There is a frequency at which electronics stops behaving the way the drawing says, and the engineers who work above it are a shrinking population.
When the wire stops being a wire
At low frequency a connection is a connection. Two points joined by copper are at the same potential, near enough, and the schematic is an honest description of the circuit.
As frequency rises this stops being true, and the transition is not gradual in its consequences.
A trace becomes a transmission line. Once its length is a meaningful fraction of a wavelength, the voltage is not the same along it. It has a characteristic impedance, and a mismatch at either end reflects power back down it.
Every piece of metal is a component. A via is an inductor. A gap between planes is a capacitor. A ground connection has impedance, and at high frequency that impedance can matter more than the component it is grounding.
Components stop being what they are labelled. A capacitor has lead inductance and becomes inductive above its self-resonant frequency. An inductor has parasitic capacitance and becomes capacitive above its own. A resistor does both. Above a few gigahertz, the part number tells you less than the package does.
Coupling happens whether you designed it or not. Adjacent traces couple. Enclosures resonate. Cables radiate.
In RF design the layout is not an implementation of the circuit. The layout is the circuit.
This is the reason the discipline resists automation. In digital design, tools handle most of the translation from intent to physical reality. In RF, the physical reality is where the design lives, and the judgement that separates a working layout from a marginal one is built by doing it, being wrong, measuring, and understanding why.
What the direction covers
The scope: radio frequency circuit design, antennas, wireless protocols, 5G and 6G systems, satellite communications and electromagnetic compatibility.
Four areas.
Radio frequency circuits. Amplifiers, mixers, filters, oscillators, matching networks and the measurement methods that characterise them.
Antennas and propagation. Radiation, gain, polarisation, path loss and the environment the signal actually travels through.
Systems and protocols. Modulation, coding, multiple access, network architecture.
Compliance. Emissions, immunity, exposure limits and spectrum regulation.
The link budget, which is the whole argument
The single most useful tool in the direction, and the one that turns a wireless question into an engineering answer.
A link budget accounts for every gain and every loss between transmitter and receiver, and compares the result against what the receiver needs to work.
Transmit power, transmitter losses, antenna gain at both ends, path loss, additional losses from obstructions, weather and body proximity, receiver noise, and the signal to noise ratio the chosen modulation requires. Add it up and the answer either closes with margin or it does not.
Three things this exposes that intuition does not.
Range is decided long before the radio is chosen. Path loss rises with distance and with frequency. No amount of firmware improves it.
Higher data rates cost sensitivity. More bits per symbol requires a higher signal to noise ratio, so a link that works at a low rate may not close at a high one. This is why real systems adapt their rate rather than fixing it.
Margin is not optional. A link that closes with two decibels to spare fails when it rains, when someone stands in front of the antenna, or when a component sits at the wrong end of its tolerance.
The practical use is in the negotiation before the product exists. Someone specifies a range, a data rate and a battery life. A link budget shows within an afternoon whether those three are compatible, and it is far cheaper to discover that on paper.
Where this sits in the domain
RF, wireless and telecommunications is the seventh of nine directions in Astra Trainer's semiconductors, electronics and quantum domain. It connects to electronics engineering, where the same layout discipline appears in gentler form, and to semiconductor physics and devices, since RF power depends heavily on compound semiconductor behaviour.
It also connects outward to space and mobility for satellite communications, to robotics for wireless control links, and to advanced manufacturing for industrial wireless. You can see the nine directions here.
Why higher frequency is not simply more bandwidth
Millimetre wave frequencies are attractive because wide contiguous spectrum is available there, and wide spectrum means high data rates. The trade is real and so is the cost, which is frequently underplayed.
Path loss is higher. For a given antenna size the free space loss rises with frequency, so the same power covers less ground.
Obstruction is severe. Millimetre wave signals are heavily attenuated by walls, foliage, rain and human bodies. A hand over a phone can drop a link. Coverage becomes close to line of sight in many environments.
Beams replace coverage. The usual answer is phased array antennas that concentrate energy in a direction, which recovers link budget and introduces a new problem: the beam must find the receiver and follow it. Beam management is continuous, and it is a significant part of why these systems are complex.
Everything gets harder to build. Tolerances shrink with wavelength. Packaging, interconnect and test all become more demanding and more expensive, and measurement at these frequencies requires equipment and technique that many teams do not have.
