|
|
|
|
In this section, we will describe how a mobile station (MS) registers with the GPRS network and becomes known to an external packet data network (PDN). We will show how packets are routed to or from mobile stations, and how the network keeps track of the current location of the user. Attachment and Detachment Procedure Before a mobile station can use GPRS services, it must register to a SGSN of the GPRS network. The network checks if the user is authorized, copies the user profile from the HLR to the SGSN, and gives a packet temporary mobile subscriber identity (P-TMSI) to the user. This procedure is called GPRS attach. The disconnection from the GPRS network is called GPRS detach. The detach operation can be initiated by the mobile station or by the network (SGSN or HLR).
Session Management and PDP Context To exchange data packets with external PDNs after a successful GPRS attach, a mobile station has to apply for one or more addresses used in the PDN, e.g., for an IP address in case the PDN is an IP network. This address is called PDP address (Packet Data Protocol address). For each session, a PDP context is created, which describes the characteristics of the session. It contains the PDP type (e.g., IPv4), the PDP address that is assigned to the mobile station (e.g., 129.187.222.10), the requested QoS, and the address of a GGSN which serves as the access point to the PDN. This context is stored in the MS, the SGSN, and the GGSN. When the PDP context is active, the mobile station is "visible" for the external PDN and is able to send and receive data packets. The mapping operation between the two addresses, PDP and IMSI, enables the GGSN to transfer data packets between PDN and MS. Users might have several simultaneous PDP contexts active at a given time. The allocation of the PDP address can be static or dynamic. In the first case, the network operator of the user's home-PLMN permanently assigns a PDP address to the user. In the second case, a PDP address is assigned to the user once activating a PDP context. The PDP address can be assigned either by the operator of the user's home-PLMN or by the operator of the visited network. The home network operator decides which of the possible alternatives should be used. In case of dynamic PDP address assignment, the GGSN is responsible for the allocation and for the activation/deactivation of the PDP addresses. Figure 1 below shows the PDP context activation procedure. Using the message "activate PDP context request," the MS informs the SGSN about the requested PDP context. If dynamic PDP address assignment is requested, the parameter PDP address will be left empty. Afterwards, usual security functions (e.g., authentication of the user) are performed. If access is given, the SGSN will send a "create PDP context request" message to the affected GGSN. The affected GGSN creates a new entry in its PDP context table that enables the GGSN to route data packets between the SGSN and the external PDN. Afterwards, the GGSN returns a confirmation message "create PDP context response" to the SGSN that contains the PDP address in case dynamic PDP address allocation was requested. The SGSN updates its PDP context table and confirms the activation of the new PDP context to the MS ("activate PDP context accepted").
Figure 1: The PDP context activation procedure GPRS also supports anonymous PDP context activation. In this case, security functions as shown in figure 1 are skipped, and thus, the user using the PDP context remains unknown to the network. Anonymous context activation can be employed for pre-paid services, where the user does not want to be identified. Only dynamic address allocation is possible in this case.
Data packet routing in a world of roaming users One of the main functions of the GGSN involves interaction with the external data network. The GGSN updates the location directory by using routing information, which is supplied by the SGSNs, about the location of a MS and routes the external data network protocol packet (after its encapsulation) over the GPRS backbone to the SGSN which is currently serving the MS. It also decapsulates and forwards external data network packets to the appropriate data network. After that it also collects charging data that is forwarded to a charging gateway. GPRS operators allow roaming through an inter-operator backbone network. The GPRS operators connect to the inter-operator network via a border gateway (BG), which can provide the necessary internetworking and routing protocols (for example, Border Gateway Protocol [BGP]). It is also likely that GPRS operators will implement QoS mechanisms using the inter-operator network to ensure service-level agreements (SLAs). The architecture's main benefits are its flexibility, scalability, interoperability, and roaming.
