Advertisement
Articles
Advertisement

The Fundamentals of Network Business Intelligence

Tue, 05/31/2011 - 8:45pm
Dave Caputo, Co-Founder and CEO of Sandvine

It allows ops to improve the quality and cost-effectiveness of their networks.

Broadband Internet consumers are spending more time using Internet applications for activities such as streaming video and online gaming, which require a high-quality network. At the same time, Internet penetration is reaching a saturation point in mature markets, especially in fixed access networks, so service providers cannot depend primarily on new subscribers as a source of revenue growth.

Operators must also ensure that their networks have the capacity to meet future demand with a high quality of experience for consumers, while keeping costs restrained in the face of network demands that change over time. This environment creates new challenges and opportunities for Internet service providers.

The good news is that service providers are in a unique position to mine an unparalleled amount of information about how and when the Internet is being used. Specifically, service providers have a wealth of data available to them in their billing and operations support systems (B/OSS) and from the equipment deployed in their network infrastructure. The problem is that some of this data is stored in databases, other data is measured and discarded, and other data is measured only on-demand.

Sample Analytics Dashboard - (click to enlarge)

Data analysis is often performed to answer particular business questions, and this approach can often be a labor-intensive process of extracting data from various sources, merging it together by hand and summarizing it into a useful form. Instead of waiting for data to be collected and analyzed based on a specific need, wouldn’t it be advantageous for service providers to have relevant, up-todate information available that tracks key metrics and trends and draws attention to problems and anomalies?

A network business intelligence platform is one that streamlines and simplifies complex analysis across many detailed data sets. The result is actionable data that leads to measurable business impact. What follows is an overview of the key components of an effective network business intelligence platform.

NETWORK POLICY CONTROL AS DATA SOURCE
The core enabling technology, network policy control, is a network and service management technique that supports all access technologies, all users and all applications to identify convergent network conditions, including subscriber, application and location information; to evaluate business logic rules in real time to make policy decisions locally; and to enforce policy actions locally and remotely that affect network conditions and subscriber experiences.

Traditionally, the information gleaned from Sandvine’s Policy Traffic Switch (PTS) device was predominantly used to identify and dynamically manage instances of traffic congestion, but now the PTS has become the backbone for collecting an accurate, broad set of network information that is vital for making business decisions. This network data now fuels the analytics engine, which leverages any additional external sources of subscriber data in order to present an aggregated view in concise dashboard format that is ideal for business analysis.

ANALYTICS ENGINE
For an ISP to truly reap the benefits of network insight, the data must be analyzed quickly and efficiently to present the actionable, real-time intelligence that is critical to effective business decisionmaking. It must also be represented in legible ways – ideally through a set of dashboards that offer different graphical representations of data collected.

In addition, extracted data can also be augmented by importing data from different sources and performing aggregate operations using highly scalable algorithms. The transformed and summarized data is ideally loaded into a data warehouse, from which the analytics platform can build a data model that is used for the generation of reports.

Other live data warehouse sources can also be queried to build a more comprehensive data model. Since detailed raw data is stored efficiently, requests for new metrics for analysis can be satisfied and displayed easily at any time. Historical trending data is retroactively analyzed and loaded into the data warehouse and the platform’s data model, and it’s made available for presentation to end users.

Business intelligence starts with aggregate information...

SUBSCRIBER DATA REPOSITORY
Another component of network business intelligence includes subscriber data management (SDM), the management of subscriber-oriented data from a variety of data sources such as B/OSS systems, including provisioning, billing and customer care, that act as a centralized repository for per-subscriber data. This subscriber-based data is leveraged by the platform’s analytics engine to complement the network data with relevant business information.

A high-performance process that imports and aggregates data, transforms it into a usable format, and loads it into the network business intelligence platform is required in order to fully leverage the plethora of information stored by such B/OSS systems. The world’s largest networks require this process to scale to collect billions of records per day. Efficient design dictates that these records are stored offline in compressed format so that they can be reused in the future to build past trends on new information.

It is equally important that the platform facilitate information-sharing for the export of data. Simplifying bidirectional information flow ensures that important data is available within many systems across the organization, and it eliminates potential bottlenecks that arise when data is concentrated within a single system.

DASHBOARD PRESENTATION
Concise visual presentation of data is mandatory in order to effectively communicate information to busy executives; however, for a complete context, it is often necessary to show multiple pieces of information concurrently. A dashboard is thus the ideal presentation mechanism since it allows multiple visualizations of related data to be displayed together on one page, which presents a much more complete view of network operation.

Highly customizable dashboards are possible due to the wealth of raw data stored within the SDM repository. Furthermore, an interactive dashboard, which is responsive to user preferences for views, zoom, trending, statistical analysis and other GUI options, is able to convey the data at the level of detail most appropriate for the end user.

NETWORK POLICY CONTROL AS ACTION AGENT
Finally, it comes around back to network policy control to serve as the agent for enforcement of the policy actions that will affect the desired business outcome.

Whether the business goal is traffic optimization, quota management, service creation or mobile packet core offload, the PTS implements the required action via policies that are enforced on the IP traffic flow whenever the defined network conditions are identified. And more IP traffic data is collected, which feeds into the analytics platform, and so the cycle of analysis continues.

CONCLUSION
A successful Internet business strategy requires sufficient data to make informed decisions regarding service plans, management policies, capital investments and premium services.

A network business intelligence solution provides operators with the opportunity to proactively improve the quality and cost-effectiveness of their network, while simultaneously creating new revenue streams by better understanding consumer needs and, subsequently, spotting opportunities for service creation and optimization.

E-Mail: info@sandvine.com

Advertisement

Share This Story

X
You may login with either your assigned username or your e-mail address.
The password field is case sensitive.
Loading