Transcript Ensuring Service Quality While Increasing Revenue

SP-11: Ensuring Service Quality While Increasing Revenue
February 4, 2009
Daniel Teichman
Senior Product Marketing Manager
Voice Service Assurance
Assuring Performance of Carrier-Class
Networks and Enterprise Contact Centers
Ensuring Service Quality While
Increasing Revenue
•
•
•
•
•
•
Service Quality Measurement
Getting from Measurement to Monitoring
Service Level Agreement (SLA) Benchmarking
SLA Reporting
VoIP Peering Examples
Summary
Testing and Monitoring Lifecycle
Pre-service
lab testing
Test service logic and operational functionality
Pre-service testing
In-service monitoring
Field service testing for performance
Benchmarking service quality & performance
Ensuring ongoing network and service quality
Service Quality Measurement
• Signaling
• Media
• Passive and / or Active Methodologies
Signaling Quality Measurements
• Real-time analysis / reporting of call completion statistics
– Post-Dial Delay, Call Attempts, Calls Successfully Completed, Failed
Calls, Answer Seize Ratio, Answer Bid Ratio, Network Efficiency
Ratio, Call Setup time, Call Disconnect time, Minutes of Use
• Real-time analysis / reporting on protocol statistics:
– SIP packets, SIP Invites, SIP Errors, SIP 1xx / 2xx/ 3xx etc, SIP
Registers, etc.
• Per call capture of signaling quality measures, retained
storage for diagnostic and analysis purposes
Media Quality Measurements
• More difficult to capture than signaling quality measures
– Includes RTP, as well as DTMF (RFC 2833) and fax (T.38)
• Real-time analysis / reporting of packet and call statistics
– Packets Received, Lost, Duplicate, OOS
– Jitter, Burst; Gap, Mean Opinion Score (MOS)
– Instantaneous or Worst point in call as well as Average per call
• Per call capture of media quality measures, retained
storage for diagnostic and analysis purposes
Getting From Measurement to Monitoring
• Measurement gives you data – lots of it!
• Monitoring create value from measurement data. It gives
you information for diagnosis, analysis and reporting
• Monitoring enables pro-active behaviors, enhances reactive
behaviors
• Correlation, data segmentation and automatic notification
of service issues are the three keys to success
Correlation per Call / Session
• Correlate signaling across both trusted and untrusted sides of SBC
• Complicated call flows, e.g. SBCs can/will change call Ids
• Measure RTP across both trusted and untrusted sides of SBC
Data Segmentation
• Network-wide assessment is not sufficient for specific SLAs
• Data segmentation is a mandatory requirement
– Per network segment or component
– Per interconnect point
– Per customer (typically large enterprise)
• Should be part an integral part of the database structure as
opposed to a reporting tool
– Integration is a must for responsive reporting (when you have
million of records to analyze)
– Segmentation has to be easy to apply – could have thousands to
manage
Data Segmentation Example
• Data segmentation by “tagging” Call Detail Information
• “Tag” assignment by multiple parameters, such as:
–
–
–
–
Phone number digits (country code)
IP address (source and/or destination)
SS7 address (originating and/or destination point code)
VLAN Id
• Combine “tags” to capture multiple “views” of traffic
Automatic Notification of
Service Level Issues
• Pro-active integration with “northbound” Network
Management systems
• Both network and segmented data based alerts
– Network-based addresses internal metrics
– Segmentation-based alerts enable isolation to particular
interconnect, peering site, or Enterprise
• Alerts on both signalling and media metrics, such as:
– Call Failure rates exceed 1.0% threshold
– Average MOS falls below 3.0
SLA Benchmarking
• Done by testing live network conditions
– For the network
– Per interconnect partner / Enterprise
• Done for specific Key Performance Indicators, such
as:
– Average / Minimum MOS
– Average / Maximum Post Dial Delay
– Answer Seize Ratio
• Should be normative across network if possible
SLA Reporting
• SLA verification via ongoing monitoring or
via periodic or on-demand testing
– For the network
– Per interconnect partner / Enterprise
• Trend analysis indicates potential problems
or future areas for re-negotiation
VoIP Peering Example 1
• Event: Carrier routing table update process fails to complete properly
• Problem: Standard network level alerts will not recognize this problem
• What service quality monitoring does:
– Identification of abnormal traffic conditions – low call completion rates
– Pro-active alert to Network Management System upon threshold violation
– Ability to isolate problem to specific peering partner or peering sites
• Remediation: Re-routing of traffic to alternate carrier
• Net impact: Minimize or avoid customer service disruptions  reduce
churn  minimize revenue loss
VoIP Peering Example 2
• Event: SLA compliance dispute between carrier and customer
• Problem: Customer believes VoIP quality falls below agreed SLA
• What service quality monitoring does:
– Analysis and reporting on quality measurements specific to customer
– Accommodate multi-site, multi-country reporting
– Trend analysis over time
• Remediation: Set up periodic capture specific quality metrics from
customer sites, using SIP loopback or SIP VQ reporting capabilities
• Net impact: Minimize or avoid future customer disputes  maximize
revenue potential
Summary
• Service Quality Monitoring is MUCH more
than just measurement
• Service Quality Monitoring provides the ability
to perform SLA verification and reporting
• More accurate handling of SLAs means more
valuable peer and customer relationships