User and Service Provider Relationship in GPRS

User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/997
4.3 Organization of the thesis
The thesis is organized as follows: Chapter 5 gives an overview of the GPRS network. It is
intended for those that are not familiar with GPRS. Chapter 6 analyzes with some detail how
Authentication, Authorization and Accounting are managed by the GPRS network. This will be
useful for the rest of the thesis. Chapter 7 describes which are the possible paths that the packets
can follow in an end to end connection where GPRS is involved. In other words, it examines
which are the possible situations at the network layer. Chapter 8 studies the existing scenarios for
IP end to end communication and chapter 9 compares these scenarios with the ones that will be
possible in GPRS. Chapter 10 illustrates the different entities that the customer has to pay. In the
last chapter the conclusions are presented.
4.4 Terminology
Some terms that are used in the rest of the thesis have a precise meaning that must be exactly
defined:
AAA protocol: there are three independent aspects (see <draft-ietf-nasreq-nasmodel-00.tex>
[34]):
• Authentication is the confirmation that a user is a valid one via the presentation of
appropriate credentials, for example password, digital certificate or phone number.
• Authorization is the granting of a specific type of service to the users, based on their
authentication and the current system state.
• Accounting is the tracking of the consumption of resources by users. It may be used for
management, planning, billing or other purposes.
Often another activity is present together to these three:
• Auditing is the tracking of the activity of the users, for example who they are, what host
are accessed, what protocols are used, etc.
Customer: it is a synonym of user, meaning the person who makes use (and pay for) of the
services provided by other entities.
Wireline: opposite of wireless; a wireline network is a network where the users are not mobile. In
particular, in this thesis it is a network where the GPRS service is not involved.
ISP: the Internet Service Provider is the company that offers the transport service to the users. In
this thesis we will treat it as a different entity  from the GPRS operator (and it is from a
logical point of view), but  it is possible, and also likely, that they belong to the same
organization.
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/998
5 Introduction to GPRS
This chapter will be a brief introduction to the main concepts of General Packet Radio Service
(GPRS); those who are already familiar with them can skip chapter 5 and star  directly from
chapter 6.
GPRS is a standardized extension of the GSM network intended to provide both packet switching
and circuit switching services. With GPRS, the traditional GSM radio interface resources can be
shared dynamically between speech and data services as a function of service load and operator
preference.
GPRS is designed to support from intermittent and bursty data transfers to occasional transmission
of large volumes of data. Several qualities of service profiles are supported.
5.1 Why GPRS ?
GPRS presents new opportunities for GSM operators to increase their profits by providing new
packet data services. It allows them to take a share in the rapid growth of the Internet usage. GPRS
provides the GSM operators with:
• New business opportunities in wireless data communications
• An evolution path to 3rd generation mobile services
• Easy way to introduce a packet switching service to an existing GSM network
• Efficient use of scarce radio resources
• Coexistence GSM and GPRS without disturbance of existing services
For end users GPRS gives the possibility to:
• Be connected all the time to IP network or X.25 networks
• Performance of up to 115 kbp/s
• Fast set-up/access time
• Be charged on the basis of data traffic volume (not connection time)
GPRS coverage is established quickly and efficiently by adding additional packet switching nodes
to the existing GSM infrastructure. GPRS will enable mobile connectivity to the worldwide
Internet or to a corporate Intranet. Several end-users will share GPRS radio resources, resulting in
much better channel utilization than with circuit switched data communication. The user may
remain connected as long as desired, but he/she is only charged for the data volume received
and/or sent.
The fact that GPRS will operate at much higher speeds than current wireless networks should
provide a huge advantage from a software perspective finally making multimedia content,
including graphics, voice and video available.
User data can be transferred transparently between the Mobile Stations (see 5.2.7) nd the xternal
data networks across an encapsulation and tunneling mechanism. Data packets are equipped with
GPRS-specific protocol information and transferred between the MS and the edge of the GPRS
Network. This mechanism guarantee that every protocol that lays over IP is transparently
transported over the GPRS PLMN.  Thus every other protocol or application that lays over the IP
level can be utilized by GPRS users.
