US 8798953

I'll search for authoritative information on US patent 8798953, including any litigation or CAFC docket activity.

US Patent 8,798,953 — Analyst Summary

Sourcing note

I searched for this patent number across general web and patent-database results. The authoritative source is the full patent text you supplied (fetched from Google Patents, US8798953B2), which I cross-checked against uspto.report's grant record for 8,798,953. I found no district-court or CAFC docket activity for this patent in my searches. My search tools cannot query PACER, Docket Navigator, or Westlaw directly, so "no litigation found" reflects the absence of results in open-web searching rather than a certified litigation clearance. I flag this as a limitation, not a conclusion.

Bibliographic data

Assignment recorded 2011-09-01 to Yuan Ze University, assignor Chien-Chang Huang (Reel/Frame 026842/0193; the recorded effective date on the cover page is 2010-10-28).

Abstract (as issued)

The patent describes a measurement structure comprising a transmission line segment calibrator, an offset series device calibrator, an offset shunt device calibrator, and a tested object measuring instrument. The transmission lines of the offset series and offset shunt calibrators are equal in length to those of the tested object measuring instrument, so all three share identical "error boxes." Once the error-box scattering-parameter matrix is obtained by the calibration method, an electronic device under test can be connected and its uncorrected measurement data processed to yield the device's RF scattering parameters.

Independent claim overview

Claim 1 — the sole independent claim — is drafted to a measurement structure, not to a method, notwithstanding the title and abstract referring to a "calibration method." It recites an apparatus using a microwave probe (having at least a ground end and a signal end) as the contact interface, comprising four elements:

1. Transmission line segment calibrator — two transmission lines with a transmission line segment connected between them; the transmission lines connect to the probe signal end so the segment's device characteristics can be measured.
2. Offset series device calibrator — two transmission lines with both an offset transmission line and a series resistor connected between them; the transmission lines connect to the probe signal end.
3. Offset shunt device calibrator — two transmission lines with both an offset transmission line and a shunt resistor connected between them; the transmission lines connect to the probe signal end.
4. Tested object measuring instrument — two transmission lines with a tested device connected between them; the transmission lines connect to the probe signal end so the device's characteristics can be measured.

In plain terms: a set of on-substrate (or on-board) test structures — a through-line, a series-resistor standard, and a shunt-resistor standard, each with defined offset line lengths — that plug into the same probe interface as the device under test, enabling self-calibration without an Impedance Standard Substrate (ISS).

Dependent claims

• Claim 2 — requires the transmission lines of the offset series calibrator and offset shunt calibrator to be equal in length to those of the tested object measuring instrument (codifying the shared error box).
• Claim 3 — the offset transmission line of the series calibrator may be equal to or different from the offset transmission line of the shunt calibrator.
• Claim 4 — the calibrators and measuring instrument may use a microstrip or coplanar waveguide as the connection transmission line.
• Claim 5 — the microwave probe is a high-frequency probe characterized as G-S-G-S-G, G-S-S-G, G-S-G, or G-S.

Points worth flagging

• Claim/specification divergence on offset line lengths. The specification states the offset transmission lines of the series and shunt calibrators "differs from" one another (and the detail section repeats "differs"). Claim 3 broadens this to "equal to or different from." This is a genuine intrinsic-record inconsistency a litigator would examine; I'm reporting the literal texts, not resolving them.
• Method vs. structure mismatch. The title, abstract, and much of the specification describe a calibration method with self-calibration equations (eqs. 1–5) and a Newton-Raphson solve for t(e^γl), z, y, z_tp, y_sp. However, no method claim issued — all five claims are directed to the measurement structure. The claimed scope therefore does not itself cover the computational method.
• Related family. A sibling application by the same inventor/assignee issued as US 2014/0118004 A1 ("Measurement Structure for Radio Frequency Scattering Parameter Measurement Applying Two Calibrators and Calibration Method Thereof"), and corresponding Taiwanese cases TWI426289B and TWI463147B appear in the family/similar-documents listings. Treat these as related, not as the same patent.
• Cited prior art of note. US 5,608,330 A (Rohde & Schwarz, seven-term calibration), US 2010/0176789 A1 (Univ. of Missouri), and US 2011/0151596 A1 (TSMC, cascaded-based de-embedding), plus non-patent literature including Williams/Wang/Arz (2003), Engen et al. TRL (1979), and Huang & Lin (2008, 2010). The Huang & Lin references are by the same inventor and are directly on point for the "line-series-shunt" lineage of this work.

Uncertainty disclosure

I could not verify the existence or absence of litigation through authoritative court databases. The bibliographic data above is drawn from the patent text you supplied and from uspto.report's grant record, which agree. If you need a certified litigation/PTAB check, that requires a PACER or commercial docket search, which my current search capability does not reach.

I'll search litigation-focused sources for this specific patent number.

