Skip to main content
Mar 14 2014

3 Strategies for an IP Core Network Refresh

Cut costs with an IP core network refresh

Next-generation core routing platforms give service providers new opportunities to rethink and refresh IP core networks . These platforms take advantage of recent breakthroughs in 100 Gigabit Ethernet (GbE) IP transport and 400 Gb/s network processors[1]. As a result, they allow service providers to move to higher capacities and more cost-effective network architectures. A recent Bell Labs study with a Tier 1 European service provider highlighted the key role that a next-gen core routing platform plays in an IP core network fresh. The service provider was operating a multi-tier, dual-plane IP core network with more than 100 node sites and faced 3 main challenges:

  • IP core network links and node capacities were underutilized.
  • The IP core network was expensive to operate.
  • Upcoming deregulation of IP peering points would massively increase traffic volumes and change traffic patterns through the IP core.

To deal with these challenges and work toward cost-cutting goals, the service provider asked Bell Labs to help them identify strategies for capacity and cost improvements. After detailed analysis, the joint team recommended 3 strategies:

  • Increase link utilization targets to recover stranded capacity.
  • Reduce the number of IP backbone nodes to optimize and simplify the IP core architecture.
  • Adopt a next-generation core routing platform for more scalable and cost-effective transport .

Strategy 1: Increase link utilization Increasing link utilization targets will help the service provider run hotter networks and save transport costs with only marginal impact on quality of service (QoS) for best-effort (BE) traffic (Figure 1). Premium traffic is protected and unaffected. The study found that the service provider could increase overall link utilization targets:

  • From 50% to 67% to save 23% in transport costs with 12% of best effort traffic affected.
  • From 50% to 80% to save 30% in transport costs with 18% of best effort traffic affected.

Strategy 2: Reduce nodes in the IP core Reducing the number of IP backbone nodes in the IP core will help the service provider save on capital expenditures (CAPEX). It will also help to save on space and power costs while meeting service availability targets (Figure 2). The study found that reducing the number of IP backbone nodes from more than 100 to 40 would allow the service provider to:

  • Save 23% in router CAPEX over 5 years
  • Save 48% in router space and power costs over 5 years

Strategy 3: Adopt a next-gen core routing platform Adopting the Alcatel-Lucent 7950 Extensible Routing System (XRS), which is a next-generation core routing platform, will give the service provider more scalable and efficient transport. The study found that the faster the service provider moves to the new platform, the lower the total cost of ownership (TCO) over the long-term (Figure 3). The new routing platform requires only a single chassis at most IP core node sites across the five years study period, as opposed to having to keep adding new chassis in the case of legacy router , resulting in far less deployment churn and maintenance cost down the road.

Moving immediately to next generation IP core nodes with higher capacity, port density and performance, along with a smaller footprint, would allow the service provider to:

  • Save 37% in TCO over 5 years (FMO2 in Figure 3) compared to the 30% saved by moving gradually to the next-gen core routing platform (FMO1 in Figure 3).
  • Save 60% in space and power costs (both FMO1 and FMO2 in Figure 4).

While the results of the study are specific to this service provider’s network, the rationale behind the study and the methodology used can be applied to IP core networks in general.

It’s time for an IP core network refresh

Many service providers are facing similar evolution requirements as the use case described in this study, and could benefit from a rethink and refresh of their IP core network: Service providers, like the one described in this article, are rethinking their IP core network architecture and evolution requirements because:

  • Traffic is increasing. On average, service providers will face a 3 to 8-fold traffic increase in IP core networks over the next 5 years[1]. Bell Labs predicts that Internet video consumption alone will grow more than 6 times by 2017.
  • Core network architectures are not designed for emerging services requirements. Older 10G technology and lower capacity platforms are not optimized for the traffic patterns of cloud and highly distributed applications.
  • There is too much stranded capacity. Past network upgrades, economic constraints and ad-hoc network engineering mean traffic routing is not optimal.
  • Network equipment is approaching end-of-life. Legacy equipment often suffers from inefficient capacity expansion, power consumption and sparing. A lack of data or control plane scaling means further investments are not an effective way to control networks costs.
  • Forced events force a rethink. Network transformation and consolidation due to acquisitions, mergers, government policy or regulatory changes affect traffic patterns. Networks need a redesign to streamline operations and costs.
  • Competition is fierce. Service providers want to make the best possible use of network capacity and to continuously reduce per-bit transport costs.

Evaluate your IP core evolution

There are a few practical steps to evaluate the options for an IP core network refresh. Through engagements with network planning and operations teams of several service providers, Bell labs developed the following approach:

  1. Determine baseline traffic and growth trends. Anticipate future traffic patterns and volumes for broadband, enterprise and mobile services. Collect enough measurement data to forecast traffic demands and to characterize current traffic demands for traffic modeling purposes.
  2. Incorporate new design requirements. Think about future traffic patterns and volumes. In some cases, that may mean anticipating government or regulatory changes that will affect traffic routing. Traffic analysis must consider these types of challenges and it must work with incomplete traffic data sets to create a long-term traffic forecast.
  3. Explore all viable network architecture and traffic reengineering options:
    • Optimizing the existing IP core network design to reflect changing traffic patterns.
    • Adjusting traffic engineering objectives to recover stranded network capacity from the existing network design
    • Consolidating high-volume demands into more cost-effective 40 GbE/100 GbE technology
    • Consolidating central office resources where there are cost advantages
    • Moving to smaller nodal topologies; larger topologies are less efficient and more expensive
  4. Evaluate next-generation routing technology and compare network transformation options. Quantify and contrast the scalability and costs of adopting a next-gen core routing platform versus upgrading and evolving the existing network.

Get the right tools and expertise

Bell Labs can assist service providers with the tools and expertise to achieve a more nimble IP core to reduce costs and maintain network performance. Figuring out the best way to get there requires:

  • Advanced traffic modeling and forecasting tools
  • Traffic engineering and network optimization algorithms
  • Expert techno-economic analysis capabilities

With the insight these tools and expertise offer, service providers can identify the right steps required to:

  • Rethink IP core networks and reduce costs.
  • Grow network capacity in a cost-effective way without increasing their resource footprint.

More Information

Read the Financial White Paper Rethinking Core Economic

To contact the authors or request additional information, please send an e-mail to networks.nokia_news@nokia.com.

Footnotes

  1. [1] Based on the results of IP core network studies Bell Labs has conducted for various service providers.
About Ben Tang
Ben’s work focuses on data networking (IP/MPLS, Ethernet, IPv6, mobile, IP/optical) architecture, capacity design and optimization, network modeling and tool development, evolution planning and economic analysis. Ben has led the assessment of next-gen data network architectures and technology selection strategies for numerous global network operators in a wide range of application domains including IP backbone, metro BNG/CDN, mobile backhaul, enhanced packet cores, and cloud data center interconnects.