US 9065609

I'll search for the specific patent number and any related litigation/docket records.

Let me run a few more targeted searches to check for any 2026 litigation or docket activity specifically naming this patent number.

US Patent 9,065,609 — Verification Summary

Important caveat on scope of my search: I located the patent's face record (Google Patents, Justia, Golden wiki, Patents-Review) and performed targeted searches for litigation/PTAB/CAFC activity. I found no CAFC 2026 docket entry, no district court complaint, and no IPR/CBM petition naming U.S. Patent No. 9,065,609 specifically. Search hits containing "'609" in docket contexts refer to different patents — e.g., U.S. 9,890,609, 11,129,163, 9,106,609, 8,676,609, and 9,147,609 — and must not be conflated with 9,065,609. Treat the litigation-negative result as "not found in the sources retrieved," not as an affirmative confirmation of zero litigation.

Bibliographic Data (face of the patent and Google Patents record)

Security-interest history (per Google Patents legal events): Wells Fargo Bank, N.A. took security interests in 2020 (Reel 052935/0001) and 2021 (Reel 057280/0028); both were released on Oct 11, 2023 (Reels 065231/0466 and 065653/0463). The release instrument lists both National Instruments Corporation and Phase Matrix, Inc. I did not find an assignment transferring title away from National Instruments.

Abstract (verbatim)

"A mechanism for blind estimation of parameters for correcting I/Q impairments. Complex samples of a complex baseband signal are received from a receiver. A cross-correlation is computed between an I component and a Q component of the complex samples. A mean square value is computed for the I component of the complex samples; and a mean square value is computed for the Q component of the complex samples. A cross-channel gain estimate is: computed based on the cross-correlation value and one or both of the mean square values; and used to apply a cross-channel gain correction to the complex samples. An estimate of an I/Q gain imbalance is computed based on the mean square values. The gain imbalance estimate is useable to correct an I/Q gain imbalance present in the complex samples. The parameters may be supplied to the receiver, enabling the receiver to apply online corrections."

Plain-Language Overview of the Independent Claims

Claim 1 — Method (blind estimation of receiver correction parameters).
A receiver is calibrated without knowing what the transmitter sent ("blind"). The method:

• (a) obtains complex baseband samples from a receiver;
• (b) computes a time-domain cross-correlation between the I and Q components;
• (c) computes mean-square values for I (I_MS) and Q (Q_MS);
• (d) derives an estimate of cross-channel gain k from the cross-correlation and one/both mean-square values;
• (e) applies a cross-channel gain correction using that estimate to produce modified samples;
• (f) derives a gain-imbalance estimate from I_MS and Q_MS, usable to correct I/Q gain imbalance in the samples.
  The claim then adds DC-offset handling: estimate an I-channel DC offset and a Q-channel DC offset and apply DC-offset corrections to the samples.

Key limitation to note: claim 1 expressly requires the cross-correlation to be a "time-domain cross correlation" — a limitation that matters for any invalidity/eligibility analysis, since several family-adjacent references (e.g., US 8,477,889) describe frequency-domain or correlation-independent blind estimators.

Claim 17 — Non-transitory computer-readable memory medium. Essentially the same (a)–(f) algorithm plus the I/Q DC-offset estimation-and-correction steps, implemented as stored program instructions executed by a processor.

Claim 19 — System. A processor plus memory storing program instructions causing the processor to perform the same (a)–(f) algorithm and the DC-offset estimation/correction steps. (Note: the claim's preamble/body grammar — "cause the processor to: (a) receiving…" — is carried verbatim from the issued patent; I am not correcting it.)

Notable dependent claims: claim 5 (apply I/Q gain-imbalance correction to get corrected samples); claim 6 (display the corrected samples); claim 8 (DC-offset correction precedes steps (b)–(f)); claim 9 (steps (a)–(f) performed by a computer external to the receiver, with parameters stored in its memory); claim 10 (transfer k and gain-imbalance estimates from computer memory to the receiver to enable online correction — the commercial heart of the disclosure); claim 11 (receiver has a programmable hardware element, e.g., FPGA, to apply corrections); claim 12–13 (supply complex samples to a transmitter and direct transmission; digitally-modulated baseband signal); claim 14 (steps performed by the receiver itself); claim 15 (transmitter is a base station); claim 16 (receiver in a wireless communication device).

Specifications of Interest (from the description)

• The math is explicitly stated: k̂ = R̃_iq(N) · I_MSINV, with Q_M(n) = Q_T(n) − k̂ I_T(n); gain imbalance Â_QI = σ̂_Q(N)/σ̂_I(N); correction Q_C(n) = Q_M(n)·(1/Â_QI). DC offsets Ī_S and Q̄_S are simple sample averages.
• The model (FIG. 5) treats I/Q impairment as a composite of transmitter + receiver distortions — gain imbalance A_QI, cross-channel (I→Q) gain k, and per-channel DC offsets.
• Characterization data: for a 60 GHz system, ~18 dB image-rejection improvement (FIG. 7A) and ~51 dB carrier-feedthrough reduction (FIG. 7B). The patent expressly states the method is not limited to 60 GHz or to RF.
• Two algorithms are distinguished: an online blind algorithm (over-the-air input, no images assumed) and an offline LO/transmitter-tone calibration algorithm (closed-form solution using a log detector).
• Cited prior art of record: US 6,785,523 (Atheros), US 2007/0058755 (Husted), US 8,415,895 (Broadcom), US 8,385,457 (Intel); non-patent literature: Shafiee & Fouladifard, "Calibration of IQ Imbalance in OFDM Receivers," IEEE ICC 2003, pp. 2081–2085.
• Forward citations of record include National Instruments' own later OTA/mmWave test patents (US 10,790,915; US 11,515,950; US 11,982,699) and US 11,050,496.

Uncertainties / Limitations of This Report

1. No CAFC 2026 docket activity found for this patent. My searches of the CAFC September 2026 argument calendar and general docket references returned nothing tied to 9,065,609. I cannot rule out a docket under a different naming convention or one too recent to be indexed; PACER/Docket Navigator would be needed for certainty.
2. Current ownership is presumed, not confirmed. The 2023 filings are releases of security interests, not assignments, so "National Instruments Corporation" appears to remain owner of record, but I did not retrieve a current USPTO PatentCenter title report to confirm.
3. "609" ambiguity risk is high. Numerous unrelated patents end or begin with "609." Any docket or PTAB hit phrased only as "the '609 patent" should be verified against its full 8-digit number before being attributed to this patent.
4. Examiner name is blank in the sources retrieved; I will not guess it.