The first results are conflating with an unrelated patent (Neonode's '879 = 8,095,879). Let me run precise searches on the exact number and assignee.

Let me check litigation-database portals directly and rule out any enforcement activity.

Bottom line

No litigation is known involving U.S. Patent No. 8,798,953.

I searched litigation-oriented sources (Unified Patents litigation portal, PACER-derived aggregators via CourtListener storage, Docket Alarm, and general web/patent-database queries) for the exact number 8,798,953 and for the patent's identifying features (Yuan Ze University; Chien-Chang Huang; "L-OS-OT" / three-calibrator RF scattering-parameter calibration). I found no district court action, no ITC Section 337 investigation, no PTAB proceeding (IPR/PGR/CBM), and no CAFC appeal naming this patent. There are therefore no entries to tabulate under the requested fields (plaintiff, defendant, jurisdiction, case number, filing date, outcome/status).

This builds on — and is consistent with — the earlier-generated summary's "no litigation found" finding. It is not a repeat of that section; the searches below were run specifically against litigation caselists rather than general bibliographic sources.

Critical disambiguation warning (do not conflate)

Practically every litigation-hit my searches returned for a "879 patent" refers to a different patent entirely:

Per the operating rule to interpret identifiers literally and not auto-correct: 8,798,953 ≠ 8,095,879. The Neonode litigation is not litigation against this patent, and I am not reporting it as such.

What I checked and what those checks show

• Google Patents legal events for US 8,798,953 (the authoritative text you supplied) list only: assignments (to Yuan Ze University, Reel/Frame 026842/0193), the 2014 grant, and maintenance-fee payments (4th yr 2018-01-24; 8th yr 2022-01-24; reminder mailed 2026-03-23). No litigation or adverse-proceeding entries appear.
• PTAB / Unified Patents caselists — no petition, institution, or final written decision for 8,798,953 returned. (Compare: the CN103399286B, TW I426289B/I463147B and US 2014/0118004 A1 items in the family listings are patent-family relations and citations, not litigation.)
• CourtListener / PACER-mirrored dockets — no complaint, answer, or order referencing 8,798,953.
• Ownership posture — assignee is a Taiwanese university (Yuan Ze University), an academic institution with no apparent assertion program. The Families Citing this family list shows later-generation filings (e.g., the two-calibrator sibling US 2014/0118004 A1, Tektronix US 2021/0170790 A1 / US 11,041,880 B2, CETC-13 US 2022/0099736 A1 / US 11,971,451 B2, ITRI US 12,163,989 B2) — these are citations/continuations, and none of those records reflects an infringement suit over the '953 patent.

Mandatory uncertainty disclosure

• The user-specified date (April 26, 2026) conflicts with my environment date (September 25, 2026); my retrievable sources appear to extend into mid/late 2026 (e.g., docket entries dated March 2026 and case records updated September 2026). I flag the discrepancy rather than resolve it.
• I could not directly render https://portal.unifiedpatents.com/litigation/caselist or query PACER/Docket Navigator/Westlaw/Lexis as authoritative live sources. "No litigation found" therefore reflects the absence of results across open-web and aggregator searching, not a certified litigation clearance.
• Because the "879" short-form collision is so pervasive in this number space, any future hit should be verified against the full number 8,798,953 and the assignee (Yuan Ze University) before being attributed to this patent.

If you need a certified clearance, that requires a paid PACER full-text search plus a Docket Navigator/Unified Patents portfolio pull — outside what my current tooling reaches.

I'll verify the structured data against open-web sources before reporting.

Let me run additional targeted checks for any Board proceedings or litigation touching this patent or its owner.

One more check on the owner's assertion history and any reexamination of this patent.

Proceedings overview

Zero AIA trial proceedings. The structured "PTAB proceedings on file" block (the canonical USPTO Open Data Portal list) returns no IPR, PGR, or CBM for US 8,798,953 — no proceeding numbers exist to report, and I will not manufacture any. My independent open-web checks (Google/Bing-type queries on the patent number, the individual claim set, the assignee "Yuan Ze University," PTAB petition text, and CourtListener) surfaced no petition, no institution decision, no Final Written Decision, no appeal docket, and no ex parte reexamination touching this patent. The defensive posture is therefore "untested patent" rather than either of the two outcomes in the task template: no claims are canceled, and critically, no claims have been validated by the Board either. A defendant faces a patent whose five claims have never been through an adversarial validity challenge — and, per the previously generated sections, appears never to have been asserted in district court at all.

⚠️ Template contradiction flagged: the "Proceedings overview" instruction offers "the patent has survived two IPRs and is hardened" as sample phrasing. That is illustrative boilerplate only. Nothing in the structured data or my searches supports any IPR history for this patent. Do not let that example language migrate into a brief or opinion.

