# Remote PHY Device: What Changes at the Node and in the Plant **Title tag:** Remote PHY Device: HFC Plant Design Guide 2026 **Meta description:** Remote phy device guide: DOCSIS physical layer, CCAP core, digital Ethernet transport, timing, node sites, fiber routes and powering for HFC plant design. **Author:** Omar Molina **Published:** November 4, 2026 **Last updated:** November 4, 2026 **Category:** ISP & Carrier Networks / Data Center **URL:** https://draftech.com/blog/remote-phy-device **Primary keyword:** remote phy device **Word count:** 2760 **Read time:** 11 minutes ![A technician in a cap plugs a green fiber connector into an open metal enclosure hung from a strand above a town.](/blog/img_remote_phy_device.webp) ## In This Article - [What Does a Remote PHY Device Do?](#rpd) - [What Changes in the HFC Plant?](#plant) - [Remote PHY Device Siting, Space and Power](#node) - [What Design Must Settle Before Construction](#design) - [Choose the Next Step for Your Role](#next) ## What Does a Remote PHY Device Do? A remote phy device (RPD) is the node-mounted unit that moves the DOCSIS physical layer out of the headend or hub while the CCAP core keeps the MAC. It connects 2 transport domains: digital Ethernet over the hub-to-node fiber and radio-frequency signals on the coax serving subscribers. That split is the starting point for cable operators, independent systems and ISPs with HFC plant. We treat the RPD as a change in where signal processing happens, with direct consequences for the enclosure and the route feeding it. The existing coax still matters. Designers and builders need to know which physical interfaces change before they can draw a workable installation. ### The conversion happens at the node Downstream, the device converts digitally transported content into RF for the coax network; upstream, it converts received RF into digital traffic for the return toward the core. CableLabs describes that conversion in its Remote PHY architecture publications. Our design concern is the boundary where optical connectivity meets the node assembly, including the physical room required to terminate each connection. The MAC stays in the core. CableLabs' Data-Over-Cable Service Interface Specifications, MHAv2 Remote PHY Specification, CM-SP-R-PHY-I21-251211, issued December 11, 2025, defines the RPD architecture and interfaces. We use that architecture to distinguish the operator's active-network responsibilities from the outside-plant work needed to support the selected equipment. Operators configure and provision the RPDs and the CCAP core, and they operate the headend. We require their approved equipment requirements before resolving the physical installation. A construction drawing should show what must be installed without implying that a crew can choose network settings to make an incomplete design work. ### Why operators adopt distributed access architecture Distributed access architecture moves functions traditionally housed centrally closer to subscribers. CableLabs' Distributed Access Architecture overview identifies reduced headend hardware requirements as one reason for that shift. Remote PHY is one implementation of the idea: the operator can retain centralized MAC functions while relocating the physical layer into the access network. For our design review, the practical question is whether the operator can support that placement at the proposed node. Space moves with the equipment. A hub may shed physical-layer equipment while a field enclosure takes on different thermal and powering requirements, so the selected assembly must be reviewed in its actual environment. We ask the operator to state the reason for choosing Remote PHY before field design begins. That answer guides the scope of the survey and prevents an enclosure replacement from quietly expanding into unrelated plant changes. A limitation of this approach is that digital transport to the node leaves defects in the retained coax plant to be diagnosed and corrected separately. ### Remote PHY versus Remote MACPHY Remote PHY moves the PHY to the node and retains the MAC in the CCAP core; Remote MACPHY moves both MAC and PHY functions into the distributed device. The difference changes where DOCSIS processing resides and what the operator must support. CableLabs describes these distributed choices through its Remote PHY and Flexible MAC Architecture work. We confirm the selected architecture before drawing the enclosure, because a generic distributed-access label does not establish the actual equipment requirements. ## What Changes in the HFC Plant? The conversion changes the hub-to-node transport from analog optics to digital Ethernet. Fiber may be reused where its condition and the selected optical design allow it, but reuse needs evidence. We start with the installed path. A cable shown on a map does not establish the optical loss of the connected circuit or prove that its terminations are available. The table separates the physical changes from the plant that remains. It is a design review map, rather than a promise that every existing node can accept an RPD module. Each row needs an answer tied to the operator's chosen assembly before construction drawings can be issued. | Plant element | Remote PHY implication | Design decision | | --- | --- | --- | | Hub-to-node fiber | Digital Ethernet replaces analog optical transport | Verify the optical path and termination arrangement | | Timing | The remote physical layer needs a supported timing path | Obtain operator transport and synchronization requirements | | Node enclosure | Selected electronics must fit and reject heat | Confirm housing compatibility and service access | | Power feed | The installed assembly determines the electrical load | Review feed capacity and power insertion | | Coax cascade | Existing amplifiers and passives remain unless separately changed | Identify retained plant and unresolved RF defects | ### Digital fiber transport Our route review follows the proposed connection from its hub termination to the node entry. We reconcile the splice records against field identification and look for portions that exist only as planning lines. Where the physical path differs from the drawing, the design must resolve that difference before the construction package relies on it. Connectors deserve an explicit check. The selected optical interface determines what can be mated at the node, and an existing pigtail should never be treated as suitable merely because it reaches the port. We show the required termination arrangement and leave equipment selection