LightSquared and the GPS Innovation Alliance disagree about the merits of the Department of Transportation’s draft test plan for developing interference tolerance masks for GPS receivers in the 1559-1610 megahertz band, with LightSquared blasting the proposal and the alliance generally praising it.
In its comments, which were filed in DOT-OST-2015-0099 by an Oct. 16 deadline, LightSquared Subsidiary LLC called the DoT draft test plan “fundamentally flawed,” and it said that the FCC and the National Telecommunications and Information Administration should conduct such testing.
“Not only does the test plan contain significant omissions that leave the scope and schedule of the proposed testing uncertain, it also fails to respond to or even consider concerns raised repeatedly by LightSquared and other stakeholders on critical issues such as device selection, transparency, independence of the testing process, testing procedures, the repeatability and verifiability of the testing regime, and the evaluation of test results,” LightSquared added. “Most significantly, the proposed test plan establishes that DOT refuses to measure what counts: the impact of any interference on performance of the device. That is what the FCC defines as ‘harmful interference.’ Given this failure to use the expert agency’s metric, the test will not actually add to the dialog that is needed to ensure that this vital spectrum can be used by both GPS and next-generation broadband services, and the long delay caused by the DOT testing will be for naught.”
LightSquared complained that “[t]he DOT study intends to measure a change in the noise floor, specifically a 1 dB change in C/N0, and to report on any such change caused by LTE use. Choosing to measure a 1 dB change may have been justified in the last century, when testing equipment was not as advanced as it is today and so interference analysis was based on changes in the noise floor—not because it was necessarily a useful metric but because that was all that could be measured. But to rely on this old, and flawed, metric, which has no correlation with actual performance of the GPS device, is misguided when modern tools are available to measure what consumers care about: whether interference causes the GPS device to give an end user faulty position or timing information. For this reason, we advocate that any testing should focus on position and timing error.”
LightSquared added “that DOT is not the right agency to evaluate compatible uses of spectrum, particularly when there are expert agencies—the FCC and NTIA—who grapple with these issues regularly, and who are statutorily charged with the responsibility of regulating spectrum usage.”
The filing continued, “LightSquared’s Comments on the draft test plan are an attempt to ensure that if DOT proceeds with its testing, the results will be useful to the FCC and NTIA as the ultimate decision-makers on compatibility issues. Importantly, LightSquared urges DOT to ensure that its testing bears the hallmarks of high-quality scientific experimentation—i.e., that the tests themselves and the results obtained are repeatable, verifiable, and can be proven as false or true. The test plan that Roberson and Associates presented at the recent DOT workshop (attached herein) meets those criteria and is designed to contribute to the vital debate on this important mid-band spectrum.”
LightSquared said that among the questions that the draft test plan leave unresolved are projection of power, multi-GPS receivers, mobile satellite service-augmented receivers, and spectrum bands. It also said that the plan should provide details concerning device selection and information on the testing process, including procedures, timeline, test facility and personnel, and an audit of the results.
LightSquared also reiterated its concern that testing 100 devices at the same time could result in interference when so many devices are placed near each other.
But the GPS Innovation Alliance said that it “supports DOT’s Adjacent Band Compatibility [ABC] process and its overall objective of establishing technical guidance for successful coexistence of GPS and other proposed uses of spectrum in nearby bands; it also welcomes the inclusion in the study of receivers that use Mobile Satellite Service (‘MSS’) augmentation signals. Because of spectrum scarcity, GPSIA and its predecessor organization have long supported efforts to evaluate how all proposed spectrum uses can be productively accommodated and coexist with established services. Except for a handful of suggestions and points discussed below, GPSIA believes that the Draft Test Plan is a commendable first step in advancing the goal of ensuring such harmonious coexistence, and it encourages DOT to move quickly toward the next steps in the process.”
The alliance said that “the Draft Test Plan’s proposed collection of extensive proprietary design information … is unnecessary for derivation of the Adjacent Band Masks,” and it endorsed “the proposed use of a 1 dB standard in testing.”
The alliance added that “the 1 dB standard has a long and well-established history in both international and domestic regulatory contexts as the appropriate interference protection criteria (‘IPC’) for GPS receivers. While DOT has proposed recording other performance metrics, such as loss of signal lock or degradation of pseudo-range or position accuracy, GPSIA believes these are inappropriate metrics for interference assessment since their inherent basis is an interference level that seriously degrades the RNSS spectrum environment and causes significant disruption to GPS receivers.