No proceedings to itemize

There is no {PROCEEDING_NUMBER} — {Petitioner} v. {Patent Owner} entry to write. For the record, the fields would be empty for every category:

Search methods and their limits (stated so you can calibrate confidence): I queried for the patent number in quoted and unquoted forms, for the title, for the assignee, and for PTAB petition text; I checked the uspto.report grant record and Google Patents' "Cited By / Citations" sections, which for this patent list only patent and NPL citations — no post-grant proceedings. I cannot directly query PTAB E2E (https://ptacts.uspto.gov/ptacts/), the ODP API, or PACER from this environment; the structured block is doing that work, and it corroborates my searches. If you need a certified clearance, hit PTAB E2E directly under "Patent Number 8798953" and confirm zero results.

Strategic summary

Claim status of US 8,798,953: all five claims are UNTESTED, not SUSTAINED. No claim has been canceled, narrowed by certificate, or amended in a § 316 motion to amend. The claim set is exactly as issued: claim 1 (independent, measurement structure), claims 2–5 (dependent). Whether claim 1 is valid over the Huang & Lin line-series-shunt literature (2008/2010) and Engen TRL (1979) — all of which the examiner cited as NPL — has never been adjudicated by the Board. Note the previously flagged intrinsic-record defect (claim 3 permits the two offset lines to be "equal to or different from," while the specification says "differs") — that is a § 112 written-description/claim-support and indefiniteness argument that an IPR petition cannot reach, since IPR is limited to § 102/§ 103 on patents and printed publications (35 U.S.C. § 311(b)). It has to be litigated in district court or raised in a PGR — and PGR is now time-barred, since the 9-month § 321(c) window closed in 2015.

Estoppel landscape: § 315(e)(2) is a blank slate. With no IPR ever filed, no petitioner, real party in interest, or privy is estopped from anything. Any defendant today may file an IPR on any § 102/§ 103 ground supported by patents or printed publications without § 315(e) exposure from earlier Board proceedings. Two practical consequences: (1) there is no § 315(b) one-year bar running because, per the previously generated sections, no infringement action appears to have been served — but confirm that independently, since service of a complaint is what starts the clock; and (2) the strongest art here is publication art. Huang & Lin's IEEE T-IM 57(5) 2008 paper and IEEE T-MTT 58(9) 2010 paper are the closest prior art, are by the same inventor (which raises § 102(b)/derivation-flavored and self-disclosure considerations for art within one year of the 2011-09-01 filing), and the Dec. 2009 APMC conference paper "Applying Line, Offset-Series, Offset-Shunt calibration for on-wafer CMOS device characterizations" is an on-point printed publication by the same group. Whether any of these qualify as prior art to the 2011-09-01 filing date, and whether they disclose the claimed structure (as opposed to the algorithm), is the entire ballgame — and it has never been argued to the Board.

Pattern signals: none of the usual indicators are present. No serial petitioner, no Unified Patents–style defensive aggregator (which would typically target a patent being asserted against many companies), no PTAB appeal by the owner, no reexam. The owner is a Taiwanese university, and the sibling patent US 8,552,742 (same inventor/assignee) and the two-calibrator sibling US 2014/0118004 A1 show the same pattern: academic portfolio, no assertion program, no Board history. The absence of IPRs is best explained by the absence of assertions, not by the patent being unassailable. Well-asserted patents almost invariably attract IPRs; a patent challenged by nobody tells you nothing favorable about its validity.

Recommended next steps

1. Verify the zero-count yourself. Run the patent number through PTAB E2E at https://ptacts.uspto.gov/ptacts/ and CourtListener's docket search at https://www.courtlistener.com/?type=r&q=%228798953%22. My result is "no proceedings found," not "certified clean."
2. If you are a defendant, do not look for an FWD to cite — there isn't one. Unlike a patent with a canceled claim, US 8,798,953 offers no free kill. Your validity attack must be built from scratch, but you have a genuinely favorable runway: no estoppel, no § 315(b) clock (verify service), and a rich vein of same-inventor IEEE printed publications that the Board has never been asked to consider against the issued structural claims.
3. Attack the claim/specification mismatch. Claim 3's "equal to or different from" versus the specification's consistent "differs from" is a § 112 issue worth preserving in invalidity contentions. It is unavailable in an IPR (§ 311(b)) and PGR is time-barred, so it needs to live in the district court case.
4. Watch the maintenance-fee clock. Per the previously generated section, a 12th-year maintenance-fee reminder issued 2026-03-23, with adjusted expiration 2032-10-09. The patent has roughly six years of term left. A lapsed fee would moot the question entirely — worth a docket entry on your calendar.
5. Do not over-read the emptiness. "No PTAB activity" is a neutral fact about enforcement history, not a validity finding. Frame it that way internally; frame the inverse (no assertion, university owner, no licensing program) if you are assessing whether a demand letter you received is even credible.

