Journal of Global Positioning Systems (2004)
Vol. 3, No. 1-2: 57-62
Differential LORAN for 2005
Benjamin B. Peterson, Kenneth Dykstra
Peterson Integrated Geopositioning, LLC, 30 Pond Edge Drive, Waterford, CT 06385, USA
e-mail: [e-mail] Tel: + 0118604428669; Fax: +01l
Kevin M. Carroll, Anthony H. Hawes
U.S. Coast Guard Loran Support Unit, 30 12001 Pacific Avenue, Wildwood, NJ 08260, USA
e-mail: [e-mail] Tel: + 0116095237321; Fax: +01l6095237320
Received: 15 Nov 2004 / Accepted: 3 Feb 2005
Note: The views expressed herein are those of the authors and
are not to be construed as official or reflecting the views of the
Commandant, U.S. Coast Guard, U.S. Federal Aviation
Administration or the U.S. Departments of Homeland Security
or Transportation.
Abstract. A multimodal group of engineers, scientists,
and industry representatives, including the U.S. Coast
Guard (USCG) and Federal Aviation Administration
(FAA) completed a major effort to define and analyze the
performance of a new Enhanced Loran system as a
backup for the navigation and timing services provided
by the NAVSTAR Global Positioning System (GPS)
provided services. Each mode of transportation has
defined requirements that the new Enhanced Loran must
meet to be acceptable in the radionavigation mix of
systems. The group developed a set of requirements for
Loran maritime navigation in terms of availability,
accuracy, integrity and continuity for the Harbor Entrance
and Approach (HEA) requirements defined in the Federal
Radionavigation Plan (FRP). This paper discusses the
goals of the Loran Support Unit for Fiscal Year 2005
(FY05), and the program to support these goals. The
factors related to achieving the objective of moving
Differential Loran from the proof-of-concept stage to an
operational status will be discussed. Also covered are the
results of an initial survey of the Inner Harbor at Boston,
MA, USA.
Key words: Loran, radionavigation, GPS, timing.
1 Introduction
The Loran Integrity Performance Panel (LORIPP) and
Loran Accuracy Performance Panel (LORAPP)
determined that an improved version of the LORAN-C
system, called Enhanced LORAN, could meet the
operational requirements of the HEA for maritime
positioning use and the FAA-derived Required
Navigation Performance of 0.3 NM (RNP 0.3). The U.S.
Department of Transportation (DOT) Volpe Center
completed a benefit-cost analysis covering this move,
with favorable results. Both reports were completed and
delivered to the Office of the U.S. Secretary of
Transportation in March of 2004. At the time of this
writing, the Loran community awaits a public decision
regarding the future of the LORAN system.
Although a definitive direction for LORAN has not been
decided, the USCG Loran Support Unit (LSU) has
continued research and development into the Enhanced
Loran architecture. Having completed the
aforementioned reports, a transition is underway from the
proof-of-concept stage to a quasi-operational status,
which will promote receiver development and other
LORAN research.
2 Differential LORAN
The basic concept of Differential LORAN is to provide
two sets of phase corrections to improve the navigation
accuracy from the current 0.25 NM level to
approximately 20 meters. One set of corrections is called
Additional Secondary Factors (ASFs) which are defined
as the phase differences between an all seawater
propagation path and the actual propagation path and are
functions of the ground conductivity and terrain along the

58
Journal of Global Positioning Systems
path. These ASFs will be obtained by detailed surveys of
the coverage area. In addition, there are temporal
changes in the observed phase caused by changes in
index of refraction along the propagation path and
variations in transmitter bias. These variations will be
measured at a fixed local monitor site, and communicated
to users via modulation of the LORAN signal. For a
detailed description of this data channel the reader is
referred to Peterson et al (2004).
3 Goals
There are two main goals for FY05. The first goal is to
establish Differential LORAN on a 24/7 real-time basis
for selected areas of the Northeastern U.S. Previous tests
were done either in post-processing or during limited
time periods in which Differential LORAN data was
broadcast over the air waves from the experimental
transmitter at the LORAN Support Unit. While these
relatively short broadcasts were useful to demonstrate
that the technology was feasible, continuous broadcasting
of real-time data is needed in order to refine the
implementation. This has the added potential benefit of
promoting receiver development.
