Unlock all 25 lessons of Telco CloudFull course ₹40,000 · paid
EPC Nodes, Interfaces and Control User Plane Split
A free lesson from Skill-Lync's Telco Cloud course, on Grow by Skill-Lync · Last updated 16 Nov 2026 · Session: “Week 1 Session 2”
What will you learn in this free Telco Cloud lesson?
The instructor describes the MME, serving gateway, PDN gateway, PCRF and HSS, then goes through the EPC interfaces including S1, S11, S10, S6a, S5 and S8, with a home routed roaming example. The lesson ends with CUPS, which separates control plane and user plane so each can scale on its own.
- MME, S-GW, PGW, PCRF and HSS functions
- S1, S11, S10 and S6a interface roles
- S5 versus S8 in roaming, home routed
- Why CUPS lets planes scale independently
What comes after lesson 2?
The next free lesson is lesson 3, “Virtualized EPC, Cloud Native Functions and the 5G Core” (19 min). Later in Skill-Lync's Telco Cloud course (paid): lesson 4, Week 2 Session 1; lesson 5, Week 2 Session 2; lesson 6, Week 2 Session 3. The full course has 25 lessons (12 h 32 min of video); see the course overview or its price and syllabus.
Lesson transcript
Machine transcript of the lesson audio (lightly cleaned). Technical terms may contain errors; the video is the reference.
Mobility Management Entity or MME plays a pivotal role in managing the UE control plane handling. The signaling exchange between the MME and UE is called as non-access stratum or NAS signaling. The signaling is transferred transparently via E node B between these two nodes. Authentication of UE, security mode establishment or temporary identity assignation to the UE are some example of NAS signaling. MME controls the subscriber attach and detach to from the network. MME will let the UE attach with the network only if the UE passes the authentication challenge. The entire PLMN is divided into small cells in cellular network. These cells are grouped into multiple tracking areas. These tracking areas are managed by the MME. When the UE changes its tracking area, it is required to inform and update the MME so that core network is aware of the UE's where about. In case there is mobile terminated call, MME can then page the UE. MME also performs the signaling management for bearer setup. MME instructs the S gateway to set up the default or dedicated bearer for the UE.
MME also interacts with the home subscriber server. The HSS provides the subscription information about the UE to the MME. MME uses the same to decide upon if the services are to be provided to the UE or not. When there is a mobile terminated packets waiting to be delivered to the UE in an idle state, they are buffered by the S gateway. S gateway will initiate data notification to the MME. MME initiates the paging as a result. UE identity TIMSI is paged in the tracking area. UE upon receiving the paging message will set up the radio connection and then downlink data can be delivered to the UE.
If the LT coverage gets limited, the UE is handed over to the UMTS or GSM network. To facilitate this handover, signaling is required between two networks. MME performs signaling management for the inter-system handover to the SGSN. Serving gateway or S gateway is a very important node in managing the user plane in EPC. This network element acts as a local mobility anchor. When handover is performed between two nodes, the user data is anchored at the S gateway. When the UE is in idle state and there is incoming downlink data, the data packets are buffered at the S gateway.
It then sends notification to the MME to initiate the paging. Serving gateway forwards the packet from the E node B to the PDN gateway and from PDN gateway to the E node. Serving gateway is capable of performing deep packet inspection. It is performed by legal authorities to check the content accessed by the user. It maintains charging data records for packets that is routed or forwarded to and for the UE. Packet data network gateway or PDN gateway or PGW is a gateway to an external packet data network.
PGW is a mobility anchor when inter-system handover is performed between LTE and CDMA 2000 networks. PGW is connected to PCRF which provides the policies that governs the charging and QoS of bearers that are being established. PCRF provides the policies but these policies are enforced by PGW. Similar to S gateway, the PDN gateway does the deep packet inspection and is aware of the data content flowing through it. PGW enforces an aggregate maximum bit rate per access point, hence throttling the data rate when required. PDN gateway also assigns the dynamic IP address to the UE during the initial attach process.
Since it is an edge of the network, it has firewall to securely connect it to the external packet data network. Similar to S gateway, PGW also performs packet routing in the uplink and downlink direction. Policy charging rule function When a UE requests a service, it expects that service to meet the quality of service requirement. The policy charging rule function or PCRF coordinates between the EPC and external packet data network. It provides the policies to PDN gateway to define how packets for a particular service are treated and charged.
The home subscriber server or HSS is an important element of LTE and IMS core networks. It is a master user database that is stored in centrally. It allows communication service provider to manage customers in real time and is most cost effective manner. A centralized database of subscriber information ensures that signaling is separated from the policy. Subscriber information remains in the HSS and the policy information in PCRF. As a result, it helps CSPs ease their data center load so that they can achieve more efficient performance. HSS allows CSPs to perform specialized functions such as barring of services and functions, activating and deactivating of SIM cards and hierarchical segregation of subscriber base on their subscriptions.