I'll gather evidence on the assignment chain and any NPE-relevant signals before writing the section.

Ownership & Assignment Analysis — US 8,798,953

Sourcing note (read first)

I could not query USPTO Assignment Center interactively from this environment; my tools are open-web search only. What I could confirm comes from the Google Patents legal-events record in the authoritative full text you supplied, cross-checked against uspto.report for grant 8,798,953. Both agree on a single recorded assignment. The correspondent of record is not exposed in either rendering of the record, and I could not surface it in open-web search — that field is a genuine gap, and I say so rather than guess. If you need a certified chain, run the patent-number query at https://assignmentcenter.uspto.gov/ (legacy mirror: https://assignment.uspto.gov/patent/index.html).

Inventors

Pattern assessment: No unusual departures to flag. A single inventor who assigned to his university employer pre-filing is the most common academic profile and is the opposite of the "inventors all leave within 12 months" precursor to a portfolio fire-sale. The related lineage of work (the Huang & Lin 2008 and 2010 non-patent-literature citations in the file) also indicates the inventor remained research-active in this exact field through at least 2010.

Original assignee

• Entity on the issued patent: Yuan Ze University (Taoyuan, Taiwan). Google Patents lists "Current Assignee" and "Original Assignee" both as Yuan Ze University; uspto.report states the grant "is currently assigned to Yuan Ze University."
• Primary line of business: Private research university (higher education and academic research). Not a product company.
• Product embodying the claims: No. The claimed subject matter is a set of on-substrate measurement/calibration test structures (through-line, offset-series, offset-shunt standards) probed with a G-S-G-class RF probe. This is laboratory metrology hardware of the kind a university cleanroom builds for research use; there is no commercial product line.
• Current status: Operating. No acquisition, dissolution, or bankruptcy is evidenced anywhere in the record. Maintenance fees were paid at the 4th year (2018-01-24) and 8th year (2022-01-24), and the 12th-year fee reminder was mailed 2026-03-23 — an assignee still actively maintaining the patent under its own name.

Assignment timeline

Chronological list of every recorded assignment. There is exactly one.

• 2010-10-28 (executed) / recorded 2011-09-01 — Reel 026842 / Frame 0193
  • Conveyance: Assignment — "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)" (Reel 026842/0193)
  • Assignor: Huang, Chien-Chang
  • Assignee: Yuan Ze University (Taiwan)
  • Correspondent: Not determinable from the sources available to me. The Google Patents legal-events feed and uspto.report's grant page do not render the correspondent field, and open-web search did not surface it. Because this chain has only one link, the "repeat correspondent" tell is in any event not testable here — recurrence requires two or more recordings.
  • Context: Original pre-filing assignment by the sole inventor to his university employer — a routine institutional/employment conveyance, not an acquisition, fire-sale, or reorg. Note the effective date (2010-10-28) precedes the filing date (2011-09-01) by about ten months, consistent with a standing university IP policy obligation rather than a transaction at filing.

Post-issuance assignments: none recorded. There is no transfer to an LLC, no security agreement, no merger, no change of name, no license recordation, and no release in the record. Per your instruction, I state this plainly: the original assignee still owns the patent.

Related (not part of this chain, flagged to prevent confusion): the same inventor/assignee family includes US 2014/0118004 A1 ("…Applying Two Calibrators…", filed 2012-10-30) and Taiwanese counterparts TWI426289B and TWI463147B. These are separate properties, not links in US 8,798,953's chain of title.

Timeline diagram

timeline
    title Ownership of US 8798953
    2010 : Huang executes assignment to Yuan Ze University
    2011 : Assignment recorded to Yuan Ze University
         : Reel 026842 Frame 0193
    2013 : Pre-grant publication US 2013 0060501 A1
    2014 : Patent granted 5 Aug
    2018 : 4th year maintenance fee paid
    2022 : 8th year maintenance fee paid
    2026 : 12th year fee reminder mailed

NPE / troll-pattern signals

Caveat on naming: per your constraint, I have not treated "university assignee" as an affirmative NPE finding in either direction. It is a factual ownership statement. Universities do sometimes assert through exclusive licensees and can appear on the Stanford NPE Litigation Database's "University Heritage or Tie" category (Category 2) — but no litigation or license recordation exists for this patent, so there is nothing to categorize.

Verdict

Insufficient data (only the original assignment).

Justification: The complete recorded chain consists of a single conveyance — Huang, Chien-Chang → Yuan Ze University, executed 2010-10-28 and recorded 2011-09-01 at Reel 026842 / Frame 0193 — and the patent has remained with the university since, with maintenance fees paid at the 4th and 8th years and a 12th-year reminder issued 2026-03-23. There are no post-issuance transfers, no LLC assignees, no security interests, no recorded licenses, and no litigation, so none of the eight NPE signals can be marked present on evidence. This is the benign "original assignee still owns it" profile, not an arranged assertion chain; the only unresolved field is the correspondent of record, which the available sources do not expose.

