Ethos-Networks

When Synchronous Digital Hierarchy (SDH) started paving its way into the access networks somewhere in the mid-‘90s, I met the transport department manager of a Tier-2 service provider. As a young, enthusiastic engineer, I excitedly introduced him to our new, innovative access-SDH solution, only to be astonished by his immediate, negative response: “I prefer a simple PDH link that never fails, than a complex SDH device I need to monitor. It will also cost me a lot more money, not only for the premium-priced equipment, but also to build the software management systems required to utilize all its capabilities.”

In a few short years, the simple inherent architecture of SDH did its thing, and after inevitable cost reduction and development of proper management systems, this person “converted” to become a devoted SDH prophet.

I learned an important lesson from this experience, something that has since proven to be true again and again. The most important factors in introducing new technologies into the transport domain are simplicity and low total cost of ownership (TCO). Unlike core networks, transport and access networks are widespread, often located in remote, unmanned sites and deployed in significantly larger volumes. Every dollar spent on equipment, every hour of training to technicians or every additional cost in provisioning new services must be multiplied by tens of thousands, if not more.

This is one of the main reasons that ATM did not evolve to become a true transport alternative. The dream of a single cell-based technology to carry all services could not fly – burdened by the weight of complexity and cost. Even when the industry has recognized that switched ATM would not become the next big thing and have settled for permanent connections (PVCs), it has not recovered from the fall. It never took a major role in the transport domain and T1s were carried predominantly directly over SONET rather over ATM.

In the move toward the all-packet network, there has been a gap in the standards to allow a simple, low-cost and reliable technology that can be used for transport networks. I have déjà vu when I see a core-network technology trying to make its way into the transport domain. MPLS routers were proposed to be the transmission vehicles for the metro/aggregation network, despite their complexity and higher cost. This was not driven by their technical merits, but rather by the lack of a standardized, simpler alternative.

Therefore, it is an important step for the transport networks industry that the IEEE has completed the standardization process of the carrier Ethernet suite, or as some would call Carrier Ethernet 2.0. The IEEE 802.1Qay, aka Provider Backbone Bridging – Traffic Engineering (PBB-TE), has been approved, enabling connection-oriented Ethernet services with guaranteed performance and transport quality. This standard joins its parent Provider Backbone Bridging standard (802.1ah) that enhances older Q-in-Q (802.1ad) for improved scalability and customer/provider domain separation, and the Connectivity Fault Management (802.1ag) that enables monitoring of the path connectivity.

Carrier Ethernet is a pure Layer 2 solution, which is designed to transport packets reliably at high quality; nothing more, nothing less. It is the goal of our industry to provide simple, cost-effective transport solutions, enabling the all-packet networks of the future.

Providing service providers with simple, yet comprehensive solutions that enable all-packet migration without compromising the simplicity experienced in TDM networks will be critical moving forward. Key drivers that will be important in helping service providers make this change include solutions with a combination of connectionless and connection-oriented Ethernet, as well as user-friendly domain management systems, in order for service providers to have the same confidence level and “ease of use” experience they have become accustomed to in their legacy SONET network.

- Yossi Saad, VP Marketing, Ethos Networks

For the original blog post click here