And the final product can still have:
or
Unexpected Performance Loss Inside the Final Enclosure.
Why?
Because wireless performance is created by the entire system:
365PCB RF & Wireless System Design approaches all of these domains as one connected product-development problem.
Wireless Performance Is a System Result.
Start With the Communication Requirement
A common mistake is:
“We want Bluetooth.”
or:
“We want Wi-Fi.”
Those are technologies.
They are not yet product requirements.
The first questions should be:
What Data Must Be Transmitted?
A few bytes?
Sensor telemetry?
Audio?
Video?
Large firmware images?
How Far?
10 centimeters?
10 meters?
100 meters?
Several kilometers through a public network?
How Fast?
Kilobits?
Megabits?
Gigabits?
How Often?
Continuous?
Periodic?
Event-driven?
How Much Latency Is Acceptable?
Seconds?
Hundreds of milliseconds?
Milliseconds?
What Is the Power Budget?
Mains-powered?
Rechargeable?
Coin cell?
Multi-year battery target?
What Network Exists?
Router?
Phone?
Gateway?
Cellular operator?
Private network?
Where Will the Product Be Sold?
North America?
Europe?
Asia?
Worldwide?
Only after these are understood should wireless technology be selected.
Choose the Wireless Technology for the Product — Not the Product for the Wireless Technology.
Wireless Requirements Engineering
A mature wireless product specification can define:
Coverage
Range
Throughput
Latency
Packet Loss
Connection Time
Roaming
Device Density
Battery Life
Security
Interoperability
Regional Availability
Update Requirements
Certification
Instead of a vague requirement such as:
“The product should have good wireless range.”
write something measurable.
For example:
Maintain the required application data rate under defined test conditions at the specified operating distance and orientation.
Wireless Requirements Must Eventually Become Measurable Tests.
Wireless Technology Selection
Modern products have many possible wireless technologies.
High data throughput and direct IP connectivity.
Low-power personal-area connectivity and smartphone integration.
Low-power IPv6 mesh networking.
Application-layer interoperability for connected devices using IP transports such as Wi-Fi and Thread.
Mature low-power mesh ecosystem for many IoT applications.
Longer-range or specialized low-rate communications where appropriate.
Low-power wide-area network architecture for low-rate IoT data.
LTE / 5G and appropriate IoT categories for wide-area operator-network connectivity.
5G RedCap / eRedCap
Reduced-complexity cellular architecture targeting applications that need more capability than very-low-rate IoT but less complexity than full-performance 5G.
Satellite-based positioning and timing.
Often the correct answer is not one technology.
It may be:
Multi-Radio Architecture.
Technology Selection Matrix
A professional architecture can compare candidate technologies across:
How far must communication work?
How much useful data must move?
How quickly must the application respond?
How much energy per transaction?
Router, gateway, phone or operator network?
Point-to-point, star or mesh?
Fixed or moving?
What trust model is required?
Which frequencies and regulatory rules apply?
How long must the ecosystem remain supportable?
Wireless Selection Is a Trade Study — Not a Brand Preference.
Single-Radio vs Multi-Radio Architecture
Many modern products use several radios.
Example:
for high-bandwidth networking.
for commissioning and smartphone interaction.
for location.
Another product might use:
Matter application layer.
Another might use:
for a mobile connected device.
Every added radio creates additional:
RF Coexistence
Antenna
Power
Firmware
Certification
and
complexity.
Every Additional Radio Adds Capability — and Interaction Risk.
Wi-Fi System Design
Wi-Fi is appropriate where products need:
Modern Wi-Fi architecture may involve:
2.4 GHz
5 GHz
and, where regional rules and product architecture permit:
6 GHz.
Wi-Fi CERTIFIED 7 introduced capabilities intended to improve throughput, latency and reliability for demanding applications.
But deploying Wi-Fi 7 is not automatically necessary.
The correct generation depends on:
Required Performance + Cost + Power + Host Processor + Antenna + Product Lifecycle.
Use the Wi-Fi Performance the Product Needs — Not the Highest Number on the Box.
Wi-Fi 2.4 GHz vs 5 GHz vs 6 GHz
Each band presents different trade-offs.
2.4 GHz
Generally provides stronger propagation through many common obstacles but has substantial coexistence and congestion challenges.
5 GHz
Offers more spectrum and potentially higher throughput.
6 GHz
Provides additional spectrum for compatible Wi-Fi generations and can support high-performance architectures, but propagation, regional availability, antenna and certification considerations remain important.
Band choice influences:
Range
Channel Availability
Antenna
Power
Coexistence
Regional SKU Strategy.
More Spectrum Does Not Automatically Mean Better Coverage.
Wi-Fi Throughput Engineering
Customers may look at a chipset's maximum PHY data rate.
But application throughput is lower.
Real performance depends on:
Therefore:
PHY Rate Is Not Product Throughput.
The performance requirement should be specified at the layer the customer actually experiences.
Wi-Fi Latency
High throughput and low latency are not identical goals.
Latency can depend on:
airtime contention
network load
power-save behavior
router behavior
application stack
cloud path
For latency-sensitive products, engineering needs to measure:
and potentially:
rather than only average ping time.
Wireless Latency Is Statistical.
Wi-Fi MIMO Architecture
Modern Wi-Fi can use multiple spatial streams.
A MIMO system may contain:
and
or more chains depending on implementation.
The antennas need appropriate:
isolation
orientation
spatial characteristics
Simply installing two antennas does not automatically create good MIMO.