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/999
5.2 GPRS Architecture
The GPRS architecture is simply an extension of the GSM architecture. Two new elements will be
added to the existing infrastructure to keep packet data traffic separated from traditional GSM
speech and data traffic. The Serving GPRS Support Node (SGSN) handles packet data traffic of
users in a geographical area. The Gateway GPRS Support Node (GGSN) connects to outside data
networks and to other GPRS networks. The nodes are interconnected via an IP backbone network.
The existing GSM nodes will be upgraded with GPRS functionality. In Figure 1 (from ETSI GSM
03.60 [4]) an overview of the GPRS logical architecture is shown. The main nodes are briefly
described in the following paragraphs.
Gf
D
Gi
Gn
Gb
Gc
CE
Gp
Gs
Signalling and Data Transfer Interface
Signalling Interface
MSC/VLR
TE MT BSS TEPDN
R Um
GrA
HLR
Other PLMN
SGSN
GGSN
Gd
SM-SCSMS-GMSCSMS-IWMSC
GGSN
EIR
SGSN
Gn
Figure 1:GPRS logical architecture
The following symbols appear in Figure 1.
A Interface between a BSS and a MSC/VLR.
C Interface between SMS-GMSC and HLR.
Gb Interface between an SGSN and a BSS.
Gc Interface between a GGSN and an HLR.
Gd Interface between a SMS-GMSC and an SGSN, and between a SMS-IWMSC
and an SGSN.
Gf Interface between an SGSN and an EIR.
Gi Reference point between GPRS and an external packet data network.
Gn Interface between two GSNs within the same PLMN.
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9910
Gp Interface between two GSNs in different PLMNs.
Gr Interface between an SGSN and an HLR.
Gs Interface between an SGSN and an MSC/VLR.
E Interface between MSC/VLR and SMS-GMSC.
R Reference point between a non-ISDN compatible TE and MT. Typically this
reference point supports a standard serial interface.
Um Interface between the mobile station (MS) and the GPRS fixed network part.
The Um interface is the GPRS network interface for providing packet data
services over the radio to the MS. The MT part of the MS is used to access the
GPRS services through this interface.
5.2.1 Serving GPRS Support Node (SGSN)
The SGSN provides packet routing to and from its service area. It is serving all GPRS subscribers
that are physically located within the SGSN service area. A GPRS subscriber may be served by
any SGSN in the network depending on location. The SGSN provides:
• Authentication, ciphering and International Mobile station Equipment Identity (IMEI) check.
• Mobility management
• Logical link management towards the Mobile Station
• Packet routing and transfer
• Collection of data for charging purposes (S-CDR, see 6.3)
• Connection to the Home Location Register (HLR), Mobile Switching Center (MSC), and
Base Station Controller (BSC).
5.2.2 Gateway GPRS Support Node (GGSN)
The main functions that the GGSN has to provide are:
• The interface towards any other external network: IP and X.25 networks as well as other
GPRS networks.
• Mobility management: limited to setting up a logical link between the MS and the Access
Point.
• Access server functionality.
• Transparent transport of the IP packets from any external network to inside the GPRS PLMN
and the other way around.
The transparent connection with the IP networks is obtained extracting the IP packets that travel
over the Gi and Gp interface in Figure 1 from their layer 2 frame (such as ATM or Frame Relay)
and encapsulate them inside the GPRS Tunneling protocol packets (see Figure 2). From the
external IP network’s point of view, the GGSN is seen as Edge Routers of any private network,
i.e. every packet sent to a MS is routed across the Internet toward the GGSN. Functionality for
routing to the right SGSN and protocol conversion is also provided by the GGSN.
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9911
5.2.3 Mobile Switching Center/Visitor Location Register (MSC/VLR)
The MSC is the SGSN counterpart for the circuit switched GSM connections. An i terface
between the MSC and SGSN is provided to co-ordinate signaling for mobile stations which have
both circuit switched and packet switched capabilities.
5.2.4 Home Location Register (HLR)
The HLR contains GPRS (and GSM) subscription data and routing information. The HLR also
maps each subscriber to one or more GGSNs. The HLR is accessible from the SGSN and the
GGSN.
