The Distributed Access Architecture (DAA) is defined by the cable industry as an architecture to meet the industry’s 10G goal, to deliver the capacity and services, and to evolve toward a software-based network infrastructure. DAA solutions call for disaggregating the access network by either moving the PHY layer to the optical node (Remote PHY or R-PHY) or by moving the PHY as well as the MAC layers to the optical node (Remote MACPHY or R-MACPHY).
Despite its significant promise, DAAs deployments have been slower than anticipated. This is largely due to:
- To date, operators had to select and commit upfront to a DAA variant (R-PHY or R-MACPHY) without an easy upgrade path if their needs changed and for the most part had to use the same configuration across their footprint to reduce complexity.
- Limited interoperability: The deployable solutions to date have been largely bookended (from the same vendor).
- DAA is more operationally complex: Unlike the integrated CCAP that can serve multiple service groups, the distributed nature of the DAA means that an access network element, such as a R-MACPHY device, typically serves a small number of service groups; in the case of R-PHY, because the CCAP core is still in the headend, the R-PHY devices need to be associated with the CCAP cores. Furthermore, moving intelligence to the access network means more maintenance is needed, which could add operating costs.
Fortunately, the challenges that have somewhat slowed the field adoption of DAAs are being mitigated, while the recent, massive uptick in bandwidth demand has brought to light the long-term limitations of the current, hardware-based, integrated architectures:
- Cable operators now have more solutions from which to select. With the recent market introduction of CommScope’s RD2322and similar solutions, operators no longer have to commit to R-PHY or R-MACPHY; for example, RD2322 is essentially an RxD, meaning that it can be deployed as an R-PHY device or a R-MACPHY device and can be changed from one configuration to the other with a software upgrade.
- More interoperable solutions. Of note is the recent announcement of Vecima Networks and Vector Technologies regarding their interoperable R-PHY solution for Liberty Global.
- Vendors and operators are increasing the operational intelligence (OI) of their equipment, reducing recurring opex costs and unleashing the power of automation and eventually AI to mitigate the complexity of the new operating environment, while improving customers' satisfaction. This is done through a combination of added OI via external software and embedded OI in the equipment itself. The added OI is usually in the form of an external software system, usually called an EMS, domain controller or SDN controller (good examples are CableOS from Harmonic and vManager from CommScope). The embedded OI, in the case of DAA, provides enhanced equipment and service monitoring via advanced telemetry or embedded measurement equipment. This enables remote monitoring, error mitigation by proactive predictions of incipient failures or degradations, and control of the equipment.
These functions typically include:
- Automatic adjustments and alignments,
- Automated configuration of the network elements,
- Remote monitoring of the active equipment,
- Remote measurements of the ingress entering through passive equipment.
Increasingly, these equipment-related functions are evolving to also include more information about the services that traverse the equipment, including service-level QoS (via service-level monitoring or synthetic traffic) and information about the network slices that are supported in the equipment. The OI, whether embedded or added, adds to the first cost of the equipment but pays for itself in recurring opex expense reduction and increased customer satisfaction.