8.10 ANALYTICAL AND COMPUTATIONAL METHODS
Direct measurement is often the preferred way to characterize an accessible RF field, but it is not the only assessment method. Conservative analytical calculations may be sufficient for simple sources and geometries. Numerical computation becomes valuable when the source, field region, surrounding environment, exposure geometry, or required dosimetric quantity cannot be represented adequately by a simple calculation or direct measurement.
The method should be no more complex than necessary, but it must be capable of answering the assessment question with adequate confidence. A more sophisticated model is not automatically more accurate. Its reliability depends on the validity of its assumptions, the quality of its input data, the correct implementation of the method, and appropriate verification, validation, convergence, and uncertainty evaluation.
8.10.1 Analytical And Simplified Models
Analytical methods use established electromagnetic relationships to calculate fields or exposure quantities for sources and geometries that can be represented mathematically. Examples include far-field calculations based on equivalent isotropically radiated power, antenna gain or pattern, feed losses, distance, and duty factor, together with simple models of dipoles, loops, apertures, and plane waves.
These methods are fast, transparent, and useful for screening, defining preliminary exclusion boundaries, locating likely field maxima, and checking more detailed measurements or simulations. They can deliberately use conservative source powers, gains, or geometries so that a result below the applicable criterion provides a clear compliance margin.
A simplified result is valid only within the assumptions of the model. A far-field equation should not be applied in the reactive or radiating near field; a point-source approximation may be inappropriate close to a large antenna or array; and free-space calculations may not represent reflections, shielding, scattering, or conductive structures. If these limitations could affect the conclusion, measurement or a more detailed computational method is required.