a total cost of ownership (tco) analysis and comparison of ... · pdf filea total cost of...

A Total Cost of Ownership (TCO) Analysisand Comparison of Edge Aggregation

Network Architectures

FUJITSU NETWORK COMMUNICATIONS INC.2801 Telecom Parkway, Richardson, Texas 75082-3515Telephone: (972) 690-6000(800) 777-FAST (U.S.)us.fujitsu.com/telecom

Network Strategy Partners, LLC (NSP), management consultants to the networking industry, helps serviceproviders, enterprises, and equipment vendors around the globe make strategic decisions, mitigate risk andeffect change through custom consulting engagements. NSP’s consulting includes go-to-market strategies,development of new service offers, pricing and bundling, and infrastructure consulting. As respectedthought-leaders in the networking industry, NSP’s consultants influence the industry through confidentialengagements for industry leaders, public appearances, and trade magazine articles. These interactionsassure NSP’s clients that they will be among the first to know the latest industry concepts and emergingtechnology trends. Each consulting engagement is uniquely structured—no forced methodologies orcanned reports are employed. NSP’s consultants’ collective experience is derived from leading firms across abroad spectrum of professional disciplines including management consulting, engineering, marketing,financial analysis, and IT management. Contact NSP at www.nspllc.com.

1


Executive SummaryIn 2004, Frame Relay and ATM generated more than $25 billion in revenue for service providers worldwide.For Tier 1 service providers in the United States, Frame Relay and ATM still represents the largest and mostprofitable revenue stream with respect to data services. However, as service providers expand theirofferings of Ethernet and Private IP services (such as IP VPNs) while initiating the migration of multiservicetraffic to IP/MPLS backbones, questions are beginning to arise regarding the most economic and reliablemeans to aggregate this traffic while maintaining customer SLAs.

Industry forecasts indicate IP VPNs and Ethernet-based services are poised for rapid growth, and a recentsurvey of several major carriers indicates that by the end of 2006, 27% of Layer 2 traffic will migrate fromnative ATM networks to IP/MPLS core backbones. In response to these planning requirements, three distinctaggregation approaches have been developed and/or implemented. These approaches can be simplydescribed as follows:

The “Upgrade Approach” – Upgrade the installed base of legacy ATM switches. At a minimum, thisincludes enhancements to the MPLS control plane and line card upgrades for new services such as PWE3VLLs. This may also include the introduction of a ‘Next Generation’ ATM switch with higher capacities andimproved port densities.

The “Layer 2.5 Approach” – Install a new breed of edge aggregation switch. Layer 2.5 combines thebest of Layer 3 application awareness with the best of Layer 2 service interworking, operations, andeconomics. A key aspect of the Layer 2.5 approach is a purpose-built architecture designed to efficientlyaggregate and groom PWE3 virtual leased lines along with traditional Frame Relay, ATM, and Ethernetservices.

The “Layer 3 Approach” – Install an MSER. Although these products lack a Layer 2 control plane andspecific migration capabilities, MSER vendors position these platforms for the termination of both legacyand new services at the MPLS edge.

The objective of this document is to provide an in-depth economic analysis comparing these threeimplementations for aggregating and grooming a portfolio of data services including Ethernet, IP VPNs,Frame Relay, and ATM at the edge of a service provider’s network. NSP, LLC analyzed the TCO of eachimplementation, comprised of CAPEX and OPEX, based on modeling a hypothetical mid-sized network’sgrowth over a five-year period.

With five-year revenue forecasts for each of the above services, NSP compared the TCO of four ”real-world”edge device alternatives: the Fujitsu FLASHWAVE® 6400 Layer 2.5 aggregation switch, the leadingincumbent Layer 2 multiservice switch (Lucent CBX 500), the incumbent’s next-generation Layer 2multiservice switch (Lucent CBX 3500), and one of the leading Layer 3 MSERs, (Juniper M320).

For your convenience, a list of acronyms can be found at the end of this document.

2


Based on our analysis contained in the body of this paper, NSP can definitively reach the following threeconclusions:

1) Using Layer 3 MSERs for edge aggregation is consistently the most capital-intensive CAPEX and incursthe highest OPEX for a service provider. These conclusions occurred in all scenarios studied.

2) The FLASHWAVE 6400 Layer 2.5 aggregation switch is the most cost-effective edge aggregationsolution—by a very significant margin, for both CAPEX and OPEX—among the alternatives for allscenarios analyzed in this paper.

3) The FLASHWAVE 6400 Layer 2.5 aggregation switch TCO is consistently at least half the TCO of theMSER alternative in all scenarios analyzed in this paper.

NSP reached its conclusions by evaluating and comparing two general approaches to building an edgeaggregation infrastructure: edge switches backhauling to routers, and an all-router network. For the edgeswitch approach, carrier edge switches such as the FLASHWAVE 6400 Layer 2.5 aggregation switch are usedto terminate Frame Relay, ATM, and Ethernet services. IP VPN services are aggregated on these edgeswitches and backhauled over an MPLS network to an MSER located closer to the network core. In the all-router approach, Layer 3 MSERs are used at the edge of the network to terminate Frame Relay, ATM,Ethernet, and IP VPN services. The Layer 3 routers are interconnected via an MPLS core network.

This study modeled a mid-sized hypothetical network with 50 POPs and 10 Super-POPs, and analyzed thefollowing scenarios:

• Aggregation of Frame Relay, ATM, Ethernet, and IP VPN services;• Aggregation of Frame Relay, ATM, and Ethernet services;• Aggregation of Frame Relay and ATM services only; and• Migration of Frame Relay, ATM, Ethernet, and IP VPN services.

For the purposes of using this paper in evaluating a ”greenfield build” scenario, the reader can utilize themigration scenarios. In a ”greenfield build”, our assumption is the incumbent will use their next-generationswitch.

Based on the results from this paper’s analysis, NSP would make the following recommendations to serviceproviders:

1) Using Layer 2.5 switches for aggregation and MPLS backhaul of IP VPN traffic to Gigabit Ethernetenabled Layer 3 edge routers is the most cost-effective approach to transitioning the network.

2) The Layer 2.5 approach provides greater flexibility in access technology choices, significant QoSassurance, OAM&P options, and reduces the complexity of provisioning new services. Given thecompetitive nature of today’s marketplace, it is imperative for service providers to use a moredifferentiated and cost-effective aggregation and backhaul architecture to reach profitability sooner.

The body of this paper presents the details of NSP’s assumptions, economic analysis, and results.

