Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

Technical Overview—Broadband Edge as a Service Node Within Seamless MPLS Network Designs

 

The Juniper Networks Broadband Edge solution enables providers to deliver traditional multiplay residential services over a simpler, collapsed network design that consolidates management and service activation points. This approach enables faster service rollouts and differentiated service offerings with greater operational efficiency, along with service velocity, service agility, and faster time to revenue for new subscribers and services.

The architectural elements of the broadband edge IP/MPLS service node design that enable consolidation of services at the edge include network topology, access protocols, IP/MPLS overlay, and interoperability with the metro/aggregation network.

This Network Configuration Example document focuses specifically on the use case in which MX Series broadband edge is deployed as an IP/MPLS service node within a seamless MPLS network architecture. The following sections provide an overview of this broadband edge solution:

The Challenges Faced by Service Providers in Today’s Market

The telecommunications market is increasingly dominated by IP packet services. The developments in this area are occurring so fast, it is difficult for service providers to keep up while staying profitable and retaining a strong client base. Their business and operational models are challenged as they struggle to manage the customer bandwidth demands brought on by increasing data services, and especially video. Broadband developments with wireline and 4G/LTE access offer much higher bandwidths to end users, and are quickly becoming key drivers for new service developments and network designs. Future bandwidth requirements and service mix uncertainties make the choice of network and service architecture even more complicated.

Network and operational complexity do not lend themselves easily to change, yet the ability to change rapidly is critical to service provider success, and it represents a significant competitive advantage. A streamlined network architecture is needed–one that reduces CapEx (by eliminating multiple service-specific elements) and OpEx (by significantly reducing the number of service provisioning points). A streamlined network architecture that converges residential, business, and mobile network infrastructures, and provides the agility to adapt to changing requirements is underway.

Seamless MPLS with pseudowire head-end termination provides a converged packet network and service architecture based on IP/MPLS. This architecture allows service providers to manage rapidly evolving service mix and bandwidth growth. It is based on a simple principle of expanding the dynamic IP/MPLS network to reach across core, metro, and access network infrastructure, enabling any-to-any packet connectivity and flexible service overlay.

This use case is a simplified backhaul architecture in which a Juniper Networks MX Series service edge provides a combination of residential and business provider edge services.

Seamless MPLS

Seamless (no boundaries) MPLS is a network architecture in which all forwarding of packets within a network, from the time a packet enters the network until it leaves the network, is based on MPLS. Seamless MPLS is critical to effective network convergence.

Figure 1 illustrates the traditional network architecture without convergence. Note the separate wireless and wireline MPLS cores and the multiple Ethernet and MPLS aggregation points.

Figure 1: Traditional Network Architecture Without Convergence
Traditional
Network Architecture Without Convergence

Figure 2 shows network convergence with seamless MPLS. Service providers are moving from the traditional access and core-facing networks to this converged residential, business, and mobile backhaul design model to manage new service introduction velocity and service node flexibility, and to ensure that end-to-end service is available regardless of network changes.

Figure 2: Network Convergence with Seamless MPLS
Network Convergence
with Seamless MPLS

Seamless MPLS with pseudowire head-end termination offers a solution to the following technical challenges:

  • Building a service-agnostic packet network at scale

  • Ensuring end-to-end service availability in the face of any failure scenario

  • Offering network service level agreements that accommodate diverse services

  • Managing new service introduction velocity

  • Coping with the operational complexities resulting from all of these challenges

The ultimate benefit of a converged seamless MPLS network is simplified service delivery as shown in Figure 3.

Figure 3: Simplified Service Delivery
Simplified
Service Delivery

In today’s Ethernet/MPLS environment, L2 and L3 functions no longer need to be separated. Seamless MPLS provides a flexible infrastructure to extend access toward service edge functions such as BNGs and mobile gateways, services that can be centralized or not, depending on scaling and services take rates.

Taking MPLS to the access, and enabling MPLS packet forwarding end-to-end across the network, requires new functionality and features and a systematic architecture that can scale to tens of thousands of nodes. Juniper Networks seamless MPLS architecture enables a smooth migration from traditional networks to a scalable, end-to-end MPLS network.

Building Seamless MPLS Networks

Taking MPLS to the access and building MPLS networks results in:

  • Service flexibility, simplified provisioning, simplified operations

    Seamless MPLS architecture is a systematic way of enabling MPLS end-to-end between access nodes, with all forwarding based on MPLS labels. This means that all service provisioning and operations are MPLS based. There is a clean separation of control plane, management plane, and data plane operations throughout the network that allows decoupling of the service provisioning plane from the underlying transport technology. It also makes way for optimizing and simplifying service provisioning and operations, making it possible to minimize the number of service provisioning points.

  • Network resiliency with deterministic, subsecond, end-to-end convergence for services

    MPLS has significant traffic engineering capabilities, enabling end-to-end service restoration. The ability to do traffic engineering based on real-time network conditions supports strict service level agreements, guaranteed service availability, and subsecond restoration of services with fast reroute mechanisms in the event of link or node failures. The goal of seamless MPLS is to extend the same benefits end-to-end across the access network.

  • Scale to the order of 100,000 nodes network-wide without compromising any of the benefits

    Seamless MPLS enhances the capacity to scale as needed. WANs or core typically consist of 100 to 1,000 L3 nodes, but the metro access and aggregation networks could each contain that many, with hundreds of access and aggregation subnetworks in one single network. The result is a network that can scale to over 100,000 nodes.

Seamless MPLS with Pseudowire Head-End Termination

Similar to MX Series PE routers, the MX Series BNG application integrates broadband subscriber management functionality with IP/MPLS service node functions. Supporting pseudowire head-end termination and subscriber management enables broadband network services to be transported over IP/MPLS and flexibly terminated on any MX Series service node within the IP/MPLS network. Figure 4 shows a number of services that are converged on this common architecture, including residential services (such as video, data, and wholesale L2TP) and business services (such as L3VPN, VPLS, L2VPN, and business Internet). With direct termination on the BNG for converged residential and business services, service providers are afforded a simplified design from a provisioning perspective (fewer provisioning points).

Figure 4: Seamless MPLS with Pseudowire Head-End Termination
Seamless MPLS with Pseudowire
Head-End Termination