Page 44 - ITU Journal: Volume 2, No. 1 - Special issue - Propagation modelling for advanced future radio systems - Challenges for a congested radio spectrum

ITU Journal: ICT Discoveries, Vol. 2(1), December 2019

distribution parameters. For off-body propagation, The parameters were derived from full-wave
due to the inability to de-embed the antennas’ numerical simulations, and from wideband
radiation patterns from measurements, the measurements using physical liquid phantoms with
measured loss corresponds to the system loss. The one antenna fixed at 3.5 cm from the phantom’s
model parameters represent a typical indoor office surface while the other was moved in the air over a
environment [1-2], for static, quasi-dynamic spatial grid [4], or in vivo. The in to off-body channel
(mimicking walk in a given position) and dynamic was parameterized based on in vivo and phantom-
user scenarios (approaching and departing from the based measurements [5].
AP), with both co- [2] and cross-polarized antennas
[1]. The table also gives the model parameters for 3. MASSIVE MIMO BELOW 6 GHz
measurements inside a dome-shaped discotheque
within a ferry, with the user walking towards and The COST 2100 MIMO channel model [6] is a
spatially consistent geometry based stochastic
away from a fixed AP [3].
channel model (GSCM) that uses the concepts of
Table 1 – Summary of MPL and fading parameters for the off- clusters and visibility regions. In COST IRACON the
body channel at 2.45 GHz extension of the model to massive MIMO channels
Large-scale Small-scale [7] includes: 1) introduction of the visibility regions
Mean path loss at the base station, 2) introduction of a gain function
Env. fading fading
( ) [dB] [dB] ∈ Ω ∈ of individual multipath components MPCs, and 3)
0
0 [dB]

generalization to full 3D geometries. A MATLAB
Office 32.0-50.0 1.71 0 1.2-3.0 0.9-19.5 1.0-2.0 implementation is freely available [8].
Ferry 25.2-64.7 1.69 0 1.7-6.5 0.8-1.5 1.5-2.1
The visibility regions at the base station is
Two variants of the MPL model are proposed for in motivated by the fact that when the antenna array
and into-body channels: a log-distance model (2) gets physically larger, the radio channel cannot be
and a linear model (3), i.e. seen as wide-sense stationary (WSS) over the array,
[ ] = 0, ( ) + 10 ( / ) (2) clusters appear and disappear, which means that
0
0
both the angular spread and the delay spread
[ ] = 0, ( ) + 10 [ / ] [ ] (3) change over the array. This effect is not captured by
0
conventional MIMO channel models.
where is the Tx-Rx distance, 0, / ( ) is the MPL Measurements in [9] indicate that individual
0
at the reference distance usually taken at 1 cm, multipath components (MPCs) have a limited
0
and and are slopes of the MPL models. lifetime within the cluster when the user equipment

moves with different MPCs of a cluster active at
The in-body channel corresponds to links
established between two implants (in2in). Into- different locations within a visibility region. This is
body covers implants, e.g. an ingested capsule modeled by a gain function, with a Gaussian shape
streaming video to a receiver placed on the patient’s in the spatial domain with its peak randomly
body (in2on) or off-body next to it (in2off). Table 2 located within the cluster for each MPC. These gain
summarizes the model parameters for two channels: functions act as weighting functions for the MPCs so
between two implants (in2in) and in to off-body that depending on where the user equipment is
(in2off), and for in to on-body (in2on) for the located in the visibility region it sees different
3.1 - 5 GHz band. The path-loss standard deviation weighted combinations. This retains the spatial
around the MPL ( ), and applicable distances ( ) consistency of the model, which is also important
are also provided. when the model is used for more advanced forms of
radio-based localization and navigation.
Table 2 – Summary of the MPL model parameters for in2in,
in2on, and in2off-body channels at 3.1-8.5 GHz The generalization to 3D geometries and support of
polarimetric channels allow for advanced antenna
( )
0, 0 0, ( ) arrangements to be used for massive MIMO. The
0
Ch.
dB dB dB/cm dB cm base stations, individual antenna locations, user
equipment, scatterer locations, and visibility
in2in - - 45 4.6 4.3 3-8
in2off 70.4-71.5 0.7-1.4 - - - 4-50 regions can all be described by their 3D coordinates.
in2on -12.2-35.8 5.8-9.3 14.8-53.4 4.5-7.4 5-5.7 2.8-8 Antenna gain patterns are included for each
antenna. The COST family of channel models
inherently captures spherical wave front effects as

28 © International Telecommunication Union, 2019