3


IntroductionToday, legacy data services such as Frame Relay and ATM dominate the data services market and generatethe vast majority of service providers’ data revenues and margins. However, the emergence of “any-to-any”connectivity requirements and new productivity applications has triggered exponential demand for IP-based data services. As a result, this activity has driven a core transition to IP/MPLS convergence. In orderfor carriers to remain competitive in this dynamic marketplace, it is essential that they reduce capitalexpenses and operational expenses while maintaining the agility to support new and emerging services.However, cost reductions and new service offerings cannot come at the expense of disrupting ordiscontinuing legacy data services.

One of the major challenges facing service providers today is architecting the network edge such that itsupports new IP and Ethernet services as well as legacy Frame Relay and ATM services. An additionalchallenge is migrating legacy traffic from the legacy ATM core network to the new IP/MPLS core network.A recent survey of several major carriers indicates that by the end of 2006, 27% of Layer 2 traffic willmigrate from native ATM networks to IP/MPLS core backbones.

In response to these planning requirements, three distinct aggregation approaches have been developedand/or implemented. These aggregation approaches can be simply described as follows:

The “Upgrade Approach” – Upgrade the installed base of legacy ATM switches. At a minimum, thisincludes enhancements to the MPLS control plane and line card upgrades for new services such as PWE3VLLs. This may also include the introduction of a ‘Next-Generation’ ATM switch with higher capacities andimproved port densities.

The “Layer 2.5 Approach” – Install a new breed of edge aggregation switch. Only one vendor, Fujitsu,offers a platform architected specifically for Layer 2.5 aggregation. That platform is the FLASHWAVE 6400Layer 2.5 aggregation switch.

The “Layer 3 Approach” – Install a MSER. Although these products lack a Layer 2 control plane andspecific migration capabilities, MSER vendors position these platforms for the termination of both legacyand new services at the MPLS edge.

NSP compared these three approaches in a variety of different scenarios. In each scenario, service revenues,gross profits, capital expenses, operating expenses, cash flows, ROI, TCO, and payback periods are contrastedfor each of the alternatives. In all the scenarios analyzed, the Layer 2.5 edge aggregation architecture hasproven to be the lowest cost approach for offering both legacy and emerging data services. Furthermore,the FLASHWAVE 6400 Layer 2.5 aggregation switch proves to be the best economic solution withsignificant capital and operations cost advantages over the alternative solutions.

4


Architecture OverviewThe analysis set forth in this white paper results from a financial model developed by NSP. This model isflexible and can be configured to analyze different types of networks, different levels of service demands,different pricing models, and supports multiple network architecture alternatives and sizes. For thepurposes of this white paper, we have modeled a mid-sized generic network with 50 POPs and 10 Super-POPs. Our assumptions for service growth and pricing are presented in the next section.

The results presented in this paper are based on the alternative architectures described earlier. Thefollowing paragraphs will describe each of these architecture alternatives in detail to provide a betterunderstanding of our economic analysis.

The first alternative, the FLASHWAVE 6400 Layer 2.5 aggregation switch, is depicted in Figure 1.

Super-POP

FRATM

Ethernet

POP

OC-12 POS orGigabit Ethernet

OC-12 ATM

Core RouterEdge Router

POP

FRATMEthernet

FRATMEthernet

CoreATM

SwitchFLASHWAVE

6400

FLASHWAVE

6400

FLASHWAVE

6400

Figure 1: Example of the Layer 2.5 Approach

In the economic model, it is assumed that the network consists of POPs and Super-POPs. In each POP thereare one or more FLASHWAVE 6400 Layer 2.5 aggregation switches that can terminate Frame Relay, ATM, andmetro Ethernet traffic. The FLASHWAVE 6400 Layer 2.5 aggregation switch connects to both the ATM andIP/MPLS core and performs service interworking between Frame Relay, ATM, and metro Ethernet. The ATMcore is used for transport of legacy traffic and service interworking. The IP/MPLS core is used in support ofnew IP services (such as IP VPNs). Over time, legacy PVCs are migrated from the ATM to the IP/MPLS core.

Each Super-POP also hosts a FLASHWAVE 6400 Layer 2.5 aggregation switch that connects directly to thecore ATM and IP/MPLS switches over either an OC-12 ATM or Gigabit Ethernet connection. In thisarchitecture, an MSER located in the Super-POP provides IP VPN services. VPN traffic is backhauled to theMSER using Frame Relay, ATM, or Ethernet connections over MPLS. The RFC 2547bis IP VPN is then initiatedat the MSER and carried across the IP/MPLS network.

5


The second alternative (Layer 2 incumbent multiservice switch) is depicted in Figure 2.

Figure 2: Example of the Layer 2 Incumbent Multiservice Switch Approach

Super-POP

FRATM

Ethernet

POP

OC-12 ATM


POP

FRATM

FRATM

Ethernet

Ethernet

CoreATM

Switch

IncumbentLayer 2Switch

IncumbentLayer 2Switch

IncumbentLayer 2SwitchMetro

EthernetMetro

Ethernet

MetroEthernet

The incumbent multiservice Layer 2 switches are used to terminate Frame Relay and ATM traffic.Additionally, Layer 2 Ethernet switches are required to terminate metro Ethernet traffic at the networkedge. The incumbent switches are connected to the ATM core and are also connected to the IP/MPLS coreusing OC-12 ATM connections. It is assumed that the incumbent switches will migrate over time from theATM core to the IP/MPLS core using ATM connections to MPLS routers. In this scenario, the MSER in theSuper-POP also provides IP VPN services.

6


The third alternative (Layer 2 next-generation multiservice switch) is represented in Figure 3. This Layer 2alternative is very similar to the incumbent solution in Figure 2. The primary difference is that the edgeswitch is a next-generation switch with a larger switching fabric and higher port densities than theincumbent’s legacy system.

Super-POP

FRATM

Ethernet

POP

OC-12 ATM


POP

FRATM

FRATM

Ethernet

Ethernet

CoreATM

Switch

IncumbentNextGen

Switch

IncumbentNextGen

Switch

IncumbentNextGen

SwitchMetroEthernet

MetroEthernet

MetroEthernet

Figure 3: Example of the Layer 2 Next-Generation Multiservice Switch Approach

The final alternative studied in this paper (Layer 3 multiservice edge router) is presented in Figure 4. In thisarchitecture, Layer 3 MSERs terminate all edge services: Frame Relay, ATM, Ethernet, and RFC 2547bis IP VPN.The MSER connects to the legacy ATM network via an OC-12 ATM connection, and it connects to theIP/MPLS network via either a Gigabit Ethernet connection or an OC-48 POS connection. Legacy FrameRelay, ATM, and metro Ethernet services are carried over MPLS LSPs. IP VPNs are provisioned at the networkedge. Each of the architecture models described above is used in the TCO analysis.