MIMO Performance Begins With Antenna Geometry.
Wi-Fi Multi-Link Architecture
One important modern Wi-Fi direction is the ability to coordinate communication across more than one link/band.
This can potentially improve:
throughput
latency
reliability
depending on implementation and network support.
But it also increases:
antenna
firmware
power
and
validation
complexity.
More Links Require More System Engineering.
Bluetooth LE System Design
Bluetooth LE is widely used for:
wearables
peripherals
commissioning
control
smartphone connectivity
low-power devices
As of 2026, Bluetooth SIG's current Core Specification is Bluetooth Core 6.2, adopted in November 2025. It includes features such as shorter connection intervals, reducing the minimum connection interval to 375 μs for appropriate applications.
This shows how BLE continues expanding beyond simple low-rate sensor links.
Bluetooth LE Is Becoming a More Capable Real-Time Wireless Platform.
Bluetooth Connection Architecture
BLE behavior depends on parameters such as:
Advertising Interval
Connection Interval
PHY
Data Length
MTU
Connection Latency
Transmit Power
These affect:
vs.
Battery Life.
For example:
Shorter intervals may reduce latency.
But they generally increase radio activity.
BLE Performance Is a Duty-Cycle Trade-Off.
Bluetooth PHY Selection
Depending on hardware and standard support, BLE systems can use different PHY modes optimized toward:
higher throughput
robustness
longer range
The correct PHY depends on:
Range + Airtime + Power + Data Requirement.
A longer-range mode does not automatically mean better product performance in every environment because airtime and throughput change.
Bluetooth Channel Sounding
Bluetooth continues moving into more advanced ranging capabilities.
Bluetooth Core 6.x includes Channel Sounding, intended to improve secure and accurate distance-awareness capabilities within Bluetooth devices.
By Core 6.2, Bluetooth SIG has also added additional resilience work around Channel Sounding against certain amplitude-based attacks.
This expands Bluetooth from:
toward:
Connectivity + Spatial Awareness.
Bluetooth Pairing & Provisioning
A product should define:
first-use commissioning
authentication
bonding
reconnection
lost-device handling
factory reset
ownership transfer
These are product experiences, not merely protocol-stack functions.
Wireless Security Begins With Who Is Allowed to Join.
Thread System Design
Thread is an IPv6-based low-power mesh networking technology intended for connected-device ecosystems.
Thread 1.4 introduced enhancements around:
credentials sharing
improved Internet connectivity
diagnostics
infrastructure integration
commissioning
mesh robustness
and current Thread Group resources now reference a Thread 1.4.1 specification.
Thread architecture commonly includes:
Thread Makes Low-Power Mesh Part of the IP Architecture.
Thread Mesh Engineering
Mesh networking can improve:
coverage
resilience
route flexibility
But mesh is not magic.
Performance depends on:
node placement
router availability
network density
traffic patterns
Battery-powered sleepy devices behave differently from always-powered routing devices.
A Mesh Network Needs an Architecture — Not Just More Nodes.
Matter Product Architecture
Matter is primarily about interoperable application behavior across IP-connected smart devices.
A product may use:
or
with BLE frequently involved during commissioning.
As of August 2026, CSA has released Matter 1.5.1. Matter 1.5 expanded the standard into categories including cameras and additional energy-management capabilities; 1.5.1 added further camera and media improvements.
Matter Does Not Replace the Radio.
It Defines How Connected Products Work Together Above the Network.
Matter Commissioning Architecture
A Matter product needs a deliberate commissioning experience.
Engineering considerations can include:
device identity
onboarding information
network credential transfer
fabric membership
factory reset
multi-admin behavior
The commissioning UX should be designed together with security and manufacturing provisioning.
The First Wireless Connection Is Part of the Product Architecture.
Matter Device Identity
Connected devices need trusted identity.
That can connect:
This makes factory provisioning an important part of the product's cybersecurity model.
Identity Can Begin on the Production Line.
Zigbee System Design
Zigbee remains relevant in many low-power mesh ecosystems.
Architecture may include:
Coordinator
Router
End Device
depending on network structure.
Engineering considerations include:
device role
network size
channel planning
coexistence
power
interoperability
Technology should be selected based on ecosystem needs rather than assuming every new smart product should use the newest protocol.
Mature Technology Can Be the Correct Technology.
2.4 GHz Coexistence
One of the hardest practical wireless problems is that many systems share the 2.4 GHz spectrum.
Common technologies include:
Wi-Fi
Bluetooth
Thread
Zigbee
When several radios operate inside one product, interference may occur even when each radio works perfectly alone.
Wireless Coexistence Is a Scheduling + Frequency + Antenna + RF Isolation Problem.
Radio Coexistence Architecture
Coexistence strategies can include:
Use non-overlapping or better-separated channels where possible.
Prevent radios from transmitting at conflicting times.
Some radio platforms support coexistence interfaces.
Reduce direct RF coupling.
Control out-of-band energy.
Coexistence Should Be Designed Before the Radios Begin Fighting Each Other.
Receiver Desense
A receiver may technically have excellent sensitivity.
But when another subsystem becomes active, sensitivity may degrade.
Possible aggressors include:
Nearby Transmitter
DDR
Display Clock
DC/DC
USB
Processor Clock
This is:
Receiver Desense.
An especially important validation method is:
Measure receiver performance with:
then with:
The Complete Product Active.
Measure the Receiver in the Product's Noisiest State.
