Signal level interpolation for coverage area prediction

Predicting the received signal level is important in the design of cellular communication systems. In this paper, signal level interpolation around measured signal levels is introduced as a means to predict coverage areas. For signal level interpolation, surface fitting methods are required and three methods have, been evaluated: moving least squares (MLS), Delaunay triangulation and averaging for interpolation (AFI). The first two methods interpolate signal level values over a wide area with samples from only a few measurement points, while AFI requires closely spaced, evenly scattered points. MLS and Delaunay triangulation are described in the literature while AFI is a novel method developed by the author. The interpolated signal levels have been compared with measurements. The mean of the standard deviation of the error for seven test cases is 6.1 dB with the MLS method and 5.1 dB with triangulation. A comparison between all three methods revealed that in cases of closely spaced, evenly scattered points AFI gives the smallest mean value of the standard deviation (3.0 dB) for four cases.