Advertisement
Articles
Advertisement

In-Building Enterprise Connectivity – The Last 100 Yards

Wed, 02/08/2012 - 1:18pm
Matt Squire, former CTO of Overture Networks, and Andy Paff, managing director of Headwaters

MSOs leverage a variety of technologies to reach a larger base of enterprise customers.

When connecting a business customer with carrier Ethernet, the last 100 yards is critical. Whether the customer is a multi-tenant SMB or a single large enterprise, connectivity to the network inside the building presents numerous challenges.

No single technology solves every problem, so limiting in-building access to one approach dramatically and needlessly reduces the available market. Project profitability is made or broken with the choice of technology used, and some options can and do work better together.

Except for the smallest customers that can be served through the DOCSIS residential high-speed data platform, the underlying technology is carrier Ethernet.

However, there is often a need to address existing client technologies such as PRI (Primary Rate Interface) and DS-n TDM equipment, and this may require a transparent approach to legacy equipment and standards. Favorable customer and operator economics are both imperative to achieve the required value proposition.

Carrier Ethernet is clearly the best converged service delivery platform, giving the MSO a number of technology choices in the last mile, including:

  • Ethernet over DOCSIS (see CableLabs L2VPN)
  • Ethernet over fiber (e.g., Ethernet rings, GPON)
  • Ethernet over copper (bonded copper to deliver high-capacity, carrier-grade Ethernet services)
  • Ethernet over NxDS-1 (e.g., manage IP network through existing TDM facilities)
  • Ethernet over wireless (point-to-point for backhaul or building access)
  • DS-N/PRI over Ethernet (e.g., circuit emulation).

Cost, reliability and scalability are the driving factors in deciding which last-mile carrier Ethernet approach fits best. The decision relative to the last-mile approach must also factor in the need to seamlessly morph into higher capacity and managed quality as demand grows. For example, an Ethernet over Copper (EoC) solution might eventually give way to a dedicated fiber solution without service interruption or performance compromise.

CONNECTIVITY IS THE FIRST CHALLENGE
Getting network connectivity to the telecommunications room in a commercial building is a big step, but getting from the telecom room to the customer is often just as difficult. Rewiring a commercial building may be prohibitively expensive, even if the building owner consents.

The problem is aggravated in the multi-tenant commercial building, as there is not usually a common access point, and everything must be either “home run” configuration from the central telecommunications location or carefully architected to ensure the proper demarcation between customers. The lack of space in the existing building conduit may be a limiting factor, or the building owner may want an exorbitant fee to access the fiber the management company put in place.

In many cases, there will be significant in-building, voice-grade copper wiring available that can be accessed at a fraction of the cost of alternatives. This is the ideal situation for use of an EoC solution. Individual copper pairs can provide high-speed symmetrical service, and multiple pairs can be bonded together to provide symmetrical 100 Mbps data service or more in an extremely reliable, robust, carrier-grade system. EoC represents state-of-the art technology that has been proven in very harsh outside plant environments, and thus often thrives in-building, where distances are short and pairs aren’t exposed to rain, temperature and other external forces.

The EoC approach requires minimal initial costs to the service provider and provides even more efficient scale as individual client bandwidth grows. Individual customers can be configured and managed separately through the single aggregation platform in the central telecommunications room. The carrier network interface may be copper, fiber or whatever is available to reach to the customer.

To optimize this approach for both the customer and the network operator, voice, video and data offerings should be converged into Ethernet at the client location.

For example, the voice service could provide an opportunity for the carrier to sell its commercial VoIP platform and other hosted or managed services that have value for the customer. Not only should the network operator enjoy significant cost advantage, they can manage the bandwidth to the customer premise through their larger network and the in-building EoC link.

This will eliminate or dramatically reduce the “best-effort” problems associated with standalone providers of cloud services. If there is a need to maintain support for a traditional PBX with a PRI circuit, this could be managed at the customer location with Pseudo-Wire Emulation (PWE). TDM traffic can be integrated with the native Ethernet traffic, resulting in further economies of scale and simplified network management.

Figure 1: Ethernet over Copper ...Establishing a flexible in-building presence has another benefit: The MSO’s sales force can target the entire building. Once a foothold is established, and a reliable mechanism for reaching any tenant of a building is established via a combination of access technologies, the MSO can then target each and every customer in that building while having a very low build-out cost to reach those new customers. The economics become easy.

When considering a large single-tenant building, other variables must be considered. Typically, the customer’s central telecommunications room will provide access to the customer’s LAN and their internal TDM network, if they exist separately. This may be the ideal situation to implement an Ethernet over Fiber (EoF) solution, integrating the high-speed data and the voice applications to provide better performance and improved economics for the user.