Sub-GHz Wireless Design
Sub-GHz technologies can offer advantages for:
longer propagation
lower-rate sensing
building penetration
specialized IoT
But frequencies and permitted operating conditions vary geographically.
System architecture must therefore understand:
Regional Certification.
Sub-GHz Is Inherently Regional.
Proprietary Wireless Protocols
Some products benefit from a custom or proprietary RF protocol.
This can allow engineers to optimize:
timing
packet structure
power
network behavior
But proprietary protocols create responsibilities for:
Interoperability
Security
Maintenance
Gateway
and
Lifecycle Support.
Custom Wireless Gives Freedom — and Ownership of the Entire Stack.
LoRaWAN System Design
LoRaWAN is designed around low-power, wide-area IoT networks.
The LoRa Alliance's current 1.0.4 certification package includes the LoRaWAN Link Layer Specification 1.0.4, with regional implementations including EU868, US915, AU915, AS923 and other regional plans.
LoRaWAN is well suited to many products transmitting:
over:
with:
Low Average Power.
But it is not designed to replace high-bandwidth Wi-Fi or cellular.
Long Range Does Not Mean High Data Rate.
LoRaWAN Regional Architecture
A global LoRaWAN product cannot simply use one RF configuration everywhere.
Regional parameters affect:
channels
frequencies
power
channel masks
data rates
The LoRa Alliance maintains explicit regional-parameter specifications for this reason.
Therefore international products may require:
or
depending on hardware and certification strategy.
Global Wireless Products Need Regional RF Engineering.
Cellular Wireless Architecture
Cellular becomes attractive when products need:
wide-area coverage
mobility
operator infrastructure
independent Internet connectivity
A cellular product can involve:
But cellular adds substantial complexity around:
bands
antenna
network certification
operator testing
power
software
regional variants
Cellular Connectivity Is a Product Ecosystem — Not Just a Modem.
5G RedCap
Full-capability 5G NR can be unnecessarily complex for some IoT devices.
3GPP introduced RedCap — Reduced Capability NR — in Release 17 to provide a reduced-complexity 5G architecture, then refined the concept in Release 18 through enhanced RedCap — eRedCap.
This targets an important middle space between:
and
Full eMBB 5G.
Possible applications can include:
industrial devices
video-adjacent connected products
gateways
wearables
monitoring equipment
RedCap Changes the Question From “Do We Need 5G?” to “How Much 5G Capability Do We Actually Need?”
Cellular Band Architecture
Global cellular devices may support many frequency bands.
Every additional band can increase:
antenna complexity
filtering
tuning
certification
validation
The product may choose between:
Broad band coverage.
or:
Fewer bands optimized per market.
Global SKU Simplicity Can Create RF Complexity.
SIM / eSIM Architecture
A cellular product needs an identity and network subscription architecture.
Possible approaches include:
eSIM / eUICC
depending on product and service model.
This decision can affect:
enclosure
field provisioning
manufacturing
operator flexibility
software
Connectivity Architecture Includes Commercial Provisioning Architecture.
Cellular Power Design
Cellular radios can have large transient current demands during transmission.
A battery system can appear to have sufficient average capacity but still fail because of:
Voltage Droop During TX Bursts.
Power architecture must therefore consider:
not only:
Average Current.
Wireless Power Design Must Follow the RF Duty Cycle.
GNSS Architecture
GNSS can provide:
position
velocity
timing
A modern receiver may support multiple satellite constellations and bands depending on platform.
But GNSS signals arriving at Earth are extremely weak.
That makes GNSS particularly sensitive to:
antenna
filtering
digital noise
GNSS Is a Receiver-Desense Problem Waiting to Happen.
GNSS Antenna Placement
GNSS antennas need appropriate view and orientation relative to the environment.
The product should avoid placing them near major:
clocks
processors
switching regulators
displays
where possible.
A Great GNSS Receiver Cannot Recover a Signal That the Product Electrically Buried.
GNSS + Cellular / Wi-Fi Coexistence
Multi-radio products may have GNSS active while other radios transmit.
Strong local transmit energy can couple into GNSS front-end circuitry.
Architecture may therefore require:
filtering
placement
antenna isolation
frequency planning
temporal control
The Weakest Receiver Often Defines Coexistence Difficulty.
Link-Budget Engineering
A wireless link should be engineered using:
Link Budget.
At a conceptual level:
Transmit Power
TX Antenna Gain
−
−
System Losses
RX Antenna Gain
=
The received signal then needs adequate margin relative to receiver requirements.
Wireless Range Begins With Mathematics Before It Reaches the Test Chamber.
Link Margin
A theoretical link that works with:
0 dB margin
is not a robust product.
Real systems experience:
orientation changes
fading
obstacles
manufacturing variation
interference
Therefore the architecture should allocate:
Link Margin.
Range Without Margin Is a Demonstration — Not a Product.
Free-Space Path Loss
RF signals lose power as they propagate.
Higher frequencies generally incur greater free-space path loss for a given distance under the same formulation.
But real environments introduce many additional effects.
Therefore a simple free-space calculation is:
A Starting Point.
Not a complete range prediction.
Indoor Propagation
Inside buildings, wireless signals encounter:
walls
floors
furniture
people
metal structures
Signals may:
Absorb.
A product tested in an open office may behave differently in a concrete building.
Range Is an Environment-Specific Specification.
Multipath
Multiple reflected versions of a signal can reach the receiver at different times and phases.
This is:
Multipath.
