SAR Testing

SAR Testing of IEEE 802.11a/b/g Devices

Aaron Sargent and Chris Zombolas

Regulators worldwide are making progress toward harmonizing standards covering portable communication devices used close to the body as well as close to the head.

Regulations setting limits for the specific absorption rate (SAR) of radio-frequency (RF) transmitting devices used in close proximity to the human body have been mandated in many countries. Lately, though, the proliferation of wireless local-area network (WLAN) and wireless fidelity (Wi-Fi) devices in portable digital instruments and laptop computers has called for an update of the current SAR testing methodology and procedures. Current methods of measurement were intended only for devices used at the ear and cannot always be applied in the testing of devices meant to be used near other parts of the body.

The broad scope of human-exposure regulations mandated by the Federal Communications Commission (FCC) in the United States and by the Australian Communications Authority (ACA) captures most RF transmitting devices, including devices employing Institute of Electrical and Electronics Engineers standards IEEE 802.11b and IEEE 802.11g (2.4 GHz) and IEEE 802.11a (5.2/5.8 GHz) as well as those used at the ear or other parts of the body. Owing to the lack of harmonized SAR measurement standards covering these devices, and also to the technical difficulties involved with SAR testing above 3 GHz, FCC introduced SAR measurement guidelines in its 2001 document OET 65C 01/01.¹ The FCC SAR measurement guidelines were largely adopted by ACA as part of the Australian regulations limiting human exposure to RF fields from portable RF transmitting devices.

This article describes the application of FCC and ACA SAR measurement methods in the testing of devices employing IEEE 802.11a/b/g, Bluetooth, and similar transmitters. It focuses on issues related to SAR measurements at 2.4–5.8 GHz while providing a general background in SAR testing procedures.

SAR Limits, Regulations, and Standards

SAR is a measure of the rate at which energy is absorbed by a certain unit mass of human tissue, usually expressed in watts per kilogram but also in milliwatts per gram. The rate of temperature increase of the body is proportional to the SAR. SAR limits vary among regulators (see Table I). FCC has a limit of 1.6 W/kg, measured in a 1-g cube of tissue mass, while in Europe the SAR limit is set at 2 W/kg, measured in a 10-g cube of tissue.

The basic European restrictions on human exposure are given in the recommendations of the International Commission on Non-Ionising Radiation Protection (ICNIRP).² The SAR limits are given in the European Norm (EN) standard EN 50360, with measurement specified by the methodology described in EN 50361.^3,4 While European and other international standards set SAR limits for exposure to all or parts of the body, current standards include SAR measurement methodologies only for devices used at the ear, such as mobile phones. No published standards exist for SAR measurements on body-worn devices.

In the United States, FCC specifies both SAR limits and methods of measurement in OET 65C.

In Australia, the ACA electromagnetic radiation standard known as EMR Standard 2003⁵ has adopted the basic restrictions given in the Australian Radiation Protection and Nuclear Safety Agency (ARPANSA) standard,⁶ which has adopted the ICNIRP recommendations for the basic restrictions. The SAR measurement methodologies are given in Schedules 1 and 2 of ACA EMR Standard 2003, which also calls up EN 50361.^7.8 However, Schedule 1 expired on March 1, 2005, and may no longer be used.

Current Measurement Standards. The scope of current standards EN 50361 and IEEE 1528 for SAR measurement has been limited to devices used at the ear, and the applicable frequency range is 300 MHz to 3 GHz.⁹ There are no harmonized measurement standards specifically for measurements of the SAR of body-worn devices, or for devices not used at the ear. For frequencies above 3 GHz, the limitations have been due primarily to difficulties in designing SAR measurement probes of sufficient accuracy and spatial resolution, and, to a lesser extent, to difficulties in formulating tissue-simulating liquids.