7


Super-POP

FRATM

Ethernet

POP

Core Router (T640)

POP

FRATM

FRATMEthernet

Ethernet

CoreATM

Switch

POP

FRATM

Ethernet

Figure 4: Example of the Layer 3 Multiservice Edge Router Approach

Service ProfilesSome of the key financial metrics presented in this study (such as ROI, payback period, and discounted cashflows) are closely tied to the service revenue model. Over a five-year period, there is a high degree ofuncertainty associated with service revenue growth rates. Some analysts have projected low levels of FrameRelay growth and high levels of Ethernet and IP VPN growth. Other analysts have projected higher levels ofFrame Relay growth due to a large-scale migration of circuit voice to VoIP. Today there are a large number ofT1 voice circuits that could migrate to T1 Frame Relay circuits as enterprises migrate to VoIP. However, forthe purpose of this white paper, NSP decided to take a conservative position on Frame Relay and ATMgrowth rates as specified in Figure 5. It should be stressed that if Frame Relay revenues grow at an annualrate higher then 2%, then the financial results for the Layer 2.5 aggregation solution will improve. Ethernetand IP VPN growth rates are more aggressive and are based on NSP’s engagements with service providersand the MEF.

Figure 5: Service Price Depreciation and Revenue Growth Rate Assumptions

Service Price Depreciation Rate Revenue Growth RateFrame Relay 10% 2%ATM 10% -20%Metro Ethernet 15% 19%IP VPN 12% 61%

8


The assumptions for access line demand are combined with pricing assumptions to calculate servicerevenues as shown in Figure 6.

Figure 6: Estimated Service Revenue Projections

The chart in Figure 6 represents the service revenues from Frame Relay, ATM, metro Ethernet, and IP VPNover a five-year period allocated to the edge equipment ROI calculation. In this model, it is assumed that50% of the total revenues from these services will be allocated to calculate the ROI on the network edge.The other 50% of the revenues are allocated to other network resources such as the core network. In thisprojection, Frame Relay remains relatively static, ATM is shrinking, and Ethernet and IP VPNs are growing.The distribution of revenue is presented in Figure 7. In this scenario, IP VPNs will become the largestrevenue generating service by year five.

0

50

100

150

200

250

300

350

2004 2005 2006 2007 2008

Year

$ M

illio

ns

IP VPN

Metro Ethernet

ATM

Frame Relay

9


Figure 7: Service Revenue Breakdown

0

20

40

60

80

100

1 2 3 4 5

Per

cen

t

IP VPN Metro Ethernet ATM Frame Relay

Economic Comparison of Edge ArchitecturesThe economics of the four edge network architectures presented in the Architecture Overview will becompared in several different deployment scenarios. NSP evaluated two different deployment approachesfor new edge platform deployments in existing networks:

1. New Aggregation2. Replace and Migrate

In the New Aggregation model, the old switches are run in maintenance mode, and new services are turnedup on the new platforms. In the Replace and Migrate model, old switches are swapped out and replaced bynew equipment. All services are migrated to the new network. The deployment scenarios considered in thispaper include both New Aggregation and Replace and Migrate.

Additionally, NSP will consider deployment scenarios with the following service combinations:• Frame Relay and ATM only• Frame Relay, ATM and Ethernet• Frame Relay, ATM, Ethernet and IP VPN

10


New Aggregation: Frame Relay, ATM, Ethernet and IP VPNIn this deployment scenario, a new aggregation strategy is employed by the service provider. All theservices (Frame Relay, ATM, Ethernet and IP VPN) are offered in accordance with the service profiles andrevenue models discussed earlier. Given the set of assumptions specified in the architecture and serviceprofile sections above, NSP used its economic model to create a five-year forecast. The results depicted inFigure 8 are cumulative revenues and expenses over the five-year period.

Figure 8: Cumulative Revenue and Expenses for New Aggregation: Frame Relay, ATM, Ethernet and IP VPN

Fujitsu Incumbent Incumbent NextGen Router Assumptions

Allocated Revenue Frame Relay 581,737,911$ 581,737,911$ 581,737,911$ 581,737,911$ ATM 37,607,167$ 37,607,167$ 37,607,167$ 37,607,167$ Metro Ethernet 59,481,128$ 59,481,128$ 59,481,128$ 59,481,128$ IP VPN 338,997,635$ 338,997,635$ 338,997,635$ 338,997,635$ Cumlative Allocated Revenue 1,017,823,841$ 1,017,823,841$ 1,017,823,841$ 1,017,823,841$ 50% AllocationCost of Revenue 712,476,689$ 712,476,689$ 712,476,689$ 712,476,689$ 30% Gross Margin

Cumulative Allocated Gross Profit 305,347,152$ 305,347,152$ 305,347,152$ 305,347,152$

Capital Cost

FLASHWAVE 6400 Layer 2.5 Aggregation Switch 37,895,000$ -$ -$ -$ Incumbent switch -$ 62,848,000$ -$ -$ Incumbent NextGen switch -$ -$ 48,286,000$ -$ Multiservice Edge Router 5,450,000$ 5,450,000$ 5,450,000$ 91,712,500$ MPLS Core Router 6,250,000$ 9,850,000$ 9,850,000$ 6,250,000$ Core ATM Switch -$ -$ -$ -$

Metro Ethernet Edge Switch -$ 8,640,000$ 8,640,000$ -$ Installation 4,959,500$ 8,678,800$ 7,222,600$ 9,796,250$ Spares 1,983,800$ 3,471,520$ 2,889,040$ 3,918,500$

Cumulative Capital Cost 56,538,300$ 98,938,320$ 82,337,640$ 111,677,250$

Network Operations Expenses

Capacity Planning 366,923$ 1,033,846$ 670,385$ 394,615$ Network Administration 1,626,923$ 3,627,692$ 2,537,308$ 5,919,231$ Service Provisioning & Configuration 2,622,115$ 6,853,846$ 4,517,308$ 4,906,731$

NOC Support 13,240,385$ 33,552,692$ 22,337,308$ 29,440,385$ Field Support 8,930,769$ 23,365,385$ 15,421,154$ 16,823,077$ Environmental 6,177,853$ 21,097,369$ 11,209,715$ 11,898,544$ Financial Penalties for Flash Cut Outages -$ 1,321,607$ 1,321,607$ 1,321,607$ EMS Service Migration Expenses 185,000$ -$ 570,000$ 570,000$ Training 1,607,227$ -$ 272,901$ 1,724,521$ OSS Service Migration Expenses 5,643,633$ -$ 1,410,908$ 1,410,908$ Hardware Support 2,826,915$ 5,580,800$ 4,116,882$ 5,583,863$ OSS Support 267,871$ 684,335$ 454,835$ 574,840$