Verify at: https://assignmentcenter.uspto.gov/ — search by patent number 8,798,953, or by Reel/Frame 026842/0193.

I'll search for the patent and each of its cited references to build a grounded prior-art analysis.

Let me gather details on the non-patent literature citations and verify the patent record.

Prior Art Analysis — US 8,798,953 B2

Sourcing note (read first)

I was asked to "search the USPTO database." My available tooling is open-web search, not a direct portal session into USPTO Patent Public Search (Patent Public Search / PatFT / PatFT-AppFT). In my searches the authoritative full text was the Google Patents record for US8798953B2 (https://patents.google.com/patent/US8798953B2/en), which is the same text supplied to me and which I treat as controlling. I therefore report the citation list exactly as it appears in the patent's own "Patent Citations" and "Non-Patent Citations" sections, cross-checked against the individual reference records on Google Patents, FreePatentsOnline, and IEEE Xplore metadata where those were retrievable. I did not find any additional USPTO-recorded references beyond the three patent citations and five non-patent citations below. Where I could not independently verify a detail, I say so.

A structural point that governs the whole § 102 analysis, carried forward from the earlier section: the issued patent has no method claim. Claim 1 is drawn to a measurement structure (apparatus), and claims 2–5 are structural dependents. That mismatch matters because all three cited patents are method references. A method reference can anticipate an apparatus claim only if it discloses the apparatus as claimed, element for element.

The complete citation list as recorded on the face of US 8,798,953

Patent citations (3)

Non-patent citations (5)

The Google Patents record marks three of the five with an asterisk (cited by examiner):

Reference-by-reference analysis

1. US 5,608,330 A — Rohde & Schwarz (1997-03-04)

• Full citation: US 5,608,330 A, "Method for calibrating a network analyzer according to the seven-term principle," Rohde & Schwarz GmbH & Co. KG. Priority 1994-10-05; filed 1995-10-04 (App. 08/539,087); granted 1997-03-04. German/EP counterpart EP 0 706 055 (B1). URL: https://patents.google.com/patent/[US5608330A](/patent/US5608330A)/en
• Brief description: A 7-term (seven-error-term) VNA calibration in which three standards are used and the first standard is a line whose characteristic impedance is known but whose complex propagation constant γ may be unknown, with γ solved afterwards from all three measurements. The reference recites several standard families:
  • LZY — the second and third measurements use the same two-port standard consisting of a series impedance Z of known value plus a lumped shunt impedance Y to ground of unknown value, with the two-port re-connected ("rotated") for the third measurement;
  • LZN — second measurement is a two-port consisting only of a series impedance (no shunt), third is a symmetric-reflection two-port (N);
  • LMS/LMO — second/third use known one-port Match or Short/Open standards using commercial substrates.
    It expressly frames this as on-wafer calibration of planar strip-line (microstrip) standards and notes γ·l can be obtained from equation (16).
• Relevance to the claims: This is the most structurally proximate patent citation. Its LZY standard — a series lumped element plus a shunt lumped element — is the conceptual ancestor of the claimed offset-series and offset-shunt device calibrators, and its "known-γ-line" philosophy is the same self-calibration idea. But it discloses a method, and the standards are described as a single two-port Y/Z lumped network reused and rotated, not as two distinct calibrator structures (each with its own two probe-interface transmission lines plus a distinct offset transmission line).
• Potential § 102 anticipation: None of claims 1–5, in my assessment. A § 102 reference must disclose every claimed element in a single disclosure. US 5,608,330 does not disclose (a) a transmission line segment calibrator "consisting of two transmission lines and a transmission line segment," (b) a separate offset-series device calibrator with two lines + offset transmission line + series resistor and an independent offset-shunt device calibrator, nor (c) a "tested object measuring instrument" as a claimed structural element, nor (d) a microwave probe having ground and signal ends as the claimed contact interface. At most it is strong § 103 art, and primarily against the method subject matter that did not issue as a claim. I flag this as an assessment, not a certified novelty opinion.