5.2.5 SMS-GMSC and SMS-IWMSC
The SMS-GMSC and SMS-IWMSC are connected to the SGSN via the Gd interface to enable
GPRS MSs to send and receive SMS over GPRS radio channels.
5.2.6 Base Station System (BSS)
The BSS is divided into 2 parts: the Base Station Controller (BSC) and the Base Transceiver
Station (BTS). The BSC contains new functionality for controlling the packet channels: a packet
control unit (PCU). There is also new functionality for mobility management in GPRS as well as
functionality for handling GPRS paging.
The new traffic and signaling interface from the SGSN is terminated in the BSC. Some co-
operation exists between elements of the current GSM services and GPRS. On the physical layer,
resources can be reused and some common signaling exist. New protocols supporting packet data
for the air interface and functions for slot and channel allocation are implemented in the BTS.
GPRS uses the same pool of physical channels as speech. This makes it possible to mix GPRS
channels (PDCH) with circuit switched channels (TCH) in one cell. A TCH is allocated to a single
user. In contrast, several users can multiplex their traffic on one and the same PDCH. The existing
GSM Abis transmission interface towards BSC is reused.
5.2.7 GPRS Mobile Station
A GPRS MS can run in one of three modes of operation. The mode of operation depends on the
services that the MS is attached to, i.e., only GPRS or both GPRS and other GSM services, and
upon the MS's capabilities to operate GPRS and other GSM services simultaneously.
• Class-A mode of operation: The MS is attached to both GPRS and other GSM services, and
the MS supports simultaneous operation of GPRS and other GSM services.
• Class-B mode of operation: The MS is attached to both GPRS and other GSM services, but
the MS can only operate one set of services at a time
• Class-C mode of operation: The MS is exclusively attached to GPRS services.
The MS can be in three different states: idle, standby or ready, i.e. the three different Mobility
Management (MM) states. Each state describes a certain level of functionality and information
allocated. The information sets held at MS and SGSN are denoted MM context.
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9912
5.3 Transmission plane
The transmission plane consists of a layered protocol structure providing user information
transfer, along with associated information transfer control procedures (e.g., flow control, error
detection, error correction and error recovery). The transmission plane independence of the
Network Subsystem (NSS) platform from the underlying radio interface is preserved via the Gb
interface. The following transmission plane is used in GPRS:
Relay
Network
Service
GTP
Application
IP / X.25
SNDCP
LLC
RLC
MAC
GSM RF
SNDCP
LLC
BSSGP
L1bis
RLC
MAC
GSM RF
BSSGP
L1bis
Relay
L2
L1
IP
L2
L1
IP
GTP
IP / X.25
Um Gb Gn Gi
MS BSS SGSN GGSN
Network
Service
UDP /
TCP
UDP /
TCP
Figure 2: Transmission plane
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9913
6 GPRS Authentication, Authorization, Accounting
This chapter analyzes how AAA is carried out by the GPRS network both at the link layer and at
the network layer. In particular, it will try to answer the following questions:
• which protocol is used for AAA purposes in GPRS ?
• can the user be authorized for all the services available on the Internet through the GPRS
network ?
• is it possible to reuse the authentication/authorization procedure carried out by the GPRS
operator ?
The purpose of the authentication security feature is to protect the network against unauthorized
use. It enables also the protection of the GPRS subscribers by denying the possibility for intruders
to impersonate authorized users.
Within GPRS there are two different levels in which authentication is carried out: the logical link
layer and the network layer.
According to Figure 2, the logical link is established between the MS and the SGSN by means of
the LLC protocol. This link is independent of the underlying radio interface protocols and is
defined in ETSI GSM 04.64, “GPRS Logical Link Control (LLC)”.
The network layer, instead, is situated on the top of the transmission plane, and connects the MS
and the GGSN without any action by the SGSN. This node has the only task of activating and
managing the PDP context, without caring about the content of the PDUs, thus without taking part
in an eventual authentication process at this level.
Both logical link and network layers will be analyzed in the following subchapters.