Cumulative Expenses 43,495,614$ 97,117,573$ 64,840,309$ 80,568,321$

Cumulative Cash Flow 205,313,239$ 109,291,260$ 158,169,203$ 113,101,581$

TCO 100,033,914$ 196,055,893$ 147,177,949$ 192,245,571$ROI ([Revenue-Expenses]/Capital) 463% 210% 292% 201%Payback Period in months 5.42 10.68 7.95 11.70

11


This analysis shows that over a five-year period, the edge switch architecture is far more cost-effective thanthe edge router from both a CAPEX and OPEX perspective. Also, the FLASHWAVE 6400 Layer 2.5aggregation switch is the most cost-effective solution out of all the edge switch solutions studied. Thecumulative TCO is presented in Figure 9, and the CAPEX for each of the alternatives over a five-year periodis presented in Figure 10.

Figure 9: Cumulative TCO

0

50

100

150

200

250

FLASHWAVE6400

IncumbentNext-Gen

Incumbent Router

$ M

illio

ns

Cumulative Expenses

Cumulative Capital Cost

Figure 10: Annual Capital Expense

0

5

10

15

20

25

30

35

1 2 3 4 5

Year

$ M

illio

ns

FLASHWAVE 6400

Incumbent Next-Gen

Incumbent

Router

12


There are two drivers of TCO: CAPEX and OPEX. The primary reasons that edge switch CAPEX is less thanMSER CAPEX is that the cost of multiservice interfaces on high-speed edge router platforms are moreexpensive. This cost factor tends to make MSERs more expensive than edge switches on a per-port basis.The incumbent switch, while less expensive then the MSER, is more expensive than the other alternativesbecause of its architectural limitations in switch fabric size and its one-to-one mapping between theprocessor cards and the physical interface ports. The FLASHWAVE 6400 Layer 2.5 aggregation switch hasthe lowest CAPEX due to its bandwidth pooling architecture, high port densities, scalability and low costper port.

These results also show OPEX to be lower for the edge switch solution and lowest for the FLASHWAVE 6400Layer 2.5 aggregation switch. Because routers are more complex devices to configure, operate, andtroubleshoot, the MSER’s OPEX tends to be higher than the switch’s OPEX. The skill level required tosuccessfully operate MSERs tends to be higher, which often translates to higher salary costs in the NOC andin the field. The complexity also tends to lead to more service outages, which impacts SLAs and the revenuestreams associated with the services affected by the unplanned outages.

For the switching functions in networks, OAM&P is typically handled by lower salaried employees thanthose required to support MSERs. The architecture of the FLASHWAVE 6400 Layer 2.5 aggregation switchalso provides significant savings in that the platform can support higher port densities due to bandwidthpooling and the utilization of universal service modules (“any service, any port”). The FLASHWAVE 6400Layer 2.5 aggregation switch is the only edge device on the market with bandwidth pooling, whichsupports the most efficient aggregation economics, especially with a mix of low-speed Frame Relay trafficand higher-speed ATM and Gigabit Ethernet traffic.

In order to provide a deeper understanding of the economic analysis and results, the following paragraphswill provide a detailed description of each of the line items presented in Figure 8.

Allocated RevenueThe allocated revenue section of the model calculates the gross profit that will be used to calculate thecumulative cash flows, ROI, and payback period for capital investments. The top line revenue calculation isdescribed in detail in the Service Profiles section of this paper. The key facts are that the revenue numbersare calculated based on unit demand, pricing and a 50% allocation of total revenues to edge capital.Therefore, the revenues projected in Figure 8 are 50% of the total revenues generated from the servicesspecified in the demand matrix. The model also assumes a gross margin of 30% to calculate the cost ofrevenue and the gross profit. Cost of revenue is made up of multiple components including sales,marketing, transport costs and access charges. The gross profit numbers are then used to calculatecumulative cash flow, ROI and payback.

Capital CostsCapital costs are calculated by configuring network elements in the POPs and Super-POPs and adding upthe costs of all the components using the published U.S. list prices of the corresponding BOM created bythe assumed configuration. These costs include edge switches, processor cards, switch fabric modules,software, and the line cards in core routers/switches necessary to accommodate new traffic. Eachcomponent of the CAPEX is described in the following paragraphs.

13


Edge SwitchThere are three line items in Figure 8 for edge switches: FLASHWAVE 6400 Layer 2.5 aggregation switch,Incumbent switch, and Incumbent NextGen switch. In each of the scenarios, these line items constitute thecapital expense for all edge switching equipment. The capital cost of the FLASHWAVE 6400 Layer 2.5aggregation switch is significantly lower than the presented alternatives in all the reviewed scenarios.

Multiservice Edge RouterThis line item specifies the CAPEX of a multiservice edge router. In the Layer 3 solution, MSERs are used toterminate both Layer 2 and Layer 3 services. The other three alternatives use edge switches to terminateFrame Relay, ATM, and Ethernet services. IP VPN services are backhauled to an MSER in a Super-POP, whichdoes the IP VPN processing. This MSER is deployed as a ‘one-armed router’ (also known as a “router on astick” approach) connected to a core router via one or more Gigabit Ethernet connections. The financialanalysis in Figure 8 demonstrates that this is a much more cost-effective way to provide IP VPN servicesthen terminating IP VPNs at the edge of the network. This is because it is more expensive to terminate alarge number of DS1 and subrate DS1 connections on an MSER as compared to aggregating those serviceson a switch and then backhauling those services over an MPLS-enabled Gigabit Ethernet trunk circuit thatterminates on the MSER.

MPLS Core RouterThis line item specifies the CAPEX required in the IP/MPLS core network to accommodate new edgedevices. It only accounts for line cards required to support new interfaces. It assumes that an IP/MPLS corenetwork is already in place so it does not account for the costs of building the core network. Incumbentswitches result in higher IP/MPLS core costs than the FLASHWAVE 6400 Layer 2.5 aggregation switchbecause those switches must use ATM conneections to the MPLS core routers while the FLASHWAVE 6400Layer 2.5 aggregation switch uses Gigabit Ethernet interfaces. ATM interfaces are more expensive thenGigabit Ethernet and therefore the core routing line card expenses are higher for the incumbent solutions.

