driving-service-assurance-with-analytics

Driving Service Assurance with Analytics

Analytics solutions are beginning to close the network intelligence gap and provide more customer-centric and business-focused functionality.

To cope with the growing service assurance issues in massive mobile data networks, operators are increasingly turning to analytics solutions. Tradtional, network-element-centric assurance systems were never really designed to look at the quality experienced by individual users.   Even today, customer calls to the contact center remain one of the best sources operators rely on to pinpoint network issues.

But by combining big data platforms with DPI feeds that tie every mobile network event to a handset, analytics solutions are beginning to close the network intelligence gap and provide more customer-centric and business-focused functionality.

Service Assurance via Analytics

Rather than relying on thousands of alarm triggers to infer root cause, analytics compiles statistics to show how things are trending.  Or it uncovers a root cause by triangulating multiple data streams such as handset type, RAN performance, and cell towers.

If root cause is the reactive side of service assurance, then capacity planning delivers the proactive side.  And here we can compare the role of analytics to the architectural stress tests performed on city buildings to find which structures and construction materials will best withstand major earthquakes.

The network equivalent of an earthquake is a major sports event, live concert, or city parade that stretches the limits of mobile infrastructure. To plan for those events properly requires the kind of meticulous simulation and testing for which analytics is well-suited.

Still another analytics focus is user behavior.  If an operator understands how its customers are using mobile broadband, they can identify congestion points and determine who they affect. Armed with this type of insight, an operator can improve its network investment metholodogies to enhance customer experience and even to ensure that premium customers are treated as such.

5 Inputs to Guide Analytics Investments

So, as operators examine strategies for investing in and evolving analytics capabilities to solve service assurance problems, they may consider the following perspectives as useful inputs:

  1. Understand that owning big network data does not always make one an expert at analyzing it: Know where you expertise lies.   Is it running networks, or working with Hadoop and MapReduce?  Better to focus on rolling out LTE and leave the analytics to experts.
  2. Question whether do-it-yourself analytics is worthwhile:   In the network, time is money, so can you really afford to wait for a home-grown analytics capability to mature?  A vendor solution not only delivers a faster ROI, it allows you to skirt around internal politics and direct your energy to managing the cost/performance of the external analytics partners.
  3. Evaluate whether your organization still relies on old analysis practices: In the days before big data, carriers would manage network capacity based on simplified measures like a customer's home address.   But today you can affordably analyze a customer’s real usage across many locations, so the old practice needs to be discarded and new sources of data such as LTE deployment plans must be added.  
  4. Trust analytics as the best way to sift through multiple dimensions: Network data is especially tricky because it crosses so many dimensions such as technologies, layers, vendors, devices, and customer value.  If one out of 55 mobile devices is causing a problem, then big data’s ability to quickly sort and aggregate statistics is useful to determine, for example, that “this Android device is causing problems with these switches.”
  5. Identifying stranded assets can make for a quick win:  Most operators own network elements that have never been turned up completely. Similar, often a customer has moved on, yet the operator is still paying for leased facilities that supported them. Identifying and rectifying these kinds of problems can demonstrate analytics’ ROI rapidly in many cases.

A Final Note: Invest in Data Visualization

The human mind best grasps multi-dimensional complexity when data is presented visually. Operators should pay close attention to a solution’s visual capabilities as they are the keys to making sense of the data being analyzed and guiding and effective course of action.

Photo by NEC Corporation of America with Creative Commons license