Mobile operators expanding their voice over LTE services in 2015 want to make sure it’s not a replay of the massive LTE data outages that plagued the industry from 2011 to 2014. One way they’re accomplishing that goal is by revamping their monitoring approach to accommodate the new infrastructures, which are larger and more complex than ever.
Clearly, demand for VoLTE is growing as operators switch on the service in more and more markets and offer an expanding assortment of VoLTE-enabled smartphones. Popular new models like the Samsung Galaxy S6, the HTC One M9 and others unveiled at the recent Mobile World Congress 2015 trade show will only add to this demand. LG has even introduced the first smartwatch with an LTE chipset, the LG Watch Urbane LTE, that lets users make VoLTE calls directly without a smartphone connection.
Not surprisingly, operators’ VoLTE strategy includes building out a muscular infrastructure to handle demand, which is expected to soar in the latter half of 2015. That means plenty of capacity margin in VoLTE servers in the IMS core, and in a host of other hardware and software elements, including multiple vendors, stretching from the core to the base station where smartphone users actually connect.
Just as important, network operations groups are revamping their approach to monitoring multivendor, large-scale, nationwide networks. Many of these groups are now insisting on the following:
• Real-time data from all the elements in the call path
• Baselines for normal performance behavior for every monitored metric
• Reliable alerts, without false positives, when behavior changes
• Fast visibility into changes and their causes
New capabilities, new opportunities
With VoLTE, voice calls are sent as prioritized packets over the same LTE data network as smartphone-based Web surfing or online gaming. That’s a key change compared to “over-the-top” voice over Internet protocol services such as Skype. Prioritization means the operator can now maintain calls and call quality. Many operators are also introducing wideband audio codecs with VoLTE, creating a distinctively clear, crisp call that resists “contamination” from surrounding noise created by traffic, leaf blowers, conversations, etc.
Currently, VoLTE users are only a fraction of all subscribers. Yet, comments in online forums show that even though there have been complaints of dropped calls and clicks and hisses, overall they like what they hear.
The appeal of VoLTE can be seen in the response of users of the first VoLTE-enabled iPhones – the iPhone 6 and 6 Plus. Calls over 3G links for these users are only one-fifth as numerous as users of earlier iPhone models, which used 3G for voice, according to data released by Newfield Wireless.
The move toward VoLTE includes today’s major mobile operators. AT&T Mobility and T-Mobile US began phasing in service early in 2014. And, last fall, Verizon Wireless introduced its VoLTE-based Advanced Calling 1.0, supporting both HD Voice and video calling. Other U.S. mobile operators have also started phasing in VoLTE in selected markets. While some subscribers have complained of dropped VoLTE calls, hissing, clicking and other odd noises, there have been no major outages.
For mobile operators, VoLTE signals the start of a new era – one that phases out older, less efficient and more costly circuit-switched voice networks, and brings voice and data together on the all-IP LTE infrastructure. It’s a shift that will mean big efficiencies and cost savings moving forward.
Complexity and scale that drives rethinking, retooling
VoLTE brings a whole new level of complexity to today’s mobile operators. To handle it, they’re rethinking and retooling their approach to infrastructure performance monitoring.
To make packet prioritization a reality for VoLTE, network operations groups must be able to manage VoLTE as a seamless service so they can track the health of that service from one end to the other. This includes the initial call request by the smartphone, cell tower, backhaul links, routers, switches, servers and a staggering array of network services embedded in the IMS core. All of these systems are based on gear from a couple dozen vendors, and all have to work together within microseconds.
Because these technologies are so new, there is a general lack of standards and best practices for monitoring and managing VoLTE services, for both the operators and their suppliers. As a result, the blueprint for VoLTE is still being created.
For example, SNMP is only one source of the data needed for effective monitoring of complex IT services. Today, monitoring platforms must be able to quickly tap into proprietary data formats (including performance logs), import structured data files, normalize them, meaningfully integrate multiple sources of data and report results via “dashboards” that can reveal developing problems at a glance.
In addition, as the number of VoLTE subscribers grows, so does the load of application layer transactions, driven in part by a surge in call flow messaging. At the same time, network operations teams need to be able to forecast the impact of that growth. Therefore, they must be able to create common transaction metrics that can run across an array of different vendors’ products and platforms. This will let them see real-time transaction behavior and also trend the subscriber distribution of specific transactions – like those among the different SIP servers in the IMS call center control function.
This data – on how many subscribers are using each CSCF server and how much traffic they’re generating – can have a big impact. Not only does it let IT confirm that the CSCF servers aren’t overloading, enabling them to hold equipment vendors accountable for promised performance levels, it also lets them forecast future demand based on subscriber growth then plan for the infrastructure changes to support those higher CSCF transaction rates.
More performance data, more new challenges
Importing and integrating data creates new challenges. First, performance-monitoring systems need a design that can physically handle such massive data flows: collecting, processing and presenting in real time for time-sensitive voice services used by millions of subscribers. Alerting and reporting functions must be equally fast. In fact, a common problem cited by IT staff is that existing tools become painfully slow. Sometimes management data is available – CPU utilization on a server – but it can’t be collected, analyzed and reported fast enough. Often this delay is all that’s needed for a service to bog down or seize up.
Secondly, monitoring systems must be able to determine what’s normal behavior for each key element in the end-to-end VoLTE call path. These baselines, factoring in time of day and even seasonal shifts, become the bedrock for alerts and alarms that can distinguish between acceptable variations and true key performance indicator anomalies that signal emerging problems.
Cooperation: the key to monitoring VoLTE system performance
Instrumenting these key capabilities across many products from many vendors calls for a new and vastly more intimate degree of cooperation among operators, equipment suppliers and monitoring tool vendors. Some systems are so new, both operators and suppliers are still figuring out exactly what metrics are needed to manage them effectively. When new equipment management metrics are added or changed, monitoring tools must adapt just as quickly to incorporate them.
Effective performance monitoring workflows do more than track the real-time health of the VoLTE service. They identify areas that are underperforming, reveal areas of congestion and let network operations staff compare vendor claims and service-level agreements, with actual performance. It’s this insight that allows for continued improvements in, and optimization of, critical new services like VoLTE.
Editor’s Note: The RCR Wireless News Reality Check section is where C-level executives and advisory firms from across the mobile industry share unique insights and experiences.