with the operator's approved design authority. Optical acceptance should refer to the intended circuit, including the connections that will remain after installation. A test of a loose fiber segment answers a narrower question. We ask for test endpoints to be named on the drawing so that a passing record can be matched to the path the operator will use. Fiber-count allocation belongs in the operator-approved hub-to-node connection plan, developed through [OSP design for cable TV headend connectivity](/blog/osp-design-for-cable-tv-headend-connectivity). ### Timing crosses the physical boundary Timing is part of the architecture. The Remote PHY specification suite includes the Remote DOCSIS Timing Interface, which addresses synchronization between the core and remote device. A functioning optical connection alone cannot establish that the operator's timing requirements have been met across the transport network. We ask the operator's transport team for the approved timing requirements before treating the route as ready for service. Our drawings identify the physical path. They also identify construction dependencies affecting it. The operator determines the timing design and validates active-network behavior, including what happens when the normal transport path changes. > **Keep the acceptance boundary visible.** Record physical fiber acceptance separately from the operator's transport and timing acceptance. A crew's completed optical test should never be used as the sign-off for an active-network function that the test did not measure. This separation gives a designer something useful to request when a handoff is incomplete: either evidence about the installed path or a decision from the operator. It also lets construction close a physical defect without claiming that every service condition has been verified. The operator retains service acceptance. ## Remote PHY Device Siting, Space and Power A node location needs more than a coordinate. We evaluate the assembly where it will be mounted, with enough access to install it and return for maintenance. The enclosure must remain reachable after nearby work is complete, and the drawing should not depend on temporary access that disappears when the construction area is restored. ### Verify the housing before specifying reuse Housing reuse requires confirmation. We request the exact existing housing identity and the operator-approved conversion arrangement, rather than inferring compatibility from an exterior photograph. When the identification cannot be read safely during survey, the unresolved item stays explicit until the operator provides a reliable record or arranges an authorized inspection. The physical review includes the space needed for cable entry and internal routing. A module may fit within a housing while its connection arrangement interferes with closing the enclosure or accessing another component. We want that conflict resolved on the selected assembly before the bill of materials assumes a simple replacement. Thermal suitability depends on the assembly. Our package should identify the approved enclosure arrangement and any installation restrictions provided for that equipment. We do not assign a generic temperature allowance to every RPD installation, because the housing and installed components determine what the assembly can support. A useful field record shows the node in relation to its support and the surrounding access area. Close views establish equipment identity; wider views explain how a managed crew can reach it. Both views should refer to the same asset identifier used on the plan, so an attractive detail never becomes a drawing for the wrong node. These physical conflicts are part of our [HFC cable network engineering and construction scope](/services/hfc-cable-network): node architecture and fiber deep routes are coordinated with power insertion and construction packages. We keep those decisions tied to the operator's approved equipment requirements so the package can be built without transferring unresolved design choices to the field. ### Calculate power for the installed arrangement Power needs a fresh review. We base the assessment on the proposed node assembly and the equipment that will remain on its supply path. Existing supply records are useful inputs, but the design must account for the changed load and the actual path delivering power to the enclosure. The review should establish the available supply capability and the voltage expected at the node under the operator's design conditions. If the feed cannot support the proposed assembly, the drawings need an approved power change before the node detail is released. A note telling the installer to verify power leaves the engineering decision unresolved. Power insertion deserves its own drawing detail wherever the arrangement changes. We show the intended connection and identify the retained plant affected by that work. The operator must establish the permitted interruption and restoration conditions; a designer should not infer permission to disturb an energized feed from a general construction authorization. Backup expectations come from the operator. We request them early enough to check whether the proposed installation can accommodate the selected equipment, while keeping any backup-duration claim tied to the actual load assumptions and accepted battery design. A battery enclosure on a site photograph does not establish the available reserve. > **Check maintenance clearance with the enclosure open.** A mounting location can look workable in plan view while blocking the access needed to service the assembly. Resolve that geometry during field design, before the approved detail becomes the crew's installation instruction. ## What Design Must Settle Before Construction We recommend settling node sites and their physical dependencies before issuing the installation package. The operator selects the architecture. That decision establishes the equipment to be supported. Our design then has to show where it can be placed and how the approved connections reach it, with unresolved conditions visible to the person authorizing release. ### Keep the retained coax visible The coax cascade stays unchanged unless the project separately replaces or rearranges it. We label retained equipment clearly so an RPD detail cannot be mistaken for approval to remove an amplifier. Where existing records disagree with the survey, the operator needs to resolve the installed arrangement before the package depicts it as confirmed. RF problems still need their own evidence. We