8.5 IP Transport Example


As stated previously, the GPRS network provides underlying mobility-aware transport for packet data between the MS and GGSN. IP is supported as a PDP type by the network. In this section we look into the details of the procedures that occur in the GPRS network to service an IP application. The following assumptions are made for this scenario:

  • The MS has a subscription for IP as a PDP type in the HLR.

  • The MS has already performed an attach procedure.

  • The MS uses a static IP address.

In our application scenario, the SMTP application is considered for sending an e-mail message from a GPRS MS (Figure 8-12). When the user executes an SMTP command to send an e-mail, the application layer in the MS triggers the SM layer to start PDP context activation. For carrying any SM layer message, the MS first needs to establish a packet channel over the radio interfacel. The MS sends a packet channel request on PRACH and the PCU assigns the resources and sends a packet uplink assignment message. The PCU allocates a TFI for UL (TFI_UL) and sets up a TBF for carrying the SM message in the UL direction.

Figure 8-12. E-mail transaction over GPRS network.

graphics/08fig12.gif

The activate PDP context request message is converted into radio blocks by RLC/MAC and transported over the air to the PCU using TFI_UL on a PDTCH. The MS includes a set of parameters to specify completely its request in the SM message. It uses its static IP address as a PDP address in the message. For a dynamic PDP address, the MS needs to send the message with an empty PDP address field. The MS includes the desired QoS profile. It also includes the access point name (APN) in the PDP context setup request. The APN is used to determine the serving GGSN for the MS by the SGSN.

After receiving the PDP context request, the SGSN carries out a user authentication procedure to verify the user access. It also sets up ciphering with the mobile for encrypting data over the air interface. Before sending the authentication and ciphering request message to the MS, the PCU needs to establish a DL TBF. In Figure 8-12, DL TBF establishment is combined with the UL ack that is sent for the UL radio blocks carrying the activate PDP context request message. In the packet uplink ack, the PCU allocates TFI_DL. Similarly, for sending an authentication and ciphering response, the MS sends a UL resource request in the packet downlink ack message.

At the time of attach procedure, the SGSN has already established an MM context for the MS and has downloaded its subscription information. This subscription information is used by the SGSN to validate the activate PDP context request. It verifies that the user is allowed to use IP as PDP type and has the associated IP address. It also validates the APN and maps the APN to a GGSN, which serves the desired external data network. If no GGSN address can be derived, the SGSN rejects the PDP context activation request message.

The GGSN creates a GTP tunnel ID for the requested PDP context by combining the IMSI stored in the MM context with the NSAPI received from the MS. The SGSN may restrict the requested QoS attributes given its capabilities, the current load, and the subscribed QoS profile. The SGSN sends a create PDP context request to the GGSN, including the negotiated QoS. The GGSN uses the APN to find the IP network, if it is connected to multiple networks. The GGSN creates a new entry in its PDP context table and generates a charging ID for charging. The GGSN may further restrict the negotiated QoS given its capabilities and the current load. The GGSN then returns a create PDP context response message to the SGSN. In case of a dynamic address, the GGSN sends the PDP address.

The SGSN inserts the NSAPI along with the GGSN address in its PDP context. If the MS has requested a dynamic address, the PDP address received from the GGSN is inserted in the PDP context. The SGSN selects radio priority based on the QoS accepted by the GGSN and returns an activate PDP context accept message to the MS. The SGSN is now able to route PDP PDUs between the GGSN and the MS.

After the signaling has established the radio connection and the GTP tunnel, the MS starts sending user plane data to the target. SMTP uses TCP for reliable transportation. The sender SMTP client processes the IP packets using SNDCP (header and payload compression). SNDCP also segments the IP packet into LLC PDUs. The LLC layer provides ciphering and reliable transmission over the radio link. Using the established uplink TBF, the packets are transferred to the BSS on a PDTCH. The data received by the SGSN are tunneled via the GTP tunnel to the GGSN. The GGSN from the PDP context determines the IP network. In this way, the SMTP mail send command eventually arrives at the SMTP server.

We can see from the flow diagram of Figure 8-12 that a lot of messages are exchanged over the air interface. There are multiple reasons for this. One reason is that a reliable transmission is needed to make sure that all the radio blocks are transferred correctly. Another reason is that TBFs are only established in one direction. Also, a TBF is active only while there are some queued LLC frames or a corresponding guard timer has not expired . The radio interface requires segmentation and ciphering of the IP packet. The GPRS attach procedure helps reduce the layer 3 messages needed to establish a session. The attach procedure downloads the user profile from the home network to the visiting SGSN. This helps in reducing the signaling for authenticating user and verifying access. Another point to note is that although the QoS is set up in the GPRS network, there is no QoS between PCU and the MS. This is solved later in UMTS by provisioning QoS in the radio bearers .



IP in Wireless Networks
IP in Wireless Networks
ISBN: 0130666483
EAN: 2147483647
Year: 2003
Pages: 164

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net