2. US 2010/0176789 A1 — Curators of the University of Missouri (2010-07-15)

• Full citation: US 2010/0176789 A1, "High frequency analysis of a device under test," The Curators of the University of Missouri; inventors Reza Zoughi, Mohamed A. AbouKhousa, Sergiy Kharkivskiy. Filed 2009-01-15; published 2010-07-15. Continuation issued as US 9,081,045 B2 (App. 13/486,606, filed 2012-06-01, granted 2015-07-14). URL: https://patents.google.com/patent/US20100176789A1/en
• Brief description: A vector network analyzer architecture that generates a standing wave on a transmission line coupled to the DUT, varies its phase with a phase shifter, samples the standing-wave magnitude at each of several phase shifts, and computes the DUT's reflection coefficient (magnitude and phase) from those samples — expressly avoiding heterodyne receivers and directional couplers. It discusses prior art slotted-line, sampled-line, and perturbation-two-port (PTP) methods with multiple PTP networks and a multi-step calibration routine, and the transmission line may be a waveguide, coax, or microstrip.
• Relevance to the claims: Technically adjacent but conceptually distant. It shares only general elements (a transmission line; possible microstrip) and the notion of calibrating a scalar/vector measurement. It is not a calibration-standard structure reference and does not disclose series or shunt resistor calibrators, offset transmission lines, or "error box" self-calibration with three on-substrate standards.
• Potential § 102 anticipation: None of claims 1–5. Even claim 4 (which recites "microstrip or coplanar waveguide as a connection transmission line") is not anticipated, because US 2010/0176789 does not disclose the calibrator/measuring-instrument structure of which the microstrip/CPW is the connecting line. Its examiner-cited inclusion reads like an artifact of the "high-frequency DUT measurement" field, not a substantive anticipation.

3. US 2011/0151596 A1 — TSMC (2011-06-23)

• Full citation: US 2011/0151596 A1, "Cascaded-Based De-embedding Methodology," Taiwan Semiconductor Manufacturing Company, Ltd.; inventor Hsiu-Ying Cho. Priority/filed 2009-12-17 (CN family 201010212769.1 filed 2010-06-24, priority date listed as 2009-12-16); published 2011-06-23. Issued as US 8,436,626 B2 on 2013-05-07; CN counterpart CN 102103167 B. URL: https://patents.google.com/patent/US20110151596A1/en
• Brief description: A de-embedding method: form a primary structure and an auxiliary structure in a semiconductor chip, where the auxiliary structure replicates a portion of the primary structure; measure both; determine a transmission (ABCD) matrix for each; and extract the matrix of a component by multiplying the primary-structure matrix by the inverse of the auxiliary-structure matrix. The structures are illustrated as symmetric dual transmission line structures with GSG pads, and the reference notes the transmission line may be microstrip and the parameters convertible to S-parameters. It touts reduced silicon area versus prior methods that required four to six dummy structures.
• Relevance to the claims: Relevant to the de-embedding context and to the pad/transmission-line/G SG environment of claims 1 and 4. It is not a calibration-standard reference: it discloses no series-resistor or shunt-resistor calibrators, no offset transmission lines, and no three-calibrator error-box scheme.
• Potential § 102 anticipation: None of claims 1–5. It does not disclose the claimed combination. It may be marginally useful as § 103 background (e.g., the use of GSG pad-replicating transmission-line structures and microstrip in on-wafer characterization).

Non-patent literature — relevance and the one item that matters most

• Huang & Lin (2008) — A Novel Calibration Algorithm With Unknown Line-Series-Shunt Standards, IEEE TIM 57(5), May 2008, pp. 891–896 (DOI 10.1109/TIM.2007.915121). This is the most substantively relevant item in the entire citation list, because it is the direct antecedent of the patent's "L-OS-OT" framework: it calibrates with unknown line, series, and shunt standards using an eight-term error model and a Newton-Raphson self-calibration solve of the line propagation constant and the series impedance / shunt admittance. It is authored by Chien-Chang Huang, the sole inventor of US 8,798,953, at Yuan Ze University. It is an examiner-cited reference. Note for the record: it predates the 2011-09-01 filing by more than one year, so under pre-AIA § 102(b) it is prior art regardless of common inventorship, and it is not saved by any grace period.
• Huang et al. (2010) — Accuracy Improvement for Line-Series-Shunt Calibration…, IEEE MTT 58(9), Sep. 2010, pp. 2497–2503 (examiner-cited). Same inventor; the immediately preceding refinement of the line-series-shunt method with on-wafer characterization. Also more than one year before the filing date.
• Engen et al. (1979) — TRL, IEEE MTT MTT-27(12), Dec. 1979 (examiner-cited). The foundational self-calibration/TRL reference; the patent's own background criticizes multiline TRL for wafer-area consumption.
• Williams, Wang & Arz (2003) — An Optimal Vector-Network-Analyzer Calibration Algorithm, IEEE MTT 51(12), Dec. 2003, pp. 2391–2401. Flagged inconsistency: the specification's Background (“Description of Related Art”) describes this Dec. 2003 MTT vol. 51 work as "a Multiline Thru-Reflect-Line (TRL) calibration method developed by NIST… multiple transmission line segments," whereas the citation list gives the title as "An Optimal Vector-Network-Analyzer Calibration Algorithm." The journal/volume/page anchors match a NIST-authored paper, but the title attributed in the citation list is not a TRL paper title. Treat the Background characterization as loose. Also note the citation list's page range "23941" is a typographical artifact for "2391."
• Wei et al. (2007) — A General 4-Port Solution for 110 GHz On-Wafer Transistor Measurements With or Without ISS Calibration, IEEE TED 54(10), Oct. 2007, pp. 2706–2714. This is the five-dummy-structure (Open, Short, Thru, Left, Right) one-tier de-embedding paper the Background cites. Purely a method reference; no structural calibrators.

