Reader Forum: Tracking VoLTE performance

Editor’s Note: Welcome to our weekly Reader Forum section. In an attempt to broaden our interaction with our readers we have created this forum for those with something meaningful to say to the wireless industry. We want to keep this as open as possible, but we maintain some editorial control to keep it free of commercials or attacks. Please send along submissions for this section to our editors at: [email protected].
It’s still early days for voice-over-LTE technology, but Strategy Analytics predicts that the number of VoLTE users will jump from nine million in 2013 to more than 800 million within five years. Infonetics also forecasts that the combined revenue from over-the-top mobile voice-over-Internet protocol and VoLTE services will reach $16 billion by 2018. The rapid growth trajectory for VoLTE is driven by clear-cut benefits over traditional circuit-switched voice networks.
VoLTE promises huge costs savings, better quality of service and even reduced power consumption in consumer devices. All over the world, operators are conducting equipment trials, load-testing infrastructure, running VoLTE migration scenarios and more. And from these early tests, service providers are beginning to learn what it will take to manage the complexity of VoLTE out in the field.
Performance visibility is critical for ensuring that systems are optimized at every point along the network. The financial stakes are high and operators need to be prepared. The better providers can track performance, the more successful and seamless initial VoLTE deployments will be.
Lessons from the field
No. 1: Collect everything
Rolling out VoLTE technology today is exponentially more complex than deploying LTE service was only two to three years ago. Part of the challenge stems from the complexity of the application itself, and part from the increasingly intricate network infrastructure needed to support mobile data traffic. On the application side, voice delivery requires extremely low levels of latency and jitter, which means that having clear visibility into the hand off between mobile device and mobile network is crucial. Operators also have to establish and track service prioritization so that packets of voice traffic are accorded higher priority than traditional data traffic.
As a consumer experience, VoLTE should meet roughly the same quality guidelines as an enterprise VoIP implementation. However, the environment for VoLTE traffic is both much larger and much more fragmented. Mobile broadband traffic today starts at the LTE radio access network, moves up into aggregation and backhaul points at cell tower sites, extends into the evolved packet core, moves across backbone networks and arrives at the IMS core. Some segments of this path are directly under a mobile operator’s control, but many are not.
In addition, mobile broadband networks are growing more heterogeneous. New hardware and software vendors are emerging alongside industry stalwarts, and all of this technology not only has to be interoperable, but operators need a single pane of glass through which to monitor system-wide performance.
Given all of the elements at play in service delivery, it’s become clear in VoLTE trials that wireless carriers need to collect a vast amount of data in order to optimize both the consumer experience and operational efficiency. To understand the big picture of performance, operators need to be able to see every piece of the puzzle and be able to know what impact each piece has on the whole.
No. 2: Know your KPIs
Because VoLTE is new, many of the metrics for measuring performance haven’t been well established. There are basic key performance indicators that should be applied to any voice service like call initiation volume, call drop rates, call set-up times, mean opinion score, latency and jitter. However, the end-to-end call experience depends on so many factors and so many different equipment types that determining what impacts each of these metrics is difficult. Call drops, for example, could be the result of a fault anywhere along the network path, or even of a particular handset.
In order to understand first that an anomaly has occurred, operators need to baseline normal performance and set up alerts for when activity moves beyond defined thresholds. If the number of packet errors received relative to the total number of packets sent exceeds a certain ratio, for instance, that’s an indicator that there’s a problem somewhere between point A and point B.
Baselines and alerts do not have to be specific to an individual piece of hardware or a software application. Consumers don’t care that a network switch is working properly if it still means their calls aren’t staying connected. In addition to collecting data on the performance of each hardware and software system, operators should consider KPIs that help indicate not only that a problem has occurred, but also why.
For example, if latency levels are high in a particular geographic region, that should be a trigger for operators to pull up a set of metrics related to the factors most likely to be causing the problem. Whether the issue might be coming from a particular cell site or from a software update installed across a select group of routers, operators should know in advance of any VoLTE service roll out how different variables could impact the customer experience. By understanding those variables and having a mechanism in place to monitor KPIs, it’s possible to mitigate risk and increase both performance and efficiency.
No. 3: Be prepared to scale
For all of the different types of data mobile operators need to collect, there is also the challenge of collecting that data across long distances and numerous network devices. A VoLTE service trial may be contained in a small geographic area, but ultimately a provider’s voice network could extend for thousands of miles. Across that territory, a tier-one wireless carrier may have anywhere between 20,000 and 75,000 cell sites, each one paired with a backhaul router, aggregated to larger backhaul routers numbering in the hundreds and connected back to a core distribution network comprised of tens of thousands of further devices. That adds up to a lot of equipment and potentially millions of objects that need to be monitored at one time.
With large-scale networks, the volume of data involved becomes a challenge, but so too does the length of time it takes to run performance reports. If an operator recognizes a problem, there has to be a quick and easy way to understand the cause and take action. The solution is implementing a monitoring system that distributes both data collection and the processing power required for data reporting. An operator can’t afford to wait for hours after sending an information request. Data must be available almost instantly no matter how far the network extends and how large the service deployment grows.
For data management purposes, it’s also critical that service providers have access to information all in one place. Without a centralized view as service deployments scale, it becomes ever-more difficult to identify where an error has occurred. Given the complexity of mobile networks, even a minor configuration change can have a major impact on traffic flows, dropped connections and failed authentications.
VoLTE trials are already well underway, and the pace of deployments is likely to accelerate as carriers begin targeted market launches and handset makers introduce new VoLTE-enabled mobile devices. Once those dominoes fall, VoLTE will gain significant momentum and operators around the world will need to be prepared in order to stay competitive. That process begins now by testing and implementing strategies for monitoring network and service performance. Performance visibility will play a critical role in VoLTE’s commercial success.
Matt Goldberg is senior director of service provider solutions at SevOne (www.SevOne.com), the leader of scalable performance monitoring solutions to the world’s most connected companies.