Page 237 - From GMS to LTE

Long Term Evolution (LTE) and LTE-Advanced Pro 223

router to the Internet, and which is discussed in the next section. On the radio network
side, it terminates the S1‐UP GTP tunnels, and on the core network side, it terminates
the S5‐UP GTP tunnels to the gateway to the Internet. S1 and S5 tunnels for a single
user are independent of each other and can be changed as required. If, for example, a
handover is performed to an eNode‐B under the control of the same MME and S‐GW,
only the S1 tunnel needs to be modified to redirect the user’s data stream to and from
the new base station. If the connection is handed over to an eNode‐B that is under the
control of a new MME and S‐GW, the S5 tunnel has to be modified as well.
Tunnel creation and modification are controlled by the MME, and commands to the
S‐GW are sent over the S11 interface as shown in Figure 4.1. The S11 interface reuses
the GTP‐C (control) protocol of GPRS and UMTS by introducing new messages. The
simpler UDP protocol is used as the transport protocol below instead of SCTP, and the
IP protocol is used on the network layer.
In the standards, the S‐GW and the MME are defined independently. Hence, the two
functions can, in practice, be run on the same or different network nodes. This allows
an independent evolution of signaling capacity and user data traffic. This was done
because additional signaling mainly increases the processor load, while rising data con-
sumption of users requires a continuous evolution of routing capacity and an evolution
of the number and types of network interfaces that are used.

4.2.5 The PDN‐Gateway
The third LTE core network node is the PDN‐GW. In practice, this node is the gateway
to the Internet and some network operators also use it to interconnect to intranets of
large companies over an encrypted tunnel to offer employees of those companies direct
access to their private internal networks. As mentioned in the previous section, the
PDN‐GW terminates the S5 interface.
On the user plane, this means that data packets for a user are encapsulated into an S5
GTP tunnel and forwarded to the S‐GW which is currently responsible for this user.
The S‐GW then forwards the data packets over the S1 interface to the eNode‐B that
currently serves the user, from which they are then sent over the air interface to the
user’s mobile device.
The PDN‐GW is also responsible for assigning IP addresses to mobile devices. When
a mobile device connects to the network after being switched on, the eNode‐B contacts
the MME as described above. The MME then authenticates the subscriber and requests
an IP address from the PDN‐GW for the device. For this purpose, the S5 control plane
protocol is used. The procedure is similar to the procedure in GPRS and UMTS, where
the SGSN requests an IP address from the GGSN, as described in Chapters 2 and 3. If the
PDN‐GW grants access to the network, it returns the IP address to the MME, which in
turn forwards it to the subscriber. Also part of the process is the establishment of corre-
sponding S1 and S5 user data tunnels. A full message flow is presented in Section 4.6.2.
In practice, a mobile device can be assigned several IP addresses simultaneously.
Several IP addresses are necessary in cases where the device is Voice over LTE capable.
The device thus needs to connect not only to the Internet but also to the network opera-
tor’s internal network to access the IP Multimedia Subsystem (IMS). At operating‐sys-
tem level of a mobile device, connectivity to the Internet and connectivity to an internal
network for IMS services is represented by two independent logical network interfaces.