4.1 Overall Architecture
Figure 4.1 describes a simplified view of the EPS architecture. Not all the network nodes and interfaces are represented here (as this was already done in Chapter 2). This picture rather focuses on E-UTRAN/EPC interactions and user signalling and data connectivity and architectural aspects which will be developed in this chapter.
Starting from the E-UTRAN part, the X2 interface shall be seen as a meshed interface rather than a point-to-point between two specific E-UTRAN nodes. This optional interface has been defined for the main purpose of forwarding packets between eNodeB so as to limit packet loss for intra E-UTRAN user mobility.
The S1 interface shall also not be seen as a simple interface between one eNodeB and one MME/Serving gateway (represented above as a single box for simplicity), since an eNodeB can possibly be connected to more than one MME. This flexibility is known as S1-flex (an equivalent of the Iu-flex 3G/UMTS option) described later in this chapter.
When MME and Serving GW are deployed as two separate physical boxes, the S1 interface is split into two parts:
- The S1-U (for User plane) – which carries user data between eNodeB and Serving GW.
- The S1-C (for Control plane) – which is a signalling-only interface between the eNodeB and the MME.
On the S5 side, a serving GW may also be linked to different PDN GW, corresponding to the fact that EPC may provide user connectivity to several different and separated IP networks.