1
SESSION-BASED ACCOUNTING
FIELD OF THE INVENTION
The inventive concepts relate to techniques for accurately 5 accounting and billing for session based packet data communication over one or more networks, for example for a mobile wireless terminal device.
BACKGROUND 10
Digital data communication is fast becoming a ubiquitous aspect of modern life. The increasing development of technology has advanced dissemination of information through telecommunication. The Internet, as well as private local 15 area and wide area data networks, enable transfer of large volumes of data at ever increasing speed and reliability. The Internet is now being used for interactive communication as well as for its original objective of information access.
Concurrent with the explosion in demand for data ser- 20 vices, there has been an ever-increasing demand for mobile communications. The number of cellular telephone customers, for example, has grown exponentially. As part of the technical development of the networks to meet the demand for mobile communications, carriers have migrated from an 25 analog-based technology to several digital transport technologies, wherein digital data is "packetized" and transmitted across digital networks. Newer versions of digital wireless communication networks, or mobile packet switched data networks, support a variety of data communication 30 services that are intended to extend the common data communication capabilities of the wired domain to the wireless mobile domain. The current trend is toward the Third Generation of Wireless Telephony (3G) networks (e.g., 3G-lx networks). The 3RD Generation Partnership 35 Project 2 (3GPP2) standard entitled Wireless IP Network Standard, 3GPP2 P.S0001-A, Version 3.0.0, ©3GPP2, version date Jul. 16, 2001 (the "3GPP2 Standard", a.k.a. the IS-835 Standard") codifies the use of mobile IP in a 3G-lx packet data network, also referred to as a code division 40 multiple access (CDMA) or CMDA2000-lx packet data network.
To communicate via a packet switched data network, each device must have a packet protocol address. In common forms of data networks the address is an Internet Protocol 45 (IP) address. Most often, data is transferred using Transmission Control Protocol/Internet Protocol (TCP/IP). With current versions of IP, the IP addresses are a scarce commodity, typically administered and assigned through an Internet service provider (ISP). 50
To send a data packet (i.e., packetized digital data) over an IP packet network, the source device must know its own address plus the address of the destination device. It is possible to assign the address when the user's device first logs-in, or when the user requests data after an idle time, 55 assuming the user's device initiates a new connection after power-up or re-connection after time-out. Generally, when a device sends a request for data to a server, the server obtains the currently assigned IP address from a request packet sent by the device, hence the server can properly address one or 60 more response packet(s) to the requesting device, whether that device is a fixed (i.e., line connected) device or mobile (i.e., wireless) station.
The addresses can be reassigned when the terminal goes off-line or remains idle for some extended period. In the 65 wireless domain, the IP addresses are assigned and reassigned by a network element. For example, after establish
2
ment of a point-to-point protocol (PPP) connection between the mobile station and a packet data serving node (PDSN) in the network, an IP address can be assigned by the PDSN. With this approach, addresses are stored in a database of the PDSN. Alternatively, an external Dynamic Handshake Challenge Protocol (DHCP) server or a home agent (HA) server can assign the IP address to the mobile station. There are two somewhat different versions of IP management techniques and services being deployed for mobile station subscribers, Simple IP (SIP) and Mobile IP (MIP). These IP address services are discussed, for example, in the IS-835 Standard.
SIP is a service in which the user is assigned a dynamic IP address from the serving PDSN. A service provider network (e.g., a radio network (RN)) provides the user's mobile station with IP routing service. The user's mobile station retains its IP address as long as that it is served by the RN, which has connectivity to the PDSN that assigned the IP address to mobile station. With the SIP, there is no IP address mobility beyond that PDSN, and as a result, there is no handoff between PDSNs.
MIP is a service in which the user is provided IP routing service to a public IP network and/or secure IP routing service to predefined private IP networks. A mobile station is able to use either a non-zero static IP address or a dynamically assigned IP address belonging to its home IP network HA. The zero HA address for an mobile station indicates that a dynamic HA allocation is required for the mobile station. If the mobile station requests a dynamic HA, then the mobile station should also request a dynamically assigned home address along with dynamic HA allocation. The mobile station is able to maintain a persistent IP address even when handing off between radio networks connected to separate PDSNs. A network (including a HA) may be any of a variety of known networks, such as a home IP network, private network, home access provider network (e.g., a IMT-2000 cellular network), or publicly accessible ISP network.
A remote authentication dial-in user service (RADIUS) is also included in the network and interfaces with the PDSN and IP network to perform accounting for a user's sessions. To aid in the accounting process, a PDSN is configured to support a PPP network session inactivity timer for each PPP session, which is typically set at 10 minutes in the prior art, but other settings may also be used. At present, there are many situations under which a PDSN user exits a call or loses radio coverage and, consequently, loses the packet session connection context. In such a case, that user could re-initiate another packet data call. If the re-initiated call is placed after expiration of the network session inactivity timer, each call is billed individually. That is, billing is accomplished by subtracting the start time from the disconnect time for each session.
Unfortunately, a problem occurs when the user re-initiates a new packet data call before the network session inactivity timer expires, i.e., prior to 10 minutes, in our example. In such a case, the two sessions can be viewed by the system as one continuous session having a start time corresponding to initiation of the first session and a stop time corresponding to termination of the second session, regardless of the downtime between sessions. The current IS-835 Standard provides that the packet session time, used in accounting, is determined based on the start time and stop time of a packet data session, which are stored according to the 3GPP2 Protocol. The disadvantage is that the user is billed incorrectly for the time between exiting the first PPP session and successfully getting a new second PPP session connection, as the first PPP session and second PPP session are treated
|