“Moreover, for some applications, the primary receiver output may not be position, but rather time, velocity, acceleration, or other parameters,” the alliance added. “Degradation of accuracy or otherwise attempting to determine effects on the ‘user experience’ are not practicable interference metrics, and DOT should rely upon the 1 dB protection criteria in derivation of the Adjacent Band Masks. GPS receivers are used in a tremendous range of end user applications beyond simple navigation. It is unclear how it would be possible to determine whether there has been ‘material degradation’ in the functioning of this wide range of GPS applications, much less what constitutes degradation that is ‘material.’ The amount of data produced for all various measures of ‘degradation’ to a ‘user’s experience’ across the large number of use cases would be administratively staggering and very unlikely to demonstrate universal trends or measurement standards.”
The alliance offered comments on the plan for radiated testing.
“First, since there will be a very large group of receivers under test, GPSIA believes that it is important that the test-bed be carefully constructed and calibrated to ensure that the GPS receivers (a) are accessible to test personnel and (b) do not interact with each other in a way that influences the results,” the alliance said. “Given the wave-lengths of GPS signals, this means that the devices should be spaced so that they are at least two feet apart. Dry-runs of the test should focus on confirming that the spacing is adequate, and the test plan should allow for increased spacing if necessary. GPSIA also prefers horizontal test-beds, which require less construction and time to establish.”
The alliance also noted that “the list of GNSS signals proposed to be simulated does not include any L2 signals. GPSIA requests that GPS L2 and GLONASS G2 signals be included in the constellation. … Third, with the large number and variety of GPS receivers under test, GPSIA encourages DOT to ensure that an adequate dwell time is used to ensure collection of accurate results.
“The objective of the proposed testing is to collect data to determine the interference tolerance masks for GNSS receivers,” General Motors LLC noted in its filing. “General Motors believes that a reasonable way to limit interference would be for DOT to recommend that the Federal Communications Commission (‘FCC’) require all 4G LTE licensees operating on Band 24, and any licensees operating in any future band classes that use the L Band spectrum, to limit noise power in the adjacent L1 band.”
GM added that it “recommends that DOT consider using a 4G LTE simulator in lab tests. Using a simulator would add credibility both to the testing and to the final test results.” The car maker also said that DoT’s proposal to use a testing grid with 100 GPS antennas “could be successful if all of the tested antennas are GNSS patch antennas. By contrast, if DOT incorporates any antennas into the test grid configuration that are combinations, such as combination cellular/GPS antennas with metal structures, there is a significant potential that radiation patterns will be affected by this configuration. In addition, the simultaneous testing of 100 GPS/GNSS receiver antennas at once could create a significant potential for receiver-on-receiver interference. General Motors recommends that DOT modify the Test Plan to address and minimize the potential for effects from configurations incorporating combination antennas and receiver-on-receiver interference.”
Logan Scott, president of LS Consulting, said, “Interference tolerance masks developed using the methodologies of the ABC draft test plan may prove dangerously permissive. The GPS L1 C/A code signal used as the primary, and often sole, navigation and timing signal in most civil applications has numerous well known structural vulnerabilities that will not be stimulated under this test plan. Specifically, C/A codes are short and repeat themselves every 1 msec leading to a line spectrum vulnerability. Low level interference can cause sustained biases in measured pseudorange and carrier phase, often with no measurable change in observed C/No.”
“The ABC draft test plan proposes using white noise as a stand-in for LTE signals,” Mr. Scott added. “White noise is the most benign form of interference there is for a GPS L1 C/A code receiver. Waveforms 100 times weaker can cause larger errors in position and time and, without altering C/No. White noise has no periodicities and it is not capable of inducing measurement bias effects.
“Navigation and timing receivers have objectives that are fundamentally different from communications receivers. Testing for adjacent band compatibility should not rely solely on SNR / C/No depression, especially since C/No is often ill-defined in cases where waveform on waveform structural interaction dominates,” Mr. Scott added. “Specific type compatibility testing is needed. If the objective is to measure adjacent band compatibility with LTE signals, you need to test with LTE signals. White noise is an inadequate and overly benign stand-in which can lead to overly permissive interference tolerance masks.” – Paul Kirby, paul.kirby@wolterskluwer.com
Courtesy TRDaily