##

**Special
Study Group 1.182:
**

## "MULTIPATH MITIGATION"

######

**Introduction**

The
precision of raw carrier phase observations recorded by modern GNSS
receivers is generally at the sub-millimetre level. However, in all
but the most benign environments, the achievable resolution of GNSS
positioning is one or more orders of magnitude worse. This discrepancy
between the theoretical hardware-dependent precision of the raw
observations and the practical accuracy of GNSS position solutions
can, in part, be attributed to the effects of site-dependent
electromagnetic scattering of incoming GNSS signals. If millimetre
level (or better) GNSS accuracies are to be routinely achieved in the
future, these electromagnetic scattering effects (commonly referred to
as multipath and diffraction) must be eliminated. The website of the
Special Study Group 1.182 is http://www.gmat.unsw.edu.au/snap/gps/iag_section1/ssg1182.htm

**Objectives of the SSG 1.182**

The goal of the SSG 1.182 is to study GNSS multipath detection and
mitigation techniques with the aim of improving existing high
precision positioning accuracies. In the context of this SSG,
multipath is loosely defined as the systematic errors in raw GNSS
observations that are due to any signal scattering effect caused by
the local environment surrounding an antenna. Furthermore, this SSG
will focus on carrier phase and code-based multipath in terms of
effects on receiver operation for high precision applications.
Finally, within the scope of the group, the term GNSS is defined to
encompass any type of global positioning system (for example, GPS,
Glonass-GPS and Galileo-GPS), or systems simulating GNSS signals (such
as in the case of pseudolites). The objectives of the group can be summarised as:

Evaluate
and compare existing and developing algorithms and techniques for
multipath detection and mitigation.

Quantify
and document the effectiveness of commercial receiver-based multipath
mitigation techniques for high precision positioning.

Investigate
and document the properties of multipath in a variety of environments
(particularly high risk environments).

Provide
information and guidelines for multipath detection and elimination for
high precision applications.

**Members
and Corresponding Members
**

** **Members:
Mike Stewart (Chair, AUSTRALIA), Penina Axelrad (USA), David Betaille
(UK), Mike Braasch (USA), Luisella Giulicchi (THE NETHERLANDS), Cythia
Junqueira (BRAZIL), Guillermo Ortega (THE NETHERLANDS), Jayanta Ray
(CANADA), Angela Reichert (USA), Rodney Walker (AUSTRALIA), Andreas
Weiser (AUSTRIA).

Corresponding
Members: Joao Batista (BRAZIL), Paul Cross (UK), Xiali Ding (HONG
KONG), Minghai Jia (AUSTRALIA), Domenico Sguerso (ITALY).

**Activities
of the SSG 1.182
**

**The
primary activities of the group since its inception in January 2000
have been:
**

*a.
*Define
the terms of reference and objectives.

*b.
*Compile
review of relevant and available literature. A list of some 120
multipath related papers is located at: http://www.cage.curtin.edu.au/~mike/ssg1.182/biblio.htm.

*c.
*Compile
review of relevant web sites. A set of links to relevant web sites can
be found at: http://www.cage.curtin.edu.au/~mike/ssg1.182/links.htm.

*d.
*Compilation
of data archive for multipath data. The SSG is in the process of
compiling a data archive to provide multipath researchers with easy
access to a variety of different data types from different
environments. The archive will also provide reference to the multipath
analysis performed by the group who supplied the data, enabling direct
comparison between different techniques and different research groups.
The archive should be on-line by late 2001.*
*

*e.
*Define
core research areas within the SSG's terms of reference. As the terms
of reference are rather broad, a number of core research sub-sections
have been defined. Individual group members have been encouraged to
monitor developments in the sub-sections relevant to their personal
fields of expertise. These
include:

·
multipath
characterisation and attitude determination;

·
multipath
mitigation developments in receiver hardware;

·
semiparametric
and parametric multipath modelling techniques;

·
weighting
and SIGMA models for multipath mitigation;

·
multipath
in space-based applications;

·
multipath
mitigation using multi-antenna arrays, time stacking and crossing
points; and

·
electromagnetic
propagation modelling for multipath analysis.

Below
is a brief summary of the technical developments being covered by this
SSG. Full reports from group members can be found at the SSG 1.182 own
website: http://www.cage.curtin.edu.au/~mike/ssg1.182/recent_reports.htm.

**Multipath Mitigation Developments in Receiver
Hardware**

A
variety of so-called multipath-mitigating receiver architectures have
been developed over the past decade.

Narrow-correlator
(NovAtel); Edge correlator (Ashtech) - The narrow-correlator concept
involves moving the traditional 'early' and 'late' correlators closer
together. The peak of the pseudo-range multipath error envelope is
reduced in direct proportion to the correlator spacing. Ultimately the
finite bandwidth of the GPS signal places a practical lower bound on
the correlator spacing. Correlator spacings of 0.1 and 0.05 chips are
commercially available, thus providing approximately a factor of 10 to
20 reduction in the peak of the error envelope.

Multipath-Estimating
Delay-Lock Loop (MEDLL) (NovAtel)-* *The MEDLL uses multiple
correlators (6 – 10) per channel in order to determine the shape of
the multipath-corrupted correlation function. The MEDLL software
determines the best combination of direct and multipath signals (that
is, amplitudes, delays, and phases) which could have produced the
measured correlation function.

Strobe
correlator (Ashtech*) - *The strobe correlator was developed by
Ashtech in 1996 and involves a linear combination of two narrow
correlator discriminator functions. The result is a discriminator
function which is very narrow and thus is significantly less
susceptible to medium and long delay multipath.

