High-Altitude Platform (RAP) is the most promlsmg infrastructure for wireless communication in the near future. An essential problem need solving for the design of RAP communication systems is to establish an effective channel model for predicting the characteristics of the communication link. There have been many papers on this problem and several channel models have been proposed. The main work of this paper is to summ arize and analyze the channel models available. Here, the existing models are divided into three groups according to their respective specialties and each model is briefly introduced with emphases on analysis of their strengths and weaknesses. Some open issues are also pointed out.The rest of this paper is organized as follows: In Section 2, the channel models available are discussed by dividing them into three categories. Finally, a conclusion and some open issues are given.
CHANNEL MODELSThis section will analyze the presently existing channel models by dividing them into three categories as strictly statistic-based models, GBSB (geometrically based single bounce) models and K-based models. For simplicity, all these models assume that the transmission frequency is lower than 10 GHz, so there is no necessity to consider attenuations caused by rain and atmospheric scintillation.
Strictly statistic-based channel modelsIn this section, the strictly statistic-based channel models proposed in [1]-[3] are discussed. References [1] and [2] present a "switched" channel model for RAPCS links that is largely based on the ITU-R Recommendation P.681-6. In this model, it is assumed that the configuration of the channel varies from one condition to another with the changes in environmental properties and that each condition can be characterized by some sort of distribution. Besides, the state transition is [16]. While in [14], the Ricean factor K is computed from a geometrical point of view as is shown in Eq.(2). K(a) = 2010g( dws /decho)dB , (2) where d ws is the length of the direct way and d eeho is related with the echo path in the two ray ground reflection model. It is noted that the model presented in [14] is over optimistic in its evaluation of the HAPCS channel. Besides the adjustment of the level of the power from the platform and the effects of the shadows are not taken into account in [14], the horizontal distance between the reflectors and the receiver is constant. So this model is inflexible. While in [15] and [16], there are few mathematical formulas for characterizing the channel. Furthermore, only total received power was collected with no information regarding LOS power and multipath scattered power during measurements. Since this narrowband wireless channel model is established totally based on experiments performed in the specific area, it is not clear whether it can be applied to other communication environments. As a whole, both models are similar with an exception that the methods for computing the Ricean factor are different.
CONCLUSIONSThis paper presents a survey of the available chan...