Legacy VLR and Its Relationship to the 5G Core
In circuit-switched 2G/3G networks, a Visitor Location Register (VLR) works with an MSC to hold temporary subscriber/location data. The 5G Core uses different service-based functions, including the Access and Mobility Management Function (AMF); AMF is not simply a renamed VLR. These network functions may be hosted on general-purpose telecom servers, whose PCB requirements depend on the actual hardware design rather than on the VLR database role.
VLR, Authentication and Equipment Identity Are Network Functions
The 3GPP network architecture specification defines VLR as a location register for circuit-switched services used by an MSC. In 2G/3G circuit-switched operation, it exchanges subscriber/location information with the HLR to support calls in the visited area. LTE/EPC uses functions such as MME and HSS; a legacy MSC/VLR may participate in circuit-switched interworking, but VLR is not the general 4G packet-core mobility database.
Authentication Center (AuC) and Equipment Identity Register (EIR) are also logical functions. Subscriber authentication data and equipment-identity checking serve different purposes and do not require a particular “Authentication Center PCB.” Their host-platform storage, security and networking requirements must come from the actual equipment design.
The 5G System architecture separates functions including AMF, SMF, UPF, UDM and AUSF. AMF handles access/mobility-related control functions; UPF carries user-plane traffic under the session-management architecture. The 5G functions distribute responsibilities differently from MSC/VLR/HLR, so AMF should not be treated as a renamed VLR or a dedicated board type.
From a Network Role to a Hardware Specification
Dedicated equipment, virtual network functions and containerized functions can use different server or appliance platforms. Cloud-native deployment does not by itself prescribe more layers, a particular CPU count or an accelerator. Establish the workload and host requirements first:
- Compute and memory: Subscriber context, signaling load, software, virtualization and recovery behavior determine the platform resources. Package escape, memory channels and the complete board determine routing density and layer count.
- Network and internal I/O: Select NIC, Ethernet, PCIe and memory interfaces from the host design. AMF signaling load is different from UPF packet-forwarding throughput; neither implies that every core board needs multiple 100/200/400 GbE ports or PCIe 5.0/6.0.
- Availability: Define the service target, outage accounting, redundancy, maintenance and failover tests. As an explicitly hypothetical annual target, 99.999% availability permits about 5.26 minutes of unavailability in 365 days before any contractual exclusions. That is a service-level calculation, not a PCB reliability rating, and cannot be established by laminate choice alone.
HILPCB can align multilayer fabrication, selected high-speed/HDI constructions and assembly with the released platform. Separate core control boards, user-plane accelerators, management modules and small-cell radio boards in the fabrication and test package.
5G Access, User-Plane and Control-Plane Roles
gNB, Small Cells
UPF and data-network path; central or edge deployment
AMF, SMF, UDM, AUSF, PCF and supporting functions
These are role groups, not a serial packet path: RAN signaling reaches AMF, while the user-plane path reaches UPF. SMF manages sessions and controls UPF; the deployment determines where functions run.
High-Speed Signal Integrity for the Selected Interfaces
A server hosting core software can carry fast Ethernet, PCIe and memory channels, but the actual PHY, reach, connectors and package determine the channel budget. Distinguish aggregate port bandwidth from per-lane signaling, and confirm the required line rate and modulation rather than deriving them from a “5G” label.
- Materials: Compare the specified grade's Dk/Df, copper profile, glass construction and pressed thickness with the complete loss budget. Megtron or Tachyon grades can be candidates; a short control link may use a less costly construction. High-speed PCB options are selected by the channel, not a universal ban on FR-4.
- Routing and transitions: Model the critical package/trace/via/connector path with the selected interface guidance. Back-drilling can reduce a stub where the structure warrants it; it does not eliminate every reflection. Define residual stub, drill tolerances and reference-plane transitions in the release data.
- HDI: Microvias and other HDI constructions can solve package-escape or transition problems. Stacked/staggered paths, fill/cap requirements and lamination sequence need review; HDI is not mandatory on every AMF-hosting board.