Assessment of the NPL under § 102: none of the five anticipates claims 1–5, because each is a method/algorithm publication and none discloses the claimed measurement structure (the transmission-line-segment calibrator plus distinct offset-series and offset-shunt resistor calibrators sharing a probe interface with the tested-object measuring instrument). The Huang & Lin (2008) and Huang (2010) papers are the most dangerous art in the set — but for method subject matter, and no method claim issued. To the extent any issued claim were ever construed broadly toward the calibration method, those two same-inventor publications (each >1 year pre-filing) would be the first references an examiner or challenger would reach for under pre-AIA § 102(b)/§ 103.

Bottom line

1. There are exactly three patent citations and five non-patent citations on the face of US 8,798,953. I found no others.
2. No cited reference, singly, anticipates any of claims 1–5 as I read them. Claims 1–5 are apparatus claims to a specific four-element measurement structure (transmission-line-segment calibrator; offset-series calibrator with a series resistor; offset-shunt calibrator with a shunt resistor; tested-object measuring instrument) sharing a microwave-probe interface. Every cited patent is a method/system reference and omits at least the distinct offset-series/offset-shunt calibrator structures, the offset transmission lines, and the claimed shared-interface measuring instrument.
3. Closeness ranking:
   • Most relevant for § 103 / method-background: Huang & Lin (2008) and Huang et al. (2010) — same inventor, directly on the "line-series-shunt" lineage; then US 5,608,330 A (Rohde & Schwarz) for the three-standard, unknown-γ, series-plus-shunt-standard concept.
   • Peripheral/field-relevance only: US 2010/0176789 A1 (Missouri) and US 2011/0151596 A1 (TSMC) — cited by examiner but structurally remote from claim 1.
4. Litigation/validity check remains open. As flagged in the earlier section, my tooling cannot reach PACER/Docket Navigator, so "no anticipation identified" is an analytical assessment from the cited record, not a certified clearance or a substitute for a full invalidity search that would also run non-cited art (e.g., Marks' multiline TRL, LRM/LRRM literature, and later on-wafer de-embedding patents).

Caveats and open items: (a) I could not open a native USPTO Patent Public Search session to confirm the examiner's list is exhaustive; the list above is from the patent's own record as mirrored on Google Patents. (b) The page-range/title discrepancy for the Williams/Wang/Arz citation is reported literally, not corrected. (c) Whether US 5,608,330 or the same-inventor Huang publications rise to § 103 invalidity of any specific claim requires an element-by-element claim chart against each, which is beyond what the citation record alone supports.

I'll ground this obviousness analysis by verifying the cited prior-art references and their technical content.

Now let me examine the three patent citations and the foundational NPL references.

Let me verify the remaining references (TSMC de-embedding, NIST multiline TRL, and Engen TRL).

Obviousness Analysis — U.S. Patent No. 8,798,953 B2

Analyst's note: This builds on the two previously generated sections (patent summary and litigation summary) without repeating them. The prior-art set analyzed here is the one appearing in the Prior Art / Citations section of the patent page (uspto.report and Google Patents listings): three patent citations, five non-patent citations, plus the admissions in the patent's own Background. I flag source limitations where they exist. This is a technical analyst's § 103 assessment, not a legal opinion or a validity conclusion.

1. Legal framework and the governing statute

The application was filed 2011-09-01, i.e., before the 2013-03-16 effective date of the AIA first-inventor-to-file provisions. Accordingly the pre-AIA versions of 35 U.S.C. §§ 102 and 103 apply, and the Graham v. John Deere framework governs:

1. Scope and content of the prior art;
2. Differences between the prior art and the claims;
3. Level of ordinary skill in the art;
4. Secondary considerations (objective indicia).

Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), a claim is obvious where the elements were known in the art and combined according to known methods to yield predictable results, where a "design incentive" or "market pressure" existed, or where the combination was "obvious to try." The Federal Circuit's In re Keller / In re Merck line permits a single reference to supply the motivation to combine with another, and the "reference must teach away" defense requires an express teaching, not mere silence. Importantly, the claims here are apparatus/structure claims — so the § 103 question is whether the physical test structures would have been obvious, not whether the equations in the specification (which are not claimed) were novel.