In this section, we will discuss about evolved packet core interfaces or EPC interfaces. The S1 interface in LTE is used between eNodeB and EPC, specifically the MME and S-Gateway. In the user plane, this interface will be based upon GPRS tunneling protocol user data . The S1 interface includes S1-C for control plane and S1-U for user plane. S1-C is interface that connects eNodeB with MME and is also called as S1-MME interface and is control plane interface only. S1-U is the interface between eNodeB and serving gateway and hence is a user plane interface only.
The signaling transport on S1-C will be based on stream control transmission protocol. This is a modified version of TCP making it more reliable, apt for signaling. The signaling protocol for S1 is called S1-AP. S1-U interface uses user datagram protocol or UDP to transport data packets. As UDP is a connectionless protocol and hence suitable for quick delivery of data packets from one end to another. S11 interface exists between MME and serving gateway. S11 is a control plane interface only. It carries control messages from MME to serving gateway or S-gateway.
Single MME can handle multiple serving gateways, each one with its own S11 interface. It is used to coordinate the establishment of SAE bearers within the EPC. See the bearer setup can be started by the MME or by the PDN gateway. The S11 interface application initiates event from the MME node as create session, delete session, create default bearer, delete default bearer, create dedicated bearer, delete dedicated bearer. It can add a rule to create, modify or update a dedicated bearer. It allows performing X2 based UE handover with S-gateway relocation.
Perform S1 based UE handover with S-gateway relocation. S10 interface is the control interface between the MMEs, which will be very similar to S3 interface between the SGSN and MME. The interface is based on GPRS tunneling protocol for control plane with additional functionalities. S10 is many to many interface. The S10 interface is used within the evolved packet core for LTE. It is used during the MME tracking area updates and handovers. In the case of inter-MME TAU or tracking area update, one MME can contact to other MME where the user had been registered before to retrieve data about the identity, security, IMC, security information like security context, authentication vectors and active SAE bearers, PDN gateway to contact and quality of service information.
Obviously, no user data runs on it. The S6A interface is between MME and HSS. S6A interface is between MME and HSS. It enables transfer. Transfer of subscription and authentication data for authenticating, authorizing user access to the evolved system. AAA interface between MME and HSS. It uses Diameter Protocol for interaction between the MME and HSS. Diameter Protocol which support transferring of subscription and authentication data for authenticating, authorizing user access to the evolved system. between MME and HSS. Diameter is defined in RFC 3588. The S5-S8 interface exists in two flavors, one based on 3GPP based interface and other will use IETF specified proxy mobile IP, PMIP for mobile control with additional mechanisms to handle quality of service.
The motivation for the S5-S8. The motivation for PMIP flavor of S5-S8 has mainly come from WiMAX or CDMA-2000 operators and vendors interested in interworking with eUTRAN, GE-RAN or UTRAN. It has been agreed in 3GPP that the usage of PMIP or GTP on S5-S8 should not impact RAN behavior or impact the terminals. The usage of PMIP or GTP on S5-S8 will not be visible over the S1 interface or in the terminal. In the non-roaming cases, the S-Gateway and PDN-Gateway functions can be performed in one physical node.
S5-S8 is a many-to-many interface. When both S-Gateway and PDN-Gateway belong to the same network, it is referred as S5 interface. If they belong to different P elements, it is called as S8 interface. S5 interface is the interface between serving gateway and PDN-Gateway. When S-Gateway and PDN-Gateway both belong to the same network, we refer it as S5 interface. For example, in case of a UE in roaming, when breaks out to the internet using the PDN-Gateway and S-Gateway of the visited network, that is why if you open a website in US, it opens the local homepage there.
Same is true in the home network. When the serving gateway and the PDN-Gateway belong to the same home network, it provides user plane tunneling and user management between serving gateway and the PDN-Gateway. It is used for serving gateway relocation due to UE mobility and if the serving gateway needs to connect to the non-co-located PDN-Gateway, for the required PDN connectivity interface is referred as S8. We will see the example in the successive slide. Please note the interfacing between visiting network MME and the HSS for home subscriber information exchange and interfacing of visiting network PCRF and home network PCRF for exchange of policies to be implemented.
S8 Interface Inter-PLMN reference point providing user and control plane between the serving gateway in the VP element and the PDN gateway in the home PLMN. S8 is the inter-PLMN variant of S5. Let us understand with this example given on the slide. Here the UE is in the visiting network and it is connected to the RAN, MME and the serving gateway of the visiting network. But since the UE is trying to access a file present on the server on the home network, PDN gateway of the home network will be used.