The honest summary is that millimetre wave is excellent for short range, high capacity, fixed or dense deployments, and poor as a general replacement for lower frequency coverage. Systems that need both use both.
Spectrum is a legal object
An aspect engineers from other disciplines consistently underestimate.
Radio spectrum is allocated and licensed by national regulators, and transmitting outside what you are permitted to transmit is an offence rather than a design flaw. Allocations differ between countries, which means a product sold in several markets may need different frequencies, different power limits or different behaviour in each.
Licence exempt bands, which most consumer wireless uses, are not unregulated. They carry limits on power, on duty cycle, on channel occupancy and often on the technique used, and they are shared, which means interference from other users is a design assumption rather than a fault.
Human exposure limits apply to transmitting equipment and are enforced. Compliance is demonstrated by accredited testing against the relevant standard, not by argument.
The consequence for a development schedule is that regulatory testing is a gate with a lead time, and a failure at that gate can mean a hardware change. Teams that plan for it treat it as a design input. Teams that do not meet it as a surprise at the end.
The roles, named
RF design engineers. Circuits and subsystems.
Antenna engineers. Design, simulation and measurement, including integration into products where space is hostile.
RF integrated circuit designers. Radio blocks in silicon.
Wireless systems engineers. Protocols, modulation and network architecture.
RF test engineers. Measurement, over the air testing, chamber work.
Electromagnetic compatibility engineers. Emissions, immunity and compliance.
Radio network engineers. Planning, optimisation and deployment.
Satellite communications engineers. Ground and space segment.
Spectrum and regulatory specialists.
Who can be trained into it
Electronics engineers. The main conversion, and it is a real one rather than an extension. The move is from lumped thinking to distributed thinking, and it is best taught with measurement alongside, because the concepts do not become intuitive until you have seen them on an instrument.
Radio and telecommunications technicians. Strong practical foundation in installation, alignment and fault finding, missing the analytical layer.
Military and aviation communications staff. Frequently excellent, holding systems experience, discipline and real world propagation intuition.
Network engineers. Into wireless systems and radio network planning, where the protocol knowledge transfers directly.
Test and measurement engineers. Into RF test, where instrument discipline is already present.
Physics graduates. A short path, since electromagnetics is the foundation.
Amateur radio operators from any background. Worth naming explicitly. Many already hold practical understanding of propagation, antennas and interference that takes years to acquire otherwise.
Transmission, exposure and licensing. Radio transmission is regulated by law in every jurisdiction, with binding limits on frequency, power, duty cycle and human exposure, and with requirements for accredited testing and, in many cases, an operator or station licence. Requirements differ by country and change. High power transmitters also present radiation and electrical hazards requiring controlled access and site specific procedures. Training builds engineering capability. It does not confer a licence, an authorisation to transmit, or compliance with any national regulation.
What to take from this
Above a certain frequency the layout stops implementing the circuit and becomes the circuit, which is why the skill takes years and why tools have not replaced it.
The link budget answers whether a wireless idea is possible before anyone builds it, and it settles most arguments about range, data rate and battery life in an afternoon.
Millimetre wave buys bandwidth with range, reliability and manufacturing difficulty. It is a good tool for dense short range capacity and a poor general replacement for coverage.
Spectrum compliance is a legal gate with a lead time, and failing it late can mean new hardware.
And the population that can do this work is ageing. If an organisation has experienced RF engineers, the transfer of what they know is a project with a deadline attached to it.
Why is RF design harder to automate than digital design?
Because the physical layout is the circuit rather than an implementation of it. Traces, vias, gaps and enclosures all become components, and the judgement that separates a working layout from a marginal one is built through measurement and experience.
What does a link budget do?
It adds every gain and loss between transmitter and receiver and compares the result against what the receiver needs. It settles questions about range, data rate and power before hardware exists.
Is millimetre wave simply better?
No. It offers far more bandwidth and pays for it with higher path loss, severe blockage by walls and bodies, a continuous beam alignment problem, and tighter manufacturing and measurement tolerances.
Why does spectrum regulation matter to engineers?
Because transmitting outside permitted frequency, power or exposure limits is unlawful, allocations differ between countries, and compliance is demonstrated by accredited test near the end of development, when a failure can require hardware changes.
Who converts well into RF engineering?
Electronics engineers, radio and telecommunications technicians, military and aviation communications staff, network engineers into wireless systems, test engineers into RF measurement, and amateur radio operators from any background.