Core ATM SwitchThis line item is similar to the MPLS Core Router line item except that these are the line cards associatedwith the legacy ATM network. In the New Aggregation model, all new traffic is connected to the IP/MPLSnetwork so these costs are zero. In the Replace and Migrate model, these costs are compared for thealternative solutions.

Metro Ethernet Edge SwitchNone of the incumbent edge switches support Gigabit Ethernet for Ethernet services. Therefore, it isnecessary to use separate metro Ethernet switches to provide these services. These are the capital expensesassociated with these switches.

14


InstallationThese are the installation costs associated with all of the capital equipment. These costs include all the costsassociated with truck rolls, site preparation, wiring, environmental preparation, chassis installation, and cardinstallation. The installation costs were calculated as a percentage of CAPEX.

SparesThese are the costs associated with spare line cards for all capital equipment. Spares costs were calculatedas a percentage of CAPEX.

Network Operations ExpensesNetwork operations expenses are calculated using a model that has been refined by NSP over a series ofprojects with both vendors and service providers. The various components of the operations expense arecalculated using data from equipment configurations and NSP’s database. In each of the operationsexpense categories, the number of employees, skill levels, and salary grades are calculated anddifferentiated between the switch and router architectures. The operations expenses considered in thismodel are only those expenses related to the network edge. Operations expenses associated with thenetwork core, transport, access, and other network components are not considered in this paper becauseour objective is only to compare and contrast edge aggregation alternatives.

The components of OPEX are described in the following paragraphs.

Capacity PlanningNetwork capacity planning is an engineering function. It involves analyzing measured statistical data andusing capacity planning tools to estimate the need for more transmission bandwidth, switches, line cards,and physical capacity.

Network AdministrationNetwork administration is an operational process associated with maintaining switches and routers withthe correct software revisions and the correct configurations. Administration also includes performanceand fault monitoring.

Service Provisioning & ConfigurationAs customers order new services, the operations group needs to provision circuits and configure routers.This is the process of large-scale provisioning and configuration to add new customers and services to thenetwork.

NOC SupportNOC support is the operations costs associated with running a NOC, a 24 x 7 call center that deals withLevel 1 and Level 2 support. Level 3 support is usually escalated outside the NOC. Both technician-level andengineering-level staff is included in NOC support.

15


Field SupportField support costs are associated with truck rolls. These costs include field technicians and field engineersproviding service to remote central offices.

EnvironmentalEnvironment costs include floor space, power, cooling, and battery backup. These costs are calculated usingthe network configurations.

Financial Penalties for Flash Cut OutagesThe migration of PVCs from the ATM network to the IP/MPLS core network is a complex activity. TheFLASHWAVE 6400 Layer 2.5 aggregation switch is designed as a migration platform to allow for the smoothmigration of PVCs without incurring network outages. The FLASHWAVE 6400 Layer 2.5 aggregation switchuses a dual ATM and MPLS control plane and a ‘bridge-and-roll’ software feature to facilitate non-disruptiveATM-to-MPLS PVC migration. The other solutions studied in this paper all require service outages to allowfor PVC migration. These are the penalties paid on SLAs for migration-related PVC outages.

EMS Service Migration ExpensesAs new platforms are added to the network, service providers must incur an EMS migration expense.This is the expense associated with integrating a new EMS into the OSS infrastructure. The only platformthat does not incur this expense is the incumbent solution. The Pegador EMS uses standards-basedinterfaces (SNMP, CORBA, and XML), and, unlike the incumbent or MSER, has one integrated solutionsupporting both legacy and Ethernet services to simplify EMS migration. Therefore, EMS migration costs are less for the FLASHWAVE 6400 Layer 2.5 aggregation switch than the router or NextGen switch.

TrainingAdding new platforms to a network requires training engineers, technicians, and operators. Training is anexpense associated with integrating each of the new platforms. The incumbent, of course, does not incurtraining expenses.

OSS Service Migration ExpensesWhile part of the expense of integrating a platform into the network is accounted for in the section on EMSexpenses, the greater component of cost is associated with full integration with the service provider’s ownOSS, and may include costs to integrate with NMA, TIRKS, and multiple other systems. For the purposes ofthis analysis, it is assumed that there is no cost associated with OSS integration for the incumbent and thatthe incumbent NextGen switch and the router already have some level of integration.The FLASHWAVE 6400Layer 2.5 aggregation switch requires complete OSS integration. Consequently, it incurs the highest OSSmigration expense.

Hardware SupportThis is the expense the service provider incurs with hardware support contracts.

OSS SupportThis is the expense the service provider incurs with OSS support contracts.

16


The following sections present different New Aggregation and Replace and Migrate scenarios. The lineitems in each of these tables are the same as those described above.

New Aggregation: Frame Relay, ATM and EthernetIn this deployment scenario a New Aggregation strategy is employed for Frame Relay, ATM, and Ethernetservices. It is assumed that IP VPN services are not offered in this scenario. The Frame Relay, ATM, andEthernet services are offered in accordance with the service profiles and revenue models discussed earlier.Given the set of assumptions specified in the architecture and service profile sections of this paper, we usedour economic model to create a five-year model. The results, depicted in Figure 11, are cumulative resultsover the five-year period.

Figure 11: Cumulative Revenue and Expenses for New Aggregation: Frame Relay, ATM and Ethernet


Allocated Revenue

Frame Relay 581,737,911$ 581,737,911$ 581,737,911$ 581,737,911$ ATM 37,607,167$ 37,607,167$ 37,607,167$ 37,607,167$ Metro Ethernet 59,481,128$ 59,481,128$ 59,481,128$ 59,481,128$ IP VPN -$ -$ -$ -$ Cumlative Allocated Revenue 678,826,206$ 678,826,206$ 678,826,206$ 678,826,206$ 50% AllocationCost of Revenue 475,178,344$ 475,178,344$ 475,178,344$ 475,178,344$ 30% Gross Margin


Capital Cost

FLASHWAVE 6400 Layer 2.5 Aggregation Switch 9,715,000$ -$ -$ -$ Incumbent switch -$ 15,666,000$ -$ -$ Incumbent NextGen switch -$ -$ 10,434,000$ -$ Multiservice Edge Router -$ -$ -$ 29,887,500$ MPLS Core Router 2,500,000$ 3,800,000$ 3,800,000$ 2,500,000$ Core ATM Switch -$ -$ -$ -$

Metro Ethernet Edge Switch -$ 5,760,000$ 5,760,000$ -$ Installation 1,221,500$ 2,522,600$ 1,999,400$ 3,238,750$ Spares 488,600$ 1,009,040$ 799,760$ 1,295,500$