Due to the fact that serving gateway belongs to a different network, a visiting network, and the PDN gateway belongs to home network, the interface used between them is now called as S8 interface. This case is referred as home routed since UE is visiting network, yet traffic is breaking out from the home network. GX Interface GX Interface GX Interface lies between policy charging rule function and policy control enforcement function in the EPS network. It enables PCRF to have dynamic control over the PCC behavior at PCEF. GS Interface GX Interface SGI Interface SGI Interface point lies between the PDN gateway and external Packet Data Network.
Packet data network may be an operator, external public or private packet data network or an intra-operator packet data network. For example, for provisioning of IMS services. This reference point corresponds to SGI for 3GPP accesses. S4 interface. S4 interface between SGSN and S gateway. S4 interface is also being used to connect UE with 3G access and PDN gateway. It provides related control and mobility support between the GPRS core and the 3GPP anchor function of serving gateway. In addition, if direct tunnel is not established, it provides the user plane tunneling. S2L interface. S2L interface between UTRAN and serving gateway for user plane tunneling when direct tunneling is established. It is based on GTPU or GPRS tunneling protocol for user plane as defined between SGSN and UTRAN or respectively between SGSN and GGSN. This provides path for tunneling of user plane when direct tunneling is used in case of handover from EPS to UMTS.
S3 is the control plane interface between MME and SGSN. It is used to facilitate user and bearer information exchange when inter-system handover is performed from EPS to UMTS or GSM. Now that we have learned about EPS interfaces, let us have a look at EPS architecture again, try to locate each interface and recall their functionality. Now let us see how EPS LTE network coexist with UMTS, GSM and non-3GPP technologies like Wi-Fi, wireless LAN. Please note that UMTS and LTE use the same HSS for subscriber database. MME is connected to serving gateway and SGSN to support inter-system handover. Here S-gateway is acting like a 3GPP anchor. That means for inter-rat handover, serving gateway or S-gateway is the node from where the data is directed. Whereas PDN gateway is an anchor for non-3GPP technologies, inter-system handover. When handover has to be performed between CDMA 2000 or wireless LAN to from LTE, data is directed from the PDN gateway. Also it is worth noting that PSTN network is connected with IP Multimedia Subsystem or IMS. IMS facilitates in handling SIP signaling used to setup voice call over the high-speed packet data network and connects it with public switched telephone network and allows voice call setup between 2G, 3G network with LTE.
control plane and user plane separations or CUPS. Control plane and user plane separation or CUPS of evolved packet core materializes scaling of control plane and user plane independent of one another and promotes a more cost-effective approach to core telecom architecture. The network for 5G. It provides an adaptable and adjustable network deployment option where operation and independent enhancement of control plane and user plane functionality is possible without affecting the functionality of existing elements subject to this development. Mobile network today is witnessing an increased variation in control plane and resource demand across the control and user planes. While earlier requirements were subject to known and calculated element of the control and user plane, some devices like IoT devices require high control plane resources or signaling. While other devices may require very little signaling resources but significant user plane and data resources. With the ever increasing technologies such that can benefit from. As an operating utility system of energy and users In particular network requirement, there is an apparent need to further bifurcate and develop the control and user plane elements individually.
Let us try to understand this with some examples. Why do we need CUPS? Since packet data has been introduced, there has been trend of increasing data. User data has been on rise from GPRS to UMTS and even more in LTE. To cater to such huge demands of bulky data transfer, there is need to scale up user plane resources. With the advent of newer and newer products like UHD TV and high definition screens need higher and higher data rates for streaming high quality content. This rise in demand for high data rate needs ramping up of user plane resources.
In the same way, introduction of large number of IoT devices in LTE network need these devices to connect with the network. Although IoT devices like smart meters, parking sensors, etc. need very small user plane data. However, due to large number of such devices, a greater capacity for control plane is expected from EPC network. With CUPS, now bifurcation of control plane and user plane resources is possible and as and when required, these resources may be added to the network. In a CUPS based architecture, a control plane function and user plane functions are separated as can be seen on the slide.
A control plane function can interact with multiple user plane functions and a user plane function can be shared by multiple control plane functions. A UE is served by a single serving gateway control plane, but multiple serving gateway user planes can be selected for different PDN connections. A user plane data packet may roam around multiple user plane functions. The control plane function controls the processing of the packets in the user plane function by provisioning a set of rules in SX sessions. Charging and usage monitoring are supported by directing the user plane function to quantify and report traffic usage using usage reporting rules.
There is hardly any change expected in the charging and policy control functions. The control plane or user plane function is responsible for GTPU, FTEID allocation. Traditional nodes, serving gateway, PDN gateway and TDF can be replaced by split node without affecting connected legacy nodes. 這些 roads,敵 damit性, PDN access, PD dollars, PDC, D&EP and can present unmántimo, D&EP, D&EP. And in this record well getting the data, כי reconstruction has been indicated and for back in the year's посмотреть is undercount and student presence.