The second goal is to develop the procedures and
working knowledge necessary to establish Differential
LORAN in an area. Knowledge gained from the marine
and aviation surveys can be integrated in support of this
goal. In addition to scientific concerns, some practical
considerations may drive the final shape of the new Loran
system.
4 Program
Differential LORAN is a technology that is applied to
both timing and navigation applications. Consequently,
two types of monitor sites have been identified: 1) Tier I
sites which possess a GPS independent, highly accurate
source of absolute time (within 10 ns of UTC(USNO))
facilitated by one or more atomic time standards
disciplined using Two-Way Satellite Time Transfer
(TWSTT), and Tier II sites which have a less accurate
and possibly GPS dependent source of absolute time.
Tier I and Tier II sites are nominally called “timing” and
“navigation” monitor sites, respectively. The Tier I sites
will support both timing and navigation users. If GPS
service is lost the Tier II sites will revert to pseudorange
vice absolute corrections whereby one correction will be
set to zero, all others calculated as relative corrections,
and the corrections will be useful to navigation users but
not to timing users. The message format includes bits to
notify users of the type of base station the corrections
come from and whether on not the GPS time reference is
available.
The Northeastern U.S is the area of the country with the
highest seasonal variation in phase propagation. Planned
monitor sites include: U.S. Department of Transportation
Volpe National Transportation Systems Center (USDOT
Volpe), to support some marine surveys in the Boston,
MA area, The USCG Loran Support Unit (LSU),
Wildwood, NJ, The USCG LORAN monitor site at
Sandy Hook, NJ, due to its proximity to the metropolitan
New York City area, and the United States Naval
Observatory (USNO), where official time for the U.S. is
maintained.
Boston Harbor will be the initial location for a marine
survey. A navigation monitor site has been established at
the Volpe Center to support surveys in the area. Once the
Boston survey research is complete and as time permits,
it is desired to apply the newly refined procedures to
another metropolitan area such as New York.
5 Issues
Communications Network: Due to the topography of the
areas surveyed, monitor sites may be placed in remote
areas and at locations with varied methods of access to
the Internet. This requires the establishment of an ad hoc
network in which data sources can be added, removed, or
moved easily. This capability requires a specialized
computer network structure. A next-generation IT
network for the Enhanced LORAN system is being
developed at the USCG Loran Support Unit, however it is
not due to become operational until FY 2007. An interim
solution that will allow for real-time data broadcast is
being developed at the LSU.
Monitor Site Density: The seasonal variation in phase
propagation is region-dependent. Differential LORAN
technology reduces the error due to this variance.
However, for a given area and a given location within the
area, the accuracy achieved using the correction from a
monitor site degrades with distance from the site.
6. Survey Considerations
There are several factors to consider when executing a
marine survey. Some of the most important ones are
discussed here.
(1) Geographic Survey Boundaries: The single
most basic question to answer in conducting a marine
survey is: what are the boundaries of the area to be
surveyed? As an example: consider the Chesapeake Bay,
VA area, which is large and has many tributaries and
other waterways connected to it. A decision needs to be
made concerning the areas of a waterway that require
Differential LORAN.

Peterson et al.: Differential LORAN 2005 59
(2) Seasonal Variations: The phase of the signal
from a given LORAN station and a given observation
point varies temporally. When conducting a marine
survey, it is necessary these temporal changes be
measured at the local monitor site and that these
variations in phase be taken into account in processing
the survey data. Once a survey has been completed, a
table of geographic points and associated nominal ASF
values are calculated. Once calculated, this table or
“grid” is loaded into a user receiver module. A
navigation monitor site sends out the temporal corrections
for the area covered by the grid. In the receiver, the
temporal corrections are used to increment or decrement
the base offset for the grid values as a whole. This
method is effective as long as the phase variation is
relatively uniform throughout the geographic region that
the grid covers. It is assumed that the temporal variations
in phase are constant over the coverage area of a
particular monitor site. To verify that this is valid for a
particular coverage area it is necessary to survey the area
at multiple times during the year.