If the voice application does utilize a traditional TDM approach (e.g., PBX), then integration with carrier Ethernet will require the PWE configuration. However, enterprise customers of all sizes are now looking for more sophisticated unified messaging services, which require IP connectivity to an on-premise or hosted server environment. This represents another good opportunity to upsell services, either immediately or down the road as the operator develops a closer relationship with the customer.

The customer edge device, centrally located at a common telecommunications room, may also require a fiber interface for both the customer LAN and the carrier network. In many cases, the customer will want some level of redundancy in the customer-located equipment, and they will certainly require that the network operator support their existing internal redundancy.

These larger customers will have increased performance expectations, and they may impose service-level agreements on the network operator that involve an agreed-upon reporting method to ensure compliance. The customer may also have multiple physical locations and the need to support internal communications, as well as external. The edge device must be able to handle these requirements from both a capacity and a network management perspective.

Carrier Ethernet edge devices exist that can provide efficient bandwidth scale and exceptional performance monitoring and can support both LAN and network redundancy. The best products provide various physical layer configurations (e.g., copper and fiber) that operate on a common chassis. This not only provides better reliability through fewer network elements, it also simplifies the operator’s life considerably from the standpoint of network homologation, software upgrades and inventory.

Figure 2: Optical and Ethernet over Copper ...NEXT CHALLENGE: EFFICIENT MANAGEMENT OF THE NETWORK
Cable MSOs must manage a “network of networks” to provide their extensive array of residential and commercial services. When considering just the commercial offerings, this will likely include a variety of physical layer assets, such as HFC/DOCSIS, optical Ethernet, EoC, PWE, Ethernet over TDM and wireless.

Compounding the complexity is the fact that the carrier does not own all of the separate transport networks. There are “off-net” leased assets, as well as “on-net” owned assets. The trend toward serving larger enterprise customers that are looking for complete solutions will further drive the need to find leased solutions to augment the on-net network.

In some cases, a project threshold internal rate of return may be reached initially by leasing copper or circuits (TDM or carrier Ethernet) on a wholesale basis, driving toward an internal build down the road when volume and revenue dictates.

The in-building configuration is similar to the last mile in this regard. However, the end user interface and the building aggregation unit can provide some important information relative to the whole network in the same way that the DOCSIS EMTA provides important information regarding the health of the residential HFC network.

The challenge of provisioning and monitoring the commercial services network is central to overall viability of the commercial services business plan. MSOs have done a remarkable job developing provisioning systems to deliver a wide array of products for their residential customers through the HFC platform.

The CableLabs DOCSIS Provisioning of EPON (DPoE) initiative is designed to ensure that existing MSO provisioning systems will enable EPON services and ensure multi-vendor compatibility. CableLabs is also working with the Metro Ethernet Forum (MEF) to ensure compatibility with the MEF standard relating to Ethernet Services Definition, Phase 2. This is an important step, particularly as it gets extended to non-GPON Ethernet architecture.

Figure 3: TDM Psuedowires ...Network monitoring and customer support is also a critical element in the success of the MSO commercial services business. To integrate carrier Ethernet in-building edge and CPE equipment, existing trouble ticket and network management system compatibility is a fundamental prerequisite. Much of the heavy lifting has been done, and MSOs have been able to adapt large third-party systems to fit their residential and commercial service systems environment.

This is a huge advantage and requirement to scale effectively, even though commercial and residential customers are often managed separately. Fewer systems result in better, more efficient performance and scale.

The same can be said for the last 100 yards. The fewer the edge devices, the better the reliability, and the easier to manage effectively. The days of point-to-point isolated Ethernet circuits are ending. MSOs must think about networks and compatible “networks of networks” in order to scale economically and provide the higher levels of service required by small, medium and large businesses as they adapt various hosted and managed IP solutions to their service offerings.

MSOs will look at total cost and compatibility as major factors when choosing solutions that reside in the building to serve enterprise customers. The last 100 yards cannot become the bottleneck for revenue growth, as there are a lot more elements found, in total, in this environment than elsewhere in the operators’ networks. Touching this equipment is expensive. There are also many approaches to the physical layer transport in the last 100 yards, including both fiber and copper, and technology flexibility is key to maximizing the service footprint.

The good news is that there is proven carrier-grade equipment that is both technically and economically scalable to fit a wide variety of the last 100 yard situations. The right equipment will also reduce complexity by consolidating functionality into fewer network elements, and it will fit nicely into established network management and provisioning processes.

Advertisement

Share This Story

X
You may login with either your assigned username or your e-mail address.
The password field is case sensitive.
Loading