Large-Scale Parameter Estimation in Channel Sounding with Limited Dynamic Range

In this paper, we examine large-scale parameter (LSP) estimates for spatio-temporal mobile radio channels in the case when the measurement dynamic range is too poor to derive the LSPs for all the measurement locations. Conventionally the LSPs are derived with, e.g., 20 dB dynamic range from the strongest detected multipath component (MPC), but such a measurement dynamic range cannot be guaranteed in all measurements as some of them suffer from large path loss (PL) and/or variation due to shadow fading (SF). We will show that the LSPs, specifically PL, SF, delay spread (DS), and angular spread (AS), can be considered as incomplete data. Incomplete data can be described either by a maximum or a minimum value (or both) when the measurements are with poor dynamic range. We show that the incomplete data should be taken into account in deriving the LSP statistics using the maximum likelihood estimation (MLE) rather than omitted as "outage". Outdoor-to-indoor (O2I) measurements at the 14-14.5 GHz range are used as an example.