I'll search for litigation involving US patent 9065609 across litigation-tracking sources.

The initial searches surfaced the patent and unrelated '906 patents. Let me run more targeted searches on litigation-tracking resources.

Known Litigation Involving U.S. Patent No. 9,065,609

Result: No litigation is known or identifiable involving U.S. Patent No. 9,065,609.

I searched litigation-tracking and patent-docket sources, including the Unified Patents litigation portal, Justia Dockets, CourtListener/RECAP, RPX Insight, and Docket Alarm, using the exact identifiers "9065609," "US9065609," and "9,065,609." I found no district court complaints, ITC investigations, PTAB proceedings, or Federal Circuit appeals in which this patent was asserted, challenged, or otherwise at issue.

What the search did return (and why it is not the patent you asked about)

Per your instruction to interpret identifiers literally and not substitute similar numbers, I want to be explicit that the apparent "hits" were false positives:

• "9065609" as a serial/document number in unrelated contexts — e.g., a Russian electromagnetic flowmeter calibration certificate (pressure transducers "зав. №№ 9065609, 9065600"), a Brazilian court dispatch number, and an IEEE paper DOI (10.1109/COBEP/SPEC44138.2019.9065609). These have nothing to do with the patent.
• Eolas v. Adobe/Google/etc. — references to a different "906 patent" (U.S. 5,906,854, the Eolas '906 patent), docket 6:09-cv-00446 (E.D. Tex.). Not U.S. 9,065,609.
• Various National Instruments cases — NI has been a frequent litigant (e.g., National Instruments v. The MathWorks, the LabVIEW patents 4,901,221 / 4,914,568 / 5,301,336; Cian IP v. National Instruments; SoftVault v. National Instruments; TurboCode v. National Instruments, No. 7:2025-cv-00084 (W.D. Tex.)), but none of these asserted U.S. 9,065,609.
• The patent's own Google Patents record lists no litigation and shows only three forward citations to later National Instruments patents (US 10,790,915; US 11,515,950; US 11,982,699) — these are citations, not enforcement actions.

Patent identification (for the record)

Caveats

• Absence of evidence is not conclusive proof of absence. Not all district court activity is fully captured in free databases, and dockets can lag PACER. A definitive check would require a PACER-fee-based search and a Westlaw/Lexis or Lex Machina / Docket Navigator litigation search.
• The patent is owned by National Instruments (with security-interest/release records with Wells Fargo and Phase Matrix, Inc. and no re-assignments of title shown). A non-practicing entity could have acquired or asserted it, but I found no such record.
• Related NI I/Q-correction patents in the same family (e.g., U.S. 8,638,893 and U.S. 8,559,560) exist, but per your instruction I did not substitute them for the patent you specified; a search on those numbers would be a separate query.

Bottom line: Based on the sources searched, I have no record of any litigation involving U.S. Patent No. 9,065,609. I cannot supply plaintiff, defendant, jurisdiction, case number, filing date, or status because no such case was identified.

I'll verify the structured "no proceedings" default against live sources before drafting.

Let me run a few more targeted checks before finalizing.

Proceedings overview

Total AIA trial proceedings on file for U.S. Patent No. 9,065,609: 0 — zero IPRs, zero PGRs, zero CBM reviews, zero derivations, and zero terminations of any kind (breakdown: 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denials). The structured "PTAB proceedings on file" block (USPTO Open Data Portal ingest) returns an empty set, and my independent web checks of PTAB-facing and litigation-tracking sources returned no AIA trial numbers, petitioner names, or Board papers tied to this patent. Bottom-line defensive posture for a defendant: there is no PTAB paper trail to work from — the patent is completely un-challenged at the Board, all 19 issued claims (1–19) stand untested and un-narrowed, and your petition would be a first-instance test of art nobody has ever put in front of a panel. This is not the "claims 1–5 have been canceled" scenario; nothing has been canceled, and nothing has been sustained either.

Because there are no proceedings to enumerate, the per-proceeding template is reproduced below in null form so the file is complete and the absence is documented rather than merely asserted.

No proceeding — no petitioner v. National Instruments Corporation (no AIA trial ever instituted)

• Type: n/a — no Inter Partes Review, Post-Grant Review, or Covered Business Method review on file
• Filed: n/a
• Status: n/a. Structured source of record (USPTO ODP "PTAB proceedings on file" block) returned an empty list; no proceeding number exists to report. I am not supplying a proceeding number because none exists — inventing one would be fabrication.
• Judge panel: n/a — no panel has ever been assigned
• Petition grounds: n/a — no § 102, § 103, § 112, or § 101 ground has ever been pleaded against these claims
• Institution decision: n/a — no institution decision, discretionary or on the merits
• Final Written Decision: n/a — no FWD has ever addressed any of claims 1–19; I make no statement about any claim's patentability before the Board
• Settlement / termination: n/a
• Appeal: n/a — no PTAB decision, therefore no CAFC appeal of a PTAB decision
• Defensive value: Total first-mover freedom. No § 315(e) estoppel has attached to anyone, no panel has construed "time-domain cross correlation," and no claim has been canceled or narrowed by amendment. Conversely, you get no free roadmap: no institution decision, no FWD claim-level findings, and no Board reasoning to cite.

Strategic summary

Claim status: 100% untested. Claims 1–19 of 9,065,609 are all UNTESTED by the PTAB. There are no CANCELED claims and no SUSTAINED claims, because no proceeding ever reached institution, let alone a final written decision. The independent claims — claim 1 (method), claim 17 (non-transitory CRM), and claim 19 (system) — remain in their as-issued form, including the express "wherein the cross-correlation is a time-domain cross correlation" limitation that appears in all three. That limitation is the natural claim-construction battleground in any dispute, and it remains definitively unconstrued by the Board: a validly issued claim term with zero PTAB or (so far as I could find) judicial gloss. For a defendant, that cuts both ways — you have no adverse construction to work around, but you also have no panel finding that a reference discloses a time-domain cross-correlation to lean on.