6.1 Voice (Circuit Switching Service)
The voice service in GPRS is provided by the traditional GSM network. For a circuit switching
service only the link layer authentication is applicable; the network layer simply does not exist
from the GPRS point of view.
Of course only GPRS MS class-A and class-B can be authenticated, since class-C can not use
GSM-services.  Regarding the voice users, the Authentication procedure in GPRS is the same as
used in GSM, i.e. the GPRS user is regarded exactly as a GSM one when he/she uses CS services.
The authentication procedure is effectuated in one of the following cases on request of the
network, but always after the subscriber identity (TMSI/IMSI) is known by the network:
• first network access;
• an access to a specific service
• a change of subscriber-related information element in the VLR or HLR
• any other case whenever there is any doubt about the subscriber identity
It is performed with an exchange of information between the MS and the MSC/VLR, as described
in Figure 3. The steps are the following:
1. The MSC/VLR computes and transmits to the MS a non predictable number called RAND.
2. The MSC/VLR request from the Authentication Center (AuC) the Subscriber Authentication
Key (Ki).
3. Both the MS and the MSC/VLR compute the signature of RAND called SRES using
algorithm A3 and the Ki.
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9914
4. The MS sends back the SRES.
5. The MSC/VLR checks the two SRESs are equal.
The Subscriber Authentication Key Ki is allocated, together with the IMSI, at subscription time.
Ki is stored on the network side in the Home Public Land Mobile Network (HPLMN), in an
Authentication Center. A PLMN may contain one or more AuC.
The A3 algorithm is described is ETSI GSM 03.20 [3].
RAND IMSI
Ki RAND
AuC
Ki
SRES 
Um Gb
interface interface
(GPRS) (GPRS)
Equal ?
A3A3
MS RADIO PATH MSC/VLR  or   SGSN
Autentication
yes/no
3
4
21
3
5
 Figure 3: Link layer authentication
6.2 Packet
6.2.1 Link layer
The authentication at this level is the same as above in 6.1, except that the protocol is
implemented by the SGSN instead of MSC/VLR. Note that in general the MSC/VLR shall not
authenticate the GPRS MS via the SGSN upon IMSI attach, but may authenticate the MS during
CS connection establishment.
6.2.2 Network layer
.
When a MS is willing to send data, it has to complete the following steps:
1. activate a Packet Data Protocol (PDP) context
2. eventually carry out the last authentication procedure to access an IP network .
6.2.2.1 PDP activation
In order to activate a PDP context the MS must have or obtain a PDP address. Here we will focus
only on IP connections and thus on IP addresses. These can be allocated to the MS in three
different ways, depending on the subscription offered by the operator:
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9915
• the HPLMN operator assigns a PDP address permanently to the MS (static PDP address);
• the HPLMN operator assigns a PDP address to the MS when a PDP context is activated
(dynamic HPLMN PDP address);
• the VPLMN operator assigns a PDP address to the MS when a PDP context is activated
(dynamic VPLMN PDP address).
For every MS zero, one or more dynamic or static IP address can be allocated and for each of
them a PDP context can be activated. When necessary the GGSN must allocate a IP dynamic
address either choosing among the ones it has available or using an external device such as a
DHCP server. This external device may be owned by an external organization like an ISP or a
corporate network. Every PDP context exists independently (all the accounting information are
collected independently for each of them) but all those with the same IMSI are associated with the
same MM context.
There are 2 ways in which the PDP activation procedure can be carried out: non anonymous and
anonymous.
6.2.2.1.1 Non anonymous
The PDP activation procedure can be effectuated in 5 steps as shown in Figure 4
Um Gb Gn
MS SGSN GGSN
                    Activate request
                   Security functions
            Create request 
            Create response
                   Activate accept
1
2
3
4
5
Figure 4: PDP activation procedure
1. The MS sends the request specifying:
• the Network layer Service Access Point Identifier (NSAPI); the NSAPI is a sort of port
that specifies in the MS the PDP Service Access Point and in the GSN nodes the PDP
context associated with a PDP address;
• the PDP type;
• its own PDP address;
• the Access Point Name: the APN is a logical name referring to the external packet data
network that the subscriber wishes to connect to;
• the quality of service requested (QoS);
• the eventual PDP configuration option.
