- Overview
- Carrier Ethernet
- Coarse Wave Division Multiplexing Solution
- Commercial Services Solution
- IP Video Surveillance
- Layer 2 Virtual
Private Networks - Network Resiliency
- OAM
- Provider Backbone Bridging — Traffic Engineering
- Service Assurance
Hard QoS - Switched Ethernet vs. TDM-PON
- Wireless Backhaul Infrastructure
Analysis and Benefits of Carrier Ethernet
in Metro Networks
Applications
With the superior set of features and performance, while maintaining its cost-effectiveness, Carrier Ethernet is outfitted for a variety of exciting, growing applications.
Service providers are deploying Carrier Ethernet for several reasons. In some cases, it is used as an access technology to deliver services directly to customers. Carrier Ethernet is also being deployed to aggregate existing access technologies. Some service providers are envision end-to-end Ethernet networks obviating the need for legacy networking technologies. The following section highlights some of the use cases of Carrier Ethernet.
Native/Active Ethernet Access
As carriers and service providers rollout Ethernet in the Metro and Access networks, more business customers are migrating from circuit-based technologies. This represents a large and growing application.

Figure - Native/Active Ethernet Access
Business customers are increasingly seeking higher-bandwidth, more cost-effective, guaranteed services. Existing technologies often lack one or more of these capabilities. Another trend is the need for secure, transparent LAN services often referred to as layer two virtual private networks. The following section describes this relatively new feature of Carrier Ethernet.
Ethernet-based L2VPNs
As depicted below, Carrier Ethernet metro networks can support a variety of layer two virtual private networks (L2VPNs).
- E-Line services provide a secure, point-to-point connection between two customer locations.
- E-LAN services enable an extension of a business LAN to multiple locations
- The emerging E-Tree service type supports multicast services (e.g., business IPTV).

Figure - Business Service Delivery (Ethernet VPNs)
Let’s take a closer look at the three Carrier Ethernet Service Types.
E-Line Service Type
The following figure shows the E-Line Service Type carrying a point-to-point Ethernet Virtual Connection (EVC) between two User Network Interfaces (UNIs).

Figure - E-Line
E-LAN Service Type
The following figure depicts the E-LAN Service Type conveying a multipoint-to-multipoint Ethernet Virtual Connection between four UNIs.

Figure - E-Lan
E-Tree Service Type
The following figure shows the emerging E-Tree Service Type carrying a Rooted Multipoint EVC between a single Root UNI and four Leaf UNIs.

Figure - E-Tree
Within each Service Type, two variations are supported. The following chart defines the six Carrier Ethernet services according to each of the three Service Types described earlier:
| MEF Service Type | Port-based (All-to-one bundling) |
VLAN-based (Service multiplexed) |
|---|---|---|
| E-Line | Ethernet Private Line (EPL) | Ethernet Virtual Private Line (EVPL) |
| E-LAN | Ethernet Private LAN (EP-LAN) | Ethernet Virtual Private LAN (EVP-LAN) |
| E-Tree | Ethernet Private Tree (EP-Tree) | Ethernet Virtual Private Tree (EVP-Tree) |
EP services vs. EVP services
Simply stated, the EP service is defined at a port level between a customer location and the provider’s network (UNI). The EVP service is defined at a sub-port (e.g., VLAN) level at the UNI.