Depending on radio architecture, multipath can:
degrade reception
create fading
or be exploited by technologies such as MIMO/OFDM
The Wireless Channel Is Not One Path Through the Air.
Fading
Wireless signal strength can change dramatically with:
location
orientation
movement
A product can move only a small physical distance and see a meaningful signal-level change.
Therefore validation should not use one orientation at one point.
Wireless Range Needs Spatial Statistics.
Receiver Sensitivity
Receiver sensitivity indicates how weak a signal can be while maintaining the specified communication performance under defined conditions.
Sensitivity depends on:
bandwidth
modulation
noise figure
required SNR
receiver implementation
But actual product sensitivity can be worse than chipset sensitivity because of:
Antenna Loss + Filter Loss + PCB Loss + Desense.
Chipset Sensitivity Is Not Product Sensitivity.
Conducted vs Radiated Sensitivity
Tests receiver performance through an RF connection.
Includes:
antenna
enclosure
orientation
complete product implementation
This distinction is essential.
Conducted Performance Proves the Radio.
Radiated Performance Proves the Product.
Transmit-Power Engineering
Transmit power affects:
coverage
current
thermal behavior
regulation
coexistence
More power is not always better.
Excess power can:
increase current
increase interference
create regulatory issues
worsen receiver coexistence
Optimize the Link — Not Just the Transmitter.
Adaptive Transmit Power
Some systems can reduce transmit power when strong links exist.
Potential advantages include:
lower energy
less interference
better coexistence
This is a system-level optimization involving:
Link Quality + Firmware + Radio Control.
Antenna Architecture
The antenna converts:
into
and vice versa.
Antenna architecture should be decided early because it can influence:
enclosure
battery position
connector
industrial design
The Antenna Needs Physical Space — Not Just a Schematic Symbol.
PCB Antenna vs External Antenna
Potential benefits:
low BOM
compact integration
no cable
but highly dependent on PCB/enclosure geometry.
Can simplify certain geometries but still requires matching and ground configuration.
May improve placement flexibility, but adds:
cable
connector
mechanical complexity
Antenna Type Is a Product Architecture Decision.
Antenna Keep-Out
PCB and chip antennas frequently need defined electromagnetic keep-out regions.
Putting:
battery
metal
display
cables
or
ground copper
into the wrong region can significantly alter antenna behavior.
An Antenna Keep-Out Is Electromagnetic Space Reserved for the Product.
Antenna Matching
Antenna impedance depends on the complete product.
Therefore it may require tuning using a network such as:
or another suitable topology.
But:
Matching Does Not Repair a Fundamentally Bad Antenna Environment.
Matching can improve impedance transfer.
It cannot create radiation efficiency that was lost because the antenna is trapped inside metal.
Antenna Efficiency
A perfectly matched antenna does not automatically radiate efficiently.
Energy may be:
accepted
but then:
dissipated as loss.
Therefore important antenna characteristics include more than S11.
Good Match ≠ Good Antenna.
This is one of the most important principles in wireless engineering.
Radiation Pattern
An antenna does not radiate equally in every direction.
The spatial distribution is its:
Radiation Pattern.
The final product enclosure and nearby structures alter that pattern.
Therefore the antenna should be evaluated in the final product configuration.
Polarization
Wireless links also depend on field polarization.
Product orientation can therefore influence performance.
This is especially important for devices that may be installed:
vertically
horizontally
randomly
Antenna Orientation Is a User-Experience Variable.
Diversity
A diversity system may use multiple antennas or signal paths to reduce fading risk.
The receiver can select or combine paths depending on implementation.
But antennas need sufficiently different channel characteristics.
Two Antennas Beside Each Other Are Not Automatically Useful Diversity.
MIMO Antenna Isolation
MIMO systems need appropriate antenna isolation and spatial characteristics.
Poor isolation can reduce:
spatial-stream independence
throughput
diversity value
Therefore MIMO needs:
Antenna + Mechanical + RF System Co-Design.
Antenna Correlation
Two MIMO antennas that respond almost identically may provide less spatial benefit.
Antenna geometry, orientation and product structure influence correlation.
MIMO Is About Independent Spatial Information — Not Antenna Count.
Enclosure Detuning
An antenna tuned on an open PCB may shift when installed into:
Plastic
Metal
Glass
Battery
Display
Cable
User's Hand
The final tuning should therefore be validated:
In the Real Mechanical Product.
Not only on the evaluation board.
Human-Body Loading
Wearable and handheld products may experience significant antenna changes when close to the human body.
This can affect:
efficiency
impedance
pattern
and also introduces applicable RF-exposure considerations.
The User Can Become Part of the RF Environment.
Metal Enclosure Wireless Design
Metal enclosures create special challenges.
Possible solutions may involve:
antenna windows
external antennas
carefully controlled slots
non-metallic regions
The industrial designer and RF engineer therefore need to cooperate early.
RF Cannot Be Added After the Mechanical Design Has Already Built a Faraday Cage.
Antenna Tuning Workflow
A practical development process can be:
Tune the Antenna in the Product It Will Actually Live In.
RF Front-End Architecture
A wireless radio may require:
RF Switch
PA
LNA
Filter
Duplexer
Matching
depending on technology and integration.
These components consume:
power
The system architect should determine what is truly needed.
Every RF Front-End Component Must Earn Its Place in the Link Budget.
Integrated Module vs Chip-Down Design
This is one of the most important ODM decisions.
