Moving Towards Analytics-Supported Network Automation€¦ · Automatable Process User: Radio Access Planning Engineer ... Automatable Process User: Radio Optimization Engineer, Radio

© MYCOM OSI Confidential

Mounir Ladki, President and CTOMYCOM OSI

Oliver Rosenberg, Group Lead PerformanceManagement NT Operations PlanningTelefónica Germany

Moving TowardsAnalytics-SupportedNetwork Automation

© MYCOM OSI Confidential

Company Overview

• Leading independent provider of Assurance, Automationand Analytics software & solutions for managing networkinfrastructure and improving customer experiences

• Deep telecoms expertise serving the world’s largestCommunications Services Providers

• Integrates network, service and customer assurance data tosupport key initiatives and use cases, such as:

– reducing churn, improving quality of service, monetizinginfrastructure and reducing costs

• Headquartered in London, UK with 250+ employeesworldwide

2016 Innovations in Big Data& Analytics

#1 in Unified NetworkPerformance Management #1 Customer Experience

Enabling Solution

2015 Product Line StrategyLeadership Award

2015 Network ManagementExcellence

2016 Outstanding Contributionto Enabling ImprovedCustomer Centricity

2016 Innovative ProjectAward

70+Customers Worldwide

8 out of 10Largest Global CSP

Groups

#1 RankedBy Industry Analysts

© MYCOM OSI Confidential

Telefonica Germany

© MYCOM OSI Confidential

Telefonica Germany

© MYCOM OSI Confidential

Digital Service Environment-Challenges

• Business and customerawareness

• Extensive automation &self orchestration ofhybrid networks

• Network as a service forindustry verticals(energy, automotive,healthcare…)

• High network availabilityand reliability for IoTmanaged services

,Network and business metrics in silos

Big Data is not actionable insight

Data analysis not structured

Investments not based on predictions

,Delayed resolution

Increasing errors and faults

© MYCOM OSI Confidential

Analytics-DrivenAutomation

© MYCOM OSI Confidential

Analytics-Driven Automation ARCHITECTURE

Policy Managem

entand O

rchestrationPolicy M

anagement

and Orchestration

Automation

(Open and Closed Loop)Autom

ation(Open and Closed Loop)

NetworkTeams

NetworkTeams

NetworkSystemsNetworkSystems

Marketing

Network Planning andService Design

AnalyticsAnalytics

Service Assurance

Business Policies Use CasesUsers

Business Driven

Personalized Offers

Business Driven

Network Optimization

Customer Experience

VIP and MVNO priorities

Real-TimeOperations

TechnicalPlanning

BusinessPlanning

• Capex efficiency through pre-integrated network, service,customer analytics

• Agile service launch usingorchestration across hybrid NFVnetworks

• Multi-team efficiency fromcollaborative platform fortechnical, marketing and careteams

• Customer insights and networkmonetization with predictiveanalytics

• Increased employee productivityby closed- loop automation

• Cloud-based, scalable, secure,network-independent platform

© MYCOM OSI Confidential

Architecture for Analytics

Big Data Cluster

Network KPI Application KPI Subscriber KPI Handset KPI

Visualization & Multi-dimensional Analysis

Analytics System

Assurance SystemsPM

FM

SQM

Billing

Marketing

Trouble Tickets

Profiling Correlation Forecasting

Data Discovery and Filtering

Assurance BridgeAssurance Bridge Big Data BridgeBig Data Bridge

Speed and ScaleData discoveryEvery 15 m/1 h

Data Processing speed>150 billion import datarecords/h

Volume of processed data1.2 terabytes per hour

Correlation performance~300 analyticscorrelations/bh

ProbesCM

DistributionProbabilistic and

Cumulative Distributionfor behavior of network

or network elements

ProfilingIdentifies intersections

(same behavioralcriteria) betweennetwork elements

CorrelationPearson’s CorrelationCoefficient to analyze

linear correlationbetween driver variableand candidate variables

in a network area

ForecastingLinear projections frompast to future periods.For entire network or

specific areas

4 Pillars of Structured AnalyticsAnalytics System Architecture

© MYCOM OSI Confidential

Customer CareValue: customer-centric, device-centric and collaborative

Benefit: ~ $ 10* million/year

Use Case: VIP/SLA monitoring, customer experiencescoring, troubleshooting, proactive messaging

Marketing/Business ServicesValue: Customer-centric, service monetization, device/OTTanalytics and collaborative

Benefit: ~ $ 10.3* million/year

Use case: Personalized services, device/apps experience,new service launches

Analytics-Driven Automation Users

Network PlanningValue: Network monetization, proactive/predictiveinvestments and collaborativeBenefit: ~ $ 16.5* million/yearUse Case: Customer-centric investments, capacitytracking, data/device usage tracking

Operations (NOC/SOC)Value: Pre-integrated, automated, employeeproductivity and collaborativeBenefit: ~ $ 9.5* million/yearUse case: Predictive faults, proactive resolution,congestion relief, data/device usage

*Figures have been based on Gartner calculationsof typical large network use case benefits

© MYCOM OSI Confidential

Benefits

Resource Optimization

Predictive Insights

Agile Service Launch

© MYCOM OSI Confidential

Analytics-Driven AutomationTop 3 Use Cases

© MYCOM OSI Confidential

Network Monetization

How long is itexpected to last?

