The main means of providing instrumental approach of aircraft for landing are beacon landing systems. It forms in space a trajectory (glide path), located, as a rule, at the angle of 3° in relation to the horizontal plane. The dependence of the glide path angle on the height of the snow cover on the underlying surface when using previously known antenna arrays in the glide slope is demonstrated in the paper. A new class of glide slope antenna arrays is proposed in the study. The requirements for the amplitude-phase distribution of currents along the antenna array of the glide slope, at which the glide path angle does not depend on the reflective properties of the underlying surface and the level of snow cover on it, are determined. We propose a procedure for constructing an antenna array with two subarrays, one of which radiates the so-called “Carrier plus sideband” (CSB) signal, and the second emits the “Suppressed carrier sideband only” (SBO) signal. The procedure of constructing an antenna array with two subarrays, one of which radiates the so-called “Carrier plus sideband” (CSB) signal, and the second emits the “Suppressed carrier sideband only” (SBO) signal are provided. In order to ensure the independence of the glide path parameters from the height of the antenna array (glide path angle and slope of the glide path), the combination of five conditions is specified in the article. A special case of the proposed antenna array is an equidistant antenna array, in which the radiating elements of the subarray of the CSB signal and the subarray of the SBO signal are partially combined. We have performed experimental studies with the 4-element equidistant antenna array as a part of a glide slope prototype. We have carried out flight measurements of the glide path angle at the airfield in the foothill with a difficult terrain for about two years: in summer, winter, next summer and next winter. The snow depth at the airport was 35–40 cm in the first winter, and 30–35 cm in the second winter. The experimental results demonstrate that the changes in the glide path angle of the stated period of time does not exceed one angle minute. This value is comparable with the error of the flight experiment. In order to simulate the influence of a higher level of snow cover on the position of the glide path, we have conducted a special experiment. When the snow cover was of 18 cm on the airfield surface, we lowered all four antennas by one meter and measured the glide path zone. The glide path angle remained the same. Flight tests of a prototype landing system with the proposed glide path beacon demonstrated that this landing system provides parameters for the III-rd category. It means that the aircraft approaching for landing and landing are possible at zero visibility.