LTE Packet Core Systems – Mobility and QoS


Long-Term Evolution (LTE) complements the success of HSPA with higher peak data rates, lower latency and an enhanced broadband experience in high-demand areas. This is accomplished with the use of wider-spectrum bandwidths, OFDMA and SC-FDMA air interfaces, and advanced antenna techniques. These techniques enable high spectral efficiency and an excellent user experience for a wide range of converged IP services. To take full advantage of these broadband access networks and to enable the co-existence of multiple technologies through an efficient, all-ip-packet architecture, 3GPP™ implemented a new core network, the evolved packet core (EPC). EPC is planned for 3GPP Release 9 and is intended for use by various access networks such as LTE, HSPA/HSPA+ and non-3GPP networks. The evolved packet system (EPS) comprises the EPC and a set of access systems such as the eUTRAN or UTRAN. EPS has been designed from the ground up to support seamless mobility and QoS with minimal latency for IP services.


The 3GPP is evolving wireless networks to become flatter and more simplified. In EPS’s user plane, for instance, there are only two types of nodes (base stations and gateways), while in current hierarchical networks there are four types, including a centralized RNC. Another simplification is the separation of the control plane, with a separate mobility-management network element. It is worth noting that similar optimizations are enabled in the evolved HSPA network architecture, providing a likewise flattened architecture.

A key difference from current networks is that the EPC is defined to support packet-switched traffic only. Interfaces are based on IP protocols. This means that all services will be delivered through packet connections, including voice. Thus, EPS provides savings for operators by using a single-packet network for all services.


A noticeable fact is that most of the typical protocols implemented in today’s RNC are moved to the eNB. The eNB, similar to the Node B functionality in the evolved HSPA architecture, is also responsible for header compression, ciphering and reliable delivery of packets. On the control plane, functions such as admission control and radio resource management are also incorporated into the eNB. Benefits of the RNC and Node B merger include reduced latency with fewer hops in the media path, and distribution of the RNC processing load into multiple eNBs.


Between the access network and the PDNs (e.g., the Internet), gateways support the interfaces, the mobility needs and the differentiation of QoS flows. EPS defines two logical gateway entities, the S-GW and the P-GW. The S-GW acts as a local mobility anchor, forwarding and receiving packets to and from the eNB where the UE is being served. The P-GW, in turn, interfaces with the external PDNs, such as the Internet and IMS. It is also responsible for several IP functions, such as address allocation, policy enforcement, packet classification and routing, and it provides mobility anchoring for non-3GPP access networks. In practice, both gateways can be implemented as one physical network element, depending on deployment scenarios and vendor support.


The MME is a signaling-only entity, thus user IP packets do not go through the MME. Its main function is to manage the UE’s mobility. In addition, the MME also performs authentication and authorization; idle-mode UE tracking and reachability; security negotiations; and NAS signaling. An advantage of a separate network element for signaling is that operators can grow signaling and traffic capacity independently. A similar benefit can also be accomplished in HSPA Release 7’s direct-tunnel architecture, where the SGSN becomes a signaling-only entity.


An important aspect for any all-packet network is a mechanism to guarantee differentiation of packet flows based on its QoS requirements. Applications such as video streaming, HTTP, or video telephony have special QoS needs, and should habitat receive differentiated service over the network. With EPS, QoS flows called EPS bearers are established between the UE and the P-GW. Each EPS bearer is associated with a QoS profile, and is composed of a radio bearer and a mobility tunnel. Thus, each QoS IP flow (e.g., VoIP) will be associated with a different EPS bearer, and the network can prioritize packets accordingly. The QoS procedure for packets arriving from the Internet is similar to that of HSPA. When receiving an IP packet, the P-GW performs packet classification based on parameters such as rules received from the PCRF, and sends it through the proper mobility tunnel. Based on the mobility tunnel, the eNB can map packets to the appropriate radio QoS bearer.


Seamless mobility is clearly a key consideration for wireless systems. Uninterrupted active handoff across eNBs is the first scenario one typically considers. However, other scenarios such as handoffs across core networks (i.e., P-GW, MME), transfer of access technologies, and idle mobility are also important scenarios covered by EPS.



Leave a Reply

Your email address will not be published. Required fields are marked *