ICLE Comments RE: Monitoring DISH’s Compliance
We submit this letter and the attached study in response to the Public Notice seeking supplemental comments on VTel’s Petition for Reconsideration[1] of the extension of construction deadlines for certain licenses held by EchoStar Corporation. As demonstrated below in our examination of select filings, we identify evidence of technical discrepancies in the showings made by EchoStar that suggest the possibility of a much larger problem with EchoStar’s license construction notifications. In our attached appendix we present the code necessary for replicating this work across all BEAs. We believe the FCC should conduct a thorough evaluation of EchoStar’s submissions and the network it claims to have constructed. If the FCC’s analysis, as the study suggests, concludes that EchoStar failed to construct the network it agreed to build, then the FCC must take immediate action to ensure that EchoStar’s spectrum does not continue to remain fallow.
The U.S. wireless market is highly competitive because the Commission has taken meaningful steps to make spectrum available to companies that put the spectrum to work and to provide a high-speed mobile broadband service for consumers. These networks have enabled substantial growth in the U.S. economy and fostered the development of new and novel technologies that benefit all Americans. Spectrum used for modern mobile broadband networks has significant impacts on the U.S. economy and the future economic security of the country and when spectrum is not being put to its highest and best use, the impacts extend far beyond a licensee and its contractors. But the FCC’s options to allocate additional spectrum for mobile broadband service are constrained by several factors. First, the inventory of spectrum suitable for mobile broadband service must be identified and licensed. Second, there is no greenfield spectrum available; therefore, incumbent operations must be protected or moved to different bands – all of which takes years. Even then, the FCC’s authority to auction spectrum has lapsed, so the FCC does not have general authority to auction spectrum that is newly allocated for mobile broadband service.
With the dearth of options for acquiring new spectrum, the FCC must ensure that allocated and licensed spectrum is put to use. Failure by any licensee to comply with its obligations to use the spectrum should be met with swift action so the licenses can be reauctioned or sold to an entity that will make use of this valuable and highly constrained resource. This is why it is so important for the FCC to confirm whether or not EchoStar is warehousing spectrum that could otherwise be put to work.
Over many years, EchoStar has amassed substantial, valuable spectrum assets with a history of repeatedly seeking extensions of time for putting those resources to use in the public interest. In 2017 and 2018, EchoStar accepted penalties rather than meeting interim deployment milestones, effectively delaying its obligation to deploy initial service.[2] In 2019 and 2024, it sought extensions of time to allow it to further delay deployment.[3] Now there is substantial doubt as to whether EchoStar has met even its delayed obligations. In our review of the record in this docket, we are particularly persuaded by some technical discrepancies in the EchoStar data identified by Spectrum Financial Partners, LLC (Spectrum Financial):[4]
- First, Spectrum Financial noted that EchoStar demonstrated coverage for mid-band spectrum licenses exclusively using coverage data for the 600 MHz band, which has significantly better propagation characteristics than the mid-band spectrum licensed to EchoStar.[5] This approach seems inherently inconsistent with the Commission’s construction requirements.
- Second, Spectrum Financial points out that there is a question regarding whether the link budget for the coverage compliance filings made by EchoStar—which rely on a 56 kbps uplink target—are compliant with its commitments to provide 5G “broadband” service. Spectrum Financial highlights that the data used for the construction filings is much lower than the 320 kbps target used as the limit for voice coverage in EchoStar’s Broadband Data Collection filings.
- Third, Spectrum Financial provides actual field measurements throughout ICLE’s home state of Oregon showing minimal activity on EchoStar licensed bands. These measurements are not consistent with the robust network EchoStar claims to operate.
Spectrum Financial’s findings regarding EchoStar’s use of spectrum in Oregon appear to be supported by ICLE’s independent review of public data. As Spectrum Financial documents, EchoStar’s construction certifications tell—at most—half of the story. The AWS-3 I Block build showing for BEA055 attached by Spectrum Financial, for example, contains no coverage data for the license for which it is filed. Instead, it shows coverage for an entirely different spectrum band.[6] In fact, in each coverage map reviewed by ICLE, EchoStar certifies that the “Minimum Received Signal Level (RSRP) is Uplink Limited” and “value shown/plotted is for n71,” where n71 is the 600 MHz band. Thus, as the best can be ascertained, for each map reviewed by ICLE, it appears that EchoStar is not demonstrating coverage solely of its AWS-3 I Block spectrum license, but instead is showing coverage using multiple licenses. To put it another way, implicit in EchoStar’s presentation is the claim that the areas of 600 MHz uplink coverage—represented in the RSRP plots—also define the areas where service exists for its other licensed spectrum. In effect, EchoStar is asking the Commission to treat its 600 MHz uplink map as not just one component of service coverage, but as the complete intersection of uplink and downlink coverage across all bands. This would only be valid if, in every case, the downlink signal from AWS-3, AWS-4, H Block, or 700 MHz is weaker than—or at best equal to—the 600 MHz uplink signal. Given the known differences in propagation, building penetration, and spectrum characteristics, this outcome is highly unlikely.
