Page 446 - Handbook of Modern Telecommunications

Network Management and Administration 3-237

• Device-based Monitoring: Metrics defining traditional network connectivity relative to broadband
access or message delivery are an essential part of understanding service availability. Traditional
network fault and performance management information is one essential part of understanding
service availability.
• Signaling-based Monitoring: SS7 signaling metrics can provide insight about quality and perfor-
mance issues concerning the flow of messages throughout the PSTN infrastructure.
• Packet-Flow Monitoring: Gathering details of the IP packet header and sometimes payload can
help to identify IP traffic flow bottlenecks, failures, or degradations of IP-based service enablers
or service components, and conditions when IP service activity are not normal.

While after-the-fact trending and analysis is effective for service planning and tracking customer
utilization metrics, an additional requirement rests with data monitoring for those operations work
groups that must deal in real time with customer complaints and network problems. Real-time data
is necessary for isolating and recovering from service problems before they become serious. Such a
strategy sounds straightforward but is difficult to implement due to limited access of partner-supplied
information, data disparities among nonintegrated systems, and often less than optimized processes.
The ability to isolate service problems and to correlate them with network issues is critical to the secu-
rity/stability of the network and in maintaining good customer relations. In most cases, service quality
issues are expected to be identified, isolated, and often resolved in real time to prevent SLA violations
and service disruptions.

3.9.4.1.3 Customer Visibility
Maintaining the reliability and quality of the network and developing a better understanding the sta-
tus of services remain critical; however they alone are not sufficient for ensuring satisfactory customer
experience. The complexity of collecting and correlating sensory data from within a multiservice, multi-
infrastructure environment requires a broader OSS/BSS strategy that involves not only sensory data
collection, but also data correlation and filtering using specific business definitions with the flexibility
to change as business needs change.
To facilitate a closer focus on the customer, key performance indicators (KPIs) should be directly
collected from sensory networks plus additional sources—services platforms, call centers, and business
support systems. These KPIs can be combined to create customer- and operator-specific key quality
indicators (KQIs), including measures of speed, availability, reliability, security, simplicity, and flexibil-
ity in the following broad categories.

• Access: The time to establish a connection and the availability of sufficient bandwidth to use a
service includes metrics that point out:
• Security problems, retries, or poor response times. These are geared to provide warning
about possible network, service, application, or user device problems.
• Areas where service availability, coverage, or bandwidth may be insufficient. These are
geared to identify when and where a user is connected to the network, what services were
used, and how long each service was accessed.
• Usage: Beyond tracking usage for billing, there is need to understand what services, devices,
applications, and access networks are being used, when, and for how long, including metrics
that:
• Identify services or applications that are performing poorly
• Show which services customers are not using, and which might be linked to poor service
definition, lack of flexibility, inadequate support, or simplicity issues
• Point to where potential sales opportunities might exist.
• Disconnect: Determining when a user has failed to connect to a service or the connection has
failed during the user session.