July 2000

BY CHUCK HARRIS

There is a sea of change in the telecommunications industry as service providers embrace a product category that collapses a host of central office switching, routing, and transport equipment into a single platform. While the terminology for this product category is still being shaken out and the number of vendors addressing this market is quickly expanding, these platforms are now being commonly referred to in the press and industry as broadband services switches or, more descriptively, as the Collapsed Central Office (CCO).

The CCO is a next-generation switch that replaces five or six voice and data devices, including a class 5 switch, in a single platform, allowing a service provider to roll out and easily manage a converged network of voice/data services through one integrated system.

While the cost savings of a CCO are enormous (a fifth to a tenth of the cost of a traditional class 5-based central office), the underlying benefit of adopting a converged central office network architecture is in streamlining the network management. With a CCO, service providers are able to roll out more services to more cities in less time -- for significantly less money than with a traditional five- or six-box infrastructure.

Instead of having to hire and support an entire team of experts to manage separate frame relay, IP routing, gateway, SS7/voice switching (class 5), ATM, and SONET devices, a service provider can instead now have a much smaller team managing a single device that does all the switching, routing, transport, and signaling.

In a nutshell, the benefits of managing a converged network with the CCO mark a watershed moment in the future of the central office landscape.

CURRENT TELECOM BUSINESS MODEL CONSTRAINTS
Most service providers have aggressive business plans to roll out services to additional cities, but are hindered by limitations that all service providers can relate to, namely:

Management. In managing a converged network of five or six devices, the service provider needs to interface and provision/monitor a variety of technologies and devices in an integrated manner, presenting a very complicated environment to the network operations administrator. Additionally, with a multi-device network, when a software load in a single product is upgraded, the service provider needs to go back and regression test the entire network, a time-consuming and complicated process.

Time. Service providers have to roll out services (a beachhead strategy) as quickly as possible. If they are the third or fourth service provider to turn up services in a particular city, they have little chance of success. With a traditional architecture, provisioning and managing a diverse network of five or six voice, data, and transport devices severely hinders the speed with which a service provider can roll out services and taxes the management team in terms of complexity and manpower.

Available expertise. To roll out bundled voice/data services, service providers need a class 5 expert, an ATM expert, and an IP expert (to name a few), to install and provision the boxes in each central office they roll out to. These people are hard to come by in this competitive market, so labor constraints mean that they can roll out services to only a few cities at a time.

Complexity and environmentals. Every central office in each city needs extensive wiring and air conditioning, presenting untold headaches in the complexity of integrating the different boxes and dealing with the environmentals involved.

Money. A legacy central office architecture based on a class 5 switch surrounded by a number of switches, routers, and gateways typically costs $2�4 million dollars per city, a hefty price tag for an emerging service provider with an aggressive rollout strategy.

BENEFITS OF A CONVERGED NETWORK
Until now, the state of the technology dictated that service providers use an "old world" central office architecture comprised of a class 5 switch surrounded by a number of data switches/routers. This presented a huge burden of integrating, provisioning, and managing these devices and the services provided off the platforms.

However a new category of product has emerged on the scene -- the CCO -- that enables a truly converged network. The CCO integrates voice, TDM (class 5 switch), ATM, frame relay, and IP networks into a single box -- with all the associated signaling and interworking built in for advanced service features. Some solutions even feature an integrated SONET add/drop multiplexer (MUX) so the box can be put into a SONET ring without the typical cost of an additional SONET MUX. With this next-generation switch, service providers can now manage an entire central office from a single interface, enabling them to:

Roll out to more cities. The cost is almost a tenth of the cost of the old architecture as all services are collapsed into a single device.

Offer more services. Not only does the next-generation voice/data switch provide class 5 and ATM functionality off the same device, but it can also provision and provide customers with advanced IP services (e.g., MPLS, DiffServ, etc.), frame relay, and can even light up a SONET ring for data transport straight out of the box.

Spend less time. One box means less complexity and does not require a whole team to integrate multiple boxes in each city. It is as simple as connecting the switch and provisioning services via a Java-based drag-and-drop network management center.