ask the operator to identify known plant defects that could affect acceptance after installation, especially where the physical work might otherwise be blamed for an existing condition. Keeping that baseline with the project gives both parties a way to distinguish inherited problems from construction discrepancies. Service-group boundaries remain an operator-approved input, with any boundary change handled through [node segmentation engineering](/blog/node-segmentation-engineering). ### Issue details that agree with each other Drawing consistency is a construction requirement. The node detail and the route sheet should identify the same asset, and their connection labels should lead to the same termination. We check that relationship before release because a crew cannot reconcile conflicting instructions by knowing which drawing was prepared most recently. The package should distinguish removal from retention wherever both appear in the same work area. That distinction needs to survive a printed sheet or a field markup. We use clear work limits so adjacent live plant remains recognizable when the installer is looking at the actual enclosure rather than the design model. A substitution needs review before installation. If the proposed assembly changes, we revisit the physical requirements that depended on the original selection. Keeping the old detail while swapping the equipment name can conceal a different power demand or an entry arrangement that no longer fits the approved housing. Construction closeout should return the installed condition to the operator's records. We want field changes tied to the affected drawing detail, with the final asset identity visible in the as-built package. That gives the operator a usable starting point for maintenance instead of an issued design that silently differs from the plant. The overall upgrade should be evaluated through [hybrid fiber coax engineering](/blog/hybrid-fiber-coax-engineering), where Remote PHY can be considered in relation to the operator's wider plant objectives. ## Choose the Next Step for Your Role **Cable operators and independent systems:** establish the active-network architecture and approved node assembly before authorizing physical design. Give the design team the requirements that control the installation, and name who accepts transport readiness. Your operations team should retain configuration and provisioning authority throughout the work. **ISPs with inherited HFC plant:** commission a targeted field review where housing identity or supply records are uncertain. Resolve those gaps before assuming that existing node sites can be reused. Our recommendation is to carry unknown conditions as open design decisions rather than allowing a drawing symbol to turn them into assumed facts. **Designers and builders:** require an installation detail that matches the selected assembly and its field location. Return conflicts before work begins. The crew needs an approved physical solution, and the operator needs a separate acceptance path for the active network after the physical installation is complete. Draftech International provides turnkey telecom and power infrastructure delivery, with engineering performed 100% in-house. Full turnkey construction is delivered by Draftech-managed subcontract crews under Draftech's QA/QC and safety program. Our [partner and delivery overview](/about) identifies the people accountable for that work, from field survey and planning through construction drawings, permits, agency coordination, construction, as-builts and closeout. Unconfirmed node space and unresolved power feeds are the kinds of physical design gaps we help close before an installation package reaches the field. Use our [project contact page](/contact) to [describe the proposed node work and the available plant records](mailto:info@draftech.com). We can then establish the engineering and construction scope around the operator's approved architecture. For a qualifying route associated with that work, Draftech engineers the first 20,000 linear feet at no cost, from feasibility and field survey through permit approval. Our [free route design offer](/free-design) begins with the owner reviewing each request before Draftech commits the package. ## Frequently Asked Questions ### What is the difference between an RPD and the CCAP core? The RPD handles the DOCSIS physical layer at the node, while the CCAP core retains the MAC in a Remote PHY architecture. These are 2 distinct functional locations connected through digital transport. We use that division to define physical design inputs, while the operator remains responsible for configuring and provisioning the devices and accepting their active-network operation. ### Does installing an RPD remove the coax amplifiers? Installing an RPD alone removes 0 coax amplifiers from the design: the existing cascade remains unless separate work changes it. We identify retained equipment on the construction drawings so the node replacement does not imply an unapproved removal. The operator must evaluate RF conditions independently and specify any amplifier or passive changes required for its intended service plan. ### Can existing fiber be used for Remote PHY? Existing fiber can be considered when the connected path meets the selected optical design and the operator approves its use. We examine both ends of the circuit, not just 1 accessible termination, and require acceptance records tied to the intended path. The available cable record alone cannot establish that the final connected circuit will support the selected equipment. ### Who verifies timing for a Remote PHY installation? The operator verifies the timing design and active-network behavior between the CCAP core and RPD. We keep 2 acceptance responsibilities distinct: physical-path acceptance and operator network acceptance. Our construction package identifies the installed route and its interfaces, while the operator establishes synchronization requirements and confirms that the transport network supports them under the conditions required for service. ### What should an operator send Draftech to start physical design? Send the approved node assembly requirements and available plant records, including any known gaps that need field verification. We use those inputs to define the survey and the physical design scope before construction details are issued. Engineering stays 100% in-house, and any included construction is delivered by Draftech-managed crews under our QA/QC, with operator configuration responsibilities stated separately.