Budget Rose: eLORAN, NDGPS Live Another Day

Everything old is eventually new again — that could be said of LORAN. Or perhaps, like clothing styles, proponents of LORAN just had to wait long enough, and it became fashionable again.

LORAN has been given $34.5 million in the president’s proposed federal budget for fiscal 2009, and made part of the provenance of the Department of Homeland Security (DHS). The U.S. Coast Guard, which has overseen the domestic LORAN program for years — and is part of DHS — would still administer the program.

Meanwhile, it appears the U.S. Department of Transportation (DOT) has bought the Nationwide Differential GPS system (NDGPS) another year of life. The FY09 Budget Request contains a line item in the Research and Innovative Technology Administration (RITA) budget for NDGPS in the amount of $4.6 million for operations and maintenance of the current system until October 2009. There is no budget item for the planned build-out of NDGPS. The budget request is subject to approval by Congress, but most likely this will go through.

LORAN

The Long Range Aid to Navigation (LORAN) system has roots that go back to World War II. (Pictured: The LORAN station Attu in Anchorage, Alaska.) The venerable system has modern value: the greater capabilities of the new enhanced LORAN (eLORAN) make it a much-needed independent, redundant backup to GPS, and one less susceptible to interference than GPS is. The Coast Guard has looked in recent years to shed LORAN, or at least remove the responsibility for funding it from its own dedicated budget.

Now it looks as if LORAN will remain as a complementary system and backup to GPS, something its supporters have suggested for some years now. Language in the proposed federal budget stated that the administration of the LORAN-C program will migrate to the National Protection and Programs Directorate (NPPD) of the Department of Homeland Security in preparation for conversion of LORAN-C operations to eLORAN, as well.

Where’s the (Rest of the) Money? But despite that language and decrees from LORAN proponents that the system has been rescued from oblivion, that $34.5 million is essentially a stopgap measure.

Funding for eLORAN going forward is still an open question; $34.5 million hardly covers the annual operating costs of the system as it exists today. As the Coast Guard budget position statement puts it, that fiscal 2009 budget request “reflects transfer of LORAN-C operations to NPPD; however; the Coast Guard will continue operation of the system in 2009 on a reimbursable basis.”

The actual upgrade to a full-fledged eLORAN system costs considerably more. It involves adding another data channel to the LORAN system as well as enhanced infrastructure, such as high-powered transmitters and new timing equipment at LORAN stations; there are 24 scattered across the United States and its territories.

While that upgrade has begun, it is far from complete — 19 of the U.S. LORAN stations have been upgraded so far — and its funding has been in limbo, as no new money has been allocated to it since fiscal 2006. Completing the upgrade could cost as much as $400 million. Where DHS will find that money remains to be seen.

NDGPS

The NDGPS funding request is neither a thumbs-up nor a thumbs-down for the system. The FY09 $4.6 million merely means that the DOT hasn’t figured out what to do with NDGPS yet, and the pain of having to fund a decommissioning program outweighs the $4.6 million to keep it running for another year, according to GPS World contributing editor Eric Gakstatter.

“When its study is completed, I seriously doubt RITA is going to find enough transportation applications to justify continuing to fund NDGPS under the DOT umbrella,”  he adds. “Realistically, it’s going to be up to federal and state government users in the affected regions to pony up the funding. You can bet that no private entities are going to contribute significant funds, if any at all. They’ll find another solution before going down that road.

“Some of the major government supporters (or associations who represent government agencies) that submitted public comments in support of NDGPS, and others, will have to come up with at least the operations/maintenance budget of approximately $5 million annually to sustain (not build out) the NDGPS as it is today,” Gakstatter said.

These include the Department of Agriculture (Farm Service Agency, U.S. Forest Service), Department of Interior (National Park Service, Bureau of Land Management), the Association of American Railroads, and various state agriculture, transportation, and other agencies.

FIGURE 1. Maritime + nationwide DGPS coverage

FIGURE 2. Maritime DGPS if 26 NDGPS sites are decommissioned

Software Caused October Anomalies

GPS World received from the U.S. Air Force GPS Wing this account of GPS IIR satellite anomalies: six instances in early October 2007 during which Block IIR satellites broadcast anomalous L1 Course/Acquisition (C/A)-code signals. This report was prepared by John Berg and Karl Kovach of the Aerospace Corporation; the complete version is available here. Excerpts follow here.

Background. During the period from 8–10 October 2007, five GPS Block IIR satellites and one Block IIR(M) satellite experienced anomalies which caused outages that ranged in duration from over 12 hours down to about 2 hours. With the exception of the first two anomalies, which overlapped, the anomalies affected only one satellite at a time.

At no time were any of the satellite orbit, attitude control, or command and telemetry functions affected. The anomalies were strictly limited to payload operations and functions. The payload hardware itself was also not endangered or affected in any way. Any user impacts resulting from the anomalies were dependent on their particular GPS receiving equipment functionality and processing described below.

