Network Business Intelligence: Revolutionizing decision-making
Click here to give us your feedback on this webinar
Would detailed knowledge of subscriber and network trends influence your decisions?
Are you asked for network profiles, subscriber segmentation, or other usage data?
Do you want to have your finger on the pulse of your data network?
Network Business Intelligence is revolutionizing decision-making by getting data out of the network and into the boardroom.
Successful Internet business strategy requires sufficient data to make informed decisions regarding service plans, management policies, capital investments and premium services. Network Business Intelligence connects the dots by giving broadband and mobile data operators insight into network data in conjunction with information from additional business systems, delivering actionable conclusions and measurable business impact.
Speakers:
Lee Brooks, Product Marketing Manager, Sandvine
Lee directs global product marketing for several Sandvine products, including the Network Analytics platform. He also oversees Sandvine’s Internet Phenomena program, and in this capacity researches and authors Sandvine’s bi-annual studies of global Internet traffic. His passions lie in understanding subscriber behaviour and trends, and spotting opportunities for network operators to leverage convergent trends in ways that both benefit the bottom line while maximizing the subscriber experience.
Peter Dykes, Senior Analyst, Informa Telecoms & Media
Peter Dykes is currently senior analyst working on Informa Telecoms & Media’s Networks Intelligence Centre. His main focus is on support software for telecoms networks. His remit includes operational and business support systems, next-generation network architectures, service provisioning, management and delivery, revenue assurance, real-time charging, billing and customer care. Peter also follows developments in areas such as policy control, network and customer data management, network optimization and traffic management.
Peter has worked as a telecommunications journalist since 1989, specializing in mobile communication technologies, infrastructure, handsets, support systems and markets. He is currently editor of OSS/BSS Analyst research service and associated contract database, and is deputy editor of the Future Mobile Networks research service.
Tags; Archive, data, networks, Sandvine, telecoms.com
Loading ...
The Q&A session is now closed. Thank you.
typically, what amount of data is generated when monitoring a 10 Gbps link ?
I’m not quite certain how to answer your question – are you interested in database sizing, for instance for deployment scalability? I’ll provide a bit of info and please let me know if it doesn’t address your query. First, Sandvine’s deployment architecture relies on PTS (Policy Traffic Switch) units as the inline component. PTSes are usually deployed in a cluster, which behaves as one large virtual unit, and can scale to networks up to 480 Gbps. This cluster reports to a Subscriber Policy Broker (SPB), which you can think of as the database component. The Network Analytics platform pulls data from multiple SPBs…so you essentially end up with a single Network Analytics platform (which lives on a very standard server) being the window into a network-wide deployment that is potentially Terabits-per-second of throughput. If you’d like more specific information, please feel free to contact me privately (lbrooks@sandvine.com) and we can set up a discussion.
Hi Guys… thanks for the presentation. I’m wondering, service providers have long since been using OSS/BSS to tune their services and marketing campaigns – what makes Network Analytics innovative or more effective?
I think the difference is in the real-time presentation. As Lee pointed out, modern systems can produce an heuristics analysis that would take an afternoon or longer to put together manually.
There are a number of evolutionary innovations that contribute to making Network Analytics (and Network Business Intelligence in general) more effective than a traditional approach. First, a network business intelligence solution is generally built specifically to combine network data with a wealth of other related sources. The primary data source is an inspection device, in Sandvine’s case this is a Policy Traffic Switch deployment. This data is much richer than what a generic probe can provide, I mentioned in the slides about subscriber-, application-, location-, and device-awareness – much more useful than IP and Port. Now that you have a deep understanding of how subscribers are using the network, you can find out more about the subscribers (in general terms, not specific individuals) by pulling in data from a subscriber profile repository. Then, you can combine with information from, say, an on-deck service like a Video on Demand appliance. This is all supported out-of-the-box, and the dashboards are designed with this in mind, vs the massive integration and custom design effort if you just started pulling in data from servers and assembling manually.