Estoppel landscape: wide open. Because no IPR was ever instituted, 35 U.S.C. § 315(e)(2) estoppel attaches to no one. No petitioner, real party in interest, or privy is barred from raising any § 102/§ 103 ground based on patents or printed publications, whether or not it was previously raised. Two practical gates remain, and both are discretionary rather than estoppel-based:

• § 325(d) (art previously presented). The prosecution record already contains four U.S. references and one non-patent publication: US 6,785,523 (Atheros); US 2007/0058755 (Husted); US 8,415,895 (Broadcom); US 8,385,457 (Intel); and Shafiee & Fouladifard, "Calibration of IQ Imbalance in OFDM Receivers," IEEE ICC 2003, pp. 2081–2085. A petition built primarily on that art invites discretionary denial. The stronger play is new art — e.g., blind I/Q-imbalance estimators that are correlation-independent (US 8,477,889 in the neighbouring art space is instructive precisely because it teaches away from relying on I/Q correlation) or time-domain vs. frequency-domain distinctions.
• The 2025–2026 discretionary-denial climate. This patent was filed 2013-07-15 and issued 2015-06-23; it is now an old, dormant patent at a company that is not an NPE. Retrieved commentary describes a marked tightening of institution, including the Director's centralised "settled expectations" discretionary-denial practice and proposed rules requiring petitioners to abandon parallel § 102/§ 103 defences and making denial mandatory where a claim was previously upheld elsewhere. That reporting is third-party analysis, not a Board holding applied to this patent, but it makes the age of this patent a live institutional risk rather than a neutral fact. Budget for a robust § 314(a)/§ 325(d) rebuttal in the Preliminary Response cycle.

Pattern signals: none — the trail is completely cold. No petitioner has ever filed on this patent (so no serial-petition risk), the patent owner has never had to defend at the Board (so no history of aggressive PTAB appeals, and no motion-to-amend practice to read), and there is no defensive aggregator in the chain — no Unified Patents, RPX, or other third-party challenge is on file. The only third-party filings in this patent's orbit are citations: the patent's own forward-citation table lists later National Instruments OTA/mmWave test patents (US 10,790,915; US 11,515,950; US 11,982,699; and US 11,050,496), which are products of the same assignee's portfolio, not enforcement activity. Combined with the litigation section of this report (no district court, ITC, or CAFC activity found for 9,065,609), the picture is a patent that appears to have been held, maintained (4th-year fee 2018-11-26; 8th-year fee 2022-12-23), and never asserted or challenged. The absence of IPR activity here is itself the signal: this is not a well-asserted patent that has attracted challengers — it is an unasserted one.

Recommended next steps

1. Do not expect a free invalidity roadmap. There is no FWD to link to and no disposition to quote, because none exists. If opposing counsel or a demand letter implies otherwise, treat any reference to "the PTAB's findings" or "IPR2026-00xxx" on this patent as unsupported until a real proceeding number is produced and verified against PTAB E2E / PTAB Decisions. The "'609" ambiguity risk is acute — several unrelated patents share that shorthand, so verify the full eight-digit number before crediting any proceeding attributed to it.

2. Challenge-window audit (all dates YYYY-MM-DD).

   • PGR: unavailable. The window closed nine months after issuance, i.e., on or about 2016-03-23. Long expired.
   • CBM: unavailable. The transitional programme stopped accepting petitions on 2020-09-16, and in any event this patent is directed to I/Q impairment correction, not a financial product or service.
   • IPR: available. No § 315(b) one-year bar has run against anyone, since the search found no complaint served on any party. § 315(a)(1) bars only a petitioner that previously filed a civil action challenging validity.

3. If you file, lead with the claim-1 limitation. All three independent claims require the cross-correlation to be a time-domain cross-correlation. Any art that estimates imbalance in the frequency domain, or that is expressly correlation-independent, fails the limitation rather than meeting it — which is a double-edged tool (it defeats sloppy invalidity theories and it is the term the patent owner will lean on defensively). Build the ground around a reference that performs a genuine time-domain <I,Q> correlation product average, or concede the construction and attack elsewhere.

4. Pre-empt discretionary denial. Because every reference of record was already before the examiner, and because the patent is now over a decade old, plan for § 325(d) and § 314(a) arguments at the institution stage rather than after. Confirm the current ownership posture before filing — the 2023 filings of record are releases of Wells Fargo security interests (Reels 065231/0466 and 065653/0463, effective 2023-10-11) naming National Instruments Corporation and Phase Matrix, Inc., not assignments of title; I found no assignment moving title away from National Instruments. Fee status is current through the 8th-year payment.

5. Track it. The patent is Active with an adjusted expiration of 2033-08-26, leaving roughly seven years of term. Given the practical difficulty of getting older patents instituted in the current climate, the most likely future PTAB event is no event at all — but if the patent is ever asserted, an IPR filing will move the Board's docket quickly, and you should expect the statutory 1-year trial clock from institution to drive the schedule.

Caveats. (a) The empty proceedings list is sourced from the structured ODP block supplied to me plus my own searches; a definitive negative would require a direct PTAB E2E / PTAB Decisions docket query, which can surface very recent filings before they are indexed elsewhere. (b) I have not assigned an examiner name or any judge panel, because none is public on the sources retrieved. (c) The 2025–2026 discretionary-denial discussion rests on third-party commentary and policy documents, not on any adjudication of this patent; treat it as climate, not precedent.

I'll research this using the assignment records and cross-reference the entities involved.

I have the core ownership-chain facts. Let me verify the inventor's affiliation and attempt to retrieve the recording correspondent.

Inventors

Pattern note (not a fire-sale indicator here). McCoy is a serial corporate inventor, not a one-off. His published-applications profile lists [Apple Inc.](/litigations/by-plaintiff/Apple%20Inc.) (36) and Freescale Semiconductor (17) as his two largest assignee relationships alongside National Instruments (5), and he appears on 2024–2026 filings assigned to CesiumAstro, Inc. His NI-assigned applications cluster in the 2013–2018 window (including US 2015/0016571, the pre-grant publication of this patent, and NI MIMO/channel-sounding calibration filings), so the NI-assignment phase overlaps the filing of the '609 application. Career mobility across employers is visible, but there is no evidence that this patent followed him out the door — the '609 patent remained with NI and NI paid the 4th- and 8th-year maintenance fees (2018, 2022). Accordingly, the "inventor departed within 12 months → portfolio fire-sale" inference does not apply. Caveat: I could not date his exact start/end dates at NI from the sources retrieved; the employments above are inferred from assignee-of-record on his applications, not from employment records.