While SAR probes are now available to cater to the recently introduced 5.2/5.8-GHz devices, the development of SAR measurement standards is lagging for these frequencies, as well as for devices used in positions other than at the ear. Although regulators such as FCC and ACA have mandated human-exposure standards to limit the SAR of devices used in close proximity to the human body, the lack of appropriate SAR measurement standards for frequencies above 3 GHz has forced the authorities to introduce preliminary methodologies and guidelines that are not as prescriptive as industry standards. The lag in publication of suitable harmonized standards and the lack of SAR probe technology in the 5 GHz band meant that FCC was not prepared to devolve the responsibility for the certification of IEEE 802.11a (5.2/5.8-GHz) devices to its authorized telecommunications certification bodies (TCBs). (Industry generally prefers to certify devices by means of the more commercially driven and less bureaucratic TCB route.)

IEC 62209. For the United States, FCC, while having different SAR limits, generally adopts IEEE/ANSI [American National Standards Institute] standards that aim to be harmonized with the International Electrotechnical Commission (IEC) SAR measurement standards. In Australia, ACA has indicated that it will adopt international SAR measurement standards as they become available. This should lead to the adoption of IEC 62209-1, already published, and IEC 62209-2 (now in draft), when it is published.^10,11

The recently published IEC 62209-1, applicable for devices used at the ear and operating within the frequency range of 300 MHz to 3 GHz, will probably be adopted as an EN standard to replace EN 50361.

The imminent publication of IEC 62209-2 relating specifically to body-worn devices will extend the scope of standards for measurement of SAR to the frequency range of 30 MHz to 6 GHz. It will include an accurate and reproducible methodology for SAR measurement for two-way radios, palmtops, laptops, desktop computers, and body-mounted wireless communication devices, designed for determining compliance with localized SAR limits in RF human-exposure standards. (The actual exposure limits are set by other standards.) The standard will be applicable to both public and occupational exposures, and will cater to concurrent exposures from multiple radio devices. IEC 62209-2 is intended to be used in conjunction with IEC 62209-1 and will provide detailed specifications for SAR measurement instrumentation and protocols and for phantom simulation of the human body for devices not used at the ear.

Figure 1. A diagram of the SAR system shown in Figure 2.

Test Equipment and Measurements

Most commercial SAR systems use small, precision, isotropic electric-field (E-field) probes to scan a phantom containing tissue-simulating liquid. A precision multiaxis robot positions the probe within the phantom at a large number of grid points. A precision positioner allows the handset to be placed underneath the phantom, facilitating scanning of the entire volume of the phantom (see Figures 1 and 2).

SAM Phantom. The specific anthropomorphic mannequin (SAM) is the phantom specified for devices used at the ear and is called up by EN 50361, IEEE 1528, and IEC 62209-1. The SAM head phantom is based on selected dimensions from a database of human males. Thickness of the phantom shell must be 2.0 mm ± 0.2 mm except at the ear reference point, where it must be 6.0 mm ± 0.2 mm. The material must be of sufficiently low loss and low permittivity. The SAM head phantom is also specified by the FCC and ACA Schedule 2 measurement methods.

Figure 2. A typical commercial SAR system, the DASY4, from Schmid & Partner Engineering (Zurich, Switzerland).

Body Phantoms. A flat phantom is specified for body-worn devices. It must be constructed from low-loss dielectric material of 2.0-mm thickness and have a uniformity of ±0.2 mm. ACA and FCC require that the length and width of the flat phantom be at least twice the corresponding dimensions of the device under test (DUT), including its antenna. The liquid depth should be 15.0 cm ± 0.5 cm. The flat-phantom dimensions for IEC 62209-2 are not yet defined, but they are likely to be smaller than the FCC and ACA dimensions.

Tissue-Simulating Liquid. The dielectric properties of brain- and muscle-tissue-simulating liquids are shown in Table II.

The tissue dielectric values for the center frequency of the transmission band should be within 5% of the target value. Linear interpolation is used to obtain the dielectric properties at other intermediate frequencies. Examples of recipes for obtaining the stated values are given in the relevant standards. Deviations of the permittivity (e) and conductivity (s) must be within 5% of the target values.

The standards EN 50361, IEC 62209-1, and IEEE 1528 do not include SAR testing of body-worn devices within their scope. Therefore, these standards do not give dielectric parameters for muscle tissue. Muscle-tissue-simulating liquids are used for body-worn devices and all devices not used at the ear. The dielectric properties for muscle-tissue-simulating liquid are given by current FCC and ACA regulations and will be provided in the new version of IEC 62209-2 when it is published.