Enhanced
strobe correlator (Ashtech); Pulse-Aperture Correlator (NovAtel)* - *For
most practical purposes the Enhanced Strobe Correlator exhibits true
P-code-like multipath characteristics. Specifically, it is virtually
insensitive to multipath with delays longer than 50 metres. More
recently, NovAtel has released the Pulse Aperture Correlator which has
very similar performance. Other manufacturers (Leica, Navcomm) have
similar architecture.

Multipath
mitigation through modified antenna design is also an important field
of research. The most recent developments include adaptive array
techniques in which two classes of solutions have been proposed. A
first is based on the joint utilisation of a direction-of-arrival
(DOA) estimation technique together with a constrained adaptive
algorithm. A second approach uses a self-adaptive constant modules
technique, eliminating the need of a pilot signal and DOA estimator.

**
**

**Multipath Mitigation Using Functional and
Stochastic Modelling**

One
of the most important developments to date in this field are the SIGMA
models which were developed to overcome artificially introduced
periods of weak satellite geometry by proper weighting of phase
observations (SIGMA-e
model) and to reduce signal diffraction effects of the phase
observations (SIGMA-D
model). The main parameter of these models is the ratio of the power
of the GPS carrier wave C [dBW] to the noise power density N0 [dBW-Hz],
in short C/N0 [dB-Hz]. Usually, geodetic receivers provide the C/N0
measurement in the receiver internal binary format or in the NMEA $GPGSV
message. There are currently discussions in progress to standardise
and include the C/N0 observation in a future RINEX format of the GPS
observation files. Recently, researchers from Leica Geosystems have
proposed a self-calibrating SIGMA-D
weight model.

A
different approach to the SIGMA models also uses signal quality
indicators such as signal-to-noise ratio (SNR) to reduce the errors
due to multipath. Work is concentrating on direct estimation of the
size and sign of multipath errors and subsequent correction of the raw
phase measurements, and the estimation of the elements of a full
covariance matrix for the raw GPS phase data according to the likely
size of the multipath contamination and the amount of correlation of
errors between satellites.

An
alternative to traditional least squares modelling of systematic
errors in GPS data has also been proposed. The semiparametric model
and penalised least squares method describe multipath by a complex but
smoothly varying function with time. The functions, and estimated
parameters such as station coordinates and ambiguities, are decomposed
using the penalised least squares method. Multipath mitigation using
the repeatability of SNR ratios over the sidereal day at permanent GPS
receivers is based on using a residual stacking algorithm. Others
separate multipath from the carrier phase observations. The University
of Colorado has developed an algorithm to utilise the
spatially-correlated characteristics of multipath to reduce multipath
in ground and space-based applications. This algorithm will be used to
mitigate multipath in ground-based GPS reference stations.

**Electromagnetic Propagation Modelling for
Multipath Analysis**

The
European Space Agency (ESA) is using a software tool "Multipath
Virtual Laboratory" (MVL) to compute multipath effects on the GPS
observables having satellite constellation location, receiver antenna
location, positioning of surrounding structures and antenna
information as input parameters. The computation of signal propagation
uses a ray-tracing angular Z-buffer algorithm, followed by an
electromagnetic field computation using the Geometric Theory of
Diffraction (GTD). The MVL tool was used to pre-compute the presence
of multipath for the rendezvous of the shuttle Atlantis with the
Russian space station MIR.

Work on modelling the multipath environment of the International Space
Station is currently underway at the Jet Propulsion Laboratory (JPL)
using a multipath simulator previously developed in 1990. In this
recent application of the simulation model, the measurement error due
to multipath has been computed for a number of different antenna
locations. Once the multipath error for each antenna is computed, the
corresponding orbit error due to multipath is determined using JPL's
GIPSY-OASIS II orbit determination software.

A
relatively new technique that involves a numerical solution to the
Parabolic Equation (PE) has been used to solve for two-dimensional
propagation over any type of terrain. The PE provides a direct
solution of Maxwell's wave equations by approximating the Helmholtz
scalar wave equation. This technique does not rely on the study of
individual ray paths as used in the GTD. Propagation simulations from
the model accurately provide the amplitude and phase of the propagated
plane wave at all points within the model domain. The PE model was
used to study the effects of diffraction and multipath caused by
various types of terrain commonly found in an open cut mining
environment.

**Multipath
Characterisation and Attitude Determination**

Both
ESA and the NASA Johnson Space Center group have been studying
multipath effects on attitude determination in the particularly severe
environment of the International Space Station (ISS). NASA has
compared space shuttle GPS flight data to predicted results from
geometrical diffraction prediction. ESA has analysed data from
on-ground experiments and in-flight demonstrations for the rendezvous
and docking of ESA's Automatic Transfer Vehicle with the ISS.
Researchers at ESA have also studied the design of a modified patch
antenna that provides low elevation and LHCP signal rejection. Size,
weight, and other characteristics are designed for space applications
with the goal of improving attitude accuracy below 1°.

Two
academic research groups specialise in multipath mitigation for
attitude determination. One group have looked at phase map corrections
for using simulations and satellite data from the CRISTA-SPAS
Experiment. Another group proposes the use of non-dedicated receivers
for attitude determination, including using a group of closely spaced
antennas for multipath correction in RTK. Attitude accuracy is quite
poor because antenna separation is very small. Multipath reduction
with the multi-antenna array has been studied in various environments,
such as in urban canyons and under foliage. Improvement due to
multipath corrections is reported to be approximately 50%.

**
**