Network Interoperability and migration strategy for LTE

By Vivek Chennamaneni

With LTE, the migration to IPV6 is unavoidable therefore many PCC solutions will require dual stack support. Incoming IP packets may be using IPV4, but they must be conveyed to the user terminal as an IPV6 packet... according to latest reports, there will be 9.3 billion mobile subscriptions by 2019 and 65% of the world’s population will have LTE coverage.

Currently, there are 279 commercially launched LTE networks in 101 countries and 482 LTE network commitments in 147 countries (GSA March 2014).

LTE, or Long Term Evolution, is the fourth generation of mobile networks (4G). LTE offers the capacityand the speed to handle a rapid increase in data traffic. LTE provides a superior user experience, great stability, high throughput, and low latency. The increased capacity will bring new and better services to users.

General Architecture

The core network and the radio access form the Evolved Packet System (EPS) as shown in below figure 1.The core network (called EPC in SAE) is responsible for the overall control of the UE and establishment of the bearers.

The main logical nodes of the EPC are PDN Gateway (P-GW), Serving Gateway (S-GW), Mobility Management Entity (MME) The eNBs (Evolved NodeB) create the E-UTRAN. Their functions include radio resource management and control, making handover decisions, selecting the MME (Mobility Management Entity) when a new user enters the network and encryption.

The MME performs bearer control (activation/deactivation), mobility management, selects the S-GW (Serving Gateway) for the UE, authenticates the UE, allocates temporary identities, tracks area updates, stores UE context, and is the control plane function for mobility between LTE and legacy access networks (being connected to the SGSN – Serving GPRS Support Node).

The S-GW’s main function is to act as a mobility anchor, which means that during a handover it continues to forward the packets to the UE. The P-GW (Packet Data Network Gateway) realizes the interface towards external packet data networks such as the internet or IMS. It assigns the user its IP address and is also responsible for policy and charging enforcement function (QoS), packet filtering and lawful intercept.

Since LTE is a Packet Switched network, its two components (E-UTRAN and EPC) connect to each other via IP. Therefore, in order to facilitate the implementation of legacy Circuit Switched (CS) functionalities like voice and SMS, the EPC needs to connect to another network: the IP Multimedia Subsystem (IMS).

Network Interoperability and Migration

With the launch of VoLTE (Voice over LTE), the transition to the 4G all-IP network has kicked off. However, SMS, based on circuit-switched technology, is still the most preferred mobile messaging service in theworld. As wireless operators launch IMS/LTE networks alongside their 2G and 3G networks, they must provide interworking with SMS for continuity of customer service and satisfaction.

Even operators launching pure 4G networks cannot afford to ignore SMS inter working as their customers still need to A communicate with contacts on legacy networks. This requirement drives the exploration of IMS and legacy messaging interoperability.

In IMS there are two types of messages, page-mode messages and session-mode messages. A page-mode message is very similar to an SMS. It is a SIP message and requires no answer. It can be a text sent between two subscribers, or a notification. Operator can use this page-mode to send any notification or promotional messages. Using session mode message, we can achieve chat between two users.

Operator can use this method to promote VAS based Product and perform life cycle management. The interconnection between the legacy domain (i.e., GSM or UMTS) based on MAP and the LTE-EPC part based on IMS can be accomplished through a gateway (IPS-SM-GW or IP-Short-Message-Gateway as illustrated in figure 2 below) that bridges LTE/IMS and legacy networks for message interoperability.

The Main functionality of the IP-SM-GW includes identifying to which domain a certain short message has to be delivered i.e., CS or IMS. When sending a message to the legacy domain, it must connect via MAP to the HSS to find the address of the MSC/SGSN concerned and maintain a record of the correlation between the MSISDN/IMSI and the address of the associated S-CSCF. 4

Also, it must check that the sender and receiver addresses are correct in the SIP headers. On the other hand, messages originating from CStowards IMS need mapping of MSISDN with SIP address at IPSM Gateway. Network policy control and the Policy Control and Charging (PCC) architecture are fundamental to any modern data network, whether 2G, 3G or 4G. Policy control is used by voice services to dynamically request QoS to ensure toll-quality communication, which is absolutely necessary in an all-IP network.

Network policy control also allows operators to more accurately measure, manage and charge for different types of data services. As mobile operators increase data service speeds, they will begin to face many challenges. Operators will need to manage bandwidth usage fairly to prevent a few users from negatively impacting the Internet experience of many others.

They will also face bandwidth-intensive applications that cause traffic congestion and denial-of-service attacks that can prevent many subscribers from accessing the network. PCC provides a framework for enabling AF (Application Function) to signal policy requests such as granting subscribers access to applications, charging and QoS. However, this presents a challenge for operators because PCC was designed for well-designed applications such as IMS, while the current market demand is for open Internet access and OTT services such as Facebook and YouTube.

With LTE, the migration to IPV6 is unavoidable so many PCC solutions will require dual stack support. Incoming IP packets may be using IPV4, but they must be conveyed to the user terminal as an IPV6 packet. Network Address Translation (NAT) will therefore be a standard function of the P-GW. The challenges described above raise questions about how to best support the PCRF element’s role as decision-maker in the network policy control planning model.

This necessitates the role of an advancedpolicy node which can compliment PCRF by supporting the following features: Contextual awareness –policy, measurement and charging based on knowledge derived from the subscriber session such as the application, access network, device type, dynamic measurements, location and time. Map subscriber’s identity to IP address, device, network type and location for measurement, policy and charging.

Interworking with external policy nodes such as GGSN or P-GWs, or even legacy nodes. Provide uniform, technology-and vendor-agnostic dynamic measurements, policy, reporting and charging. Support dual IPV6 and IPV4 stacks, and IPv6 transition technologies.

The author is CTO, Netxcell Limited.