Effectiveness of Software Solutions in Reducing Errors Due to Multi-Path in Spherical near Field Measurements

Summary form only given. The presence of multi-path reflections is usually a source of significant error in low frequency spherical near field measurements as the test antennas typically have low gain and the cost of lining the anechoic chamber with optimal low reflectivity is prohibitive. Some earlier papers have discussed the effects of there errors on far field patterns and mitigation of some of the errors using hardware solutions and by range optimization. Many of the commercially available software suits for near-field to far field conversion provide algorithms and utilities for reducing the errors due to multi-path. Some examples of such software solutions include MARS add-on for Near Field Systems Inc. (NSI), IsoFilter technique from MI Technologies, and Spherical mode filtering routines in CASAMS and TICRA SNIFTD software. These techniques generally require either over sampling of measured data or mounting the antenna such that the phase centre is displaced with respect to the centre of measurement thereby acquiring data on a sphere larger than the minimum sphere originating at the phase centre. The error reduction algorithms then estimate and filter out the contribution due to multi- path This paper presents the results of a study undertaken to determine the effectiveness of some of these algorithms in reducing the multi-path errors in a low frequency measurement facility. A variety of antennas such as horn antennas, log-periodic, helical and phased arrays were measured in a spherical near field facility housed in a sub-optimally lined anechoic chamber. The far field data with and without the software correction was compared to the expected behaviour of the antenna calculated using numerical techniques. The results of these measurements showing the effectiveness and limitations of the techniques studied will be presented. Practical considerations in reaching optimal solutions will be discussed.