Original assignee

• Entity on the face of the patent: National Instruments Corporation (Austin, TX), per the 2013 assignment at Reel 030798/0362.
• Line of business: Automated test, measurement, and instrumentation — PXI/modular instruments, LabVIEW software, RF/microwave and (via Phase Matrix) millimeter-wave test hardware. The patent's own characterization data are for "a particular 60 GHz RF system," consistent with NI's RF/mmWave product line.
• Product embodying the claims: Yes. The disclosure describes correction circuitry placed "after the analog-to-digital conversion circuitry of the receiver" (FIG. 10), implemented in a programmable hardware element (FPGA), ASIC, or processor — i.e., functionality that maps directly onto NI's PXI/modular RF receivers and its LabVIEW FPGA codebase. NI's later OTA/mmWave test patents (US 10,790,915; US 11,505,496; US 11,515,950; US 11,982,699) cite this patent, showing continued internal use in the same product family.
• Current status: Acquired — but operating. Emerson Electric Co. (NYSE: EMR) closed its acquisition of NI for $8.2 billion equity value on October 11, 2023; NI became Emerson's "Test & Measurement" reporting segment, still headquartered in Austin. NI is not dissolved and not in bankruptcy. Notably, no assignment of title to Emerson is recorded against this patent — the record still names National Instruments Corporation, which appears to remain the holding entity (an Emerson subsidiary).

Assignment timeline

The Assignment Center record for US 9,065,609 contains one title transfer plus three lender filings (two grants, one release transaction covering two reels). All reel/frame values below are taken from the Google Patents legal-events table for US 9,065,609, which mirrors the USPTO Patent Assignment Search records. The "Correspondent of record" field was not exposed in any source I could retrieve (Google Patents legal events, Justia, Patents‑Review, patentguru); I am therefore leaving it blank rather than guessing.

1. 2013-07-15 (executed) / recorded 2013-07-15 — Reel 030798/0362

   • Conveyance: Assignment (Assignment of Assignors' Interest)
   • Assignor: McCoy, James W. (individual inventor)
   • Assignee: National Instruments Corporation, Texas
   • Correspondent: not retrievable from the sources searched
   • Context: Initial employee-inventor assignment to employer — routine, recorded on the filing date of application 13/942,065.

2. 2020-06-12 (executed) / recorded 2020-06-14 — Reel 052935/0001

   • Conveyance: Security Interest (Grant of security interest / collateral assignment)
   • Assignor (Grantors): National Instruments Corporation AND Phase Matrix, Inc.
   • Assignee (Secured Party): Wells Fargo Bank, National Association, North Carolina
   • Correspondent: not retrievable
   • Context: Securitization (loan collateral) — this is the patent-level recording of NI's Collateral Agreement dated June 12, 2020; Phase Matrix, Inc. is NI's wholly owned RF/microwave subsidiary (acquired May 20–23, 2011 for ~$38–40.7M). No title transfer; NI is pledging assets to a lender.

3. 2021-06-18 (executed) / recorded 2021-06-18 — Reel 057280/0028

   • Conveyance: Security Interest (Amended and Restated Collateral Agreement)
   • Assignor (Grantor): National Instruments Corporation
   • Assignee (Secured Party): Wells Fargo Bank, National Association, North Carolina
   • Correspondent: not retrievable
   • Context: Securitization / refinancing — restates and continues the 2020 lender lien; still collateral, not title.

4. 2023-10-11 (executed) / recorded 2023-10-13 — Reel 065231/0466

   • Conveyance: Release of Security Interest (releasing Reel/Frame 057280/0028)
   • Assignor (Releasing Party): Wells Fargo Bank, N.A., as Administrative Agent
   • Assignee/Released Party: National Instruments Corporation, Texas
   • Correspondent: not retrievable
   • Context: Lien release — payoff/termination of the lender lien. The execution date (Oct 11, 2023) is the exact date Emerson closed its acquisition of NI, consistent with a change-of-control payoff of the credit facility.

5. 2023-10-11 (executed) / recorded 2023-10-13 — Reel 065653/0463

   • Conveyance: Release of Security Interest (releasing Reel/Frame 052935/0001)
   • Assignors/Released Parties: Phase Matrix, Inc., California and National Instruments Corporation, Texas (two entries on the same transaction)
   • Assignor (Releasing Party): Wells Fargo Bank, N.A., as Administrative Agent
   • Correspondent: not retrievable
   • Context: Lien release — discharges the 2020 collateral grant as to both Phase Matrix and NI.

Net title position: After 2023-10-13 the only recorded encumbrance history is closed out. The only recorded transfer of title on this patent is the 2013 inventor→NI assignment; everything after that is lender collateral and its release. Maintenance fees were paid at the 4th year (2018-11-26) and 8th year (2022-12-23), and the patent is Active with an adjusted expiration of 2033-08-26.

Timeline diagram

timeline
    title Ownership of US 9065609
    2013 : Filed by James W McCoy
         : Assigned to National Instruments
    2015 : Patent granted
    2020 : Wells Fargo security interest
    2021 : Amended Wells Fargo security interest
    2023 : Security interests released
         : NI acquired by Emerson

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. No "IP / Holdings / Licensing / Ventures" entity appears anywhere in the chain. The only parties on record are National Instruments Corporation (operating), Phase Matrix, Inc. (operating subsidiary), and Wells Fargo Bank, N.A. (a bank acting as secured party). Reels 052935/0001, 057280/0028, 065231/0466, 065653/0463.

2. Known asserter in the chain — NOT PRESENT. No assignee in the chain matches Acacia, Marathon, Intellectual Ventures, IPNav, Wi‑LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, or any Spangenberg entity. Current/historic owner is a practicing manufacturer.

3. Repeat correspondent across the chain — UNCLEAR (data not retrieved). The USPTO Assignment Center correspondent field for this patent's recordings was not returned by the sources I could access, so I cannot state who recorded the filings or whether one attorney recurs. I decline to infer recurrence from the parties' names. (If forced to characterize: three of the four post-2013 recordings are lender-side instruments that one would expect to share a single bank-side recording agent — but that expectation is not evidence, and it is not an NPE tell in any event.)

4. Cascading transfers — NOT PRESENT. There is no chain of consecutive LLC-to-LLC assignments. The post-issuance activity is two lien grants and their releases, spread over three years (2020 → 2021 → 2023), which is ordinary secured-lending cadence, not rapid shell-chaining.