Can provide:
faster development
reduced RF design burden
potentially simpler certification path
but often with:
higher unit cost
larger size
reduced architecture freedom
Can provide:
lower BOM at scale
compact integration
custom RF architecture
but demands more:
RF + Antenna + Firmware + PCB + Certification Engineering.
Module vs Chip-Down Is a Volume, Risk and Lifecycle Decision.
Pre-Certified Modules
Using a certified radio module can reduce some certification burden.
But it does not mean:
The complete final product automatically has no compliance work.
In the U.S., FCC guidance for modular transmitters defines integration conditions, including module/host requirements and the authorized antenna/configuration framework. FCC updated its modular certification guidance in 2024.
The host product still needs appropriate review for its own configuration and applicable requirements.
Module Certification Simplifies Compliance.
It Does Not Eliminate System Responsibility.
Regional Wireless Compliance
A global wireless product may encounter requirements involving:
Radio Spectrum
EMC
RF Exposure
Electrical Safety
Cybersecurity
depending on product and market.
In the United States, most intentional Part 15 radiators generally require certification before marketing.
In Europe, radio equipment falls under the Radio Equipment Directive 2014/53/EU, whose consolidated version remains current as of May 2026.
Regulatory Strategy Should Begin During Architecture — Not After DVT.
Regulatory Band Planning
The same radio chipset may not be allowed to use the same:
in every country.
Therefore firmware and hardware may need region-aware configuration.
Software Cannot Legally Turn Every Frequency On Everywhere.
Regional behavior must be controlled.
RF Exposure
Products transmitting RF energy can be subject to RF exposure requirements depending on:
power
frequency
distance to users
product category
This should be considered early for:
wearables
handheld products
body-worn devices
RF Exposure Can Influence Antenna Placement and Transmit Power.
Wireless Cybersecurity
Wireless creates an external attack surface.
Product security should consider:
Device Identity
Authentication
Encryption
Key Management
Secure Boot
Secure Update
Debug Protection
Factory Provisioning
The radio protocol's built-in security is only one layer.
Secure Radio Does Not Automatically Mean Secure Product.
Wireless Credential Architecture
A product may contain:
Wi-Fi credentials
Bluetooth keys
Matter credentials
cloud certificates
cellular credentials
The engineering team should define:
Where Are They Generated?
Where Are They Stored?
Who Can Read Them?
How Are They Replaced?
Credentials Are Product Assets.
Device Identity Provisioning
Each production unit may require a unique identity.
The flow can be:
This links:
Security + Firmware + Manufacturing + Cloud Backend.
Wireless Manufacturing Can Become Cryptographic Manufacturing.
Secure Boot
Connected products should consider whether unauthorized firmware could take control of the radio or device.
A secure chain can be:
The Device Should Know What Software It Is Running.
OTA Architecture
Wireless products often require firmware updates after deployment.
A robust OTA process may include:
OTA Is a Reliability Feature and a Security Feature.
Update Recovery
The most important OTA question is often:
What Happens if the Update Fails?
Potential strategies include:
A/B Partitions
Recovery Image
Rollback
depending on storage and platform.
A device should not become unusable because:
power failed
network disappeared
image was corrupted
Every OTA Strategy Needs a Recovery Strategy.
Large Wireless Firmware Updates
Update size affects network architecture.
A tiny BLE product may have a small firmware image.
A Linux wireless device may have hundreds of megabytes or more of software.
This changes:
storage
bandwidth
update time
battery impact
Software Size Can Become a Wireless-System Requirement.
Wireless Firmware Architecture
Radio firmware may need to manage:
initialization
scanning
connection
roaming
reconnect
timeout
network loss
credentials
security
power state
A product should define behavior for:
Bad Networks.
Because real networks are frequently bad.
Connection-State Machine
A robust architecture may explicitly define:
This creates predictable behavior.
Connectivity Is a State Machine — Not a Boolean Variable.
Reconnection Engineering
Network loss is normal.
Products need policies for:
retry interval
backoff
alternate network
power consumption
user feedback
Aggressive reconnection can drain batteries.
Slow reconnection can create poor UX.
Recovery Behavior Is Part of Wireless Performance.
Roaming
Mobile Wi-Fi or cellular-connected products may need to change access points or network cells while operating.
Engineering should define:
handover tolerance
buffering
session persistence
application behavior
Connectivity Must Survive Movement if the Product Moves.
Packet-Loss Engineering
Wireless links occasionally lose packets.
The system should distinguish:
from:
Application-level reliability.
Some applications may tolerate loss.
Others require acknowledgement and retransmission.
Reliability Should Be Engineered at the Correct Protocol Layer.
TCP vs UDP Architecture
Some connected products need reliable ordered transport.
Others prioritize low latency and can tolerate controlled loss.
The protocol decision should follow:
not generic preference.
Network Protocol Is Product Behavior.
MQTT & IoT Messaging
MQTT is widely used in IoT products because of its publish/subscribe architecture.
IEEE's newest 1451.1.6-2025 smart-transducer work now even defines carriage of IEEE 1451 messages over MQTT, illustrating its significance in networked sensor architectures.
A complete IoT system can become:
Wireless Development Extends Beyond the RF Link.
Cloud Connectivity Architecture
A connected product often needs:
device registry
authentication
messaging
telemetry
command/control
logs
This creates an architecture:
↕
↕
↕
↕
A Connected Device Is a Distributed System.
Offline-First Product Design
Cloud connectivity should not automatically mean the product stops functioning when the Internet disappears.
Depending on product requirements, local operation may continue.
The architecture can define:
What Requires Cloud?
and:
What Must Continue Locally?