(3) Grid density: This factor is influenced by the
spatial gradient of the ASF for a given area. A spatial
gradient develops when there is a significant difference in
the land path between a given LORAN station
(LORSTA) and two points. Assuming that it is desirable
to have a uniform level of accuracy for the area that a
grid covers, the existence of a gradient is problematic
since it means that the grid points must be closer together
for the high-gradient regions of the area. Another
solution is to divide the area into sub-grids of different
point spacing, or simply restrict grids to cover areas
where the ASF gradient is below a certain threshold.
Finally the grid must be in a format amenable to receiver
manufacturers.
(4) Source of Ground Truth/Geographic datum:
There are two possible sources of ground truth for the
ASF surveys: the USCG maritime Differential GPS
system and the Wide Area Augmentation System
(WAAS) operated by the FAA. DGPS is based on the
North American Datum of 1983 (NAD 83) and WAAS is
based on the World Geodetic System 1984 (WGS 84).
Both systems have comparable accuracy. In the surveys
done this far, we have logged both DGPS and WAAS
data simultaneously and have compared the two sets of
fixes and compared the differences to that predicted by
the differences between NAD 83 and WGS 84.
7 Loran Data Channel (LDC) Test for 2005
The real-time dissemination of Differential Loran data
(i.e.: moving data from multiple monitor sites to a central
database and broadcasting the same data from a LORAN
station) will represent a major move forward for
Differential Loran, allowing more effective test of the
technology and process, and will support additional
research in the field. The success of this endeavor
depends on proper integration of specialty software and
COTS hardware.
LORSTA Seneca, NY is the planned first broadcast node
in this network. Initially, observations from monitor sites
at the US DOT Volpe Center at Boston, MA and USNO
at Washington, D.C. will be broadcast from this station.
Communications between the monitor sites, a central
server and LORSTA Seneca, NY will be crucial to the
success of this endeavor. Currently, the operational
network for the LORAN system is being used for the
present Loran data collection efforts. There are three
obstacles to using this scheme for real-time corrections.
First, the architecture of the current operational network
coupled with the protocols employed is not amenable to
the type of data requirements for research. Second, the
security policy for the operational LORAN network does
not permit adding users on an ad hoc basis and with
varying security assurance levels, and does not allow
access from the Internet. Third, the remote possibility
that a catastrophic network glitch could be caused by this
research makes using the operational network an un-
attractive option. For these reasons, it was decided that a
network other than the operational network would be
used. Due to the prohibitive cost of acquiring another
research network for this specific purpose, it was decided
to use the Internet for communications during this test
and research phase.
LSU has undertaken the effort to determine the
requirements for the next-generation Loran network,
which will support Differential Loran messaging;
however the planned operational phase is for FY2007.
An interim, Internet-based solution is being developed at
LSU to facilitate research and monitoring of the
differential messages. This communications scheme will
allow dissemination of real-time differential corrections.
8 Architecture of Differential LORAN Data Network
In general, Differential LORAN is being implemented for
this experiment in the following way: Monitor sites
(navigation or timing) are placed at strategic locations
near certain waterways. The sites produce Loran
observations at a specific reporting interval which are
immediately sent to a central computer at LSU via the
Internet. Upon arrival at LSU, the observations are
logged and immediately relayed to the applicable
LORSTA (initially LORSTA Seneca) for broadcast. So
there are three types of nodes in the aforementioned
network: monitor, central, and broadcast nodes. Only one
central node (the server) exists. The location of the
monitor nodes is influenced mainly by available
space/real-estate, proximity to desired coverage area (for

60
Journal of Global Positioning Systems
navigation sites), and proximity to existing sources of
high-quality oscillators (e.g.: cesium clocks).
