Ser. No. 10/307,335 to confirm the status of the account in question and related profile capabilities. Technicians skilled in the art will recognize that the invention of present need not be limited to the aforementioned Open Charging (OC) middleware platform and gateway system and other similar network implementations may be employed without diluting the intent and scope as such.

[0028] The Open Charging (OC) middleware platform and gateway system 250 accesses the subscriber account server (SCP), removing the access charge from the account (pre-paid). Where the account can not support the requested charge then a negative response is returned (not shown) and a negative authorization is passed 11 to the GGSN as a failed access request.

[0029] From the Unified Rating Service 240 the user profile in question is recovered. In the preferred embodiment, assuming both the account check, and the profile were positive an authorization accept 11 is returned to the GGSN 10 via the gateway system's 200 RADIUS Proxy server 260.

[0030] The gateway system 200 preloads the said user profile and loads the rating plans (detailed further in FIG. 2A and FIG. 2B), as a result, when the first user data packet is received the gateway controller 220, already has a complete user profile built.

[0031] So thus, still in reference in FIG. 1 now, whenever the mobile subscriber initiates a data session 12 and the RADIUS Proxy Server 260 triggers the gateway controller 220 with the subscriber specific identification parameters such as the IP address, MSISDN, APN etc., said controller 220 makes a token reservation request to OCG's Prepay Account Manager 250 application in order to obtain a certain amount of usage quota; 250 responds with a the requested number of credits/quota for consumption, for a particular ID.

[0032] Simultaneously, the gateway controller 220 obtains the rating and charging information 240 specific to the subscriber. The gateway controller 220 then arms the IPCE 210 with the appropriate information, including MSISDN, IP address, the rating specific inspection criteria etc.

[0033] IPCE 210 then internally arms the respective IP Flow Classification triggers (not shown) associated with the specific information. For example, the Flow Classification might specify the URL's allowed for the subscriber, the blocking of streaming service etc. The subscriber specific information is in addition to the regular usage statistic detectors. These generic IP Flow Classifiers perform accounting of usage on IP addresses, ports, protocols and maintain the user state information within the several IP flows.

[0034] Where events corresponding to these trigger points occur, the IP Flow Classifiers update the Internet Protocol (IP) charging and rating middleware platform and gateway system 200 accounting interface (not shown), which monitors the usage related to each type of data flow within the user data session.

[0035] The Internet Protocol (IP) charging and rating middleware platform and gateway system 200 decrements the usage details from the allocated quota. Periodically, these usage statistics are updated on the gateway controller 220, which applies the charging to the usage information with rating information and decrements the credits reserved. Where the allocated tokens are exhausted, a token confirmation report is sent to OCG's Prepay Server 250, in order to debit the users account balance and simultaneously reserve additional credit/quota and allocation of usage quotas.

[0036] The Packet Analyzer component of the IP Classification Engine (IPCE) instantiates and configures itself using the downloaded IP Classification Modules Registry. It creates a decision tree based on the IP Classification Modules registry for analyzing and determining packet types. FIG. 2A and FIG. 2B serve to delineate these processes.

[0037] When a packet is captured, it is passed onto the Packet Analyzer. If the source and destination IP addresses are not blocked, the Packet Analyzer obtains a copy of the packet and immediately re-releases the packet into the network. (Blocked packets are copied and dropped within the system. However, if the source or destination IP addresses or ports are defined to be free, the message is forwarded regardless of the blocking logic.) Using the copy of the captured packet, the Packet Analyzer determines the packet type using the decision tree of FIG. 2A.

[0038] For commercial purposes it must be recognized that only chargeable packets may be analyzed.

[0039] When the packet type is determined, the Rating and Charging Component of the IP Classification Engine is notified.