The values in Table II should be linearly interpolated when testing at other frequencies within the band. As it is difficult to achieve the 5% tolerance at some frequencies, FCC and ACA allow the use of 10% tolerance as an interim measure. This relaxation is generally given only for frequencies in the 5 GHz band.

SAR Scans. A typical SAR scan comprises a power reference measurement, surface check, area scan, zoom scan, and power drift measurement.

Figure 3: Typical area scan for handset.

For the power reference measurement, a single E-field measurement is performed in a defined position above the DUT. This is done at the end of the test in order to check for source drift. If power drift is excessive, the measurements should be repeated or the uncertainty should be reevaluated. Either optical or mechanical surface detection is used to identify and map the phantom surface with a high degree of precision. This is the surface check.

In the area scan, a coarse measurement grid is used to determine the approximate location of the peak SAR values. Figure 3 shows a typical area scan for a head position, and Figure 4 shows typical area scans for a laptop position. The zoom scan is a fine-resolution volume scan performed at the peak SAR location determined previously by the area scan. The zoom scan volume is used to evaluate the spatial-peak average SAR for a 1- or 10-g cube of tissue. Following the zoom scan measurement comes performance of an extrapolation from the closest measured points to the surface and an interpolation to a finer resolution between all measured points (see Figure 5).

Figure 4. Typical area scan for a laptop PC.

Measurement System Verification. FCC OET 65C and ACA EMR Standard 2003 Schedule 2 verification tests are identical. They are necessary for checking and tracking the performance of the SAR measurement system. A flat phantom and precision dipole radiating source are specified to determine whether the measurement system meets its performance requirements. The verification frequency must be within 100 MHz of the device transmit frequency. System verification checks should be performed daily within each operating frequency band of the test sample. The measured 10-g SAR must be within 10% of the expected target values for the specific phantom and validation dipole transmit antenna. System verification must be performed 24 hours prior to a compliance test in order to ensure that any setup or system errors are identified. The general verification setup is shown in Figure 6.

E-Field Probe Design. The E-field probe must be a minimally perturbing structure containing either single or multiple electrically small E-field sensors, along with the components necessary to transform the sampled RF signal into a proportional direct current or voltage. An elemental E-field probe consists of a thin dielectric substrate containing a sensor, such as an electrically short (in tissue) dipole; a diode to rectify the RF signal; and a balanced high-resistance transmission line to extract the rectified signal. The dipole elements and the high-resistance leads are usually applied to the substrate by means of thin-film techniques. Diodes and any other discrete components are bonded to the thin-film elements.

Figure 5. A SAR area scan and zoom scan during a 900-MHz validation.

An isotropic probe consists of three such devices arranged in an I beam (or H beam), or a delta (triangular) beam configuration—the latter being preferable—with the axis of each dipole orthogonal to the axes of each of the others, for example, aligned along the diagonals of a cube (see Figure 7).

Probe output depends on the following factors:

• Field polarization and the direction of incidence.

• Local field gradients at the vicinity of a measurement point where the probe tip is located.

• Media dielectric properties and boundaries between media in the vicinity of the probe.

• The frequency, modulation, and power level of the source, that is, the DUT.

• Interfering field sources, such as noise, static fields, extremely low-frequency fields, and so on, from the ambient and nearby equipment.

Figure 6. General SAR measurements system verification setup.

In addition, other physical influences, including temperature and humidity, can affect probe performance.

In compliance test applications, the highest integrated SAR value is the main concern. These values can be found mostly at the inner surface of the phantom and cannot be measured directly because of the sensor offset in the probe. To extrapolate to the surface accurately, the distance between the probe tip and the surface must be known with some accuracy. Most E-field probes employ both optical and mechanical surface-detection methods to resolve these issues.

Figure 7. Typical E-field probe structure.

The advantage of optical surface detection is that it measures the surface distance from the center of the probe tip. This gives the correct and precise distance to the phantom surface, even for angled surfaces. This is important for head scans where the surfaces are irregular. Unless the precise distance from the surface to the probe tip is known, the SAR measurement uncertainty will be degraded.