5. Pre-litigation transfer — NOT PRESENT. No infringement suit naming US 9,065,609 was located (consistent with the litigation summary already generated), so there is no transfer within 6 months of a suit. No venue/standing-motivated assignment appears.

6. Bankruptcy fire-sale — NOT PRESENT. NI exited via an $8.2B cash acquisition by Emerson (closed 2023-10-11), not a Chapter 7/11 proceeding. There is no sale-of-portfolio-in-bankruptcy event, and no NPE bidder entered the chain.

7. Privateering — NOT PRESENT. NI did not transfer the patent to an NPE to assert against competitors; it retained title itself (and still holds it as an Emerson subsidiary). No SEC disclosure, Patent Progress, or EFF coverage of an NI-funded assertion vehicle tied to this patent was found.

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. The 2023 endpoints are lien releases returning the patent to clean, encumbrance-free ownership by NI.

Verdict

Operating-company assertion.

The ownership chain is a textbook operating-company provenance: a single inventor assignment to National Instruments Corporation (Reel 030798/0362, executed/recorded 2013‑07‑15), followed only by lender collateral filings to Wells Fargo (Reels 052935/0001 and 057280/0028) and their releases on 2023‑10‑11/13 (Reels 065231/0466 and 065653/0463). There is zero NPE evidence in the record — no shell holdings LLC, no known asserter, no cascading LLC chain, no bankruptcy sale, and no assignment to a defensive aggregator — and the current owner is a practicing manufacturer (NI, now Emerson's Test & Measurement segment) whose products embody the claimed I/Q-correction circuitry.

Caveat on the label: no infringement suit naming this specific patent was located, so the "assertion" prong of this verdict is not demonstrated for US 9,065,609; read the verdict as "operating-company-held / not an NPE," the nearest available bucket. Two open items would sharpen the picture: (i) the Assignment Center correspondent-of-record field, which I could not retrieve and which is required to move signal 3 off "unclear," and (ii) confirmation whether any merger/change-of-name recording to Emerson exists that is not surfaced in the Google Patents legal-events table.

Verify at: https://assignmentcenter.uspto.gov/ (search "9065609" or application 13/942,065) — mirrored at https://assignment.uspto.gov/patent/index.html.

I'll conduct targeted searches for the patent and each cited reference of record, grounding each against the authoritative full text already provided.

US 9,065,609 — Prior-Art Analysis of the Citations of Record

Scope note: Per the operating rules, I interpret "9065609" literally and exclude similar numbers. The record (Google Patents / Justia / FPO, which mirror USPTO data) confirms US 9,065,609 B2 is unique and matches the full text already provided. I retrieved partial claim text for the cited references; where I could not verify a reference's full text I say so rather than fill gaps.

New datapoint / contradiction flag: The earlier "Patent summary" section stated the examiner name was blank. Justia's record now returns Primary Examiner: Lihong Yu. That is a correction to the earlier summary. Everything else in the earlier sections is consistent.

1. The complete prior-art "citations" set (U.S. patent + NPL) for 9065609

The full authoritative text lists exactly four U.S. patent citations and one non-patent citation. There are no foreign patent citations of record. (Note: the forward citations — US 10,790,915; US 11,050,496; US 11,515,950; US 11,982,699 — are later NI patents that cite this patent and are not prior art.)

Verbatim note on #5: the face record prints the authors as "H. Shafiee, H. and S. Fouladifard," which is internally inconsistent (the leading "H." is repeated). I am reproducing it as recorded rather than correcting it.

All five are valid prior art to 9065609. Because Appl. 13/942,065 was filed 2013-07-15 (post‑AIA), AIA 35 U.S.C. § 102 governs. Each reference was publicly available (§ 102(a)(1)) before that date; #2 and #3 also qualify under § 102(a)(2) as "effectively filed" earlier (2002 and 2009, respectively).

2. Reference-by-reference analysis

Reference 1 — U.S. 6,785,523 B2 (Atheros Communications; Husted et al.)

• Description: An AGC system that distinguishes desired in-band signals from high-power out-of-band signals. It measures power before/after FIR filtering, then verifies a candidate in-band signal by a multi-threshold comparison of a normalized self-correlation obtained by "summing the product of each sample and the complex conjugate of one of a plurality of previously obtained samples taken a predetermined amount of time earlier; and normalizing the sum." Claims 7, 11 and 13 recite removing DC offset from the received signal.
• § 102 assessment: No claim of 9065609 is anticipated. The reference is a signal-detection/AGC disclosure, not an I/Q-impairment corrector. Mapping to claim 1:
  • (b) The '523 "self-correlation" is a sample-vs-delayed-sample autocorrelation of one real signal — not a cross-correlation between the I component and the Q component as claim 1 requires.
  • (c)/(d)/(e)/(f) — no I_MS/Q_MS computation, no cross-channel gain k, no k‑based cross-channel correction, no gain-imbalance estimate from I/Q mean squares.
  • It does recite DC-offset removal, touching the DC-offset element of claim 1, but that is the only overlapping element and cannot anticipate.
• Actual relevance: Background art showing that correlation and DC-offset removal were known; usable only in a § 103 combination.

Reference 2 — U.S. 2007/0058755 A1 (Husted; Atheros) — the closest reference

• Description: Passive ("blind-ish") I/Q mismatch calibration in a quadrature receiver using normally received data — emphasis in the disclosure that the received signal is "other than a signal received solely for the purposes of calibration." It performs statistical calibration over I and Q digital signals, computes a magnitude mismatch vector and a phase mismatch vector (plus a scalar frequency-independent phase term), and its claims recite "comparing signal powers of the in-phase and quadrature digital signals" and "determining the degree to which the in-phase and quadrature digital signals are correlated with one another." Family members: U.S. 7,158,586 B2 (granted 2007-01-02) and U.S. 7,672,656 B2 (granted 2010-03-02).
• § 102 assessment: No claim is fully anticipated, but this is the reference most likely to have driven the "time-domain cross correlation" limitation in granted claim 1. Element mapping:
  • "Comparing signal powers of I and Q" is a functional analogue of claim 1(c) (I_MS/Q_MS) and touches claim 1(f)'s gain-imbalance concept.
  • "Determining the degree to which I and Q are correlated" touches claim 1(b)→(d).
  • Gap that defeats anticipation: the '755 calibration is explicitly frequency-domain/statistical (FFT-based magnitude/phase mismatch vectors, scalar frequency-independent phase term). It does not disclose (i) a time-domain cross-correlation, (ii) a cross-channel gain k computed from that correlation and the mean squares and applied as Q_M = Q_T − k·I_T, nor (iii) the DC-offset estimation-plus-correction steps required by claims 1/17/19. Because § 102 requires every element in a single reference, it cannot anticipate.
• Cross-reference to this patent's text: the published claim 1 (US 2015/0016571) did not contain the "time-domain cross correlation" clause; the granted claim 1 does. This is consistent with the applicant narrowing to distinguish a frequency-domain/statistical correlation disclosure of the '755/Shafiee type.