[0040] The IP Classification Modules (IPCM) Registry is used to specify the logic for analyzing the packets. It specifies a decision tree for determining the packet type of each packet; it also specifies the IP Classification module to be used for determining each packet type

[0041] As explicated earlier with reference to FIG. 1, to enable all the above features of the Internet Protocol (IP) charging and rating middleware platform and gateway system, the Unified Rating Service (URS) on the control plane (i.e. service logic plane) is invoked to arm Flow Classification triggers on the data plane (i.e. user traffic plane). When a subscriber starts a data session, the standard and subscriber specific rating information is loaded from the URS onto the IPCE through the Controllers.

[0042] Based on this rating profile, the appropriate Flow Classification triggers are armed within the IPCE. During regular inspection of IP flows, when a set of events corresponding to the armed triggers occurs in the data plane, the control plane service logic is triggered to take control of the IP flow to perform further analysis, accounting, routing and control of the session.

[0043] Protocols such as FTP, POP, RTP/RTSP and SMTP have well-defined state machines where the service logic session information can indicate when a state has been reached. On the contrary, HTTP does not have well-defined state machines, but can be inspected in a way such that service logic can indicate when a state has been reached. For example, the control plane service logic has been implemented to start the counting the number and type of bytes transported after a HTTP GET Response has been detected. The accounting of packets is terminated when a TCP FIN is detected. Further refined service logic has also been implemented in the service logic of the IPCE, to detect content based on Layer 7 information. For example, particular internet content may have restricted access, or may be free of charge.

[0044] Consider further the advances achieved by the IPCM in providing a unique way of identifying and classifying MMS traffic for rating purposes. Now, MMS traffic is usually transported over the standard WAP protocol stack within a GPRS network; the IP Classification Flow Monitors detect MMS traffic within the WAP protocol stack through the Content Type field within the WSP layer of the WAP protocol or the presence of the MMSC URI in the GET header of the WSP layer of the WAP protocol. All WAP messages with the same Transaction ID as the original MMS packets are considered to be part of the MMS message stream. To ensure that all packets related to a MMS message are correctly rated, session-management is required to correlate all traffic with the same Transaction ID (in the WTP layer).

[0045] Grouping by Transaction ID is terminated by either a time-based counter or the arrival of an Abort (004) PDU within the Wireless Transaction Protocol layer.

[0046] Tables 1A and 1B have been included herewith to further elucidate these advances.

TABLE 1A
MMS Traffic Identification
Layer	Field	Value
Wireless	Transaction ID	Transaction ID compared against
Transaction		existing transaction ID for MMS
Protocol		traffic streams. If the transaction
		ID matches an existing MMS
		traffic stream, it is automatically
		identified as MMS traffic.
Wireless Session	Content Type	e.g. application/vnd.wap.mms-
Protocol		message
Wireless Session	URI (where	The URI of the serving MMSC.
Protocol	Content Type is
	unavailable).
Wireless Session	PDU Type	e.g. Get (0 40)
Protocol	URI	http://MMSC URImessage-
		idXXXXXX

[0047]

TABLE 1B
MMS Traffic Data Requirements
Layer	Field	Description
Frame	Packet Length	The size of the
		packet. Value is used
		as the size of the
		packet during IP
		rating. Packet Length
		is presented in bytes.
Internet Protocol	Source/	The source or
	Destination	destination IP
		address. Used to
		identify the
		subscriber.
Wireless Transaction	Transaction ID	Assigned transaction
Protocol		ID used to correlate
		all transactions within
		the subscriber-
		initiated session.

[0048] The Internet Protocol (IP) charging and rating middleware platform and gateway system obtains the usage accounting information (not shown), applies the associated rating information to calculate the incurred charge to the subscriber and decrements that from the quota allocated by the Prepaid Application Manager, in the case of prepaid subscribers. When the allocated quota has been depleted (or lower than a defined threshold) the quota usage is confirmed with a request for more. In the event that the subscriber has no more balance in their prepaid account, it is possible to selectively drop the subscriber packets, or direct them to a top-up site, depending on the subscriber specific rating information. In the case of post paid subscribers, the appropriate usage CDR's (or similar type Event Records) are generated for transfer to the network billing system.