Maximizing multi-antenna performance in dense industrial deployment
KEY TAKEAWAYS:
- RF coexistence is a primary system challenge
- Antenna strategy is the biggest performance lever
- Robust performance requires coordinated design
A single definable area can host multiple endpoints, each vying for a strong connection. Environmental sensors, access control readers, asset tracking tags, condition monitoring nodes and mobile robots all rely on good signal strength and low noise.
Each connected device may feature Bluetooth Low Energy for provisioning, Wi‑Fi for high-rate data, a sub-GHz link for long-range coverage, and a cellular or 5G modem for fallback. The challenge is straightforward but often underestimated: how do you pack all these radios into a tight enclosure and still deliver reliable performance in a noisy industrial environment?
Multiple radios operating in and around the 2.4 GHz ISM band are the source of the problem. Bluetooth, Wi‑Fi, proprietary links and even microwave ovens share the same spectrum. Protocol features such as adaptive frequency hopping and carrier sense exist because of this congestion. Heavy traffic raises the noise floor and creates long bursts of occupancy on popular channels. Packet loss, latency spikes and reduced throughput may only appear when the site is fully populated with devices.
Adjacent channel and blocking effects are a second, more subtle problem. A high-power Wi‑Fi or 5G transmission can desensitize a nearby receiver even when its center frequency is in a different band. Imperfect filters and limited dynamic range mean that strong out-of-band signals still drive current into the low-noise amplifier (LNA). If the front end is marginal, the LNA or mixer compresses, and the sensitivity of the low-power link collapses just when it is needed most.
Inside the device, self-interference is unavoidable. RF paths share ground, supply rails and enclosure space. Power amplifier harmonics, digital clocks, DC/DC converters and common mode currents on cables all couple into antenna feeds. High-power gateways (e.g. cellular or industrial radios) may juggle several watts of RF power on one band and tiny sensor signals on another. Isolation and filtering become critical. Even in simple sensor nodes, a poorly routed antenna feed or noisy converter can swamp a weak sub-GHz signal.
The industrial setting makes all this worse. Metal racks, cable trays, HVAC ducts and machinery create a rich multipath environment. Reflections cause deep fades and rapid spatial variations in signal strength. Workers, forklifts and doors add moving obstacles. A node that sails through lab tests can behave unpredictably when installed on a steel column or inside an electrical cabinet.
To tame this complexity, coexistence needs to be treated as a primary design requirement rather than a certification afterthought. That starts with antenna choices.
Antennas for IoT deployment come in many forms
A selection of LTE MIMO antennas, Wi-Fi® MIMO antennas, plus GPS/GLONASS antennas demonstrates the design variety available to engineers. [Source: Taoglas]
Antenna design for compact multi-radio industrial IoT end points
The antenna can be the most compromised component in an industrial IoT design. Mechanical and visual constraints usually dictate enclosure dimensions and mounting locations long before an RF engineer is involved.
The temptation is to drop in a reference PCB or chip antenna and assume the datasheet performance will carry over. In practice, efficiency and radiation patterns are dominated by ground plane size, surrounding materials and the proximity of other antennas.
PCB trace antennas are the low-cost, no BOM option. They can offer solid performance when laid out on the recommended board size with a clean ground reference. Performance collapses when the ground plane is shrunk, cut by slots or crowded with tall components. On multi-radio boards, meeting all the “keep out” requirements for several trace antennas can be impossible.
Chip and ceramic antennas trade some efficiency for repeatability and a small footprint. They still demand a specific ground plane size, clearance zone and matching network, but they are less sensitive to minor layout variations. For 2.4 GHz and 5 GHz Wi‑Fi and Bluetooth links in small sensor nodes, they are a common choice. Their controlled bandwidth can simplify matching and coexistence design.
Flexible printed antennas add another degree of freedom by moving the radiating element away from the main PCB. They can be bonded to the inside of a plastic enclosure, wrapped around a non-metallic standoff or placed behind a logo window if the material is RF-transparent. This is particularly helpful when the main PCB is buried among metal objects or must be very small. Flex antennas also make it easier to add a dedicated sub-GHz radiator while keeping the main board compact.