Additionally, the size of the team needed to install the CCO is significantly smaller, so multiple cities can be provisioned at once.

Have fewer headaches. The complexity and environmentals are dramatically reduced when dealing with a single device rather than an entire room of different products.

The CCO leverages Moore's Law in reducing what used to be multiple boxes into integrated chipsets in a single platform the size of a small, two-drawer filing cabinet. As a result, the:

• ATM switching matrix is now available in a chipset;
• Entire TDM switching matrix and SS7 processing are provided in a multi-chip complex;
• IP edge router is now integrated into a single blade; and
• SONET add/drop MUX is integrated into the switching card.

With this solution a service provider can provision services to customers simply by using a Java-enabled drag-and-drop format to provision individual channels for TDM, frame relay, or IP service, and individual T1s for ATM service.

CONCLUSION
The CCO is a natural evolution of the central office and significantly simplifies the management of what used to be a five- or six-device architecture spanning voice, data, switching, signaling, and transport. For a competitive service provider that needs to manage networks in various remote cities, the huge headache and costs of provisioning, managing, integrating, and regression testing a five- or six-device network with SS7, TDM, ATM, IP, gateway, and SONET add/drop MUX is a nightmare and takes an entire team of voice and data experts months to set up. Compare this to installing a compact single switch and immediately turning up services in a new city.

Now that multiple devices can be reduced to individual chipsets and integrated into a single switch, service providers will be able to provision and manage more services to more cities in less time with more money -- significantly changing the service provider's business model.

Chuck Harris is vice president, marketing and business development, for Tachion Networks. He is responsible for all marketing activities including product management, market development, strategic alliances, and marketing communications. Tachion is the innovator of the FUSION 5000, a Collapsed Central Office product that integrates an entire central office into a single switch at a fifth to a tenth of the cost of a legacy central office architecture. For more information, visit Tachion's Web site at www.tachion.com.

[ return to the July 2000 table of contents ]

BY BRIAN MAHONY

Network operators responsible for implementing next-generation services have discovered that they are no longer just tactical operations managers responsible for overseeing an inventory of network elements and monitoring faults and congestion -- they have become strategic leaders responsible for implementing the service provider's vision and building sustainable competitive advantage. Here are four trends this new breed of network operations leader will need to stay ahead of, and the weapons of choice for managing next-generation services.

1) Infrastructure Deployment Migration
Tools: Service mediation platforms, media gateways, broadband access aggregators, billing mediation, mass provisioning, and service level management

Next-generation services enabled by service mediation platforms have a difficult business case today -- voice over packet is not widely deployed or accepted, and the network infrastructure required to keep the levels of availability and reliability of the PSTN requires a major redesign, if not overhaul, of packet backbones. However, service mediation switches can be deployed with Inter-Machine Trunks (IMTs) and ISDN primary rate interface (PRI) links to allow the offloading of Internet traffic before it hits the tandem switch, creating a six-month or less return on investment (ROI) business case. Because Internet call times (modem calls) far exceed traditional voice calls, this solution enables the PSTN switch congestion to be reduced at a price that is an order of magnitude more cost effective than deploying another tandem switch. These solutions have an immediate impact on the service provider's operations costs, but also serve as the foundation for new services in the future. Coupled with on-board or peripheral billing mediation functionality, the media gateway can turn proprietary call detail records (CDRs) into industry standard (BAF/AMA) records for use by legacy billing systems. As the service provider's back-office systems become ready and the market adopts next-generation services such as unified messaging, softswitches can be added to an existing footprint in the central office to manage universal service call control using the already-paid-for converged transport infrastructure.

In the data world, private line aggregation solutions are being deployed which also have an immediate business case based on cost reduction. Today's next-generation edge routers are far more dense, and can manage many more channelized connections at wire speed than routers based on five-plus-year-old architectures. With the added intelligence that is being built into new interface cards and software loads, these same devices will become the engine for a second business case based on generating revenues from services such as virtual private networks (VPNs), VoIP, and application service provider (ASP) applications such as video editing.