In the case of a dynamic PDP address, naturally, the MS should not specify any address.
2. An authentication protocol may be executed as described in 6.1 on r quest of the network in
the case of  any doubt about the subscriber identity (access to a specific service, change of
subscriber–related information in the VLR or HLR).
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9916
3. The SGSN can either reject the PDP activation request or transmit it to the GGSN. In the
latter case it creates a Tunnel identifier (TID) and it may restrict the quality of service
according to the actual traffic load or to the user profile.
4. The GGSN checks if the subscriber is allowed to reach the APN requested and when
necessary carries out an authentication procedure as described in 6.2.2.2.2 Then it creates a
new entry in its PDP context table, a Charging Id and a PDP address if requested. Then it
returns a Create PDP context Response message in which it specifies whether UDP or TCP
should be used in the backbone network (Gn interface in Figure 2) and the QoS that can be
further restricted.
5. The SGSN returns to the MS an Activate PDP Context Accept message, eventually containing
the PDP dynamic address The SGSN is now able to route PDP PDUs between the GGSN and
the MS and to start charging.
Of course, it could happen that the PDP Activation request is unsuccessful, for example because
the IMSI is not known or because the Access Point Name (APN) can not be reached or needs a
particular subscription. It could also happen that the network initiates the procedure to activate a
PDP context. Other specific cases are described in ETSI GSM 03.60 [4].
6.2.2.1.2 Anonymous
This case is analogous to the non-anonymous one, but no subscription is needed. The network
doesn’t know who uses the PDP context but the possibilities of the user mobility are limited to
only one Routing Area (RA). This kind of service is expected  to be used by machines with
specific function (such as a can distributor). From the activation point of view the main
differences are:
• Only dynamic PDP addressing can be used
• In step 1 the MS should use a Random TLLI (Temporary Logical Link Identity) for
identification purposes.
• The APN specified should be an external network that provides anonymous services, since it
is the only entity that can be charged.
6.2.2.2 IP access
Once the user has activated the PDP context, he/she can finally access the public or private Data
Network across the Gi interface (see Figure 1).
Interworking between GPRS and IP networks is described in ETSI GSM 09.61 [8].
From the external point of view the GPRS network looks like any other network, and all the IP
packets directed to MSs are routed toward the GGSN. This node is also supposed to interrogate a
Domain Name Server for routing purposes and to be behind a firewall, in order to eventually
restrict the usage of applications that are running over the IP layer.
Two different kinds of access are possible: transparent and non transparent.
6.2.2.2.1 Transparent
In this case the MS can simply start sending IP packets that will be routed by the GGSN. The MS
is given an IP address belonging to the GPRS operator, either static or dynamic. The MS does not
need to send any authentication request at PDP context activation and the GGSN does not need to
take any part in the user authentication/authorization process. In the transparent case the GGSN is
supposed to provide at least the basic ISP services.
The GGSN and the whole GPRS network is transparent to the application level (see Figure 2).
This means that the customer can use this service for any kind of application that lays over the IP
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9917
level, for example the popular HTTP. Basically the GPRS network does not set any restriction on
what is used over the IP layer.  Of course, also tunnels to a private IP network are possible, for
example via SSL, and, over this kind of secure protocols, a Virtual Private Network can be
created.
GPRS
network
R Um Gb Gn Gi
IP IP IP IP
Any protocol Any protocol
e.g. TCP or IPsec e.g. TCP or IPsec
L2or L2 or L2
PPP PPP GPRS
bearer
GPRS
bearer L2
or Intranet
TE MT GGSN Internet
RelayIf needed, security and 
confidentiality are ensured at 
this level
Figure 5: Transparent access to IP network
The communication between the GPRS network and the IP network on layer L2 in Figure 5 can be
performed over any network, even an insecure one like the Internet, because the point to point
security is ensured by the protocols on the upper level.