This raises significant legal and policy concerns:
- As ICLE understands it, EchoStar is using the band for which it is filing for the base-to-mobile downlink and using the 600 MHz band for the mobile-to-base uplink. In the case of paired spectrum, like AWS-3 Blocks G-J and the AWS H Block, EchoStar thus appears to not use half the spectrum for which it is licensed—the uplink portion of the band for which it is filing a construction certification.
- EchoStar’s construction filings for the AWS-3 A1 and B1 blocks claim to be uplink limited and use Reference Signal Received Power (RSRP) plots for the 600 MHz band. RSRP plots typically are used to measure downlink signal strength and quality. But the AWS-3 A1 and B1 blocks are uplink only bands. Aside from that, there is also no explanation how the coverage of the uplink transmissions from the AWS-3 band could be limited by the uplink transmissions from the 600 MHz band.
- EchoStar is asking the Commission to take on pure faith—while providing zero empirical data in support—that there is not a single spot in the U.S. where, due to some propagation anomaly or localized conditions, EchoStar’s 700 MHz, AWS-3, AWS-4 and H Block downlink coverage does not equal or exceed its 600 MHz uplink coverage (using a low edge speed of 56 kbps). The only way EchoStar’s claimed coverage boundaries could be correct is if, in every situation, the downlink signal from higher bands was always weaker than the uplink signal over 600 MHz — which defies both physics and field experience. This belies credulity, especially given the comparative network coverage included by Spectrum Financial that contrasts EchoStar’s claimed “AWS” coverage against the AWS coverage of the more mature T?Mobile and AT&T mid-band networks in the same area.
Concerned by these discrepancies, ICLE reviewed certain December 2024 ULS Construction showings by EchoStar which were filed to demonstrate EchoStar met the 80% accelerated coverage requirements. Each of these filings raised significant concerns. For example, EchoStar filed five license construction showings for BEA053—Pittsburgh, PA. Even though each filing purports to show coverage that is 600 MHz-limited (and therefore each filing should be identical), the coverage figures in the certifications ranged from 82.1% for WQTX252 to 82.6% for WQJY993 (although a variance of 0.5% may not seem large, consider that EchoStar purported to have met its national coverage obligation with a margin of only 0.08%).[7]
ICLE also compared the coverage data in those filings to EchoStar’s BDC voice coverage data.[8] ICLE overlaid the population data with Pennsylvania counties using Census Bureau GIS data, prorating the population by county overlap and H3 resolution 9 overlap. ICLE found that EchoStar’s coverage—aggregating 5G and voice—covered only 66.98% of the population in BEA053. EchoStar presumably will point to its use of a 56-kbps uplink threshold for coverage as explaining the difference in percentage of population covered. However, the incremental coverage difference between 5G with a 1 Mbps uplink and the 320-kbps voice uplink was only 0.15% of the population (4,337 people). Meanwhile, EchoStar claims that using a 56 kbps uplink threshold allows it to reach 82.6% of the population—an increase of approximately 15.62 percentage points over the 66.98% covered at the 320 kbps threshold (59.5% + 7.36% + 0.15%). It is implausible that reducing the uplink speed from 1 Mbps to 320 kbps yields only a 0.15% coverage increase, but reducing it further to 56 kbps would result in a tenfold increase in coverage. This sharp inflection suggests not a genuine expansion of usable service, but a use of technical thresholds to inflate coverage figures without meaningfully improving network performance. This inconsistency is illustrated in the chart below showing a waterfall of covered population versus uplink speed, with the line showing the decrease in speed.
