Fading Prediction on Microwave Links for Airborne Communications 1 S. Loyka, A. Kouki, F. Gagnon Department of Electrical Engineering, Ecole de Technologie Superieure 1100, Notre-Dame St. West, Montreal (Quebec), H3C 1K3, Canada Email: sergey.loyka@ieee.org Abstract- Fade depth prediction on airborne line-of-sight communication links is considered in this paper. There is no specific model for this scenario at the moment. The two ray multipath model adapted to a realistic scenario of hilly or mountainous terrain was used to account for flight and terrain geometry, and antenna parameters. Surprisingly, it was found that (i) in many practically important cases the fade depth depends on the path clearance angle only and (ii) the two-ray model predicts roughly the same fade depth dependence on the path clearance angle as the well-known Olsen-Segal model. This may be considered as a theoretical justification, as to the best of our knowledge – for the first time, of the path elevation angle factor in that model. 1. INTRODUCTION Signal fading is a severe obstacle to reliable microwave communications. It affects the system performance in several ways. In this paper, we consider the impact of fading on signal-to-noise ratio (SNR) at the receiver of an airborne high capacity line-of-sight (LOS) communication link. This effect is accounted for in the link power budget analysis using the fade margin, which quantifies the loss in SNR under fading conditions, usually – for a given fade outage probability (i.e., the probability that the given fade depth will be exceeded). The received signal power can be presented as [1]: P R T R T T R L L L G G P P = (1) where P R and P T are received and transmitted power correspondingly, G R and G T are receive and transmit antenna gains correspondingly, L R and L T are receive and transmit path losses (i.e., cable loss, loss due to aging, antenna misalignment etc.), and L P is propagation channel loss ( 1 , , > p T R L L L ). The propagation loss is usually factored out into three main components [2]: L S P L L L L 0 = , where L 0 is the average path loss (for example, free space loss), L L is the long-term fading (for example, due to shadowing), and L S is the short-term fading due to multipath. Hence, the fade margin F can be defined as L S L L F = . The transmitter power or the transmit or receive antenna gains must be increased by F to sustain the reliable link operation as compared to the case of unfaded propagation channel. The main concern for line-of-sight links operating at less than 10 GHz is the short-term fading, which may be produced by ground or atmospheric multipath (however, other atmospheric impairments may sometime contribute as well, especially at low clearance angle). An accurate estimation of fade depth is of great importance for the design of a highly reliable communication link. Most models for signal fading prediction that are used in microwave wireless link budged analysis are mainly empirical in nature and rely on extensive measured data [1,3- 6]. Consequently, these models include the aggregate effects of the different fading mechanisms and do not allow for any insight into the relative importance of these mechanisms. When theoretical models are proposed, they are often limited to particular fading scenarios and are useful when such scenarios represent the dominant mechanism of fading. When the microwave wireless link considered is covered by an existing empirical model, that model is used and the relative contribution of different fading mechanisms is not important. However, if the link considered is not covered by any existing empirical model, additional analysis of the relative contribution of different fading mechanisms is of large importance. In this paper, we consider the airborne link scenario shown on Fig. 1. In this scenario, the ground station communicates with the airborne communication node through a multipath channel, which is also a subject to atmospheric impairments. Ground station Atmospheric impairements Ground multipath Airborne node Fig. 1. Airborne communication link scenario. 2. MODELS FOR FADING PREDICTION In general, there are two main contributions to fading [1]: (i) due to ground multipath or due to reflection from ground objects like mountains or hills, and (ii) due to atmospheric 1 This work was supported in part by NSERC and BAE Systems Canada