Reference 3 — U.S. 8,416,895 B2 (Broadcom; Jensen et al.)

• Description: I/Q imbalance calibration in a receiver using a non-decision-directed (NDD) imbalance canceller followed by a decision-directed (DD) imbalance canceller (per related EP 1 703 633 A2). It evaluates an interferer/image signal metric — "a mean, a total energy, an average power, a mean square, an instantaneous power, a root mean square, a variance, a norm" — and iteratively adjusts sequence parameters (gain, phase) between a calibration mode and an operational mode to minimize imbalance.
• § 102 assessment: No claim anticipated. Its mean-square/RMS usage refers to an interferer/image metric used as a feedback cost function, not to computing I_MS and Q_MS of the received I and Q components and then deriving a cross-channel gain k as claim 1(c)–(d) require. It discloses no time-domain I/Q cross-correlation, no k‑subtraction correction, and no DC-offset estimation/correction as claimed. The calibration‑mode/operational-mode distinction touches the spirit of claims 9–10 but not the claimed subject matter. Relevant to § 103 only.

Reference 4 — U.S. 8,385,457 B2 (Intel; listed as Laudel et al. on the Justia record)

• Description: I/Q imbalance calibration for wideband zero-IF tuners implemented in a MoCA/OFDM context, using two-tone probes injected in controlled loop-back modes (Type-2 probe / MoCA 2.0 loopback). It separates frequency-dependent (baseband filter) from frequency-independent (mixer) imbalance, calibrating transmit and receive paths with digital equalizers/weighted summation.
• § 102 assessment: No claim anticipated — and it teaches away from the "blind" premise. Claim 1 of 9065609 requires estimating parameters from complex baseband samples in a blind manner (the patent's specification: "the receiver estimates the correction parameters without prior knowledge of information impressed by the transmitter"). The '457 reference is a non-blind, probe-injection, loop-back technique requiring known calibration tones. It discloses no time-domain I/Q cross-correlation, no k estimator, and no I/Q mean-square-based gain-imbalance estimate. Relevant to § 103 only (and arguably as evidence of what "blind" excludes).

Inventor caveat: I retrieved the Intel '457 abstract/description and assignee (Intel, filed 2009-09-23, granted 2013-02-26) but did not independently verify the full inventor list; the Justia citation table lists it as "Laudel et al."

Reference 5 — Shafiee & Fouladifard (IEEE ICC 2003) — non-patent literature

• Description (from the record): "Calibration of IQ Imbalance in OFDM Receivers," a printed publication on gain/phase I/Q imbalance calibration in OFDM receivers.
• § 102 assessment: It is a qualifying § 102(a)(1) printed publication (2003 ≪ 2013-07-15), and it is the only NPL citation of record — typically the reference an examiner relies on for the general concept of blind/statistical I/Q imbalance calibration in OFDM. However, I could not retrieve the paper's full text (the search step budget was exhausted before returning it), so I cannot map it to specific claim elements and will not assert it anticipates any claim. To state an anticipation position I would need the paper's actual estimator description. Based on its title/venue alone, the most likely overlap is with the gain-imbalance and blind calibration concepts of claim 1(f) — not with the time-domain cross-channel-gain k steps.

3. Bottom-line § 102 conclusion

Net: On the citations of record, no single reference appears to anticipate any of the 19 claims of US 9,065,609. The four U.S. references and the NPL function as § 102/§ 103 background and obviousness material. The single limitation that most clearly separates the granted claims from this art is claim 1(b)'s requirement that the cross-correlation be a "time-domain cross correlation between the I component and the Q component" — a clause absent from the published claim 1 and evidently added to avoid the frequency-domain/statistical calibration art of Husted '755 (and likely Shafiee). A § 103 attack would most plausibly rest on '755 (blind statistical I/Q calibration) in view of '895 (mean-square/RMS metric feedback) or '755 in view of '523 (correlation + DC-offset removal).

4. Methodology and limitations

• Sources: Google Patents, Justia Patents, FreePatentsOnline, and USPTO-mirrored records for 9065609 and each cited reference; the authoritative full text you provided is treated as controlling over search snippets.
• What I did not do: I did not substitute any similar number (e.g., 9,890,609; 9,106,609; 8,676,609; 9,147,609; or the Eolas 5,906,854 "906" patent) for 9,065,609. Every ID above is quoted literally.
• Unverified items I will not fabricate: (i) the full claim text of U.S. 8,385,457 and U.S. 6,785,523 beyond what is quoted; (ii) the full inventor list of U.S. 8,385,457; (iii) the substantive content of the Shafiee & Fouladifard paper. These gaps limit the precision of any anticipation opinion on those references and would require the actual documents (or USPTO Patent Public Search / PatentCenter full-text retrieval) to close.
• Anticipation standard applied: § 102 requires that a single reference disclose each and every claim element, arranged as claimed. Several references disclose individual elements (correlation, DC-offset removal, mean-square metrics, blind calibration), which is why they are more naturally § 103 rather than § 102 material. No reference on the record is a clean anticipatory reference.

If you want, I can next build a formal element-by-element § 103 chart for claim 1 against the '755 + '895 and '755 + '523 combinations, or attempt a focused retrieval of the Shafiee & Fouladifard paper in a fresh search budget.

I'll ground the obviousness analysis in the actual disclosures of the references of record. Let me pull the key prior-art teachings.

Let me pull the NPL reference and the family-adjacent NI patents that appear on the "Similar Documents" list.