Additionally, the analytics engine is perfect for identifying correlations, forecasting, etc – again, this isn’t typically native function on something like a cache or BSS.
I suppose at a high level, you’re comparing a comprehensive tool specifically designed for a range of targeted service provider use-cases with a hodge-podge of data sources that each present information in their own way, and insulated in their own functional bubbles.
I could go on longer, but I see I’ve already gone on a great deal already…but please feel free to e-mail me personally if you want to chat in detail.
who are the primary players in the Network Business Intelligence space?
The market for BI platforms is in its infancy in the telecoms sector, but it is well developed in the IT sector, meaning there are significant opportunities for IT vendors to make inroads. The usual suspects, such as Oracle, IBM and HP, already have mature offerings in the sector that are being focused on telecoms operators, but smaller and more specialized vendors, such as Sandvine, are increasingly making a name for themselves by offering the means to make sense of the plethora of customer and network data that is available to those service providers that are prepared to invest in data discovery. The truth of the matter is that virtually any company involved in data warehousing or data mining could quickly develop a platform to collect and display data, so companies such as KXEN, Teradata and even the traditional infrastructure players with customer- and network-data-collection offerings could also join in.
The players actually vary along a spectrum: you get giant analytics players who are in the much larger field of generic Business Intelligence, and at the opposite end there are niche companies that will focus analytics on a particular access technology (say, mobile networks) or application/service family (for example, video). Examples of the big players include SAP and Oracle, but the thing to note here is that they’re not at all focused on network business intelligence. The niche players do a fine job at their limited target set of use-cases or markets – their limitation is that they don’t cross into other markets (like all access technologies, or all services) very well. I like to think that Sandvine has been very forward-thinking in recognizing the need to bridge these gaps. We’re the only Network Policy Control vendor who offers an analytics product, so we’re perfectly positioned to help service providers with any access technology to get the focused insight that generic business intelligence cannot yield.
What is the difference between the Network demographics server that comes on the SRP platform and then the NDA – are they both just not reporting services\servers.
They’re really addressing different use-cases. While Network Demographics gives access to data that can be used by many teams within the organization, it’s particularly used by operations and engineering teams to display the past and present in very granular detail. Additionally, it’s well-suited to “on demand” targeted queries. Data is typically displayed in units like bytes or minutes, and generally the data is in raw format; that is, additional operations have not been performed.
Network Analytics is focused on delivering business insight that is usually the result of algorithmic processing. It takes the raw data, processes it, and then displays the past and present, and projects the future based on observable trends. Additionally, Network Analytics normalizes data as much as possible to show relative changes, which are more aligned with the questions posed by executives, marketing, product management, etc. For instance, “How much has per-subscirber peak-time bandwidth increased in the past six months?” is a question that Network Analytics can answer very easily.
Finally, perhaps the most obvious and important differentiation between Network Analytics and Network Demographics is that Network Analytics is designed to take in data from multiple sources, not just Sandvine elements. It’s perfectly suited to augmenting your network data with information from B/OSS, subscriber profile databases, on-deck service statistics, etc.
Is the Analytics platform\server installed with each and every PTS for e.g if we had 5 PTS then we have to buy 5 NDA platforms.
Hi there – thanks for your question. Network Analytics does not have a 1:1 installation requirement. A single analytics platform interacts with any number of data sources, and in a typical deployment it is “plugged in” to the Subscriber Policy Broker, which in turn has a 1:N ratio to the PTS units. So, you can have 1 analytics platform connected to several SPBs, which in turn are each managing the data from multiple PTS units or clusters. Please let me know if I have not sufficiently answered your question.
Your answer perfectly satisfies the question. Thank you.
Will the presentation be e-mailed out or able to download?
Hi. Yes, there will be an on demand version of the webinar available here within 24 hours. We’ll also be in touch about sending you the presentations direct.