What is theprofile

(preferences,devices, etc.) ofmy customers in

this area

Pushpersonalized

content to thecustomers at a

promotional rate

Monitor QoS,QoE, traffic,

capacity . Tune &adjust

Where do I haveexcess capacityin my network?(free spot map)

Automate

Analytics

Analytics

Analytics

Automate

© MYCOM OSI Confidential

• Predict trends of new mobile/ IoTdevices

• Identify high traffic generatingmobile devices and IoT networks

• Reduce devices that generateabnormal traffic posing security andcongestion risks

• Visualize popular devices in real-time• Bundle tariff plans and devices based

on trends and usage

Mobile and IoT device analytics

© MYCOM OSI Confidential

QoS Orchestration in NFV

Continuous analytics/datastreams from physical andvirtualized networks

VNF Manager

Orchestrator

VirtualInfrastructure

Manager

NFV MANO

EMSs

VNFs

NFVi

OSS

Detect policy violationidentifying probable causes

Trigger OSS and MANOlayers for orchestrationand correction

Execution of feedbackfrom physical andvirtualized network

© MYCOM OSI Confidential

Telefonica GermanyAnalytics based OSS Case Study

© MYCOM OSI Confidential

Telefonica’s Architecture and EcosystemTelefonica and MYCOM OSI

• Integration points betweenPM/Analytics and externalSystems

• Information schedules anddistribution between systems

• Supported business processes• Notifications to third parties• Data On-demand

© MYCOM OSI Confidential

• Intelligent RAN Expansions,optimization, maintenance andplanning– Combine network congestion

and subscriber value densitymetrics

– Isolate top network/value sites– Prioritize capacity expansions– Explore and verify suitable

expansion option

Big Data Integration with Network Data

© MYCOM OSI Confidential

Telefonica’s Value-based Capacity Planning

• Develop cell cluster profiles (cells withsimilar network/device/customercharacteristics)

• Forecast and prioritize capacity upgradesusing geo-location, ARPU, subscriber density,handset mix and OTT data

• Optimize capex planning• Telefonica is now thinking of automating

forecasting and capacity planning forweekly/monthly actions.

Data Sources: PM, CRM, Billing, Probes

User: Radio Access Planning EngineerAutomatable Process

© MYCOM OSI Confidential

Network Centric approach Customer & Value Centric resolution

Re-prioritize

• Correlate PM KPIs with location data,subscriber density, technology andservice type

• Prioritize optimization and rolloutactions (capacity increase, traffic re-routing, re-configuration etc.) based onCongestion Score

• Tactical optimization and rolloutprioritization for future marketingstrategies (e.g. Apple device launch)

• Telefonica will use the above info toautomate optimization and rolloutactions

Telefonica’s Congestion Optimization

Data Sources: PM, CRM, Billing, ProbesUser: SOC Operator, Service Manager, Business Demand ManagerAutomatable Process

© MYCOM OSI Confidential

Telefonica’s Usage-Driven Optimization• Analyze customer behavior/patterns• Develop optimization schemes to follow customer

patterns (eg a 4G user can be forced to 4G cell torelieve a congested 3G cell)

• The EMS parameter(3G to 4G) can be configured inreal-time using automation (eg via a SON system)

Handoffoptimization

through forcedattachment to 4G

sites

Data Sources: PM, CM, CRM, Billing, Probes

User: Radio Optimization Engineer, Radio Parameter Planning EngineerAutomatable Process

© MYCOM OSI Confidential

Summary

© MYCOM OSI Confidential

• Big Data and Network Monetization– Strategic investments based on real network/customer demands– Monetize Big Data and Network Infrastructure on weekly/monthly

basis

• Opex Efficiency– Capacity analysis reduced from weeks to near

real-time– Reduced MTTR– Manpower savings– Solve QoS problems in hybrid NFV networks through real-time

policies and closed-loop automation

• Agile Service Upsells– Location based insights to create new services– Introduce contextual services/products faster– Personalized marketing

Relate CustomerBehavior to

NetworkUtilization

Single Point ofPlanning andOperationsDecisions

Rapid Valuebased Problem

Resolution

Predict CustomerBehavior

Forecast NetworkProblems

SmarterProactive

Investments

Analytics-Driven Automation: Summary

CONFIDENTIAL INFORMATION, RESTRICTED TO MYCOM OSI© MYCOM OSIThis document contains forward-looking statements based on current expectations, forecasts and assumptions of the Company that involve risks anduncertainties. Forward-looking statements are subject to risks and uncertainties associated with the Company's business that could cause actual results tovary materially from those stated or implied by such forward-looking statements.

Thank you