Nothing has the potential to shake up the telecommunications access world more than the Central Office Re-architected as a Data Center (CORD) initiative, which was launched last year by ON.Lab and is now part of the Linux Foundation’s many open-source initiatives to open up communications networks.
For decades, the access portion of telecommunications networks – as well as many other parts of the network – have been the domain of highly specialized and proprietary technology. The access network is crucial because it’s where important wide area network (WAN) services such as broadband and mobile get extended out to the customer. In the last decade, this has generally meant a mixture of passive optical networking (PON) and digital subscriber line (DSL) services. Now CORD can change the way these services are deployed.
In the CORD vision, all central offices could be standardized around a generic hardware infrastructure defining racks of servers and switches. These generic hardware modules will be controlled by software that can be programmed and run using a standard, Linux-based operating system. This is a radical shift from the past when access equipment was housed in proprietary racks of either optical line terminals (OLTs) or Digital Subscriber Line Access Multiplexer (DSLAMs). The CORD model will transform this, creating standardized generic hardware platforms.
The beauty of this for access means that the telecom industry can take a data center architecture and extend it out to the access media network. There will be more flexibility to change and configure devices. Software control means that hardware can be programmed to behave differently with different services, for example perhaps allocate more bandwidth to services such as streaming media.
CORD will enable service providers to build a standard backend infrastructure for every central office, no matter which services are to be deployed, and access modules can be snapped in and scaled as line cards in a standard server rack to add new customers.
This has huge potential to transform the telecom access network into a scalable, data center model – enabling the creation of the telco cloud. In large, Web-scale data centers, racks of standardized servers and switches can be reconfigured or added on the fly to meet service needs. This model is not possible in the current access network, which is a mish-mash of legacy equipment and non-standard architectures. CORD would enable telecommunications suppliers to standardize on one architecture, bringing economy of scale to the industry.
CORD was initially created by AT&T as part of its Domain 2.0 software-defined networking (SDN) program. Many major players have joined the initiative, including Comcast, Google, Verizon, and China Mobile. CORD has defined several use cases, including Residential CORD (R-CORD), Mobile CORD (M-CORD), and Enterprise CORD (E-CORD). Cable is likely to be addressed in the future.
CORD is essential for completing the vision of SDN and Network Functions Virtualization (NFV), whereby the communications network is based on open and programmable components. The key to the access portion of the network will be a flexible OS that enables access components to be programmable and automated using standard data models such as NETCONF/YANG. CORD brings disaggregation, virtualization, and standardization to the access network.
Of course, this isn’t a done deal. There are still questions about CORD. Will enough carriers get on board to give the movement major momentum? Can appropriate orchestration standards be developed so that the access equipment from different vendors can interoperate? Will the standards fragment? Are the virtualization elements of the CORD operating systems such as ONOS scalable and stable? Carriers are testing these platforms now, but there is more work to do.
What is clear is that the carriers see a way to use CORD to leverage their investments in broadband. Having a standard, programmable architecture will simplify the complexity of managing many access services over a variety of media — fiber, copper, cable, and wireless. CORD has the potential to create one open network platform.
ADTRAN recently announced major commitments to be CORD-ready. The company is an Open Network Operating System (ONOS) Collaborator and is focused on ON.Labs initiatives, which include CORD, ONOS and Virtual OLT Hardware Abstraction (VOLTHA). ADTRAN’s CORD-based programmable access architectures are ideally suited to support residential, business and backhaul services over a converged access architecture including 10G EPON, XGS-PON, NG-PON2 and P2P 10GE, as well as copper and coax access with G.fast.
R. Scott Raynovich has been a technology analyst and journalist for 20 years. He is the founder and principal analyst of Futuriom (www.futuriom.com), a technology analysis firm.