Connectivity Failure Should Not Create Unnecessary Product Failure.
Edge Processing
Local processing can reduce:
wireless traffic
cloud cost
latency
privacy exposure
A sensor might send:
instead of streaming:
10,000 raw samples per second.
The Best Wireless Optimization May Be Sending Less Data.
Wireless Power Budget
Battery life should be modeled from:
TX Current × TX Time
RX Current × RX Time
Processing
Sensor
Sleep Current
Instead of looking only at:
radio sleep current.
Battery Life Is an Energy-per-Use-Case Calculation.
Energy per Transaction
For many IoT products, a useful metric is:
Energy per Successful Message.
This includes:
A technology with low instantaneous current can still consume high energy if connection time is long.
Connection Setup Energy
Wireless protocols differ significantly in startup/connection overhead.
For rare transmissions, connection energy can dominate actual data-transfer energy.
This is why technology selection should consider:
Traffic Pattern.
not merely:
data rate and sleep current.
RF Duty-Cycling
Low-power products may schedule radio activity.
Example:
This can dramatically improve battery life.
But duty-cycling affects:
responsiveness
latency
network availability
Power Saving Is a Product Behavior Trade-Off.
Multi-Radio Power Coordination
A product containing several radios should avoid activating all of them unnecessarily.
For example:
may be disabled after setup except when needed.
can sleep between transactions.
may use periodic fixes.
Multi-Radio Architecture Needs an Energy State Machine.
Wireless Thermal Engineering
High-throughput Wi-Fi or cellular transmission can generate significant heat.
Heat can affect:
modem
processor
battery
antenna environment
Sustained throughput should therefore be tested:
Inside the Final Enclosure at Real Ambient Temperature.
Not only on an open development board.
Wireless PCB Architecture
The PCB must integrate:
RF
Digital
Power
Antenna
Connectors
Wireless PCB architecture should consider:
antenna region
ground
matching
digital-noise sources
shielding
Wireless PCB Design Is Mixed-Signal RF Design.
RF Stack-Up
The PCB stack-up influences:
controlled impedance
transmission-line geometry
return current
antenna
The design should understand whether the RF path uses:
or another suitable structure.
RF Geometry Belongs to the Stack-Up.
Antenna Feedline
The feedline connecting radio to antenna should preserve the intended impedance environment.
Potential discontinuities include:
vias
matching pads
connector
The Antenna Cannot Correct Every Loss Created Before the Antenna.
RF Test Connector Strategy
During development, conducted RF access may be useful.
A test connector or controlled RF access path can allow:
transmitter measurement
receiver measurement
calibration
But test hardware itself adds:
cost
parasitic effects
Production strategy should determine whether it remains in the final product.
Antenna OTA Testing
OTA — Over-the-Air — validation includes the complete radiated system.
Depending on product requirements, metrics can include:
and
or other appropriate radiated performance measures.
These give a much more complete picture than S11 alone.
Antenna Match Measures Impedance.
OTA Measures Wireless Product Performance.
TRP
TRP integrates transmitted performance across spatial directions.
It includes effects from:
PA
RF Loss
Antenna Efficiency
Enclosure
and
Pattern.
Therefore it helps answer:
How much useful RF power does the complete product actually radiate?
TIS
TIS evaluates receiver sensitivity across spatial directions.
It includes:
antenna
receiver
desense
mechanical configuration
TIS Can Reveal Problems a Conducted Receiver Test Cannot See.
OTA Chamber Validation
Radiated testing may use:
anechoic chamber
reverberation chamber
other appropriate controlled test methods
depending on application and test standard.
The goal is to reduce dependence on random office testing.
Wireless Engineering Needs Repeatable RF Environments.
Orientation Testing
Wireless products should be tested in multiple orientations.
Especially for:
handheld
wearable
portable
arbitrarily installed products
because antenna patterns and user interaction vary.
One Orientation Is Not a Wireless Product.
100 — Coexistence Validation
A multi-radio product should deliberately test combinations such as:
Wi-Fi TX + BLE RX
Cellular TX + GNSS RX
Wi-Fi + Thread
depending on architecture.
The relevant question is:
Does Every Radio Still Meet Its Requirement When the Other Radios Are Active?
101 — Digital-Noise Validation
Activate major digital subsystems one at a time:
DDR
Display
USB
Ethernet
CPU load
and measure receiver degradation.
This builds a:
Desense Map.
That map can identify which subsystem is consuming RF margin.
102 — Power-Converter Coexistence Testing
Change:
operating mode
switching frequency where supported
and observe radio behavior.
A spur moving with converter frequency is strong evidence of coupling.
Wireless Debugging Should Change the Aggressor and Watch the Receiver.
103 — Wireless EMC
Wireless products must both:
and
avoid unwanted emissions.
That makes EMC particularly interesting.
The product needs:
useful intentional emissions
controlled spurious emissions
immunity to external RF
A Wireless Product Must Radiate the Right Energy — and Control the Wrong Energy.
104 — Harmonics & Spurious Emissions
Transmitters can generate:
harmonics
intermodulation
broadband noise
RF front-end filters and architecture should keep unwanted energy within applicable requirements.
This needs to be measured in the final product configuration.
105 — Pre-Compliance Engineering
Formal compliance testing should not be the first time RF performance is measured.
A stronger process is:
This can reduce late-stage redesign risk.
106 — Certification Planning
Certification planning should begin during:
Architecture.
Questions include:
Which countries?