Power Integrity from the Processor and Board Load Profile
Release processor, memory, NIC and accelerator rail requirements from their actual device and platform specifications. A signaling host and a packet-forwarding accelerator can have different power/transient demands. Review startup, workload changes, sleep, recovery and the power path at the relevant pins.
- Planes and connections: Size copper, neckdowns and vias for current, voltage drop and temperature rise. Close power/ground spacing can reduce inductance and provide distributed capacitance, but it does not replace the regulator or local decoupling model.
- Decoupling: Follow device guidance and account for effective capacitance under bias, ESR/ESL, mounting inductance and anti-resonance. More capacitors or a prescribed assortment of values do not automatically produce a better PDN.
- SI/PI interaction: Define rail noise and transient budgets, then assess return paths, simultaneous switching and any supply-induced timing effects. Correlate models and prototype measurements at the specified load states rather than assuming one universal frequency or impedance target.
The same release discipline applies to virtualized Mobile Switching Center platforms and packet-core appliances: software roles identify the workload, while selected hardware defines the board-level electrical limits.
Core Network Platform Choices Across Generations
4G (EPC)
Dedicated or virtualized EPC platforms
MME/HSS and packet-gateway roles
Interfaces and layers follow the equipment
5G (5GC)
COTS or purpose-built hardware (VNF/CNF)
Separate control and user-plane requirements
Release the actual NIC, memory and host design
6G (Future Core)
Research and evolving requirements
Evaluate accelerators/optical packaging where justified
No fixed PCB layer/rate prescription
Thermal Management from Component to Enclosure
Rack power may be measured in kilowatts, but a PCB thermal design needs the dissipation and permitted temperature of its own components, neighboring heat sources and enclosure conditions. Confirm airflow or liquid-cooling interfaces, heat-sink attachment and service loads at the platform level.
- Material and copper: High-Tg PCB material addresses thermal/assembly properties; high Tg is not high thermal conductivity. Heavy-copper routes can serve selected current or spreading needs, with geometry and assembly constraints reviewed separately.
- Thermal vias: Use the package vendor's permitted land/via arrangement and heat path. Via-in-pad can require filled/capped construction for assembly, and a via array cannot be placed indiscriminately beneath signal BGA lands.
- Local thermal features: Copper coins, thermal interfaces or specialized constructions may help a demonstrated bottleneck. Their attachment, electrical isolation, contact pressure and process route must fit the actual package; a heat pipe is not an assumed embedded-board feature.
- Layout and enclosure: Review CPU/NIC/accelerator hotspots, airflow obstruction, mechanical support and heat-sink loads together. Measure representative workloads and ambient conditions before selecting a thermal fix.
Material and Manufacturing Decisions
Use grade-specific data rather than assigning one Dk, Df or Tg to an entire loss class. The examples below are manufacturer-published typical values; confirm construction-specific design data and availability before release.
Example Laminate Data: Compare the Test Conditions
| Parameter | Isola FR408HR | Isola I-Speed | Isola Tachyon 100G |
|---|---|---|---|
| Typical Dk at 10 GHz | 3.65 (Bereskin stripline) | 3.63 (Bereskin stripline) | 3.02 (IPC-TM-650 2.5.5.5) |
| Typical Df at 10 GHz | 0.0095 (Bereskin stripline) | 0.0060 (Bereskin stripline) | 0.0021 (Bereskin stripline) |
| Tg by DSC | 190°C | 180°C | 215°C |
| Use in the board review | Compare with actual channel/thermal requirements | Compare loss, copper and construction at the required reach | Compare low-loss option with complete channel and process cost |
Sources: FR408HR, I-Speed and Tachyon 100G. Dk test methods differ as shown; values are not interchangeable design constants or lot acceptance limits. Tg does not track loss class monotonically, and its stated DSC method matters.
For dense multilayers, confirm registration, drilling aspect ratio, copper plating, filled/capped via planarity and assembly flatness with the released stackup. These manufacturing controls support the chosen host board; an Authentication Center or EIR software role does not itself establish a layer count or special material grade.