Obviousness Analysis — U.S. Patent No. 9,065,609 B2

Under 35 U.S.C. § 103 (AIA), as applied to a patent with a July 15, 2013 filing/priority date

Scope note. This analysis builds on the two previously generated sections (Patent Summary; Litigation Summary) and does not repeat the bibliographic, prosecution-history, or litigation findings set out there. It relies on the four references cited on the face of the patent, the single non-patent citation of record, and the "Similar Documents" list — all as retrieved in this pass. Where I could not verify a reference's content in this pass, I say so rather than filling the gap.

1. The operative analytical framework

Because the '609 application was filed July 15, 2013 — after the March 16, 2013 AIA cutoff — and claims no earlier priority, the AIA version of § 103 governs, with § 102(a)(1)/(a)(2) defining the prior-art universe. Under Graham v. John Deere and KSR Int'l v. Teleflex, the inquiry is:

1. scope/content of the art;
2. differences between the art and the claims;
3. PHOSITA level; and
4. secondary considerations.

The § 103 question here reduces to whether the combination of (i) an art-recognized blind/passive I/Q-calibration technique, (ii) art-recognized per-channel mean-square/statistical metrics, and (iii) art-recognized baseband correction hardware would have rendered claim 1 obvious. In my view, the answer is yes as to claim 1 and most dependents, with the strongest (but still likely surmountable) non-obviousness argument lodged in the specific ordering and algebraic normalization of steps (b)–(f).

PHOSITA (for the record)

A person having ordinary skill in the art would be an engineer with a B.S. in EE (or equivalent) and ~2–4 years of experience in RF/digital-communications receiver design, including direct-conversion receiver impairment modeling, baseband DSP for I/Q correction, and statistical signal estimation. This level matters because the core operation at issue — normalizing a cross-correlation by a mean-square — is undergraduate-level linear-algebra/least-squares material.

2. Claim 1, element by element, against the references of record

Claim 1 (method) requires: (a) receiving complex baseband samples from a receiver; (b) a time-domain I/Q cross-correlation; (c) mean-square values I_MS and Q_MS; (d) a cross-channel gain k̂ from the cross-correlation plus I_MS and/or Q_MS; (e) applying a cross-channel gain correction to obtain modified samples; (f) a gain-imbalance estimate from I_MS and Q_MS; plus I-channel and Q-channel DC-offset estimation and correction.

Summary of the gap. No single reference of record appears to disclose all of claim 1. But every element is individually disclosed, and the remaining question is whether the combination and the particular ordering would have been obvious.

3. Where the real novelty sits (and why it is likely obvious)

Two features of claim 1 are the most defensible candidates for non-obviousness:

1. The "time-domain cross correlation" limitation. This is the only express structural/procedural narrowing added over the disclosure as filed. It excludes frequency-domain estimators of the Shafiee type. But Husted '755 expressly contemplates both domains, and Atheros '523 implements the correlation in the time domain. Substituting a time-domain correlator for an FFT-bin estimator is a simple substitution of one known element for another (KSR factor (B)) yielding predictable results — and is in fact the cheaper implementation for a sample-domain receiver.

2. The decomposition and ordering: estimate k from the I→Q cross-correlation normalized by the I mean square; remove the cross-channel (skew) distortion first; then measure the I and Q signal strengths to get the gain imbalance. The '609's own specification explains the rationale — "Because the IQ cross-channel distortion impacts the relative signal strengths of the I and Q channels, this distortion may be removed prior to estimation of the received signal strengths." This is a real insight, but it is also the recognized best practice for decoupling two linearly independent impairments: you orthogonalize before measuring per-axis scale. Intel '457 discloses exactly this two-stage philosophy — it compensates frequency-dependent phase imbalance and then frequency-independent gain imbalance (Method 100, steps at 402–408), i.e., it removes one impairment before estimating the other. A POSITA with Shafiee's relations plus Intel's sequencing would arrive at the '609 ordering as a predictable variation.

4. Specific § 103 combinations, with articulated motivation

Below are four combinations I consider the strongest. I rank them by how cleanly they dispose of claim 1.

Combination A (strongest for claim 1): Husted '755 + Broadcom '895 (+ Atheros '523)

Motivation to combine (explicit and articulated):

• Same field, same problem, same assignee/inventor. Husted '755 and Atheros '523 share a common assignee (Atheros) and a common named inventor (Paul J. Husted), and both describe the same receiver architecture (802.11a, 5 GHz quadrature front end). This is the classic KSR "common ownership / common problem" motivation.
• Broadcom '895 supplies its own reason to combine. It expressly criticizes factory calibration because the receiver "may operate at temperatures that differ from the controlled operating conditions" and "at operational frequencies that differ" — motivating in-situ/online calibration. That is precisely the design incentive that drives combining a statistical (Husted) estimator with a mode-switchable (Broadcom) corrector.
• Reuse of existing receiver machinery. Broadcom's signal-metric module and Atheros '523's power/correlation detectors already exist in the receiver for AGC; reusing them for impairment estimation is a predictable design economy.

Disposition: claim 1 obvious; the time-domain cross-correlation is supplied by Atheros '523 or by Husted '755's express domain-agnostic teaching.

Combination B (strongest for the estimation mathematics): Shafiee & Fouladifard + Husted '755 + Atheros '523

Motivation: Shafiee supplies the closed-form I/Q-mismatch parameter relations and the statistical justification for estimating them from received samples; Husted '755 supplies the blind, no-dedicated-signal motivation and the digital correction architecture; Atheros '523 supplies the time-domain correlation and DC-offset pre-conditioning. The combination amounts to applying Shafiee's estimator in Husted's passive framework using Atheros' time-domain correlator — three known techniques, same art, predictable result.

Disposition: claim 1 obvious; the specific "time-domain" limitation is met by '523/Husted and is a mere substitution for Shafiee's FFT-bin approach.

Combination C (strongest for the correction hardware and dependent claims): Broadcom '895 + Intel '457 + Atheros '523

Motivation: Intel '457 expressly identifies the hardware cost/flexibility problem ("existing hardware-based DFE techniques lack flexibility and it is expensive, time consuming and challenging to modify the DFE design"), motivating digital baseband implementation; Broadcom '895 supplies the calibration-mode metric minimization; Atheros '523 supplies the sample-domain correlation/mean-square statistics. Combining them yields a digital-baseband I/Q-correction subsystem that estimates parameters from correlation/power statistics in a calibration mode and applies them in an operational mode — the architecture of claim 1 plus claims 9–11 and 14.

