Page 233 - From GMS to LTE

Long Term Evolution (LTE) and LTE-Advanced Pro 219

station, interference from neighboring base stations, etc. Achievable speeds in practice
are hence much lower. A full discussion can be found in Sauter [3].
The interface between the base station and the core network is referred to as the S1
interface. It is usually carried either over a high‐speed fiber cable, or alternatively over
a high‐speed microwave link. Microwave links are, for example, based on Ethernet.
Transmission speeds of several hundred megabits per second or even gigabits per sec-
ond are required for most eNode‐Bs as they usually consist of three or more sectors in
which more than a single carrier is usually used today. In addition, a backhaul link also
carries traffic from collocated GSM and UMTS installations. Transmission capacity
requirements for backhaul links can thus far exceed the capacity of a single sector.
The S1 interface is split into two logical parts, which are both transported over the
same physical connection.
User data is transported over the S1 User Plane (S1‐UP) part of the interface. IP pack-
ets of a user are tunneled through an IP link in a manner similar to that already described
for GPRS to enable seamless handovers between different LTE base stations and UMTS
or GPRS/EDGE. In fact, the General Packet Radio Service Tunneling Protocol (GTP) is
reused for this purpose [5] as shown in Figure 4.2(b). By tunneling the user’s IP data
packets, they can easily be redirected to a different base station during a handover as
tunneling makes this completely transparent to the end‐user data flow. Only the desti-
nation IP address on layer 3 (the tunneling IP layer) is changed, while the user’s IP
address remains the same. For further details, readers may refer to Chapter 2 on the Gn
interface, which uses the same mechanism. The protocols on layers 1 and 2 of the S1
interface are not described in further detail in the specification and are just referred to
as layer 1 (L1) and layer 2 (L2). As a consequence, any suitable protocol for transporting
IP packets can be used.
The S1 Control Plane (S1‐CP) protocol, as defined in 3GPP TS 36.413 [6], is required
for two purposes. First, the eNode‐B uses it for interaction with the core network for its
own purposes, that is, to make itself known to the network, to send status and connec-
tion keep‐alive information and for receiving configuration information from the core
network. Second, the S1‐CP interface is used for transferring signaling messages that
concern the users of the system. For example, when a device wants to communicate
using the LTE network, an individual logical connection has to be established and the
core network is responsible for authentication, for supplying keys for encrypting data
on the air interface and for the establishment of a tunnel for the user’s data between the
eNode‐B and the core network. Once the user’s data tunnel is in place, the S1‐CP

Figure 4.2 S1 control plane (a) and user plane (a) (b)
(b) protocol stacks.
S1 (CP) IP user data
SCTP GTP

IP IP
L2 L2

L1 L1
S1 control plane S1 user plane