Uploads Causing Problems. In each case the anomaly began during a navigation upload, when the Control Segment updated the satellite with new navigation message data. As noted, it affected only the Block IIR satellites, and did not occur on every upload. In fact, only about one-fourth of the uploads to the IIR satellites during this period triggered the anomaly.

Signal Characteristics. Based on analyses conducted by the satellite manufacturer and confirmed via observations by various users, it is evident that, after the onset of each anomaly, the satellite initially continued to broadcast standard L1 C/A-code, but with a navigation message that was invalid. Specifically, the navigation message data appeared to be an unpredictable sequence of 1’s and 0’s with invalid parity.

Root Cause. In every case, a navigation data upload from the Control Segment induced an upset in the on-board mission data unit (MDU) processor. The problem encountered specifically concerns the management of the uploaded navigation data in memory as managed by the software on the satellite, rather than a physical limitation or constraint on the memory itself. Depending on the exact on-board memory location into which data was being loaded, a given satellite upload message might or might not trigger the anomaly. This explains why only some uploads experienced problems and induced the anomaly (that is, approximately one-in-four).

Impact. The satellites continued to broadcast standard C/A-code immediately after the onset of the anomaly with invalid navigation message data which failed even the most rudimentary validity checks. In the course of the Control Segment’s recovery actions, the signal code varied from standard C/A-code to nonstandard C/A-code (NSC) and back to standard C/A-code, while the navigation message data varied from invalid to default to valid-but-unhealthy and eventually back to valid-and-healthy.

The presence of invalid navigation message data affected different manufacturers’ GPS receiving equipment differently. Some GPS receiving equipment continued to track the anomalous signals and generate normal pseudorange measurements while “coasting” on previously collected navigation message data. Some GPS receiving equipment stopped tracking the anomalous signals altogether. Still other GPS receiving equipment fell between these two extremes. Although the signal tracking responses varied greatly, the signal acquisition responses were nearly identical. Virtually no GPS receiving equipment would be expected to be able to fully acquire the signals after anomaly onset, due to the inability to perform subframe synch and hence to resolve the 1-millisecond ambiguity of the C/A-code. However, GPS receiving equipment and associated user system logic for reacting to these effects necessarily varied on a case-by-case basis. GPS receiving equipment that had already been tracking the C/A-code signal prior to the anomaly onset, and which continued to track C/A code despite the presence of invalid navigation message data and/or default navigation message data, could have been impacted.

Resolution Status. The root cause of the anomaly was determined in the course of analysis by the Block IIR navigation payload subcontractor. Once the problem was clearly understood, the chosen course of action was to slightly modify the Control Segment’s Block IIR upload software to prevent the problem. Since the modified code was installed, there have been and will be no further upload anomalies related to this problem. Fortunately the new GPS operational master control station, Architecture Evolution Plan (AEP), which had undergone a flawless transition on September 14, 2007, provided critical data that enabled operators to troubleshoot and resolve the problem in short order.

Future Plan. While it is clear that the immediate problem is solved, some longer range actions remain.

High on the list of planned actions is the obvious need — in light of this experience — to update GPS documentation and guidance such as IS-GPS-200 and the Standard Positioning Service (SPS) Performance Standard (SPS PS) to explicitly caution GPS receiving equipment manufacturers and users against utilizing pseudorange measurements in the presence of invalid or default navigation message data. GPS receiving equipment or associated user systems that stopped using the pseudorange measurements from the six affected Block IIR satellites while the satellite was broadcasting invalid or default navigation message data perceived nothing worse than a satellite outage. In contrast, GPS receiving equipment and associated user systems that continued using pseudorange measurements while coasting on previously collected navigation message data perceived huge pseudorange errors (tens or even hundreds of kilometers). IS-GPS-200 and the current SPS PS (3rd Edition, October 2001) simply suggest to the reader that invalid or default navigation message data received from a Navstar satellite should not be used. They are both silent regarding use of the associated pseudorange measurements while a satellite is broadcasting invalid or default navigation message data.

The status of the follow-up activities will be monitored as part of the GPS Wing’s Integrity Failure Modes and Effects Analysis (IFMEA) process. The primary objective of the IFMEA process is to analyze and document all of the possible failure modes which can directly or indirectly have an adverse effect on the integrity of the position, velocity, and time information provided to users. The corollary objectives are to reduce/eliminate the failure modes and to mitigate any remaining effects on integrity. The success of the follow-up activities will both reduce/eliminate the risk of this type of failure mode reoccurring in the future, and will guide GPS receiving equipment manufacturers and associated user system developers towards implementing validity checks to mitigate the effects if this type of failure mode ever does reoccur.

GLONASS Sweet 16

All three GLONASS satellites launched December 25 last year have been set to healthy. Satellite no. 721 is orbiting in slot 13 and broadcasting on the frequency channel -2; 722 is in slot 9 and channel -2; 723 is in slot 11 on channel 0. There are 16 operational GLONASS satellites in orbit with two of these set unhealthy and undergoing maintenance. Six GLONASS satellites are scheduled for launch in 2008 — three in September and three more in December. The first two improved GLONASS-K satellites are scheduled for launch the following year.