Research Article

Nonlinear Bayesian Tracking Loops for Multipath Mitigation

Table 1

Comparison between channel models.

JahnPérez-FontánLooSteingass/Lehner

Frequency1820 MHz (L-band)L, S, Ka bandUHF, L, S, Ka bandGPS L1, but others can be selected

Bandwidth30 MHz (wideband)Narrowband, widebandDepends on meas. campaignWideband

User(i) Handheld at 4 km/h.(i) Handheld.Depends on meas. campaign(i) Vehicle
(ii) Vehicle at 40 km/h(ii) Vehicle(ii) Pedestrian

DynamicsConstant velocityNot provided for L band, widebandNot providedVariable velocity

(i) Direct path(i) Direct signal(i) Direct signal(i) Direct Path, up to 3 diffracted LOS
Types(ii) Near echoes(ii) Multipath echoes.(ii) Multipath echoes(ii) Reflected echoes
(iii) Far echoes(iii) Diffuse components

NumberScenario dependant50 raysNot providedUp to 50

Attenuation10 to 30 dB9 to 16 dB<30 dBRice fading

DopplerNot straightforwardIncluded in generated time seriesNot straightforwardDeterministic included in the generated time series

(i) LOS(i) LOS(i) LOS(i) LOS
Shadowing(ii) Shadowing(ii) Moderate shadowing(ii) Light shadowing(ii) House front diffraction
(iii) NLOS(iii) Deep shadowing conditions(iii) Heavy shadowing(iii) Trees
(iv) Lamp posts

Elevation dependant15°, 25°, 35°, 45°, 55°15°, 30°, 45°, 60°, 70° 80°19° (other based on different references)[5°, 90°], based on 9 levels of measurement data: 5°, 10°, 20°, 30°, 40°, 50°, 60°, 70°, 80°

Azimuth dependantNone.Uniform azimuth distribution.None.Distribution extracted from measurements and depends on elevation

(i) Open(i) Open(i) Suburban
(ii) Rural(ii) Suburban(ii) Tree shadowed
Environments(iii) Suburban(iii) Urban(iii) Other based on different referencesUrban and suburban for car and pedestrian
(iv) Urban(iv) Tree shadowed
(v) Highway

(i) Different scenarios available(i) Time series are obtained prior to statistical model generation(i) Complete statistical model(i) Downloadable model available
(ii) Complete statistical model(ii) Complete statistical model for both LOS and multipath rays(ii) To be used as baseline for other statistical models(ii) Urban, suburban and rural
(iii) Good shadowing and NLOS approach(iii) Vehicle and pedestrian
(iv) Available parameters for GNSS L-wideband(iv) Realistic model
Pros.(v) Includes different elevations
(vi) Includes houses, trees and lamp posts
(vii) Low computational cost.
(viii) Takes into account generated scenario
(ix) Generates MP profile
(x) Developed for GNSS

(i) Measurements-based model parameters(i) Measurements-based model parameters(i) Measurements-based model parameters
(i) The license does not allow the modification of the code. Although own propagation data can be used to modify the model
(ii) Not possible to obtain parameters for user-defined scenarios(ii) Not possible to obtain parameters for user-defined scenarios(ii) Not straightforward multipath parameters(ii) Scenario needs to be reexecuted for each considered satellite
Cons.(iii) Constant velocity with fixed values(iii) Not enough measured data to extract significant model parameters(iii) Not enough available data to extract parameters for different scenarios
(iv) No elevations over 55°(iv) Limited data for applicable GNSS conditions: L band, wideband(iv) No detailed user dynamics description
(v) Rays Doppler not fully defined(v) Limited data for applicable GNSS conditions: L band, wideband