Platform Requirements to Define Before PCB Release
These are release inputs, not performance scores. A core control host, user-plane accelerator and conventional enterprise server can have different priorities; compare their actual configurations.
Network Slicing, Edge Deployment and Hardware Boundaries
AMF participates in access and mobility procedures within the 5G slicing architecture, alongside functions such as NSSF and SMF. Slice policy, selection and session continuity are system/software responsibilities; they do not map to a special “VLR/AMF PCB.” Distinguish the eMBB, URLLC or mMTC service requirement from the host-platform interface and resource budget.
Edge deployments can place user-plane processing and applications nearer the user; placement of AMF and other control functions follows the operator's architecture. A lower application-latency target does not automatically require AMF on every edge node.
- Form factor: Release the rack/appliance dimensions, connector positions, board support, service access and cooling interfaces. Compact rugged equipment and data-center hosts can need different assembly routes.
- Environment: Specify powered/storage temperature, humidity, vibration and contamination exposure for the actual site. Enclosure protection and selected screening are separate from material Tg.
- Timing and coordination: Network synchronization, software-state consistency and failover are distinct requirements. Oscillator, clock distribution or time-sync hardware is selected from the platform interfaces and accuracy budget, not from an assumption that every distributed AMF board needs a new clock architecture.
6G Research and Future Hardware Choices
ITU's IMT-2030 work provides a framework and evaluation process for future mobile systems. It does not prescribe a fixed core-server PCB construction. Treat AI-assisted network operation, new radio bands and optical packaging as evolving system choices, with the actual component and interface specification controlling a manufacturing release.
- Compute and interconnect: Accelerators, chiplets and co-packaged optics may change package, board and cooling interfaces. A PCB and a package substrate remain different manufacturing scopes; do not assume one automatically becomes the other.
- Optical paths: An optical module or co-packaged optical engine needs the corresponding electrical channel, power, mechanical and thermal handoff. Optical networking is not an automatic requirement for every mobility-management host.
- Power and thermal controls: Monitoring sensors and firmware can support load/cooling management when the platform defines their accuracy, limits and response. Claims of predictive self-healing need system evidence.
- Materials: Ceramic, glass or LCP options should address the intended RF, packaging or interconnect problem. A future radio's carrier frequency does not mean that a remote core-server board carries that RF frequency.
Radio-Band Context: Keep RF and Core-Server Requirements Separate
Sub-6GHz
Coverage and IoT deployments
Requirements depend on the radio system
Separate RF and digital host paths
Millimeter Wave (mmWave)
Hotspot broadband/FWA examples
RF front-end and antenna constraints
Actual channel and thermal review
Terahertz (THz)
Research for future systems
Candidate communication/sensing concepts
No assumed core-PCB RF requirement
Radio-band choices affect RF front-ends such as small-cell boards. Core-server compute, digital links and power are selected from the network workload and hardware platform; no universal exponential PCB requirement follows from the carrier band.
Quoting a Telecom Server or Core-Network Board
HILPCB's high-speed PCB route lists 2–20 layers standard and up to 48 by review, 75/75 µm standard and eligible 50/50 µm advanced geometry, and ±10%/advanced ±5% impedance options with TDR. Multilayer, HDI, backplane and heavy-copper routes are separate choices confirmed against the actual construction; they are not combined maxima or requirements of a VLR/AMF function.
SMT assembly, through-hole and turnkey assembly can coordinate sourcing, connector/package assembly, inspection, programming and customer-defined powered tests. Available first-article, profile, SPI/AOI, selected X-ray, electrical/functional and lot/serial records are confirmed with the quotation. Specify any channel/VNA measurements, screening or system failover tests separately.
Send fabrication/drill files, stackup, materials and impedance table with the BOM/CPL, host/NIC/interface part numbers, connector and thermal drawings, quantities and sourcing plan. Include critical channel/via requirements, power/thermal limits, firmware and test instructions, acceptance class and required records. These inputs let the fabrication and assembly route follow the real telecom platform and establish a controlled baseline for pilot and repeat production.