External whip or puck antennas are still common on gateways, access points and outdoor nodes. They provide higher efficiency, clearer radiation patterns and better separation from internal noise sources. The trade-offs are mechanical and visual impact, cable and connector losses, and the risk of damage or mis‑installation in harsh sites.
Multi-band and multi-antenna strategies
In multi-radio designs, the key question is whether to share antennas across bands or provide separate ones. A single wideband antenna that covers 2.4 GHz and 5 GHz (and sometimes sub-GHz harmonics) simplifies mechanical design and reduces BOM items, but it forces compromises in matching and may increase coupling between radios. Separate antennas deliver better efficiency and isolation but consume more space and require more feeds and switches.
Table 1: What to consider when selecting an antenna for IoT applications
| Aspect | Single wideband antenna | Seperate band-specific antennas |
|---|---|---|
| Footprint and BOM | Minimal board area, single feed and matching network, lowest component count | Larger board area required, multiple feeds and matching networks, higher component count |
| Antenna efficiency | Compromised efficiency across bands; hard to optimize for both 2.4 GHz and sub-GHz | Better efficiency per band; each antenna can be tuned for a specific frequency and bandwidth |
| Inter-radio coupling | Higher risk for mutual coupling and interference between radios sharing one feed | Physical and electrical separation reduces coupling and makes isolation targets easier to hit |
| Matching complexity | Complex wideband matching; sensitive to component tolerances and layout changes | Simpler narrowband matching per antenna; more predictable and repeatable performance |
| Cost and time to market | Lower device and assembly cost but may need more RF tuning and validation cycles | Higher device cost, but RF behavior is more predictable, which can shorten debug and validation |
Comparison of single wideband antennas versus separate band-specific antennas in dense multi-radio industrial IoT designs.
Sub-GHz links often benefit from a dedicated antenna. At 868 or 915 MHz, where physical size and ground reference dominate performance, a well-tuned quarter wave whip or carefully engineered compact antenna can yield several dB of link budget improvement over a squeezed-in compromise. That translates directly into range, penetration through walls and battery life.
Spatial separation and polarization diversity help reduce mutual coupling. As a rule of thumb and where space allows, aim for at least a quarter wavelength of the lowest operating frequency between antennas, use orthogonal placement where possible and keep feeding structures and grounds symmetrical. In small nodes this is difficult, but even modest separation and careful orientation can deliver measurable coexistence gains.
RF front ends, filtering and PCB layout for robust multi-radio IoT nodes
Once the basic antenna strategy is set, attention shifts to the RF front end and PCB implementation. Integrated front-end modules now combine power amplifiers, low noise amplifiers, switching and filtering in compact packages. They are widely used for Wi‑Fi and IoT radios because they save board area and take care of much of the precision RF work.
For dense multi-radio systems, several characteristics stand out:
- Linearity and blocking performance. Strong out-of-band signals can drive the LNA into compression or generate intermodulation; a higher third-order intercept indicates resilience to these effects. A more linear front end is less likely to generate intermodulation products that fall into another radio band.
- Integrated filtering. Built-in bandpass filters and harmonic traps limit spurious emissions and help prevent interference with neighboring bands. They may need to be combined with external filters to meet coexistence goals.
- Switching and isolation. Low-loss, high-isolation transmit/receive switches or antenna multiplexers enable antenna sharing while keeping leakage between chains under control. In some gateways, RF switches let several radios share a high gain external antenna without excessive crosstalk.
Discrete power amplifiers (PAs), LNAs and filters offer more control over performance but consume more area and design time. They make sense when output power, noise figure or special filtering requirements push beyond what integrated parts can offer. For most compact industrial nodes, well-chosen integrated devices strike a good balance.
Filtering strategy is tightly tied to antenna sharing decisions. Dedicated SAW or BAW filters on each 2.4 GHz and 5 GHz path can greatly reduce mutual interference. Where sub-GHz radios coexist with cellular modems, narrowband filters are often essential to stop modem harmonics from landing in the sub-GHz receive band. At system level, diplexers can let two radios share a single antenna while operating in widely separated bands.
Layout is where many coexistence ambitions rise or fall. Practical guidelines include:
- Keep RF traces short, straight and on a single layer where possible, with controlled impedance and solid ground underneath.
- Avoid routing high speed digital lines or noisy DC/DC switch nodes parallel to antenna feeds or LNA inputs; if crossings are unavoidable, make them at right angles.