Which radios?
Which bands?
Module or chip-down?
Which antennas?
What RF exposure category?
Which ecosystem certifications?
The answers can influence component selection before the schematic is released.
107 — Ecosystem Certification
Regulatory certification and technology-ecosystem certification are different things.
A product may need, depending on technology and commercial goals:
Bluetooth qualification
Wi-Fi certification
Matter certification
Thread certification
LoRaWAN certification
in addition to regional regulatory compliance.
Legally Allowed to Transmit ≠ Proven Interoperability.
108 — Wireless Interoperability
Products should be tested against more than one:
phone
router
gateway
when relevant.
Real ecosystems contain implementation differences.
The Standard Defines Behavior.
Products Reveal Edge Cases.
109 — Router Compatibility
Wi-Fi devices may encounter:
different routers
bands
security settings
congestion
network topologies
A product should not be qualified only against the engineering team's favorite access point.
Test the Ecosystem the Customer Will Actually Use.
110 — Smartphone Compatibility
Bluetooth and Wi-Fi products may interact with different:
phone vendors
permissions models
The RF link may work while the user experience fails because of application/OS behavior.
Wireless Product Engineering Extends Into the Host Ecosystem.
111 — Matter Interoperability
Matter exists specifically to increase ecosystem interoperability.
But implementation still requires:
correct device model
commissioning behavior
security
network behavior
As Matter expands—now including camera-related capabilities in 1.5/1.5.1—the amount of product/system engineering around application interoperability grows as well.
112 — Wireless Diagnostics
A good connected product should be able to report information such as:
connection state
network reason codes
reconnect count
firmware version
packet statistics
where useful.
A Wireless Failure Should Leave Evidence.
Otherwise field debugging becomes guesswork.
113 — RSSI Is Not the Whole Link
RSSI is useful.
But high RSSI does not automatically mean a good link.
The system may also suffer from:
interference
packet collisions
high noise
protocol congestion
Strong Signal ≠ Healthy Connection.
114 — Link-Quality Metrics
Depending on technology, useful metrics may include:
RSSI
SNR
Packet Error Rate
Retry Rate
PHY Rate
Throughput
Latency
The product should monitor metrics relevant to actual application behavior.
115 — Field Telemetry
Connected products can potentially report anonymized/appropriate operational metrics such as:
connectivity uptime
reconnect events
firmware version
error statistics
This can help identify real-world wireless issues across a deployed fleet.
Field Connectivity Data Can Become Engineering Feedback.
116 — Wireless Reliability Engineering
Wireless reliability includes more than hardware survival.
The system should survive:
Router Reboot
Network Loss
Weak Signal
Credential Change
Server Loss
OTA Failure
Radio Reset
A product may remain electrically healthy while becoming functionally useless because connectivity does not recover.
Connectivity Recovery Is Reliability Engineering.
117 — Watchdog & Radio Recovery
Firmware can monitor whether the radio subsystem is making progress.
Recovery may escalate:
depending on fault architecture.
But watchdog resets should not be used to hide poorly understood software bugs.
Recovery Should Be Controlled — and Failures Should Be Logged.
118 — Wireless Product Variants
Global products may require variants because of:
frequency
cellular bands
antennas
certification
A product-family architecture can use:
Design the Global Platform While Controlling Regional RF Differences.
119 — Wireless Component Lifecycle
Radios and modules change over time.
A wireless component EOL can trigger:
firmware changes
antenna tuning
regulatory testing
ecosystem certification
Therefore lifecycle risk can be especially expensive.
Wireless Component Replacement Is Often a System Redesign.
Select radio platforms with intended product lifetime in mind.
120 — Module Lifecycle vs Chipset Lifecycle
Modules can simplify development but add dependency on the module supplier.
Chip-down architectures depend more directly on semiconductor roadmaps.
Both need lifecycle planning.
Development Convenience Today Should Not Create Product Obsolescence Tomorrow.
121 — Wireless DFM
Wireless DFM must protect RF intent during manufacturing.
Examples include:
stack-up control
transmission-line geometry
antenna keep-outs
component placement
shielding
Wireless DFM Is Electromagnetic DFM.
122 — Production RF Test
Production should verify more than:
Device responds over UART.
Depending on product risk, manufacturing test may check:
receiver function
frequency
power
antenna path
identity
The exact test depth should balance:
Risk + Coverage + Cycle Time + Cost.
123 — Conducted Production Test
Where architecture provides an RF test port or suitable internal route, conducted testing can provide efficient factory measurements.
Potential parameters can include:
output power
frequency
basic receiver function
Production Test Should Measure What Manufacturing Can Realistically Break.
124 — Radiated Production Test
Some products may use a shielded enclosure or RF test chamber to confirm radiated connectivity.
This can detect problems in:
antenna
matching
assembly
that a purely digital test cannot see.
125 — Wireless Calibration
Certain RF products require calibration for:
power
frequency
antenna tuning
depending on radio architecture.
The goal should be automated and reproducible calibration where possible.
RF Calibration Should Become a Controlled Factory Process.
126 — Device Credential Provisioning
Manufacturing can also provision:
MAC Address
Serial Number
Certificates
Matter Credentials
Cloud Identity
or other required product-specific data.
Each value should be:
unique where required
traceable
verified
Manufacturing Creates Both the Physical Device and Its Digital Identity.
127 — Manufacturing Traceability
A connected product can link:
This becomes valuable during field failure analysis.
Wireless Traceability Connects Factory Data With Field Behavior.