ICLE also compared EchoStar’s six showings[9] in BEA011 (and PEA048, which comprises the identical set of counties) for the Harrisburg, PA market with its BDC filings for both December 31, 2024, and June 30, 2024. Applying the same population coverage methodology, EchoStar’s BDC data for the second quarter—three months before it asked for an extension—shows EchoStar had already met its “accelerated” requirement and that EchoStar’s coverage was only 0.06% less (738 people) than it was as of the December deadline. As much as EchoStar now states it made “significant investments” in reliance on the extension and that its extension was not a “simple plea for more time,” EchoStar’s commitment with respect to Harrisburg—and ICLE presumes many other markets—was simply the fulfillment of an empty promise.
That suggests there may have been no quid in the quid pro quo EchoStar alludes to when it represents “the milestone extensions that EchoStar received were in exchange for the acceleration of the milestones of other licenses as well as substantial new commitments” (emphasis in original).[10] If EchoStar were not making the substantial commitments it was purporting to make to justify the extension—and instead was relying on investments made in the ordinary course of business—the FCC has good cause to reconsider the extension that was previously granted.
ICLE recognizes that there may be good reason for companies to seek and obtain extensions of construction deadlines. However, the basis for which EchoStar sought the extensions and the process the FCC undertook to grant them do not give any comfort that EchoStar’s history of spectrum warehousing will stop anytime soon. Indeed, even in the best case, if EchoStar complies with each of its Commitments, the FCC (and the American people) have no guarantee that EchoStar will put the spectrum to use for another three years. Companies are starving for spectrum and the only viable course at the moment is acquisition through secondary markets. In light of the value of these assets to the American public, it is thus entirely appropriate for the Commission to reconsider its continued tolerance for EchoStar’s warehousing practices—especially in light of what appear to be wildly exaggerated claims about what it has actually accomplished and what it committed to accomplish. Spectrum is simply too important to the U.S. economy and to U.S. security interests to be permitted remain unused in EchoStar’s hands.
In the attached appendices we include the output of our code analysis, visualizations of coverage maps generated, and the Python source code that can be used to recreate this analysis for any set of BEAs in the United States.
Thank you for the opportunity to provide these reply comments.
[1] Petition for Reconsideration of Vtel Wireless Inc., WT Docket No. 22-212 (Oct. 21, 2024), https://www.fcc.gov/ecfs/search/search-filings/filing/1021167931859.
[2] See Letter from Donald K. Stockdale, Jr., Chief, Wireless Telecommunications Bureau, Federal Communications Commission to Jeffrey H. Blum, Senior Vice President & Deputy General Counsel, DISH Network LLC (Jul. 9, 2018), available at https://docs.fcc.gov/public/attachments/doc-352379a1.pdf.
[3] See Wireless Telecommunications Bureau Seeks Supplemental Comment on Vtel’s Petition for Reconsideration of the Extension of Construction Deadlines for Certain Licenses Held by Echostar Corp., WT Docket No. 22-212 (May 12, 2025), available at https://docs.fcc.gov/public/attachments/DA-25-404A1.pdf.
[4] Comments of Spectrum Financial Partners LLC, WT Docket No. 22-212 (May 27, 2025), https://www.fcc.gov/ecfs/search/search-filings/filing/10528678810841.
[5] It is not entirely clear what EchoStar’s statements concerning its coverage mean, as it uses the term “Uplink Limited” when describing them, without further explanation. At a minimum, the commission should require clarification from EchoStar before accepting its filings.
[6] See, e.g., FCC ULS File Number 0011376642.
[7] See, e.g., Letter from EchoStar Corp., WT Docket No. 22-212 (Jan. 10, 2025), https://www.fcc.gov/ecfs/document/1011082930532/1.
[8] The FCC recently released data as of Dec. 31, 2024—the exact date that EchoStar’s coverage obligations under the September 2024 extension matured. ICLE downloaded EchoStar’s Pennsylvania H3 coverage files from the BDC website for 5G-NR (35/3 mbps), 5G-NR (7/1 Mbps) and mobile voice, as well as 2023 H3 400m population data from the Humanitarian Data Exchange: https://data.humdata.org/dataset/kontur-population-united-states-of-america.
[9] These call signs were taken from EchoStar’s September 2024 request for an extension. See Letter from EchoStar Corp., WT Docket No. 22-212 (Sep. 18, 2024), https://www.fcc.gov/ecfs/document/1091867842711/1.
[10] Comments Of EchoStar Corp., WT Docket No. 22-212 (May 27, 2025), https://www.fcc.gov/ecfs/document/105280597605377/1.