- Use plenty of stitching vias around RF traces, filters and front-end modules to confine return currents and reduce coupling into adjacent circuits.
- Reserve clean ground and supply domains for the most sensitive receivers and consider low noise linear regulators for critical analog or RF blocks even in primarily switch mode systems.
Thermal management matters too. Concentrating several power amplifiers and front-end modules near one edge can cause local heating, changing device characteristics and detuning matching networks over temperature. Spreading heat sources, using thermal vias and providing adequate copper under power devices improves stability.
System-level RF coexistence strategies for large smart building deployments
Even with solid antennas and hardware, dense multi-radio deployments benefit from coordination at system level. Firmware and network planning choices can have as much impact on reliability as another dB of link margin.
The simplest lever is time domain scheduling. Radios share information about planned transmissions so that high duty cycle links do not trample over latency sensitive ones. A Wi‑Fi video stream can be throttled or briefly paused when a critical control message must be sent over a low-power, sub-GHz channel. On a single node, a coexistence manager in firmware can arbitrate access to a shared antenna or front end, making sure only one high power transmitter is active at any instant.
Frequency planning at deployment time adds another layer of control. Selecting Wi‑Fi channels that minimize overlap with neighboring networks, reserving specific channels or bands for safety critical traffic and exploiting sub-GHz channels with lower congestion all help. In large buildings, it pays to treat spectrum as a shared resource across tenants and systems rather than letting every installer pick channels in isolation.
Adaptive data rate and power control techniques further aid coexistence. Many protocols already support dynamic modulation and coding. Extending this to explicitly consider interference levels allows nodes to back off transmit power when conditions are good and save headroom for tougher scenarios. Lower average transmit power reduces both self-interference and overall RF pollution.
Diversity, beamforming and advanced RF techniques for dense IoT networks
Spatial and polarization diversity can greatly improve link robustness in reflective environments. Even a basic two antenna diversity scheme at the gateway, with appropriate separation, can mitigate deep fades caused by multipath. For Wi‑Fi and cellular links, MIMO and beamforming are now emerging in a few advanced gateway systems; mechanical and mounting decisions should let these techniques work rather than forcing antennas into cramped corners.
More advanced approaches are beginning to appear in research and high-end deployments. RF sensing and machine learning algorithms can monitor spectrum occupancy, interference signatures and traffic patterns, then adjust channel assignments or scheduling in real time. Location aware beamforming can direct energy toward specific zones or devices, reducing unwanted interference elsewhere. These ideas are still rare in small sensor nodes, but they point to where gateway design is heading.
Testing RF coexistence and antenna performance in real industrial environments
The last step is verification. Many coexistence problems only show up in the field, but a structured test plan can catch the worst issues before rollout.
Start with conducted tests to characterize transmitter spectra, receiver sensitivity and blocking performance for each radio path. Confirm that power amplifiers meet mask and harmonic requirements, filters behave as intended and LNAs are not easily driven into compression by external signals.
Move quickly to over the air testing in anechoic or mode stirred chambers, even if they are small. OTA measurements reveal antenna efficiency, radiation patterns and the combined effects of enclosure and mounting hardware. For multi radio devices, measure not only absolute performance but also how one radio activity affects another throughput or packet error rate.
Recreate realistic coexistence scenarios by combining devices and infrastructure elements in the test environment. Run high throughput Wi‑Fi traffic, continuous Bluetooth connections and intermittent sub-GHz messaging at the same time while moving antennas or test devices through representative positions. Track metrics such as latency, retry counts and effective throughput, not just low-level RF figures.
Finally, plan pilot deployments in the harshest parts of the intended environment: near machinery, in service risers, in long corridors with many fire doors or around heavily occupied workstations. Instrument endpoints and gateways to log RF conditions and performance over time. Feedback from these pilots can guide adjustments to antenna placement, channel plans and firmware coexistence policies before scaling across an entire building or plant.
By treating RF and antenna performance as a shared responsibility across hardware, firmware and deployment teams, engineers can build industrial IoT and smart building networks that keep working long after the installation crew has left site.
Read more about RF design in IoT systems:
- Antennas for RF Designs in the IoT
- 4 Keys to Selecting Antennas For Small IoT Devices
- Putting the smarts into building automation