Capacity Planning 242,308$ 626,538$ 505,385$ 263,077$ Network Administration 726,923$ 1,879,615$ 1,516,154$ 3,946,154$ Service Provisioning & Configuration 1,557,692$ 3,972,115$ 3,193,269$ 3,271,154$

NOC Support 7,476,923$ 19,066,154$ 15,327,692$ 19,626,923$ Field Support 5,296,154$ 13,551,923$ 10,903,846$ 11,215,385$ Environmental 3,641,844$ 9,887,654$ 6,591,770$ 7,932,362$ Financial Penalties for Flash Cut Outages -$ 547,145$ 547,145$ 547,145$ EMS Service Migration Expenses 185,000$ -$ 570,000$ 570,000$ Training 918,000$ -$ 188,678$ 1,149,681$ OSS Service Migration Expenses 5,441,025$ -$ 1,360,256$ 1,360,256$ Hardware Support 696,255$ 2,525,900$ 1,139,658$ 1,846,088$ OSS Support 153,000$ 390,963$ 314,463$ 383,227$




17


In this scenario, there is significantly less equipment deployed. This is because the revenue model for FrameRelay, ATM, and Ethernet has remained the same, but IP VPN revenue and access line requirements havebeen subtracted. Because IP VPN revenue was projected to grow at a high rate, a significant amount of thecapital equipment deployed for IP VPN is not needed.

One of the interesting results is that if IP VPN services are removed, then the ROI and the payback periodsfor all solutions are significantly better then in the previous scenario. This is a reflection of the assumedvolatility of IP VPN pricing over the five-year period. Therefore, the CAPEX investment in support of IP VPNstends to be high while the corresponding margins for these new services (which drives ROI) tend to getlower over the period.

In this scenario, Fujitsu’s solution is significantly less expensive then the alternative solutions, mainlybecause incumbent solutions require Ethernet switches for Ethernet services that lead to extra CAPEX andOPEX. As discussed earlier, the MSER expenses are high because of higher port costs and OPEX for routers.The CAPEX and OPEX of the alternatives are compared in Figure 12 and five-year CAPEX for the alternativesis depicted in Figure 13.


0

20

40

60

80

100

FLASHWAVE6400

IncumbentNextGen

Incumbent Router

$ M

illio

ns

Cumulative Expenses


18


New Aggregation: Frame Relay and ATMIn this deployment scenario, a new aggregation strategy is employed for Frame Relay and ATM services. It isassumed that Ethernet and IP VPNs are not offered. The Frame and ATM services are offered in accordancewith the service profiles and revenue models discussed earlier. Given the set of assumptions specified inthe architecture and service profile sections above, we used our economic model to create a five-yearmodel. The results depicted in Figure 14 are cumulative results over the five-year period.


0

2

4

6

8

10

12

1 2 3 4 5

Year

$ M

illio

ns

FLASHWAVE 6400

Incumbent NextGen

Incumbent

Router

19


This scenario is fairly similar to the previous example modeling Frame Relay, ATM, and Ethernet. BecauseEthernet services are not offered in this example, revenues, CAPEX, and OPEX are slightly lower. In this case,The FLASHWAVE 6400 Layer 2.5 aggregation switch performs better than the alternatives as depicted inFigure 15 and Figure 16. The primary reasons for this cost advantage are:

• Better port densities and lower costs per port• Bandwidth pooling• Lower environmental costs due to better port densities• Edge switch operational advantages over MSERs

Because Frame Relay’s assumed revenue growth rate is 2% and ATM revenue growth rate is –20%, theCAPEX investment is fairly low compared to the other scenarios where VPNs and Ethernet were drivingservice growth. It should also be noted that OPEX is a larger component of the expense then CAPEX.This is because service growth and subsequent CAPEX is small over the period.

Figure 14: Cumulative Revenue and Expenses for New Aggregation: Frame Relay and ATM


Allocated Revenue Frame Relay 581,737,911$ 581,737,911$ 581,737,911$ 581,737,911$ ATM 37,607,167$ 37,607,167$ 37,607,167$ 37,607,167$ Metro Ethernet -$ -$ -$ -$ IP VPN -$ -$ -$ -$ Cumlative Allocated Revenue 619,345,078$ 619,345,078$ 619,345,078$ 619,345,078$ 50% AllocationCost of Revenue 433,541,555$ 433,541,555$ 433,541,555$ 433,541,555$ 30% Gross Margin


Capital Cost

FLASHWAVE 6400 Layer 2.5 Aggregation Switch 8,805,000$ -$ -$ -$ Incumbent switch -$ 15,666,000$ -$ -$ Incumbent NextGen switch -$ -$ 10,854,000$ -$ Multiservice Edge Router -$ -$ -$ 23,287,500$ MPLS Core Router 1,800,000$ 3,050,000$ 3,050,000$ 1,800,000$ Core ATM Switch -$ -$ -$ -$

Metro Ethernet Edge Switch -$ -$ -$ -$ Installation 1,060,500$ 1,871,600$ 1,390,400$ 2,508,750$ Spares 424,200$ 748,640$ 556,160$ 1,003,500$



Capacity Planning 242,308$ 363,462$ 242,308$ 263,077$ Network Administration 726,923$ 1,090,385$ 726,923$ 3,946,154$ Service Provisioning & Configuration 1,557,692$ 2,336,538$ 1,557,692$ 3,271,154$

NOC Support 7,476,923$ 11,215,385$ 7,476,923$ 19,626,923$ Field Support 5,296,154$ 7,944,231$ 5,296,154$ 11,215,385$ Environmental 3,641,844$ 9,887,654$ 6,591,770$ 7,932,362$ Financial Penalties for Flash Cut Outages -$ 547,145$ 547,145$ 547,145$ EMS Service Migration Expenses 185,000$ -$ 570,000$ 570,000$ Training 918,000$ -$ 91,800$ 1,149,681$ OSS Service Migration Expenses 5,098,036$ -$ 1,274,509$ 1,274,509$ Hardware Support 604,485$ 2,200,400$ 792,528$ 1,429,988$ OSS Support 153,000$ 229,500$ 153,000$ 383,227$




20



0

10

20

30

40

50

60

70

80

90

FLASHWAVE6400

IncumbentNext-Gen

Incumbent Router

$ M

illio

ns

Cumulative Expenses



0

1

2

3

4

5

6

7

8

9

1 2 3 4 5

Year

$ M

illio

ns

FLASHWAVE 6400

Incumbent Next-Gen

Incumbent

Router

21


The Business Case for MigrationIn the previous sections of this paper, our analysis showed that in a New Aggregation scenario the edgeswitch architecture is more cost effective than the Layer 3 MSER architecture. This section of the paperpresents the business case for Replace and Migrate. In this scenario, it is assumed that all the incumbent’slegacy switches are replaced with FLASHWAVE 6400 Layer 2.5 aggregation switches and that all fourservices (Frame Relay, ATM, Ethernet, and IP VPN) are supported. A five-year ROI analysis for the Layer 2.5migration is presented in Figure 17.

