Evolution of Cable Networks
Multiple-system operators (MSOs) are continuously pushing the limits of their cable infrastructure. Fiber is being extended closer to the end customer in an effort to increase per-user capacity by reducing the number of active nodes in the coaxial network and limiting resource sharing on the coaxial cable. In parallel, the troublesome analog optical RF transmission from the headend to the radio node is increasingly becoming digitized.
From Fiber Nodes …
Up to now, hybrid fiber coaxial (HFC) networks have utilized a fairly simple analog media converter known as fiber node to mediate the optical analog signal into an electrical radio frequency (RF) signal. This multi-channel signal carries analog TV channels as well as MPEG video streams and high-speed data on subcarriers. It is then distributed in the coaxial network to cable modems located with residential customers or at enterprise sites.
…to Remote PHY Devices
Optical transmission is actually not very well suited for analog signals due to the impact of fiber dispersion as well as the high shot noise of photonic transmission. The latter results in poor system budgets and small margins. Due to this, there is a significant benefit in substituting the analog optical transmission between headend and fiber node with a digital optical link. A digitized fiber node is known as remote PHY device (RPD). It converts the digital signal into the analog RF DOCSIS signal for the coaxial part of the access network. In such distributed access architectures (DAAs), the distributed converged cable access platform implements cable modem termination system (CMTS) functionality partially in the central hub and partially in the RPD.
Synchronization in Distributed Access Architectures
In HFC networks, synchronization information is piggybacked on the QAM signal and transported from the CMTS at the headend to the cable modem on the customer premises. The fiber node passes this information transparently through. This approach no longer works with DAA architectures, as there is no analog QAM signal between the headend and the RPD. In consequence, a new method for providing synchronization to the edge of the cable network needs to be introduced.
IEEE 1588 Precision Time Protocol (PTP) is a well standardized and widely applied technology for time and frequency synchronization over packet networks. This protocol uses Ethernet frames or IP packets for distributing time information and can be flexibly applied for moderate timing accuracy as well as for very precise frequency and phase synchronization.
Typical DOCSIS implementations do not impose stringent timing requirements and IEEE 1588 PTP can be applied by simply adding a grandmaster as a timing source at the headend. Timing packets are distributed over the network using IP unicast as defined with G.8275.2 PTP profile. This method is sufficient for standard cable networks such as residential access and allows very cost-efficient synchronization.
Precision Timing Services
Cable networks can also be used for connecting base stations in mobile networks, which need precise timing. This requires a UTC-traceable synchronization architecture to be deployed. A central cesium clock, which is synchronized from a GNSS signal, can provide very precise time information to a grandmaster. This grandmaster sources PTP packets to the data network. Any network element in the data path needs to be able to process PTP packets in order to compensate for device-specific latency. This approach might require a fork-lift upgrade of existing data networks.
Alternatively, GNSS-sourced grandmasters can be distributed in the network. This allows precise synchronization to be introduced in a much more flexible and cost-efficient way. With this second approach, grandmasters are required in different sizes as cable operators will want to place them at aggregation nodes or even colocate them with RPDs or cable modems in case of highly accurate synchronization requirements. The combination of GNSS-sourced timing and network-based PTP timing provides a high level of resiliency.
The synchronization architecture needs to be carefully designed in line with timing requirements and operational considerations. A consistent and comprehensive grandmaster portfolio with sophisticated assurance functions enables a cable network operator to flexibly design and implement the synchronization network. In addition, MSOs will need to identify any deviation from the required time precision. That’s why synchronization delivery must be complemented by sophisticated synchronization assurance functions. Synchronization is an extremely technical topic and the innovation and expertise of an experienced supplier is essential to design and implement the most suitable solution.
The transition from HFC to DAA is impacting synchronization in cable access networks. A central grandmaster delivering PTP in unicast mode will be sufficient for standard cable networks, while significantly more sophisticated synchronization will need to be applied for precise timing services.
The wide range of accuracy requirements translates into a need for different synchronization products. In case of very high accuracy, the timing source should be sourced at the edge of the cable network and synchronization delivery needs to be complemented by sophisticated synchronization assurance functions.
Oscilloquartz provides a most comprehensive synchronization portfolio for economic, manageable and precise synchronization of any device and application.