Disposition: claim 1 obvious; claims 9, 10, 11, 14 (external/receiver-side estimation, transfer to receiver, programmable hardware element) obvious in view of Intel '457's digital-domain calibration-then-operation and the well-known use of FPGAs ASICs for baseband correction (which the '609 itself describes as routine).

Combination D (for the "blind" preamble and the application context): Husted '755 + Shafiee + Broadcom '895

Motivation: Husted '755 expressly disclaims the need for a dedicated calibration signal ("does not require that a separate calibration signal be transmitted to the receiver... The received signal is other than a signal received solely for the purposes of calibration"), which reads directly on the "blind estimation" preamble and on claim 2 ("receiver generates the complex baseband signal in response to receiving a signal transmission from a transmitter"). Shafiee supplies the parameter estimation; Broadcom '895 supplies the compensation module.

5. Dependent-claim disposition

6. The counter-arguments I would expect, and my assessment

1. "The reference combination teaches away — Shafiee is frequency-domain; the claim requires time-domain."
   Weak. Husted '755 expressly teaches time- or frequency-domain application, and Atheros '523 implements the correlation in the time domain. Substituting a time-domain correlator for an FFT-bin estimator is the sort of "simple substitution of one known element for another" KSR endorses. There is no teaching away.

2. "None of the references teaches the cross-channel gain k as a separate, correctable impairment distinct from gain/phase imbalance."
   This is the best argument. The '609 treats k (I→Q skew) as distinct from A_QI (gain imbalance) and from phase. However: Shafiee's model attributes the image to the same gain/phase mismatch and supplies a compensation block that mixes I into Q; Husted '755 explicitly frames the impairment as I and Q becoming "correlated" due to non-90° LO phase — which is precisely the physical quantity a cross-correlation measures and a correlated term corrects. A POSITA would recognize k = ⟨I,Q⟩/⟨I,I⟩ as the least-squares projection coefficient for removing the correlated (skew) component of Q onto I — textbook Gram-Schmidt orthogonalization. Expect an examiner/Petitioner to rely on the "known technique to improve similar devices" rationale.

3. "The ordering (remove k before estimating gain imbalance) is a non-obvious insight."
   Moderate. The '609's stated reason (cross-channel distortion corrupts relative I/Q signal strengths) is analytically correct, but Intel '457's two-stage, phase-then-gain compensation shows the same decoupling principle. At most this argument rescues claim 1's ordering — it does not rescue claims 17/19, which recite the same steps without a claimed order.

4. Secondary considerations. The '609 reports ~18 dB image-rejection improvement and ~51 dB carrier-feedthrough reduction (FIGs. 7A/7B). I would treat these as results flowing from known techniques, not nexus-establishing unexpected results — the 51 dB feedthrough improvement is attributable to DC-offset correction (Atheros '523/Shafiee), and the 18 dB image rejection is what the art predicts from removing a k/mismatch term. No evidence of commercial success, licensing, copying, or industry praise appears in the record retrieved.

7. Contradictions / caveats to flag against the earlier sections

• NPL title discrepancy. The '609's "Non-Patent Citations" list gives the title as "Calibration of IQ Imbalance in OFDM Receivers," but the indexed version of the Shafiee/Fouladifard ICC '03 paper (pp. 2081–2085) is titled "Calibration of IQ imbalance in OFDM transceivers." Per the operating rules I have not auto-corrected either identifier; a POSITA should treat them as the same publication, but the title on the patent face differs from the published title.
• US 8,477,889. The previously generated Patent Summary cited US 8,477,889 as a "family-adjacent" frequency-domain estimator. I could not verify that reference's content or date in this pass, and it does not appear on the '609's citation lists. I have therefore not relied on it in any combination above. Treat that prior mention as unverified.
• Family-adjacent NI documents are poor § 103 art. The "Similar Documents" list includes US 8,559,560 ("Measuring the I/Q impairments of a receiver"), US 8,638,893 ("Mechanisms for the correction of I/Q impairments"), and CN 107579941 B (offset-LO I/Q impairment correction). I could not confirm their "effectively filed" dates in this pass. Two cautions: (i) their dates must be verified before use; and (ii) if they share the '609's inventive entity (James W. McCoy) or are commonly owned by National Instruments, AIA § 102(b)(2)(C) (and possibly § 102(b)(2)(A)–(B)) would except them from being § 102(a)(2) prior art. I would not build an obviousness case on these.
• Post-dating references — NOT prior art. Appearing on the "Families Citing" list are GB 2,537,800 (Imagination Tech Ltd, priority 2014-12-22), US 9,698,917 (Nokia, priority 2015-03-31), and WO 2017/219265 (Huawei, 2016). Each post-dates the '609 (priority 2013-07-15) and cannot be used as § 102/§ 103 prior art against it. They are relevant only as evidence of the state of the later art or as potential § 101/§ 112 context — not for § 103.

8. Bottom line

• Claim 1, claim 17, claim 19 — likely unpatentable under § 103. Every element is disclosed across Husted '755, Broadcom '895, Intel '457, Atheros '523, and Shafiee & Fouladifard, and there are multiple express motivations to combine (avoiding dedicated calibration signals; factory-calibration drift; digital-baseband flexibility; common assignee/inventor between the two Atheros references; reuse of existing AGC correlation/power hardware).
• The narrowest survivable feature is the specific ordering/normalization in claim 1: k̂ = R̃_iq·I_MSINV, applied before the I/Q signal-strength measurement. This is the only place where an argument for non-obviousness has real traction, and even there Intel '457's phase-then-gain decoupling sequence supplies a strong teaching-suggestion.
• The dependent claims and medium/system claims add nothing that is not separately disclosed; claims 8 (DC-before-correlation) and 11 (programmable hardware element) are squarely met by Atheros '523 and by the conventional FPGA/ASIC baseband implementations that the '609 itself treats as routine.
• Caveat on confidence: the mappings above rest on the face-record references and the descriptions retrieved in this pass (URLs cited inline: Google Patents/patentimages PDFs for US 6,785,523, US 7,158,586, US 8,415,895, US 8,385,457; the ICC '03 paper). I did not retrieve full claim charts for the Broadcom and Intel references or a verbatim copy of the Shafiee paper's complete text. A final validity opinion should be confirmed against the full texts of US 8,415,895 and US 8,385,457 (both specifications) and the complete Shafiee & Fouladifard paper, and should verify the filing dates of US 8,559,560 and US 8,638,893 before any of them is used.