2) Rapid Service Creation Environment
Tools: Softswitches, service creation environments (SCEs), Java Advanced Integrated Network (JAIN)

In the dark ages, services were defined through a painstakingly long and costly process. Today, the model for fast and open service creation is through softswitches and JAIN-based SCEs. While the media gateways handle the physical transport of time-division multiplexing (TDM) to IP and asynchronous transfer mode (ATM) traffic back and forth, the softswitch manages universal call (or service) control, and the SCE manages the overall workflow of the network elements and Operational Support Systems (OSS) applications. It does this by defining each of these OSS applications and network elements as Java beans or "objects." Through the JAIN initiative, relationships between these objects are defined according to the supporting protocols. For network elements, this includes SS7 ISUP, SIP, H.323, and MGCP. For OSS applications this includes J-DBC, LDAP, DNS, SNMP, and CORBA.

The responsibility for service creation could be put in the hands of an operations-smart marketing/engineering team, who will drag and drop Java beans to explore different service scenarios. With this framework in place, vendors can create new services within weeks. What's more, they can open the SCE up to development by third-parties to innovate, or to service providers themselves to differentiate.

3) Service Differentiation Based on Value-Based Billing and QoS
Tools: Next-generation billing systems with complex rating, QoS, common open service policy (COPS), DiffServ, and multi-protocol label switching (MPLS)

Eventually, the service provider's network engineering/operations team will need to be able to manage the PSTN and IP worlds as one end-to-end path. This means they will need to deploy mass provisioning tools that can set up service and customer-specific permanent virtual circuits (PVCs) and ensure that if the customer is paying for a "gold" QoS level, it means the same thing (or better) to all the network elements across the path. A policy server running COPS will be needed to manage policy enforcement across these devices. Protocols such as DiffServ and MPLS will make sure connection-less IP packets mimic the QoS found on the PSTN.

This in place, the service provider can dream up new ways to differentiate based on what the customer is willing to pay for. It is not enough to offer customers flat-rate tariff models for these new services. For example, a service provider may offer one hundred different flavors of a combined broadband/ASP service to target certain customer segments. In many cases, the differences will be based on service quality parameters such as bandwidth, jitter, delay, and packet loss. In other cases, the differences will be based on how the customer wants to pay. The customer may want to pay by bandwidth, by duration, time of day, or qualitative metrics such as the customer's profession or demographic group.

4) Web Portal-Based Service Selection and Customer Care
Tools: Web portals, meta directories, data mining/decision support systems

Today's network operations leader will need to understand how the above service creation, QoS, and billing capabilities can be managed in an environment where service interactions between the provider and customer are being set up and torn down dynamically, with the customer in control. A Web-based portal will be required for on-demand service selection. Customers have a one-stop shop for all their voice and data services, and have the ability to change their profile information as often as they want. If a customer chooses a service that cannot be supported by the service provider's network, such as video streaming, the transfer to a wholesale network service provider is invisible to the customer. Service providers will increase customer retention by offering customers a choice of services to sign up for under a tariff model that best suits them.

Service providers will catalog all services being offered through the portal using a meta directory. The meta directory will manage data across disparate OSS applications to provide one integrated customer profile, service profile, and network profile. This data will be accessible to a variety of decision support tools. Using these tools, the service provider will be able to loop back to the beginning of the whole service creation process -- defining services that uniquely address the needs and usage patterns of specific customer segments.

Brian Mahony is director of strategy and business development at Unisphere Solutions. His focus is on defining and building service-focused solutions that allow rapid deployment of converged voice and data services. Unisphere, a Siemens company, designs, markets, and supports converged networking solutions that allow telecommunications providers to efficiently create and deploy innovative services. The company redefines "start-up" by offering service providers innovative and proven world-class technology developed in entrepreneurial environments. For more information, visit the company's Web site at www.unispheresolutions.com.

[ return to the July 2000 table of contents ]