Figure 2: Routing of Data Packets between a Fixed Host and a mobile MS As mentioned before, the GPRS network encapsulates all the data network protocols into its own encapsulation protocol. This protocol is called the GPRS Tunneling Protocol. This is done in order to ensure security in the backbone network and in order to simplify the routing mechanism and the delivery of data over the GPRS network. Figure 2 shows two intra-PLMN backbone networks of different PLMNs connected using an inter-PLMN backbone. The gateways between the PLMNs and the external inter-PLMN backbone are called border gateways. Among other things, these gateways perform security functions to protect the private intra-PLMN backbones from unauthorized users and attacks. The figure shows how packets are routed in GPRS. We assume that the packet data network is an IP network. A GPRS mobile station that is located in PLMN1 sends IP packets to a host connected to the IP network, e.g., to a Web server connected to the Internet. The SGSN that the mobile station is registered to encapsulates the IP packets that come from the mobile station, examines the PDP context, and routes them through the intra-PLMN GPRS backbone to the appropriate GGSN. The GGSN decapsulates the packets and sends them out on the IP network, where different IP routing mechanisms are used to transfer the packets to the access router of the destination network. That router delivers the IP packets to the host. Assuming that the home-PLMN of the mobile station is PLMN2, an IP address has been assigned to the mobile by the GGSN of PLMN2. Therefore, the MS's IP address has the same network prefix as the IP address of the GGSN in PLMN2. The related host is now sending IP packets to the MS. The packets are sent onto the IP network and are routed to the GGSN of PLMN2 (the home-GGSN of the MS). The GGSN of PLMN2 queries the HLR and obtains the information that the MS is currently located in PLMN1. It encapsulates the incoming IP packets and tunnels them through the inter-PLMN GPRS backbone to the appropriate SGSN in PLMN1. The SGSN decapsulates the packets and delivers them to the MS.
The main task of location management is keeping track of the user's current location so that incoming packets can be routed to the relevant MS. For this purpose, the MS frequently sends location update messages to its current SGSN. If the MS sends updates more often, its location (e.g., its current cell) is not known exactly and paging is necessary for each downlink packet, resulting in a delivery delay. On the other hand, if location updates happen very often, the MS's location is known to the network, and the data packets can be delivered without any additional paging delay. However, a lot of uplink radio capacity and battery power is consumed for mobility management in this case. Thus, a good location management strategy will be a compromise between these two extreme methods. Because of this reason, a state model like that which is shown in figure 3 below has been defined for location management in GPRS. A MS can be in any one of three states depending on its current traffic amount. The location update frequency depends on the state of the MS.
In IDLE state the MS is not reachable. When performing a GPRS attach, the MS reaches the READY state. With a GPRS detach it might disconnect from the network and fall back to the IDLE state. All PDP contexts will be deleted. The STANDBY state will be reached when a MS does not send any packets for a longer period of time, and therefore the READY timer (which was started at GPRS attach) expires. In the IDLE state, no location update is performed. It means that the current location of the MS is unknown to the network. A MS in the READY state informs its SGSN about every movement to a new cell. For the location management of a MS in the STANDBY state, a GSM location area (LA) is divided into several routing areas (RA). In general, an RA consists of several cells. The SGSN will be informed only when a MS moves to a new RA; cell changes will not be disclosed. To find out the current cell of a MS in the STANDBY state, paging of the MS within a certain RA must be performed. For MSs in the READY state, no paging is necessary. Whenever a MS moves to a new RA, it sends a "routing area update request" to the SGSN it is assigned to. The message contains the routing area identity (RAI) of its old RA. The base station subsystem (BSS) adds the cell identifier (CI) of the new cell, from which the SGSN can derive the new RAI. Two different scenarios are possible:
To sum up, GPRS mobility management consists of two levels: 1) Micro mobility management tracks the current routing area or cell of the mobile station. It is performed by the SGSN. 2) Macro mobility management keeps track of the mobile station's current SGSN and stores it in the HLR, VLR, and GGSN. |
|
Last updated: 02-06-2003.
|
