As the digital revolution gains pace, Communication Service Providers get flooded with big data, constantly generated from numerous internal and external sources including Data Warehouses, OSS, BSS, Network elements and the Internet.
Intracom Telecom has embarked on the big data revolution by introducing BigStreamer™, a Big Data Analytics and Complex Event Processing platform developed specifically for the telecommunications industry.
BigStreamer™ empowers CSPs to gain competitive advantage by interpreting the big data they are surrounded with and acting upon it in real time. In short:
Intracom Telecom views CSP success as depending on the CSP's ability to achieve a satisfying customer experience by managing simultaneously four key strategic elements: the Network infrastructure, the Services offered, the Customers themselves and the telco Business.
BigStreamer™ embraces all four strategic elements of a CSP's business, as shown in the Telco-specific Big Data Use Cases table.
BigStreamer™ combines six key components, as shown on the diagram and briefly described below.
A telco-oriented Data Model, accompanied with telco-specific Metrics was developed to enable data collected from various sources to match CSPs business. Both data model and metrics are driven by TM Forum's standards.
BigStreamer™ makes use of the computational and storage power of any Hadoop distribution (e.g. Apache, Cloudera, etc.) installed by the CSP or Intracom Telecom.
An ample Hadoop-enabled Telco Adapters Repository ensures that BigStreamer™ can seamlessly interconnect and capture data from virtually any telco information source, such as: Digital Exchanges, Soft Switches & IMS, Fixed Broadband Access & IPTV, OSS/BSS, LTE, 3G & CDMA, SDH & Metro Ethernet for Network Discovery and many more.
Prediction Model Repository contains KPIs and predictive analytics algorithms used by data scientists to formulate, store and utilize powerful predictive data models that give answers to pressing business challenges that CSPs face. Such big data use cases may include Self-Healing Networks, Subscriber Churn Prediction & Prevention and Real-Time Customer Care Empowerment.
Event Stream Processor module captures streaming data, correlates occurring events and automatically generates triggers or alerts for subsequent actions by other systems. Triggers are determined by applying the appropriate algorithm(s) from the prediction models repository. A telco-oriented domain specific language makes writing complex rules easy. Complex Event Processing functionality is based on Apache Storm, a distributed processing engine, and powered by Drools, a carrier-grade inference engine.
Finally, BigStreamer™ enables the users from various CSP functions to enhance their decisions by conducting ad hoc analytics and visualization. This is made possible through partnership with key technology vendors in business intelligence software, which provide rich and highly-interactive interfaces with web-based reports, dashboards and analytics.