128 — EVT Wireless Validation
Does the Wireless Architecture Work?
EVT should validate:
radio operation
antenna concept
link budget
major coexistence risks
power
firmware connectivity
The goal is to expose architectural problems early.
EVT Is the Time to Discover That the Antenna Location Is Wrong.
Not after tooling.
129 — DVT Wireless Validation
DVT should validate the mature product across:
final enclosure
orientations
temperature
battery conditions
coexistence
real networks
multiple units
pre-compliance
This asks:
Does Wireless Performance Survive the Real Product?
130 — PVT Wireless Validation
PVT asks:
Can Production Reproduce Wireless Performance?
Focus can include:
assembly
antenna match distribution
calibration
identity provisioning
firmware release
traceability
yield
Wireless engineering now becomes manufacturing engineering.
131 — Statistical Wireless Production
Across production volume, engineers can analyze distributions such as:
TX Power
Frequency Error
RSSI Calibration
Antenna Match
Test Yield
Trends can reveal:
component lots
material changes
assembly variation
Wireless Production Data Can Become RF Process Intelligence.
132 — Simulation-to-Measurement Correlation
A mature wireless development loop is:
Wireless Models Become Valuable When They Predict Real Products.
133 — Field-to-Lab Correlation
Field complaints such as:
Poor range in warehouse.
should eventually be translated into measurable engineering conditions.
For example:
signal attenuation
multipath
interference
orientation
Then reproduce them in controlled testing.
Turn Field Experience Into Engineering Evidence.
134 — Wireless Design Reviews
A high-quality program can include:
Is the selected radio appropriate?
Does sufficient theoretical margin exist?
Does the front end support the requirement?
Is the physical implementation viable?
Can multiple radios operate together?
Can the energy target be met?
Are identity and credentials controlled?
Is the market strategy understood?
Can the radio be tested and provisioned repeatedly?
Review the Wireless Product — Not Just the Radio Schematic.
135 — What Does World-Class Wireless System Design Look Like?
At the highest level:
That is the difference between:
and
Engineering a Wireless Product.
Depending on product scope, a 365PCB ODM wireless program may include:
Wireless Requirements Specification
Wireless Technology Trade Study
Wi-Fi Architecture
Bluetooth LE Architecture
Thread Architecture
Matter Architecture
Zigbee Architecture
Sub-GHz Architecture
LoRaWAN Architecture
Cellular Architecture
RedCap Architecture
GNSS Architecture
Multi-Radio Architecture
Frequency / Regional Plan
Link Budget
Range Analysis
Receiver Sensitivity Budget
TX Power Budget
Throughput Budget
Latency Requirements
RF Front-End Architecture
Radio Module / Chipset Selection
Module vs Chip-Down Analysis
Antenna Technology Evaluation
Antenna Placement Strategy
Antenna Matching Architecture
Diversity Architecture
MIMO Architecture
Coexistence Analysis
Receiver-Desense Analysis
Frequency-Planning Analysis
RF Power Architecture
Battery-Life Model
Energy-per-Transaction Analysis
Wireless Firmware Architecture
Network State Machine
Reconnection Strategy
Provisioning Architecture
Secure Boot Inputs
OTA Architecture
OTA Recovery Strategy
Device Identity Architecture
Credential Provisioning Strategy
Cloud Connectivity Inputs
Edge Connectivity Architecture
PCB RF Layout Requirements
RF Stack-Up Requirements
Antenna Keep-Out Requirements
Mechanical RF Requirements
Shielding Requirements
Conducted RF Test Plan
Antenna Tuning Plan
OTA Test Plan
TRP / TIS Test Inputs where appropriate
Coexistence Test Plan
Interoperability Test Plan
EMC Pre-Compliance Plan
Regional Certification Plan
Ecosystem Certification Inputs
EVT Wireless Validation Plan
DVT Wireless Validation Plan
PVT Production Inputs
Production RF Test Plan
Manufacturing Programming Procedure
Identity / Credential Provisioning Procedure
Wireless Traceability Plan
Field Diagnostics Strategy
Component Lifecycle Review
The exact scope should always follow:
Technology + Frequency + Product Environment + Target Markets + Product Risk + Production Volume.
Wireless performance is product-specific. Achievable range, throughput, latency, receiver sensitivity, radiated power, battery life and coexistence performance depend on the radio chipset, RF front end, antenna, PCB, enclosure, firmware, network environment, regional requirements and final product configuration.
We Don't Claim Wireless Range From the Radio Datasheet Alone.
We Evaluate the Complete Wireless Product.
You can begin with:
or simply:
Tell Us Who the Product Needs to Communicate With — How Far, How Fast and for How Long.
365PCB can help translate:
Don't Just Add a Radio.
Choose the Right Wireless Architecture.
Budget the Link.
Engineer the Antenna.
Control Coexistence.
Protect the Receiver.
Optimize the Energy.
Secure the Identity.
Design the Recovery Path.
Validate Over the Air.
Plan Certification Early.
Make Wireless Performance Repeatable in Production.
365PCB RF & Wireless System Design connects:
RF + Antenna + Protocol + Firmware + Security + Mechanical + Regulatory + Manufacturing
into one coordinated product-development process.
A Wireless Product Is Not a Radio Module With an Antenna.
It Is an RF, Antenna, Protocol, Firmware, Mechanical and Regulatory System.
[Discuss Your Wireless Product]
[Submit Your RF & Antenna Requirements]
[Request a Wireless System Engineering Review]
English
