Behind a routine announcement there is sometimes a more significant message or shift of strategy and that is the case with Arris declaring general availability for the second round of its cable CCAP platform. Specifically this is the second generation of active modules for Arris’ platform for CCAP, its E6000 Converged Edge Router, which is itself significant given Arris is the world’s leading CCAP vendor. CCAP has a vital role to play for just about all MSOs in their IP migration by combining Edge QAM video distribution with CMTS DOCSIS modem function in a single device. Over time the video can then migrate to the CMTS to go all IP, but operators will do this at different times with varying approaches.
CCAP was originally conceived as a centralized architecture where all the DOCSIS CMTS terminating functions were integrated with the Edge QAM in the core device, the idea being to save rack space and reduce hardware costs by enabling use of virtualized off the shelf processors. The industry then swung towards distributed architectures, but with varying options to suit different legacy networks, with some putting both the physical connectivity and MAC data link functions out to nodes in the network, while others retained the MAC layer centrally and only distributed the PHY.
So Remote PHY (R-PHY) distributes the PHY function to the fiber nodes while retaining central control over the MAC, while R-MAC-PHY distributes both, but still retains the most complex functions centrally.
Arris had been a cheerleader for flexibility and giving operators a choice between different options for CCAP deployment, supporting both R-PHY and R-MAC-PHY. It still pays lip service to this, but has signaled strongly with its latest E6000 router announcement that it is now putting its weight behind R-PHY and falling into line with those rivals such as Cisco who have leaned that way all along. Nokia however, as we covered in story last week, is still backing both horses about equally.
Yet momentum is shifting towards R-PHY as it wins the backing of major MSOs such as Liberty Global and also because its separation of functions lends itself to more efficient use of the network and more straightforward migration. It keeps the more numerous distributed nodes as simple as possible, which eases management, reduces deployment costs and saves power. It has become apparent that large R-MAC-PHY deployments would make it hard for operators to meet energy budgets. R-PHY also fits better with most virtualization strategies by concentrating the complex processing, demanding more memory and storage, in the cloud where the hardware is available and then allowing IP over Ethernet to be extended ever closer to subscribers.
Arris’s CCAP announcements are designed also to take advantage of steps already taken by MSOs to prepare for DOCSIS 3.1, the latest version of DOCSIS, which should include increasing spectrum to clear the way for higher bit rates and more channels. Arris estimates that MSOs will face an average 50% increase in bandwidth consumption over the next 12 months and has catered for that by making the new modules software-upgradeable so that channel density can be increased further over time in the same hardware and rack footprint.
The system comprises the Downstream Cable Access Module 2 (DCAM-2) and the Upstream Cable Access Module 2 (UCAM-2) for deployment in the fiber node, along with the centralized Router System Module 2 (RSM-2). These modules allow operators to migrate to a R-PHY Distributed Access Architecture, which will increase density a further three times over Arris’ current CCAP generation, without having to change any other hardware.
Arris has been piloting the extended CCAP router capabilities for some time and in fact claimed to have made the first commercial deployment of R-PHY capabilities with Danish MSO Stofa, the country’s second largest pay TV and broadband operator.
This was announced in July 2017 and highlighted how it is mid-range operators, rather than the largest, who have been at the head of DOCSIS 3.1, CCAP and now R-PHY deployments. Cisco also has reported most initial interest from smaller customers including Korea’s D’Live and South Florida’s Blue Stream. In fact Cisco claims to have had significant interest in its Infinite Broadband Remote PHY solution at the SCTE Cable-Tec Expo this week in Denver, USA.
It was another mid-range Danish operator that claimed to have started the world’s first DOCSIS 3.1 roll out in May 2016 using equipment from China’s Huawei, now approaching completion with 1.5 million households connected. A year earlier in June 2015 TDC and Huawei completed the first DOCSIS 3.1 field test demonstrating a measured downstream transmission rate of 840 Mbps.
DOCSIS 3.1 is an essential part of the IP migration story because of the great efficiency improvements it brings, making it easier to deliver OTT services at high QoS while preparing the ground for eventually moving QAM over to IP as well. In practice most operators will continue running QAM and IP protocol stacks side by side for years, but at least DOCSIS 3.1 allows more efficient coexistence as well as a path towards eventual elimination of QAM.
The immediate benefits are in performance and efficiency, with DOCSIS 3.1 increasing capacity for data and QAM video over given spectrum by 50% through use of OFDM (Orthogonal frequency-division multiplexing ), which exploits multiple sub-carriers in channels up to 192 MHz in width, combined with Low Density Parity Check (LDPC) forward error correction (FEC), which is more efficient than the Reed-Solomon FEC codes used in early DOCSIS versions. This combination of OFDM and LDPC allows operators to use higher QAM modulation orders up to 4096 to yield that increase in QAM video capacity. The greater density requires higher tolerance to errors during signal transmission which is where the LDPC comes in.
It is worth noting that of the four biggest US MSOs, Comcast, Charter, Cox and Altice USA, only Comcast, which is Arris’s largest customer, has deployed DOCSIS 3.1 at any scale yet. This raises two points, firstly that Comcast is well placed to ensure availability and timely supply of the components through its close ties with Arris. This began in January 2013 when Comcast paid $150 million for a stake in Arris, followed in July 2016 with an agreement for the operator to buy up to 8 million Arris’ ordinary shares over the following two years. The key point here is that the agreement specified that Comcast would buy a set percentage of products and services for network and cloud services from Arris. For Comcast this meant security of supply at a time when it was possible that fast growing demand for DOCSIS 3.1 modems would hold up deliveries. Now Comcast has expressed its interest deploying R-PHY using Arris’s latest E6000 router modules.
The other point is that leading MSOs are approaching IP migration in different ways and that DOCSIS 3.1 will not always be part of the plans. Altice for example has said it plans to skip DOCSIS 3.1 altogether and go straight to fiber-to-the-home (FTTH). This is unlikely to be feasible for many operators, but Charter has denied any immediate plans to migrate from DOCSIS 3.0 to 3.1 while it digests its Time Warner Cable and Bright House acquisitions, while Cox has muttered that it will eventually go there but not just yet.