AbstractThis paper summarizes the results of a measurement campaign, which set out to characterize Inmarsat's broadband global area network extension (BGAN‐X) L‐band land vehicular (LV) channels as a function of elevation angle and candidate antenna systems. Measurements were conducted with three types of directional antennas on motorway routes around London using Inmarsat's Indian Ocean Region (IOR), Atlantic Ocean Region‐East (AOR‐E) and Atlantic Ocean Region‐West (AOR‐W) satellite transmissions, corresponding to elevation angles of, respectively, ∼7.5°, ∼13° and ∼30°. Simultaneous recordings of data on two types of directive receivers enabled correlation of several channel parameters. Statistics were used to refine a baseline model that derived BGAN‐X LV channel parameters by scaling the statistics obtained from an omni‐directional antenna receiver. The measurements demonstrate that Ricean mean and Ricean factor vary significantly over the chosen routes at elevation angles below ∼13°. Further, the Doppler shift of the direct path can be predicted accurately. A qualitative analysis indicates that the multi‐path Doppler spread is generally (not always) caused by uniform scattering around the vehicle over short distances and undergoes spatial filtering due to the receiver antenna. A technique was established to derive statistics of a two‐state Markovian channel model at any specified fading threshold by gathering tracking system state, thereby allowing one to ascertain service expectations and operational parameters such as connection time‐out of a packet‐switched system or reliability threshold of a circuit‐switched system. The technique was used to identify and select appropriate routes, particularly for the IOR and the AOR‐W regions. Data analysis is in progress to establish the reason for the slow variations of signal mean at low‐elevation angles in Ricean conditions and to derive appropriate statistical models. Furthermore, image processing of the video recordings is expected to reveal the extent of optical and RF fades correlation. Copyright © 2008 John Wiley & Sons, Ltd.
Read full abstract