6.2.2.2.2 Non transparent
A non transparent access to the IP network means that:
• the MS IP address does not belong to the PLMN addressing space;
• the MS address can be either static or dynamic;
• in order to authenticate the user a protocol like RADIUS or TACACS must be performed
between the GGSN and the Intranet/ISP.
• in the case of a dynamic address, a protocol like DHCP between the GGSN and the
Intranet/ISP must be used.
• this is the only case where the user must have another subscription besides the GPRS one (in
order to access the external IP network); i.e. in this case some subscription data  of the user
are also stored in another register than the HLR.
The protocol at the Gi interface (L2 and layers below in Figure 6) is defined by mutual agreement
between the GPRS operator and the Intranet/ISP administrator. It can be performed over any kind
of network, even an insecure one, but somehow it must guarantee a secure link (by secure protocol
in higher levels).
From the APN (Access Point Name) the GGSN can deduce which is the correct protocol and
which are the necessary security features such as PPP or IPsec.
The MS specifies that it is willing to have a non transparent access to the APN that has to be
reached using the PDP Configuration Option field during step 1 in Fi ur  4, as described in
6.2.2.1.1 The Create PDP Context Request packet is described in ETSI GSM 09.60 [7]. It
contains the information that the GGSN needs in order to carry out the authentication procedure
according to the protocol used (RADIUS; DHCP; TACAS, etc.) with the external network,
typically the credential obtained with some authentication protocol like PAP. This procedure is
described in Figure 6.
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9918
GPRS
network
R Um Gb Gn Gi
DHCP/
RADIUS
DHCP/
RADIUS
UDPUDP
IP IPPPP
L2 L2
PPP GPRS
bearer
GPRS
bearer
or Intranet
TE MT GGSN ISP
Typical example: 
Authentication protocol 
PAP, and then IPCP 
request by the TE.
The GGSN uses the 
credential obtaind by the 
MT with PAP, and sent 
in the cofiguration option 
of the PDP activation 
procedure
Figure 6: Non transparent access to IP network: signaling plane
Once the user is authenticated by the external entity, the whole GPRS network works as a tunnel
between the MS and the entity referred by the APN. The GGSN will have the only responsibility,
for this particular PDP,  to deliver the PDUs and to collect the data for charging proposes, thus
essentially the same as in the transparent case (described in Figure 5).
6.3 Accounting
The accounting system in GPRS is described in GSM 12.15 [9].
Charging information in the GPRS network is collected separately for each specific PDP context.
They are created by both the SGSNs and GGSNs that are serving that PDP context.
Every PDP context is assigned a unique identity number called Charging ID. The Charging ID is
generated by GGSN at PDP context activation and transferred to context requesting SGSN.
The most important information that must be collected are:
1. the identity of the subscriber;
2. the usage of radio interface: the amount of data transmitted in both uplink and downlink
directions separately;
3. the used protocols;
4. the Quality of Service Requested and Negotiated;
5. the location of the MS;
6. the duration of the PDP context from the activation to the deactivation;
There are two main record types, called SGSN Call Detail Record (S-CDR) and GGSN Call Detail
Record (G-CDR). Other records are provided for mobility management in the SGSN (M-CDR)
and for SMS accounting. All these records are then post processed according to the operator
billing policy in order to generate the invoice for the users.
The following tables describe in details the information stored in S-CDR and G-CDR.
For each field a key is present, with the following meaning:
M This field is mandatory and always present. Any exceptions to this rule are explicitly
described.
C This field is only available under certain conditions. If this condition applies the field is
present. The conditions under which the field is available are individually described.
O This field is optional. The manufacturer can decide to implement it or not.
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9919
S-CDR
Field Description
Record Type M GPRS SGSN PDP context record.
Network initiated PDP
context
C Present if this is a network initiated PDP context.
Anonymous Access
Indicator
C Set to true to indicate anonymous access (and that the Served IMSI is not supplied)
Served IMSI M IMSI of the served party (if Anonymous Access Indicator is FALSE or not supplied).
Served IMEI C The IMEI of the ME, if available.
SGSN Address M The IP address of the current SGSN.
MS Network CapabilityO The mobile station network capability, as defined in GSM 04.08 
Routing area O Routing area at the time of the record creation.