Figure - UNI
The EP service is best suited for delivering a single service terminated at a single port. All frames arriving at a port are conveyed across a single EVC regardless of the VLAN ID.
The EVP service is designed for multiple services delivered to a customer. In this case, the UNI performs service multiplexing as it carries two or more EVCs simultaneously. Each EVC may contain one or more VLAN IDs based on a configured table.
The ability to simultaneously deliver a dynamic range of guaranteed Ethernet services demands sophisticated carrier-grade equipment. Each of the services has a set of performance parameters. These are listed in the following table:
| Category | Attribute | Type of Parameter | Value |
|---|---|---|---|
| EVC | EVC Type | Point-to-Point, Multipoint-to-Multipoint, or Rooted-Multipoint | Select one |
| EVC ID | Identifier for service across network | Arbitrary string | |
| UNI List | A list of UNI IDs | List of <UNI ID, UNI Type> pairs | |
| Maximum # of UNIs | Point-to-point | 2 | |
| Multipoint-to-multipoint or Rooted-Multipoint | >= 2 | ||
| Remapping | CE-VLAN ID Preservation | Maintains Customer VLAN ID across network | Yes or No |
| Remarking | CE-VLAN CoS Preservation | Maintains Customer VLAN CoS across network | Yes or No |
| Service Frame Handling | Unicast | Discard, Deliver Unconditionally, or Deliver Conditionally | Select one |
| Multicast | Discard, Deliver Unconditionally, or Deliver Conditionally | Select one | |
| Broadcast | Discard, Deliver Unconditionally, or Deliver Conditionally | Select one | |
| Layer 2 Control Protocols | Processing | Tunnel or Discard (e.g., 802.1w RSTP, 802.3ad LACP, etc) | Select one action per Layer 2 Control protocol |
| MTU | Maximum Size | Maximum size service frame allowed | Integer >= 1522 |
| Quality of Service / Performance | CIR, CBS | Committed Information Rate, Committed Burst Size | Bits per second (e.g., 35 Mb/s), Bytes per second (e.g., 8 KB) |
| EIR, EBS | Excess Information Rate, Excess Burst Size | Bits per second (e.g., 50 Mb/s), Bytes per second (e.g., 8 KB) | |
| Frame Delay | One-way transmission delay | Milliseconds (e.g., 100 ms) | |
| Frame Delay Variation | Percentage of service frames delivered across network below certain threshold | e.g., 95% | |
| Frame Loss Ratio | Percentage of service frames arriving at ingress UNI and delivered to egress UNI | e.g., 99% | |
| Availability | Percentage of time within interval that Frame Loss Ratio performance is small | e.g., 99.9% |
Business Service Delivery
An important and growing trend is that many facilities-based carriers and multiple service operators (MSOs) are leveraging fiber-assets and their close proximity to business customers by deploying Carrier Ethernet as an access technology. These service providers are installing fiber to the business or multiple-tenant unit. Using Carrier Ethernet as the service transport, providers are able to offer transparency, high-bandwidth, and superior cost-performance compared with other technologies.
In some cases, carriers are deploying dedicated optical fiber access links to individual customers. One or more guaranteed services are delivered over this link. Multiple tenant units are served with fiber to the building and twisted-pair drops to individual businesses. Carrier Ethernet supports up to several hundred services multiplexed over these lines.
L2 vs. L3
While L3-based VPNs remain popular, increasingly L2-based techniques are increasing being deployed. It is useful to explore this important industry trend. L3VPNs evolved as extensions of the L3 Border Gateway Protocol (BGP) core. As a result, L3VPNs:
- Are more difficult to provision
- Require more complicated protocols
- Demand more processing power
- Require customers to exchange routing tables (Inherent resistance and error-prone)
- Requires routing protocols between provider and customer
- Limited Quality of Service and Service Level Assurance guarantees (e.g., best-effort)
- BGP private route space and memory demands lead to higher router count which costs significantly more capital and operational expenses
Based on these disadvantages, the industry has invested in developing lower-cost L2VPN technologies. This has yielded L2VPNs which are:
- Easier to provision
- – Fewer steps
- – Requires less-sophisticated (expensive) personnel
- Simpler control-plane
- Full transparency to customer’s routing infrastructure
- Transparent to customer’s chosen security/encryption techniques
- Excellent QoS on service and transport infrastructure
- – Committed Information Rate
- – Excess Information Rate
- – Frame Delay, Frame Delay Variation, Frame Loss Ratio, Availability parameters
CWDM
With the falling cost of CWDM components, some carriers are leveraging the capacity of fiber to node and tap locations by using multiple wavelengths. For instance, up to 10 wavelengths can be cost-effectively dedicated to service 10 customers.
The following figure shows a breakout of a physical optical link into multiple lambdas and services. As shown in the diagram, an L2VPN can be configured as a discrete service/interconnect or as an aggregate of multiple services (e.g., Business VoIP, Business Intranet, Internet access, etc).

Figure - Physical and Logical Partitioning
Emerging IEEE Standards
At the forefront of the L2VPN standards arena is IEEE 802.1ad Provider Bridging. This specification standardized the technique for performing VLAN tag stacking allowing customers to retain existing VLAN configurations and partitioning. The new outer VLAN is used as the service identifier for transport across the access and metro network. In addition, this important development added a QoS bit in the L2 header indicating drop eligibility. During periods of congestion, Carrier Ethernet switches can efficiently perform tail drops minimizing the impact to loss-sensitive services.
More recently, the IEEE 802.1ah Provider Backbone Bridging project is defining a MAC header encapsulation technique to enhance the scalability of Carrier Ethernet and reduce the memory and learning requirements of provider devices. By eliminating the need to learn and manage customer MAC addresses in the interior of the network, providers can build and manage larger more efficient Carrier Ethernet-based networks.
On the service management side, IEEE P802.1ag Connectivity Fault Management (CFM) adds needed path and network-based Operations, Administration, and Maintenance features. In 2004, the IEEE standardized link-based Ethernet OAM which provides the first-order capabilities of discovery, link-based loopback, and link performance monitoring. By contrast, CFM enables an 8-layer hierarchy of management frames to traverse Carrier Ethernet networks. For instance, one layer may allow customer-generated traffic to be conveyed end-to-end transparently. At another layer, an operator or provider may convey management traffic without impacting customer service levels.
These recent and emerging developments are enabling cost-effective and large-scale Carrier Ethernet services for a variety of carrier-grade business services. Due to the rich service offering, bandwidth potential, and L2VPN support, many providers are finding a fast payback on the investment in Carrier Ethernet solutions.
While the Ethernet business services market is growing rapidly, Carrier Ethernet is being leveraged in other portions of the network.
Carrier Ethernet backhaul applications
Other providers are continuing to leverage copper or coaxial assets and choosing to use Carrier Ethernet as a backhaul technology. By associating customers/service combinations with Ethernet Virtual Connections, Carrier Ethernet can efficiently transport services regardless of the access technology. By leveraging the single and double tagged 4KVLANs, large numbers of business and residential customers may be aggregated and individually serviced. Increasingly, Carrier Ethernet is being used as a lower-cost backhaul technology for DSLAMs and wireless access (e.g., WiMax).
Residential Service Aggregation
Increasingly, Carrier Ethernet is being deployed in residential service networks. It aggregates both native Ethernet and non-Ethernet endpoints such as DSL, wireless, and HFC. The rich Quality of Service characteristics together with the service management capabilities enable Carrier Ethernet to support large scale customer counts within a single layer two domain.

Figure - Residential Service Aggregation