Mechanical surface detection uses the proximity sensor built into the probe. In angled surfaces, the probe detects the surface and then moves back until the touch condition is cleared. However, there will still be some uncertainty in the final probe position if the angle of incidence is not known. This method results in increased uncertainty for head phantoms but is not a problem for flat phantoms. Mechanical surface detection is highly repeatable when used with a flat surface.

Probe Uncertainty. The degree of E-field probe uncertainty depends on several factors, including:

Figure 8. Hemispherical and axial isotropy.

• Axial isotropy, which is the ability of a field probe to respond to a transverse electromagnetic (TEM) plane wave impinging from a direction along the probe axis equally regardless of axial orientation (see Figure 8).

• Hemispherical isotropy, which is the ability of a field probe to respond equally to TEM plane waves impinging, with arbitrary polarization, from directions around the surface of a hemisphere (see Figure 8).

• Spatial resolution, which is the ability of a probe to discriminate between two SAR peaks in close proximity. For spatially averaged SAR measurements (up to 2.45 GHz), the error is generally small as long as the probe tip diameter is less than 8.0 mm. At 5.8 GHz, the tip diameter should be at least 3 mm.

• Boundary effect, where, in the vicinity nearest to the inner surface of the phantom shell, the sensitivity of the probe deviates from that defined under normal calibration conditions. This occurs because of the capacitive coupling between the probe and the medium boundary at the liquid and phantom-shell surface. The effect can be largely compensated.

• Linearity error, a category that encompasses errors from the assessment and compensation of the diode compression effects for continuous-wave and pulsed signals with a known duty cycle.

• System detection limits, which are a source of uncertainty when the measured field strength is too close to the detection limit of the probe.

• Signal response time, evaluated as the time required by the system to reach 90% of the expected final value after cycling of the power source.

• Integration time, which (along with the discrete sampling intervals used in the probe electronics), if not synchronized with the modulation characteristics of the measured signal, may mean that the RF energy at each measurement location is not fully or correctly captured.

• Probe positioning, the procedures for which affect the tolerance of the separation distance between the probe tip and the phantom surface.

Influences on the Measured SAR

A number of things can influence the results of SAR measurements, including probe calibration and response, the properties of tissue-simulating liquids, and test sample positioning.

The SAR is the measure of the rate of energy absorption per unit mass of tissue at a specific location in the tissue medium. This absorption rate is proportional to the rate of temperature increase in a tissue, as well. It is calculated as follows:

SAR = σ/ρ • |E|²,

where σ is the conductivity of the tissue in siemens per meter, ρ is the mass density of the tissue in kilograms per cubic meter, and E is the root-mean-square electric-field strength in volts per meter.

The Actual Body. In most handheld transmitters, the antenna radiates within 1–2 cm of the user's head or body. Even at low power levels, relatively high field strengths would be expected near the antenna. The actual field strength is highly dependent on the location, orientation, and electromagnetic characteristics of adjacent objects, including the user's body.

The user of the handset is normally in the reactive near-field region of the antenna where the electromagnetic field is mostly nonpropagating. The energy absorbed by the user mainly derives from the electric fields induced by magnetic fields generated by the current flowing along the antenna and other radiating structures of the device.⁹

The RF energy is scattered and attenuated as it propagates through the body tissues. Maximum energy absorption occurs in the more absorptive high-water-content tissues near the surface of the head and near the surface of other parts of the body. To account for near-field energy coupling effects, portable transmitters are evaluated with dielectrically realistic head and body models—the phantoms.

High-Frequency Measurement Issues. SAR testing of portable computers fitted with IEEE 802.11b and 802.11g (2.4-GHz) WLANs is practical with current SAR probes and tissue-simulating liquids. Now, the recent proliferation of WLAN devices using IEEE 802.11a protocol has introduced a need for practical SAR measurement equipment operating over the frequency bands 5.2 GHz and 5.8 GHz. Unfortunately, such high frequencies exacerbate a number of problems that also exist in the lower frequency ranges but are not so troublesome there.

The main measurement issues at 5.2 and 5.8 GHz are extremely steep field gradients close to the phantom surface, the physical limitations of current E-field probe design, and the absence of published recipes for tissue-simulating liquid for frequencies above 3 GHz. These have the consequence of greater measurement uncertainty. This increase in uncertainty is due to the availability of only imprecise extrapolation, interpolation, and integration algorithms for maximum-SAR evaluation and to the degraded positional repeatability of the DUT with respect to the test phantom.