Local Area Code O Location area code at the time of the record creation.
Cell Identity O Cell id at the time of the record creation.
Charging ID M PDP context identifier used to identify this PDP context in different records created by
GSNs
GGSN Address Used M The IP address of the GGSN currently used. The GGSN address is always the same for an
activated PDP.
Access Point Name M The logical name of the connected access point to the external packet data network.
PDP Type M PDP type, e.g. X.25, IP or PPP
Served PDP address M PDP address of the served IMSI, e.g. an IPv4, IPv6 or X.121.
List of traffic data
volumes
M A list of changes in charging conditions for this PDP context, each time stamped. Charging
conditions are used to categorise traffic volumes, such as per QoS/tariff period. Initial and
subsequently changed QoS and corresponding data values are listed. Data volumes are in
Octets above the SNDCP layer and are separated for uplink and downlink traffic.
Record opening timeM Time stamp when PDP context activation is created in this SGSN
or record opening time on following partial records opened, includ  a minimum of date,
hour, minute, and second.
Duration M Duration of this record in the SGSN.
SGSN change C Present if this is first record after SGSN change.
Cause for record closingM The reason for the release of record from this SGSN, for example PDP context release,
GPRS detach, abnormal termination.
Diagnostics O A more detailed reason for the release of the connection.
Record Sequence
number
C Partial record sequence number in this SGSN. Only present in case of partial records.
Node ID O Name of the recording entity
Record extensions O A set of network/ manufacturer specific extensions to the record.
Local Record Sequence
number
O Consecutive record number created by this node. The number is allocated sequentially
including all CDR types.
Table 1 : SGSN – Charging Data Record details
User and service provider relationship in GPRS
Carlo Filippini                                                                                                                                                                                       12/17/9920
G-CDR
Field Description
Record Type M GPRS GGSN PDP context record.
Network initiated PDP
context
C Present if this is a network initiated PDP context.
Anonymous Access
Indicator
C Set to true to indicate anonymous access (and that the Served IMSI is not supplied).
Served IMSI M IMSI of the served party (if Anonymous Access Indicator is FALSE or not supplied).
GGSN Address M The IP address of the GGSN used.
Charging ID M PDP context identifier used to identify this PDP context in different records created by
GSNs
SGSN Address M List of SGSN addresses used during this record.
Access Point NameM The logical name of the connected access point to the external packet data network.
PDP Type M PDP type, e.g. X.25, IP or PPP
Served PDP AddressM PDP address, e.g. an IPv4, IPv6 or X.121.
Remote PDP AddressO List of PDP addresses of the remote host or DTE e.g. an IPv4, IPv6, or X.121 (Included if
the PDP type is X.25)
Dynamic Address
Flag
C Indicates whether served PDP address is dynamic, i.e. allocated during PDP context
activation.
List of traffic data
volumes
M A list of changes in charging conditions for this PDP context, each time stamped. Charging
conditions are used to categorise traffic volumes, such as per tariff period. Initial and
subsequently changed QoS and corresponding data values are listed. Data volumes are in
octets above the GTP layer and are separated for uplink and downlink traffic.
Record opening timeM Time stamp when this record was opened, include a minimum of date, hour, minute, and
second.
Duration M Duration of this record in the GGSN.
Cause for record
closing
M The reason for the release of record from this SGSN, for example PDP context release,
GPRS detach, abnormal termination.
Diagnostics O A more detailed reason for the release of the connection.
Record Sequence
number
C Partial record sequence number, only present in case of partial records.
Node ID O Name of the recording entity.
Record extensions O A set of network/ manufacturer specific extensions to the record.
Local Record
Sequence number
O Consecutive record number created by this node. The number is allocated sequentially
including all CDR types.
Table 2: GGSN - Charging Data Record details
The data volume record present in both the CDRs includes one or more container, each of them
includes Data Volume Uplink, Data Volume Downlink (in octets), Change condition and Time
Stamp, Quality of Service Requested and Negotiated (if changed from last container). Change
condition defines the reason for closing the container, such as tariff time change, QoS change or
closing the PDP context.