Page 234 - From GMS to LTE

220 From GSM to LTE-Advanced Pro and 5G

protocol is used to maintain the connection and to organize a handover of the connection
to another LTE, UMTS or GSM base station as required. Further details are discussed
in Section 4.6.
Figure 4.2(a) shows the S1‐CP protocol stack. IP is used as a basis. Instead of the
commonly known Transmission Control Protocol (TCP) and User Datagram Protocol
(UDP) on layer 4, the telecom‐specific Stream Control Transmission Protocol (SCTP)
is used as defined in RFC 4960 [7]. It ensures that a large number of independent signaling
connections can be established simultaneously with in‐sequence transport, congestion
management and flow control.
In previous 3GPP radio access networks, base stations were controlled by a central
device. In GSM, this is the base station controller (BSC), and in UTMS it is the RNC. In
these systems, the central controllers are responsible for setting up the radio links to
wireless devices via the base stations, for controlling the connections while they are
used, for ensuring QoS and for handing over a connection to another base station when
required. In LTE, this concept was abandoned to remove latency from the user path and
to distribute these management tasks, as they require significant resources if concen-
trated in a few higher‐layer network nodes. Packet‐switched connections especially
generate a lot of signaling load because of the frequent switching of the air interface
state when applications on the device only transmit and receive information in bursts
with long timeouts in between. During these times of inactivity, the air interface con-
nection to the mobile device has to be changed to use the available bandwidth efficiently
and to reduce the power consumption of mobile devices. Details on this can be found in
Chapter 3 for UMTS and Section 4.7 for LTE.
As a consequence of this autonomy, LTE base stations communicate directly with
each other over the X2 interface for two purposes. First, handovers are now controlled
by the base stations themselves. If the target cell is known and reachable over the X2
interface, the cells communicate directly with each other. Otherwise, the S1 interface
and the core network are employed to perform the handover. Base station neighbor
relations are either configured by the network operator in advance or can be detected
by base stations themselves with the help of neighbor cell information being sent to the
base station by mobile devices. This feature is referred to as Automatic Neighbor
Relation (ANR) and requires the active support of mobile devices as the base stations
themselves cannot directly detect each other over the air interface.
The second use of the X2 interface is for interference coordination. As in UMTS,
neighboring LTE base stations use the same carrier frequency so that there are areas in
the network where mobile devices can receive the signals of several base stations. If the
signals of two or more base stations have a similar strength, the signals of base stations
that the mobile device is not communicating with at that moment are perceived as noise
and the resulting throughput suffers significantly. As mobile devices can report the
noise level at their current location and the perceived source to their serving base sta-
tion, the X2 interface can then be used by that base station to contact the neighboring
base station and agree on methods to mitigate or reduce the problem. Details are dis-
cussed in Section 4.12 on network planning aspects.
Like the S1 interface, the X2 interface is independent of the underlying transport
network technology and IP is used on layer 3. SCTP is used for connection manage-
ment, and the X2 application protocol defined in 3GPP TS 36.423 [8] encapsulates the
signaling messages between the base stations. During a handover, user data packets can