Compliance Testing

In general, compliance testing of portable RF transmitting devices must be performed if exposure is at a range of less than 20 cm from the body and the power threshold is exceeded.

Portable Equipment. WLAN and Wi-Fi transmitters installed in laptop computers and handheld devices are classified as portable equipment under the general-public/uncontrolled-exposure category. They often exceed the power threshold specified by the ICNIRP, FCC, and ARPANSA requirements and might be used within 20 cm of the human body. Because it is not possible to prove their compliance by checking against the reference levels, an SAR test most likely will be necessary. The WLAN standards IEEE 802.11a/b/g cover the frequency ranges and modulations outlined in Tables III and IV.

If a laptop computer or portable digital device incorporates multiple transmitters, then each mode must be assessed. Device modulation, power output, and duty cycles must be taken into account to ensure that the device is operating in a worst-case mode. The intent is to make sure that the SAR evaluation is conservative.

EMR compliance of multifunction, multiband devices is becoming very difficult to achieve. Some personal portable computers use three bands—for example, 2.45, 5.2, and 5.8 GHz—and some also include the Bluetooth function. If each WLAN module has an antenna, then the cumulative SAR levels should be determined. Using preapproved transmitter modules does not guarantee compliance, because it is not known how the original module was configured for the initial compliance test. In some cases, different antennas are used, or different modules using different modes may operate simultaneously.

Due to near-field coupling effects, small changes in device positioning can lead to unexpected changes in energy absorption. In the case of devices used at the ear, the effect of device positioning on the SAR levels must be considered. Current published standards such as EN 50361, IEC 62209-1, and IEEE 1528 precisely define the positioning relative to the ear.

FCC OET 65C and ACA EMR Standard 2003 specify procedures for SAR measurements of devices not used at the ear. The phantom dimensions are related to the DUT dimensions and the wavelength of the DUT transmitter.

Two-Way Radios and Portable Phones. Portable telephones operating at 2.4/5.2/5.8 GHz are now commonplace. To comply with European (ICNIRP) and ACA requirements, these phones must be tested at the head and in the body-worn position if supplied with hands-free kits and exceeding 20 mW. The power threshold for FCC compliance is slightly higher.

Figure 9. The belt-clip position for testing hands-free portable telephone.

In the belt-clip position, the transceiver is placed underneatha flat phantom and suspended such that the belt clip touches the phantom (see Figure 9). If the device incorporates a headphone socket, then testing with the hands-free earpiece/microphone connected is required.

Figure 10. The face position for testing two-way PTT radios for compliance.

To test a PTT (push-to-talk) two-way radio in the face position, the device is placed 2.5 cm from the phantom (see Figure 10). This position is deemed equivalent to the device being placed in front of the nose during use.

Laptop and Tablet Computers. SAR test methods are constantly being updated by FCC. The OET 65C guidelines define a variety of test positions for laptop and tablet personal computers (PCs).

Figure 11. The lap-held position simulates a tablet PC on a person's lap.

The lap-held position simulates a laptop/tablet PC being used on a person's lap (see Figures 12 and 13). This position provides a conservative estimate of the actual SAR that a typical user would experience. The bottom side of the laptop is pressed against a flat phantom.

If a tablet PC has display-mounted antennas and an interactive screen display, then the arm-held (interactive-display) test configuration may be applicable (see Figure 14). The face of the tablet screen is pressed against the flat phantom. This position is designed to evaluate forearm exposure.

Figure 12. The lap-held position for testing a tablet PC.

The edge position simulates the use of tablet PCs, which are held along the edges of the screen surround (see Figures 15 and 16). The location of antennas in these regions necessitates SAR compliance. Hand exposure testing is not typically required, but such positions are designed to evaluate forearm exposure.

In addition, the back-of-lid position can be employed to account for occasional exposure to the arm or torso region.

Figure 13. The arm-held position for testing a tablet PC.

