provision, efficiently operate, and accurately bill these


Integrated management architecture leverages common
data and human resource across application functions,
different services, and heterogeneous network
technologies. The requirements for each individual
management function and service offering, however,
may vary. For example, unlike data transfer – based
Ip telephony will bring about dramatic change in the applications such as email, voice transport across an IP-
way IP services are planned, provisioned, managed, and based infrastructure demand low latency and jitter. For
billed. In order to build and retain strong customer base example, the round trip time (RTt), which is the time
for these new services, service providers need to meet, if required by a network to travel from the source to the
not exceed, the customer expectation set by today’s destination and back, including the time to process the
traditional voice services. Acceptance of IP telephony message and generate a reply, should to be less than 250
will depend on the quality and efficiency with which ms. Not only will the permanence threshold triggers be
service providers offer, deliver, and manage IP services. different, the servicing of these packets will be done
Installation, configuration, and activation must be rapid based on a priority scheduling scheme.
and error – free. Furthermore, customers will want
direct control over the reconfiguration of services and
real-time visibility into impact change has on thei
User Datagram Protocol (UDP) and real – time transport
operating costs. Once the service is activated, customers
protocol (RTP) are typically used to transfer Voip
will want the provider to guarantee service quality as
packets. UDP is a connectionless transport layer procol
defined by industry standards. Corporate customer in
in the TCP/IP protocol stack. It is a simple protocol that
particular will need to be assured that the provider is
exchanges datagrams without acknowledgment or
proactively monitoring performance to avoid problems
guaranteed delivery, requiring that error processing and
and providing them visibility into the performance data
retransmission function be handled by other protocols.
collected. This article discussed an integrated manger
RTP is an Ipv6 protocol, which is designed to provide
support system for IP-based networks illustrating the
end - to - end network transport functions for
functions needed to support the unique challenges of
applications transmitting real – time data, such as voice,
managing VoIP services. An example of a service
video, or simulation data, over multicast or unicast
management system is also described.
network services. RTP provides services such as payload
type identification, sequence numbering, time stamping,
and delivery monitoring of real time applications.

IP networking is a booming market for telecom and
datacom service providers and equipment vendors. The particular scheme used and type of traffic will
Service providers are quickly defining and bringing to determine the corrective measures needed to fix a
market differentiated IP service including voice problem. Examples of priority scheduling systems
transport, virtual private networks (VPNS), application include expedited forwarding as defined in the
policy prioritization, differentiated services (e.g., gold, differentiated services (Diffserv) framework, and
silver, bronze), multimedia, and transport Lans ensuring priority Queuing wicth class Based Weighted Fair
profitability from these services requires a Queuing. Integrated management architecture helps
comprehensive service management architecture that mediate these differences, limiting exposure to the
enables service providers to carefully plan, quickly differences for management operators and end

g. For instance. services. links). or directly over wavelength division ATM and IP networks. most important to the busuness (e. backbone link configuration management. critical backbone and network planning. The purpose of statics data are forwarded to the fault module to provide this type of study is to identify situations where comprehensive diagnosis capabilities including alarm . Capacity planning provides a long-term view of network The real time performance management process is a demands and requirements. the most profitable traffic engineering . service activation. the capacity planning function to determine the required Another example of real time performance management capacity as well as optimal equipment locations (or is constant monitoring of high priority customer services homing arrangements) based on forecast or expected (e. especially for / SDH). bottleneck network elements (e. It should be noted that the can also be offered over ATM networks. elements that affect the operation of the entire network) performance management. Finally statistics can also be used as the allowing efficient support of services such as IP basis to observe traffic trends and estimate the load for multicast.g. multiplexing (WDM). The extended is used efficiently by the mission critical aplications functions. function collects and forwards data on a regular basis to supports studies that monitor the ability of the existing the appropriate module..g. which is typically owned and controlled by a service provider. edge. on the other hand. Traffic statistics data is typically forwarded to the performance module for data Figure 1 ilustes the typical structure of an IP network analysis and detection of potential performance including core. This article discussed an extended framework of the network operators to make certain that network capacity FCAPS functions for IP based networks. network control missions. 1 include service restoration. Real time performance management networks. It can be divided into real – time The acronym FCAPS is often used in the literature to (or near-real-time/short term) and long – term refer to integrated management of various types of management. accounting management. Estimated growth factors MPLS networks integrate IP routing protocols directly can then be used to estimate the future forecasting over asynchronous transfer mode (ATM) switch. Performance management is the process of converting raw IP traffic measurements into meaningful performance measures. backbone technology. gold) as well as customers who have been demands. the data collection Long-term performance management. complaining about the performance of their services. and access subnetworks. as shown in fig. The network administrator who wants to applications sharing the sama link get their fair share do hypothetical (what – if) studies typically carries out without interfacing wicth the mission critical traffic. The exceptions (based on the collected and trending data). in this particular example. synchronous data collection functionality has been implemented in optical network / synchronous digital hierarchy (SONET several element management system. alarms and fault IP networks to meet service objective. Once the network is in place. coronations and pinpointing. is Traffic statistics as well as network topology data may assumed to be based on multiprotocol label switching also be for awarded to the planning module to estimate (MPLS) such as core and edge label switch routers. IP class of service. data collection. the base exogenous traffic (2).. traffic loads. IP services use in network engineering.. Customer sites are connected to the network. It computes network mechanism to guarantee that enough bandwidth is element growth rate and generate a long – term capacity reserved for time sensitive IP voice traffic while other expansion plan. It is intended for (1). and security management as well as mission crtical applications. It refers to the open systems interconnection typically includes snapshots of the behavior of (OSI) five functional areas: fault management. and IP VPNS.

