Design Con 2015

Voice over LTE rollouts require new ways of thinking about test

-February 06, 2013

LTE is now a market reality and is becoming firmly entrenched as a wireless network standard
in markets all over the world. A recent report from the GSA (Global Mobile Supplier Association) stated that the number of commercial LTE networks now stands at 105, deployed across 48 countries. It is anticipated there will be 195 deployments by the end of 2013.

The progress of network development has been rapid. While some operators are still in the trial phase, the first movers to the technology have already deployed commercial ‘voice over LTE’ (VoLTE) services to their subscribers. According to ARC Charts there will be over 74 million VoLTE enabled handsets in the market by 2016. So far, the delivery of LTE has been reliant on a ‘best-effort’ model to 4G service islands with a fallback in place to 3G networks. However, this will soon be replaced with a fully integrated VoLTE system with a stringent requirement for quality of service (QoS) tolerances to deliver premium quality voice.

Of course, VoLTE is where the promise of LTE is realized, providing a service that replicates all the features of traditional voice and offers the end-user a superior alternative to OTT VoIP applications, which are now readily available over wireless networks. With VoLTE, operators are able to deliver an end-to-end QoS through the evolved packet core (EPC) as well as over the radio network – which is also improved through MIMO (multiple in and multiple out) antennas. This is no small task and provides engineering as well as test and measurement challenges to mobile operators moving rapidly towards VoLTE. The vendor ecosystem has responded with the provision of big data ‘geoanalytics’ platforms that provide an operator with an end-to-end view of user quality of experience across the core and radio access networks.


Figure 1: When deploying VoLTE, engineers need a highly scalable big data engine that delivers end-to-end visibility into what's happening from the radio network to the core and IMS. (Image shown from Newfield Wireless TrueCall).

An end-to-end view
For engineering departments in wireless operators around the world, LTE is still relatively new and internal competencies are still being developed. At this point, delivering basic data services in an optimized way has still not been fully perfected. However, competitive pressures, spectrum constraints and other market conditions have forced a number of carriers to deploy VoLTE with aggressive timelines.

A fair amount of engineering considerations have to be taken into account to ensure the successful implementation of VoLTE and the delivery of VoLTE quality assurance (VQA). Real-time monitoring and network analysis tools need to be in place to provide operators with an end-to-end view of network and service performance.

Specifically, an operator needs to continuously monitor all the different key data points: the radio interface, eNodeB, evolved packet core, the policy server (policy charging and rules function; PCRF) and the Internet multimedia subsystem (IMS) itself. To optimize network performance, engineers need to have the ability to look at the spectral constraints in conjunction with the PCRF. To perform that correlation accurately, they need a highly scalable big data engine that delivers complete visibility into what's happening from the radio interface, to the eNodeBs, the MMEs and into the core network and IMS.

The LTE spectrum being used is limited and costly, and, by definition, a radio interface can be very unpredictable. It is highly variable in nature and, as a consequence, the implementation of services can be a challenge. This is true whether that is a finite, deterministic QoS-based service, or if services are based on an unpredictable and constrained channel.

If VoLTE services are to be launched successfully, one area that is crucial to get right is an understanding of the eNodeB scheduler. Engineers assigned to support and maintain VoLTE need to have visibility of the performance of the eNodeB scheduler if they are to ensure that the service is fully operational. To do this, they first need to know how the service is functioning. For example, how is the IMS is set up and what QoS attributes are assigned to the radio bearers? Engineers also need to understand how the intermediate protocols of the channel are set up by the nodes and how the eNodeB is channelizing the radio bearers over the radio interface. Once this is well understood, the engineer should be able to ensure the delivery of an effective QoS experience for the user over the radio interface.

Loading comments...

Write a Comment

To comment please Log In