It is not generally necessary to test for exposure to the hands or wrists, because the SAR limit for these areas is much higher. The highest measured SAR levels generally correspond to the location of the antennas and their proximity to the body.

Figure 14. Lap-held position for testing a tablet PC.

Australian Compliance Requirements. In Australia, the procedures for testing laptop and tablet PCs for compliance are prescribed in Schedule 2 of the ACA EMR standard. The need to evaluate the SAR depends on where the antenna is located on the portable PC—on the bottom, along the front, or on the lid—and whether it can operate with the lid closed.

Figure 15. An edge position test setup   for a tablet PC.

When a laptop is used on the lap in what is considered a typical use position and the antenna is located along the front of the device, then exposure would occur at less than 20 cm from the stomach or groin. For this reason, SAR compliance  is necessary for WLAN devices, laptops, tablet PCs, and the like.

Table S1 in the ARPANSA standard grants a test exemption for general-public exposure (unaware users) when the mean power is less than the threshold level in Table S2 and the separation between the device and the body is more than 20 cm.6 Table S2, listing threshold levels for testing, gives the nominal mean output power, expressed in watts, for the frequency range 450–1500 MHz as 3150 divided by the frequency in megahertz. Thus, at 2500 MHz, the threshold mean output power would be 3150/2500 W, or 1.26 W.

This exemption holds only if the manufacturer can prove that it is not possible for the device to be within 20 cm of the body in normal use. If the device is used at a distance closer than 20 cm and the power threshold of Table S2 is exceeded, then the only option is to carry out an SAR test as if testing for compliance. Also, in the case of a device used within 2.5 cm of the body and having output power exceeding 20 mW, SAR evaluation must be performed.

Figure 16. An edge position test setup for a tablet PC.

High-Frequency Measurement

Because of the steeper gradients of the field distribution in the 5–6 GHz frequency range, improved scanning methods and better dosimetric E-field probes are required. The draft standard IEC 62209-2 requires that probes exhibit the following minimal performance:

• A probe tip diameter no greater than 8 mm for frequencies below 2 GHz and no greater than 16 mm divided by the frequency in gigahertz for frequencies above 2 GHz (for example, no larger than 2.8 mm for 5.8 GHz).

• Maximum sensor displacement from the bottom of the probe no greater than 4 mm for frequencies below 2 GHz and no greater than 8 mm divided by the frequency in gigahertz for frequencies above 2 GHz (for example, no farther than 1.4 mm for 5.8 GHz).

A typical 5–6-GHz probe has a smaller tip diameter than a standard 900/1800-MHz E-field SAR probe, and the sensors are located 2 mm from the tip. Leading manufacturers of SAR equipment are currently developing 5–6 GHz probes with tips less than 2 mm in diameter and a distance from probe tip to sensors of approximately 1 mm. The major benefit of an optimized probe is better spatial resolution.

5–6 GHz Issues. Tissue-simulating liquids used in SAR testing at 5–6 GHz have to be made of special materials because of the unsuitability of the usual base ingredient, water, at such high frequencies. Formulating the dielectric properties required for tissue simulation at 2.4/5.8 GHz presented challenges. Much experimentation and research is necessary to identify nontoxic, environmentally friendly liquids.

The strong decay of the fields in the frequency range of 5.2–5.8 GHz requires that the zoom scan cube be physically smaller than recommended in current OET 65C guidelines. The x-y-z dimensions of a 5–6 GHz measurement cube are 30 X 30 X 21 mm, with 7 X 7 X 8 points. This compares with the recommended 30 X 30 X 30 mm with 7 X 7 X 7 points. Figures 17 and 18 diagram the x-z plane only for standard and high-frequency zoom scans.

Figure 17. Zoom scan for 900 MHz (a). Zoom scan for 5200/5800 MHz (b).

The parameters for the 5–6 GHz cube represent a trade-off between time and accuracy.

Measurement Uncertainty. One of the major factors influencing measurement uncertainty in the 5–6 GHz band is the rapid decay of fields close to the phantom surface. Figure 18 shows the decay of E-fields from the phantom surface in the frequency range of 450–5800 MHz.

Figure 18. Comparison of E-field penetration depth.