computes network element performance measures (route delay. It simplifies the service provider’s ability to validated by data collection system (or element monitor customer services by providing the status of the management systems).) based on the characteristic engineering traffic engineering is perhaps the most challenging loads. to service or performance exceptions.corrective planning is necessary. customer or organization. the last bit in a frame at network ingress to delivery  Number of packets dropped due to protocol errors. fault service – level agreement (SLA). Finally. usually better values. provide early warning of potential service degradation so that a corrective plan can be formulated before service is affected. polled element was active and running.end packet loss.9+ percent.  Response time: measures how quickly the network  Amount of time a network element is in a severe moves information. packet loss ratio. of first bit in the same frame at network egress.  Number of packets dropped due to servere  Network latency: elapsed time between receipt of congestion per inferface.  Amount of time a network element is in a mild  Jitter: variation in network latency. The fault management process is similar to the real – time performance process except that it uses the collected alarms and fault statics to detect and correct problems by pinpointing and correlating faults through Typically.  Loss ration: percentage of sent frame discarded or  Number of times a network element enters a mild not received. Example of IP SLA metric include:  Number of packets dropped due to mild congestion  Service availability: the percentage of time each 1 per interface. congestion state. element loads (utilization. One of the most critical steps in developing inherently includes all the network element that constitute it. and traffic engineering algorithms. congestion state. etc. and then provide end users and performance raw traffic measurements include: network operators the freedom to establish quality and grade of service objectives specify to their applications. The performance management process then converts the  Throughput: total traffic volume (usually in bits per validated raw measurements into meaningful network second). it compares the calculated performance result for the short . the system. jitter. average peak operation without a “hard” failure. comprehensive management methods is to define the required traffic measurements that are the basis for performance. raw traffic measurements are collected.. SLA reports are intended to be tailored to a specific  Number of packets transmitted per interface.  Number of packets received per interface. This is needed when and long terms with the service objectives to identify objectives are not being satisfied and.g. etc. The ability to monitor a service a database. the performance management process function of the management process for IP networks. where possible. end – to . congestion state. The process then than 99. Example of IP SLA reports are intended to correlate fault and performance data. and then stored in batch mode in subscribed services.). congestion state.  Mean time between failures: average down time between consecutive failures. next it calculates statistics to characterize the load  Network uptime: percent of time the network is for traffic engineering purposes (e. average and peak packet loss.  Mean time to repair: average down time (from when  Number of times a network element enters a severe an outage is detected until it is reported fixed). average busy season). delay.

While there are objects. balancing. Markov modulated Poisson process. the fractional Brownian motion procedures. On/off models with exponential on/off distributions. which typically have exponentially decaying correlation functions ). authorization. Traffic engineering is olso an essential input for capacity expansion 2 network dimensioning. . and Markovian models that exhibit only short-range dependence (e. whereas congestion avoidance are typical considerations of backbone routes or switches. and configuration management. Configuration management deals with the physical and geographical interconnections of various IP network elements such as router. and collection of potential servicing problem before the service is operational data.represents the action that the network (or network configuration management also deals with service administrator) should consider in order to relive a provisioning user profile management. and congestion control. defining quality and grade of service requirements. Traffic engineering functionality has been added through the use of tunneling Security management includes authentication. and M/Pareto / models). rerouting. which is the basis for recognizing affected. This includes the action of The development of appropriate models for traffic setting up an IP network or extending an already engineering depends primarily on clear understanding of existing network. and the threshold values.. The on/off model has been proposed to model voice – over – IP (VoIP) calls with alternating active periods (talk spurts) and silent periods. traffic engineering methods depend on the function of the network element. The models can be divided into a user is allowed to do once the user is identified. and network planning. multiplexers. It includes the procedure for initializing. traffic techniques for IP edge routers include packet classification. and modification of network elements and network resources. and taking out existing network elements. setting various parameters. engineering IP- based networks is new. allocating names to management IP statistic characteristics of the traffic. deletion. on/off model with heavy –tailed distributions for the on/off duration. two categories: those that exhibit long-range Authorization usually follows any authorization dependency (e. For instance.. more than a hundred years of experiences in traffic engineering circuit – switched networks. operating. This may include rehoming. Finally.g. setting. load change in the state of the network. model.g. and Gaussian auto-regressive models. admission control. Authentication establishes the identity of both the sender and the receiver of information. and other essential secure communication issues. Integrity checking of confidential information is often done if the Several traffic models and network dimensioning if the identity of the sending or receiving party is not methods for packet networks have been proposed in the properly established (5). The man functions of configuration management are creation. and modifying the set of parameter that control the day to day operation of the networks. Authorization establishes what literature (3) en general. mechanisms or forced route algorithms. and links. switches. The parameters of the on/off models can be estimated from actual traffic traces or by using typical default values.