FINANCE GROUPS AT MANY LARGE CARRIERS are increasingly driving technology-spending decisions. As carriers and vendors attempt to disassociate themselves from the mistakes and excesses of the “new economy,” the new era, which we'll call the “counter-revolution,” must look to marry the positives of the pre-Internet era—mainly financial stability—with the growth and innovation of the Internet era. The Internet is still growing at 80-100% per year and carriers need new services to drive revenue more than they ever have in the past. Let's examine one area that will be impacted by this counter-revolution: carrier routing.

The combination of applications such as data, telephony, and multimedia, and wireless integration or data delivered over a wireless handset has broad implications for the networks that need to support them. These applications can potentially play a role in delivering differentiated value to customers. This means, of course, that enough broadband capability has to be available and that the underlying infrastructure can support these new applications.

Existing IP platforms are neither reliable nor secure enough to support new service requirements. Current infrastructure also does not allow services to scale efficiently, without disruption, as demand materializes. Carriers are virtually unanimous on the two most important capabilities required for building out the new public network—scalability and reliability. In other words, an IP network that matches the original PSTN.

Explosive growth of IP traffic forced carriers to expand their networks quickly with equipment not specifically designed for their needs. Currently, most networks have at least three tiers of routers—edge routers, to which customers connect; aggregation routers that groom multiple lower-speed connections from edge routers onto a smaller number of higher-capacity connections; and core routers, which take traffic in from aggregation routers and send it to other core routers. Traditional routers were also notoriously unreliable and carriers needed redundant routers at each point of presence (POP) to boost reliability. The largest carriers also connect multiple routers together in their core POPs as they run out of slot capacity. The ports used to interconnect routers can cost $200,000 each and are not generating any revenue. Rather, these ports are being used to connect to another core router within the same POP. All of these routers are managed independently under the increasing constraint of fewer available resources to manage them—resulting in a costly and cumbersome IP network.

Router scalability impacts the economics of the POP by reducing non-revenue generating ports. Ports are used to support customer traffic not connecting to other routers. Reliability can also affect POP economics by reducing operating expenses and generating capex savings by eliminating the need for redundant routers. Replacing redundant routers with a single 99.999% availability router would bring the IP network closer to a voice network model.

When the goal was keeping up with traffic growth or even building out networks in anticipation of demand, carriers were less concerned about how the networks were built. Today, carriers are realizing that the way IP backbones were built in the past may be one of the reasons their IP business is under-performing in terms of profitability. A core router specifically built for carriers could pave the way for a new network architecture that would be more cost effective, more reliable, and performs better.

Carriers are looking for improvements in the next generation of routers, such as better resiliency and faster convergence in support of real-time applications. Delivering carrier-class reliability and scalability are capabilities that every router vendor is talking about, but few actually deliver. Unfortunately, for incumbent router vendors, these are not capabilities that can be easily added to systems that were not initially designed to support them. A question that is worth asking is whether established router vendors will have the ability to deliver these next-gen platforms to the market. It is a significant undertaking requiring complex software and hardware development. Achieving high availability begins with building reliable hardware and software. Opportunities for reliability improvement exist through product development, training, network design and redundancy issues, network utilization, and multi-layer protection strategies.

In order for IP Services to reach their potential, the reliability of routers must be dramatically improved from the current average of 1200 minutes of downtime per year. One of the fundamental design objectives for carrier-class core routers is to deliver 99.999%+ availability (5 minutes of downtime per year). In order to achieve this goal, vendors must take a comprehensive approach to reliability and availability, which minimizes both scheduled and unscheduled downtime in IP networks.
Redundant Hardware: Core router hardware must be designed with no single point of failure. All components have to be fully redundant including route controllers, line cards, power feeds, bay controllers, and cooling units. The switch fabric must be completely passive and provide rich path diversity for intra-router traffic. Other features such as carrier-class enclosures with integrated wire management can also minimize the risk of inadvertent fiber removal.

Robust Software: Core routing operating system software has to be purpose-built for high-volume core routing and leverage a distributed architecture for high reliability. Newly available, Non-Stop Routing NSRTM Technology further improves overall system availability by providing hot-standby route controller protection without the need for protocol extensions (see Appendix 1 - NSR Technology).

Non-Disruptive Operational Model: Core routers must be designed to minimize the disruptions to a network that typically occur with router configuration changes, software upgrades or expansion.

• A scalable architecture design allows in-service addition of line-cards or bays. • Links can be expanded without packet loss using Composite Links technology. • Software upgrades can be performed in-service (hitless) using non-stop routing technology. • All components are hot swappable.


Vendors must step up and guarantee 99.999% availability. For carriers, the immediate impact of achieving 99.999% router availability is higher quality IP services, fewer outages and reduced maintenance costs. In the longer term, carriers can realize significant network cost reduction by redesigning their networks to eliminate redundant routers and links required to compensate for the low availability of conventional routers.