Figure 17: Cumulative Revenue and Expenses for Replace and Migrate

Fujitsu Migration ROI Year 1 Year 2 Year 3 Year 4 Year 5

Allocated Revenue

Frame Relay 110,895,900$ 113,557,109$ 116,282,180$ 119,072,646$ 121,930,075$ ATM 11,112,222$ 8,928,049$ 7,173,188$ 5,763,255$ 4,630,453$ Metro Ethernet 8,147,284$ 9,697,611$ 11,542,945$ 13,739,424$ 16,353,865$ IP VPN 10,935,408$ 21,870,815$ 43,741,630$ 87,483,261$ 174,966,521$

Annual Allocated Revenue 141,090,814$ 154,053,584$ 178,739,944$ 226,058,586$ 317,880,915$Cost of Revenue 98,763,569$ 107,837,509$ 125,117,961$ 158,241,010$ 222,516,640$

Annual Allocated Gross Profit 42,327,244$ 46,216,075$ 53,621,983$ 67,817,576$ 95,364,274$Cumulative Allocated Gross Profit 42,327,244$ 88,543,319$ 142,165,302$ 209,982,878$ 305,347,152$

Capital Cost

FLASHWAVE 6400 Layer 2.5 Aggregation Switch 28,903,500$ 3,342,049$ 4,177,561$ 5,221,951$ 6,527,439$ Incumbent switch -$ -$ -$ -$ -$ Incumbent NextGen switch -$ -$ -$ -$ -$ Multiservice Edge Router 3,630,000$ 419,729$ 524,661$ 655,827$ 819,783$ MPLS Core Router 5,580,000$ 645,203$ 806,504$ 1,008,130$ 1,260,163$ Core ATM Switch -$ -$ -$ -$ -$ Metro Ethernet Edge Switch -$ -$ -$ -$ -$ Installation 3,811,350$ 440,698$ 550,873$ 688,591$ 860,738$ Spares 1,524,540$ 176,279$ 220,349$ 275,436$ 344,295$

Annual Capital Cost 43,449,390$ 5,023,958$ 6,279,948$ 7,849,935$ 9,812,419$Cumulative Capital Cost 43,449,390$ 48,473,348$ 54,753,296$ 62,603,231$ 72,415,650$


Capacity Planning 49,629$ 62,036$ 77,545$ 96,932$ 121,165$ Network Administration 212,997$ 266,247$ 332,809$ 416,011$ 520,013$ Service Provisioning & Configuration 351,129$ 438,912$ 548,640$ 685,800$ 857,250$ NOC Support 1,765,137$ 2,206,422$ 2,758,027$ 3,447,534$ 4,309,418$ Field Support 1,195,738$ 1,494,673$ 1,868,341$ 2,335,426$ 2,919,283$ Environmental 826,709$ 1,033,386$ 1,291,733$ 1,614,666$ 2,018,333$ Financial Penalties for Flash Cut Outages -$ -$ -$ -$ -$ EMS Service Migration Expenses 185,000$ -$ -$ -$ -$ Training 1,760,227$ -$ -$ -$ -$ OSS Service Migration Expenses 5,643,633$ -$ -$ -$ -$ Hardware Support 441,181$ 551,476$ 689,345$ 861,681$ 1,077,101$ OSS Support 35,746$ 44,683$ 55,854$ 69,817$ 87,271$

Annual Expenses 12,467,127$ 6,097,835$ 7,622,293$ 9,527,867$ 11,909,833$Cumulative Expenses 12,467,127$ 18,564,962$ 26,187,255$ 35,715,122$ 47,624,955$

Annual Cash Flow (13,589,273)$ 35,094,282$ 39,719,742$ 50,439,774$ 73,642,022$Cumulative Cash Flow (13,589,273)$ 21,505,009$ 61,224,751$ 111,664,525$ 185,306,547$Discounted Cash Flows (13,589,273)$ 33,423,126$ 36,026,977$ 43,571,773$ 60,585,474$Cumulative Discounted Cash Flows (13,589,273)$ 19,833,853$ 55,860,829$ 99,432,603$ 160,018,077$

TCO 55,916,517$ 11,121,793$ 13,902,241$ 17,377,802$ 21,722,252$ROI ([Revenue-Expenses]/Capital) 69% 144% 212% 278% 356%IRR 274%Payback Period in months 16.62

This analysis demonstrates the key financial metrics associated with the FLASHWAVE 6400 Layer 2.5aggregation switch capital investment over a five-year period. The revenue assumptions associated withROI and payback are the same as those discussed earlier, namely, 50% of total revenue is allocated to edgeequipment. Out of the allocated revenue, 30% gross margins are used to calculate the gross profits used tocompute ROI, Cash flows, IRR, and payback period. Given these fairly conservative assumptions, our modelshows that a service provider can replace incumbent switches and migrate to an IP/MPLS network with apayback period of 16.6 months, an ROI of 356%, and an IRR of 274%.

22


Figure 18 gives a high level view of the business case for migration by presenting ROI and discounted cashflows on the same graph.

Figure 18: Discounted Cash Flows

(40)(20)

0

20

40

60

80

100

0

50

100

150

200

250

300

350

400

120

140

160

180

1 2 3 4 5

Year

$ M

illio

ns

Per

cen

t

Discounted Cash Flows

Cumulative Discounted Cash Flows

ROI ([Revenue-Expenses]/Capital)

Now that the business case for Replace and Migrate is established, the alternative solutions are compared.Figure 19 compares each of the alternatives over the five-year interval. In the Replace and Migrate scenario,it is assumed that the FLASHWAVE 6400 Layer 2.5 aggregation switch, incumbent NextGen, and MSERsolutions replace all the incumbent switches, connect to both the ATM and IP/MPLS networks, and migratetraffic to the IP/MPLS core network. The incumbent alternative only adds new switches for new services,however, it also migrates PVCs to the IP/MPLS core using ATM trunks to connect to MPLS routers. The keypoint in this analysis is that over the five-year period the capital expense of replacing all incumbent switcheswith FLASHWAVE 6400 Layer 2.5 aggregation switches is less than the capital expense of keepingincumbent switches in place and adding new incumbent hardware to support new services. This is not trueof any of the alternative solutions. The other salient point is that the payback period of the FLASHWAVE 6400Layer 2.5 aggregation switch solution is 16.6 months while the router alternative payback period is 52.4months.