9 Required Equipment
The equipment being used for this experiment is mostly
commercial off-the-shelf (COTS). The nodes are
connected via the Internet. The central node requires the
least amount of equipment, consisting of a fast computer
running connected to the Internet and running specialized
software to relay the differential messages. The
broadcast node requires a computer to receive the
messages from the server and encode them for
transmission to the standard equipment at the Loran
Station. The computer at this node is also connected to a
Loran receiver, and a source of absolute time. Finally, an
uninterruptible power supply (UPS) will be used to
prevent unnecessary loss of power. The monitor node
requires a computer connected to a source of UTC and a
Loran receiver. A very stable oscillator is required for a
timing monitor site. A UPS is also used at this type of
node.
10 Boston Harbor Survey
An initial survey of the Inner Harbor at Boston, MA,
USA was conducted on July 17, 2004. Although
previous marine surveys have been conducted, this
survey helped bring some lingering issues to the fore.
ASFs are calculated and organized by cells in a two
dimensional grid of latitude and longitude. Cell size is a
variable to be determined, and it may vary from port to
port or even within a port. Specialized software has been
developed to perform some calculations on the raw
survey data. The software calculates and plots for each
cell:
a. Number of samples
b. Mean
c. Standard deviation
d. Maximum difference to any adjacent cell
Figures 1 through 8 illustrate the analysis for Boston
harbor. Figure 1 shows the path of the survey on a
nautical chart.
Figures 2 and 3 show the number of data points per cell
for cell sizes of 0.005 and 0.002 degrees respectively.
Figure 4 shows the mean ASF for the 9960Y signal. The
ASFs are relative or pseudo-ASFs meaning that they are
all relative to the 9960M signal which has its ASF set to
zero. The values are therefore the difference between the
9960Y (Carolina Beach) ASF and the 9960M (Seneca)
ASF and are negative due the larger portion of land in the
path from Seneca to Boston. Figures 5 and 6 show the
maximum absolute value of the difference in ASF to any
of the eight adjacent cells for cell sizes of 0.002 and
0.005 degrees respectively. Figures 7 and 8 show the
standard deviation of ASF for cell sizes of 0.002 and
0.005 degrees respectively. The intent is to determine
whether enough data was collected, the data collected is
valid, and that the cell density is sufficient such that
variations within a cell or between adjacent cells are
adequately bounded.
Figure 1. Path of Boston Harbor Survey
Figure 2. Number of Data Points per Grid Cell (Cell Size 0.005
degrees)

Peterson et al.: Differential LORAN 2005 61
Figure 3. Number of Data Points per Grid Cell (Cell Size 0.002
degrees)
Figure 4. Average ASF for 9960Y (Carolina Beach) Signal
Figure 5. Difference in ASF Between Adjacent Grid Cells (cell size
0.002 degrees)
Figure 6. Difference in ASF Between Adjacent Grid Cells (cell size
0.005 degrees)
Figure 7. Standard Deviation of ASF By Cell (cell size 0.002 degrees)
Figure 8. Standard Deviation of ASF By Cell (cell size 0.005 degrees)

62
Journal of Global Positioning Systems
11 DGPS vs. WAAS
Figure 9 shows the comparison of DGPS and WAAS
positions for the survey. The mean East difference is
0.05 m with a standard deviation 0.24 m and the mean
North difference is -1.05 m and with a standard deviation
0.26 m. The values predicted by HTDP.exe from NGS
Geodetic Tool Kit (www.ngs.noaa.gov) are 0.18 m East
and -1.01 m North.
-1 -0.5 00.5 1
-2
-1.8
-1.6
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
East - meters
North - meters
DGPS vs WAAS positions
Figure 9. Difference between GPS and WAAS positions for Ground
Truth
12 Conclusions and Recommendations
We have presented an outline of the effort to take
differential LORAN from the proof of concept stage to an
operational system. The main issues discussed include
the communications network necessary to broadcast real
time differential data and the methodology of conducting
and analysing ASF surveys.
Acknowledgements
This work was supported by the US Federal Aviation
Administration LORAN evaluation program. The program
manager is Mitchell Narins.
References
Peterson B, Dykstra K, Carroll K, Hawes A, and Swaszek P
(2004) Differential Loran-C, Proceedings of GNSS 2004,
Copenhegan, Denmark, May 2004.