The data displayed in Figure 18 were derived from the results of system validation using precision dipoles at 450, 900, 1800, 2450, 5200, and 5800 MHz. The graph illustrates the steep field gradient present in the frequency range of 5–6 GHz. If a standard cube of 30 X 30 X 30 mm were used at 5800 MHz, the SAR level at the top edge of the cube (the z-axis) would be approximately 50 dB down from the peak. This would cause gross errors in the final averaged SAR value.

The steep field gradients increase measurement uncertainty in several ways. One involves device positioning. Small variations in device placement can greatly affect the final spatial-averaged SAR level. A variation of a few millimeters can reduce the final averaged SAR level by 4 dB (approximately 130%) in the 5.8 GHz range, whereas the same variation in the 2450 MHz band would correspond to about 1 dB (a reduction of approximately 25%).

The steep gradients increase the level of error in the interpolation, extrapolation, and integration algorithms used to calculate the final spatial-averaged SAR. The algorithms may add uncertainty due to general assumptions about field behavior; therefore, they may not perfectly predict the E-field distribution in the tissue. The uncertainty is directly proportional to the spatial resolution chosen in the measurement and postprocessing methods. It is increased from only 1% for frequencies below 3 GHz, but increases to 20% for 5–6 GHz measurements. All SAR standards limit the maximum length of a scan to 30 minutes, so the accuracy of the measurement uncertainty is time dependent. Newly marketed optimized field probes have reduced the uncertainty.

Also, the boundary effect is increased from 1% for frequencies less than 3 GHz to 2% for 5–6 GHz measurements. The increase is due to placement of the probe tip closer to the phantom surface.

Probe positioning is the final factor. Owing to the small tip diameter, it is difficult to incorporate optical surface detection capability in the probe tip. With curved surfaces, use of mechanical surface detection rather than optical surface detection increases the uncertainty by approximately 2.5%. This does not have a significant effect on a flat phantom.

SAR measurement standards limit the measurement uncertainty to a maximum of 30%. While this has been achievable below 3 GHz, only the newly developed probes meeting the aforementioned specifications for the 5.2–5.8 GHz frequency range will give a measurement uncertainty of less than 30%.

Conclusion

Despite the complications added to the testing process, SAR testing of 5.2- and 5.8-GHz portable devices has proven to be practical, with only a slight increase in the measurement uncertainty. The electrically smaller probes being designed by major manufacturers of SAR equipment enable measurement uncertainties at these frequencies to come in below the 30% maximum.

While no harmonized standards now exist for frequencies above 3 GHz, FCC guidelines are constantly improving to meet the challenge of the new and more demanding SAR measurement requirements. Also, the necessary work is currently under way among international standards bodies. It is expected that internationally harmonized testing methods and procedures will soon be available to cover both head- and body-mounted devices operating at frequencies up to 6 GHz.

References

1. Federal Communications Commission, "Evaluating Compliance with FCC Guidelines for Human Exposure to Radiofrequency Electromagnetic Fields," OET Bulletin 65, ed. 97-01; supp. C, ed. 01-01 (Washington, DC: Office of Engineering and Technology, FCC, 2001).

2. International Commission on Non-Ionising Radiation Protection, "Guidelines for Limiting Exposure in Time-Varying Electric, Magnetic, and Electromagnetic Fields (up to 300 GHz)," Health Physics 74 (1998): 494–522.

3. EN 50360:2001, "Product Standard to Demonstrate the Compliance of Mobile Phones with the Basic Restrictions Related to Human Exposure to Electromagnetic Fields (300 MHz–3 GHz)" (Brussels: CENELEC, 2001).

4. EN 50361:2001, "Basic Standard for the Measurement of Specific Absorption Rate Related to Human Exposure to Electromagnetic Fields from Mobile Phones (300 MHz to 3 GHz)" (Brussels: CENELEC, 2001).

5. "Radiocommunications (Electromagnetic Radiation—Human Exposure) Standard 2003" (Melbourne: Australian Communications Authority, March 2003).

6. "ARPANSA Radiation Protection Standard No. 3: Maximum Exposure Levels to Radio-Frequency Fields—3 kHz to 300 GHz" (Sydney: Australian Radiation Protection and Nuclear Safety Agency, 2003).