23


Figure 19: Cumulative Revenue and Expenses for Each Scenario


Allocated Revenue Frame Relay 581,737,911$ 581,737,911$ 581,737,911$ 581,737,911$ ATM 37,607,167$ 37,607,167$ 37,607,167$ 37,607,167$ Metro Ethernet 59,481,128$ 59,481,128$ 59,481,128$ 59,481,128$ IP VPN 338,997,635$ 338,997,635$ 338,997,635$ 338,997,635$ Cumlative Allocated Revenue 1,017,823,841$ 1,017,823,841$ 1,017,823,841$ 1,017,823,841$ 50% AllocationCost of Revenue 712,476,689$ 712,476,689$ 712,476,689$ 712,476,689$ 30% Gross Margin


Capital Cost

FLASHWAVE 6400 Layer 2.5 Aggregation Switch 48,172,500$ -$ -$ -$ Incumbent switch -$ 62,848,000$ -$ -$ Incumbent NextGen switch -$ -$ 74,746,000$ -$ Multiservice Edge Router 6,050,000$ 6,050,000$ 6,050,000$ 128,682,500$ MPLS Core Router 9,300,000$ 16,950,000$ 16,950,000$ 9,300,000$ Core ATM Switch -$ -$ -$ -$

Metro Ethernet Edge Switch -$ 9,000,000$ 9,000,000$ -$ Installation 6,352,250$ 9,484,800$ 10,674,600$ 13,798,250$ Spares 2,540,900$ 3,793,920$ 4,269,840$ 5,519,300$



Capacity Planning 407,308$ 1,033,846$ 953,077$ 657,692$ Network Administration 1,748,077$ 3,627,692$ 3,385,385$ 9,865,385$ Service Provisioning & Configuration 2,881,731$ 6,853,846$ 6,334,615$ 8,177,885$

NOC Support 14,486,538$ 33,552,692$ 31,060,385$ 49,067,308$ Field Support 9,813,462$ 23,365,385$ 21,600,000$ 28,038,462$ Environmental 6,784,827$ 21,097,369$ 18,900,113$ 19,830,906$ Financial Penalties for Flash Cut Outages -$ 1,321,607$ 1,321,607$ 1,321,607$ EMS Service Migration Expenses 185,000$ -$ 570,000$ 570,000$ Training 1,760,227$ -$ 380,001$ 2,874,202$ OSS Service Migration Expenses 5,643,633$ -$ 1,410,908$ 1,410,908$ Hardware Support 3,620,783$ 5,983,800$ 6,084,522$ 7,865,003$ OSS Support 293,371$ 684,335$ 633,335$ 958,067$




24


Figure 20 compares the five-year cumulative TCO for each of the four alternatives in the Replace andMigrate scenario. The FLASHWAVE 6400 solution has the lowest CAPEX and OPEX of all the alternativespresented.


0

50

100

150

200

250

300

350

400

FLASHWAVE6400

IncumbentNext-Gen

Incumbent Router

$ M

illio

ns

Cumulative Expenses


25


ConclusionMSER vendors will exert considerable influence over the process of selecting the best approach for Layer 2migration, given their incumbency within the core IP/MPLS network. However, NSP believes that serviceprovider network planners will execute their initial deployment assessments based on mitigating the risk ofimpacting customer SLAs, preserving the OAM&P of their revenue-generating services, and optimizing theproposed deployment based upon the TCO.

Based on the analysis contained in the body of this white paper, NSP can definitively reach the followingthree conclusions:

1) Using Layer 3 MSERs for edge aggregation is consistently the most capital-intensive and incurs thehighest OPEX for a service provider. These conclusions occurred in all scenarios studied.

2) The Fujitsu solution using the FLASHWAVE 6400 Layer 2.5 aggregation switch is the most cost-effectiveedge aggregation solution for both CAPEX and OPEX by a significant margin when consideringalternatives for all scenarios analyzed in this paper.

3) The Fujitsu solution using the FLASHWAVE 6400 Layer 2.5 aggregation switch is consistently at leasthalf the TCO of the MSER alternative in all scenarios analyzed in this paper.

Based on the results from this paper’s analysis, NSP would make the following recommendations to serviceproviders:

1) Using Layer 2.5 switches for aggregation and MPLS backhaul of IP VPN traffic to Gigabit Ethernet-enabled Layer 3 edge routers is the most cost-effective approach to transitioning the network.

2) The Layer 2.5 approach provides greater flexibility in access technology choices, significant QoSassurance, OAM&P options, and reduces the complexity of provisioning new services. Given thecompetitive nature of today’s marketplace, it is imperative for service providers to use a moredifferentiated and cost-effective aggregation and backhaul architecture to reach profitability sooner.

Given the results presented in this paper, NSP can conclusively say that the Fujitsu solution is the besteconomic choice for service providers regardless of their service strategies and whether they adopt NewAggregation or Replace and Migrate strategies.

26


© Copyright 2005 Fujitsu Network Communications Inc. All Rights Reserved.FLASHWAVE® is a trademark of Fujitsu Network Communications Inc. (USA).FUJITSU (and design)® and THE POSSIBILITIES ARE INFINITE™ are trademarks of Fujitsu Limited.All other trademarks are the property of their respective owners.

Acronym Descriptor

ATM Asynchronous Transfer Mode

BOM Bill Of Materials

CAPEX Capital Expense

CORBA Common Object Request Broker Architecture

EMS Element Management System

FR Frame Relay

IP Internet Protocol

IRR Internal Rate of Return

LSP Label Switched Path

MPLS Multiprotocol Label Switching

MSER Multiservice Edge Router

NMA Network Monitoring and Analysis

NOC Network Operations Center

NSP Network Strategy Partners

OAM&P Operations, Administration, Maintenance & Provisioning

OPEX Operating Expense

OSS Operational Support System

POP Point of Presence

POS Packet Over SONET

PVC Permanent Virtual Circuit

PWE3 Pseudowire Edge-to-Edge Emulation

QoS Quality of Service

RFC Request For Comment

ROI Return on Investment

SLA Service Level Agreement

SNMP Simple Network Management Protocol

SONET Synchronous Optical Network

TCO Total Cost of Ownership

TIRKS Trunk Integrated Record Keeping System

VLL Virtual Leased Line

VoIP Voice over Internet Protocol

VPN Virtual Private Network

XML Extensible Markup Language

a total cost of ownership (tco) analysis and comparison of ... · pdf filea total cost of...

Documents