Definition
SAR (Specific Absorption Rate) is defined as the radio frequency power absorbed by a unit mass of biological tissue, measured in W/kg. It combines complex electromagnetic field distributions with human anatomical models and is used to assess whether mobile phones, wearable devices, and short-range communication devices exceed human safety limits at maximum transmission power. Different regions and standards (such as FCC, ICNIRP, and national standards) have differences in averaging time and method, test distance, and posture. Exported products need to be verified separately according to the target market.
Relationship with Module Integration
IoT modules themselves may not undergo separate SAR testing, but once integrated into the complete device, the distance between the antenna and the human body, grounding of metal casing, plastic thickness, and dielectric loss all affect local field strength. Scenarios such as wearable devices, children's watches, and walkie-talkie style gateways often place antennas close to the human body. If SAR pre-assessment is not performed, the maximum transmission power may have to be reduced during certification, affecting the link budget. The design phase should reserve margin for antenna position, orientation, and power control strategy.
Ideas for Reducing SAR
Includes: increasing the distance between the antenna and the human body, using low-dielectric-constant enclosure materials, reducing the duty cycle, triggering power reduction through sensors that detect wearing status, and using directional antennas to deflect the main lobe away from the human body. Attention should be paid to balancing power reduction with network coverage, and verifying that the protocol stack can remain stable in the network after power backoff. Some modules support dynamic power control (TPC) in coordination with SAR sensors.
Integration with OTA and conducted tests
SAR and TRP are not simply oppositional: sometimes, optimizing antenna efficiency to improve TRP may increase SAR when close to the human body. System-level joint optimization should be performed. Passing conducted spurious tests does not mean SAR compliance, because near-field distribution and far-field efficiency differ. Laboratory tests typically use liquid tissue simulants and probe scanning, which are costly, so early simulation and preliminary assessment are recommended.
Engineering Practice Supplement
Regarding 'SAR and exposure limits', it is recommended to differentiate in the architecture between real-time control links and non-real-time telemetry links: control and interlock should preferably be locally closed-loop, while cellular is used for monitoring, configuration, and OTA. On the cloud platform side, observable metrics (connection duration, message delay distribution, error codes, reconnection reasons) are used to drive optimization, rather than just checking 'connectivity'. Documentation delivery includes interface descriptions, error code tables, extreme operating conditions, and known limitations; test delivery includes test cases, data, and conclusions, facilitating third-party re-testing or auditing.
Tissue Averaging and Test Distance
Specific absorption rate divides RF power dissipated in tissue by mass. It describes how much a local part of the body absorbs under close illumination. It does not describe whether base-station coverage is good. Limits usually average over a small mass of tissue, then over a stated time. Markets differ on the averaging mass, the test distance, and the posture. One distance's result on a whole device cannot be applied to another market. A module alone on an evaluation board often cannot produce the final result, because the distance from antenna to skin, the plastic thickness, and the metal ground are decided by the whole device. Cellular is most likely to exceed the limit on the channel at maximum power and the highest duty cycle. That channel is not necessarily the one the service uses most.
The condition of use is a product that, in normal use, comes close to the torso or the limbs. A gateway fixed on top of a metal cabinet, with the antenna far from the body, is usually not limited by exposure. A watch, a handheld gateway used like a radio, or a reader worn against the body has to leave the distance in the mechanical design. Reducing power, reducing power only after wear is detected, and moving the antenna's main lobe are all available. Each of them has to be checked again to confirm the device can still register and still carry the service. Simulation can show where the hot spot is early. Certification still follows the body phantom and distance of the target market.
Power Backoff That Consumes Uplink Margin
The most common miss is discovering an exceedance only just before certification, then cutting maximum power as a temporary fix. Uplink then fails at the cell edge, and that power backoff was never written into the link budget. The second is a test posture that is not the real way the product is worn. The lab distance is a gap on the order of ten millimeters, and the product against the skin has no such gap. The third is to infer an exposure pass from a conducted-spurious pass or a total-radiated-power pass. The near-field hot spot and far-field efficiency are not the same picture. A housing material with a higher dielectric constant moves the hot spot. The old report cannot be reused.
How to measure over-the-air efficiency is in radiated tests and OTA. A shield and a metal ground change both leakage and the field close to the body. After the ground changes, measure the exposure distribution again; see shielding and grounding.
Certification inputs should state the target market, the test distance, the wear posture, and the maximum transmit state allowed. Missing any one of them, the report cannot be reused on the next revision. If power is reduced only when proximity is detected, verify the default when the sensor fails: reduce power, or full power. A default of full power exceeds the limit as soon as the sensor fails. After the reduction, repeat the uplink test in the same posture and confirm registration and service can still complete. Do not look only at an average in the far field. A change of housing material, thickness, or the gap from antenna to shell voids the old hot-spot location. Conducted spurious, port power, and exposure are three records. A pass on the first two cannot stand in for the third. Store the hot-spot location with the mechanical revision. The next change of a screw or a battery well then shows whether a retest is required, instead of a full run every time or no test at all. After a mechanical revision, at least retest the posture that used to be the hot spot.