2. Level of ordinary skill

A POSITA here would be a microwave measurement engineer or applied RF device engineer with: (a) a B.S. (typically M.S.) in electrical engineering or physics; (b) 2–5 years of experience with vector network analyzer (VNA) error models, on-wafer probing, and calibration/de-embedding; and (c) working knowledge of TRL/LRM families, ABCD/transmission-matrix cascading, and planar transmission-line (microstrip, coplanar waveguide) design. This is a mature, crowded art with well-developed standards — precisely the setting in which KSR's "predictable results" rationale has its strongest force.

3. Scope of the claims (the reference point for § 103)

As established in the prior summary, claim 1 is the sole independent claim and is a structure claim. Its elements:

Two observations drive everything that follows:

• The claim does not recite the calibration algorithm, the self-calibration equations (1)–(5), the Newton–Raphson solve, or the impedance transformation (6)–(9). Those are the patent's only apparent technical contribution, and they are unclaimed. The claimed subject matter is a set of three on-substrate standards plus a DUT site sharing a probe interface.
• The claim does not itself require equal transmission-line lengths — that appears only in dependent claim 2. Claim 3 is drafted as "equal to or different from," which by construction excludes nothing.

4. The prior-art set on the page

Patent citations (3):

Non-patent citations (5):

Background admissions (part of the intrinsic record, usable as prior art): the patent concedes that (i) two-tier ISS-based SOLT/LRM calibration was conventional; (ii) one-tier de-embedding with five dummy structures was known (N5); and (iii) NIST multiline TRL (N2) already achieved "completion of calibration and de-embedding process in a one-tier fashion."

5. Element-by-element mapping

Conclusion on mapping: every element of claim 1 is disclosed, in substance, across P1, N3 and N4 — the three references that appear on the face of this patent as the examiner-cited art. The mapping is direct enough that the analysis below frames both a § 102 and a § 103 theory.

6. Grounds of obviousness

Ground 1 (strongest): Huang & Lin 2008 (N3) in view of Huang et al. 2010 (N4)

What N3 teaches: a calibration kit consisting of a line standard, a series kit, and a shunt kit, implemented on a conductor-backed coplanar waveguide (a planar, probe-contacted structure), with self-calibration recovering the line propagation constant and the lumped-element impedance/admittance for later de-embedding. What N4 adds: on-wafer implementation, complete modeling of the resistive series and shunt standards (i.e., the "high-frequency parasitic effect devices" of the patent's Figure 2), and reference-impedance transformation based on the acquired characteristic impedance — which is exactly the patent's equations (6)–(9) and step 304.

Why a POSITA would combine: N4 is the direct, expressly identified accuracy improvement of N3's own method. Both are by the same inventor group, in the same journal family, using the same LSS standard taxonomy. Combining a known technique with its own published refinement is the paradigm of KSR "predictable use of prior art elements according to their established functions." There is no teaching away; N4 is framed as an improvement, not an alternative.

Result: claim 1 is rendered obvious. Note the timing point flagged below (§ 9) that N4 may independently be § 102(b) art.

Ground 2: US 5,608,330 (P1) in view of Huang & Lin 2008 (N3) / Huang et al. 2010 (N4)

P1 supplies the self-calibration architecture with exactly three standards — a line plus concentrated-component two-ports including a series impedance standard and a shunt admittance standard — and expressly teaches (a) that only the line's characteristic impedance need be known, with γ solved from the measurements; (b) that such standards are simple and economical to fabricate; and (c) that the method is "especially suited for on-wafer measurements" and realizable in planar stripline technology. In other words, P1 supplies both the technical content and the express motivation to implement three planar standards (line/series/shunt) on-substrate for on-wafer self-calibration.

N3/N4 supply the planar/CPW physical realization on GaAs or similar substrates, the resistive series/shunt kit construction, and the resistive-standard parasitic modeling. Combining P1's calibration architecture with N3/N4's planar standard layouts yields the claim 1 structure.

Motivation, articulated: P1 itself states the problem ("three simple and economical calibration standards") and the suitability for on-wafer use; N3 supplies the planar-technology implementation. A POSITA seeking to eliminate the expensive ISS — the very problem the patent frames as its objective — would have looked to exactly this combination.

Ground 3: Wei 2007 (N5) and the Background admissions, in view of N3/N4/P1

N5 teaches one-tier on-wafer characterization without an ISS, which is the stated objective of the '953 patent ("without using the Impedance Standard Substrate but requiring simply the application of three calibrators"). The patent's own Background admits N5, N2 and the ISS-based two-tier approach. Once the POSITA accepts the goal of ISS-free, one-tier de-embedding (taught by N5, and by the patent's admission), selecting the LSS standard set of N3/N4 (three standards, fewer than N5's five dummy structures, and fewer than N2's multiline set) is an obvious design choice driven by the recognized wafer-area and cost penalty the patent itself identifies. This is a KSR "design incentive" case.

Ground 4: Williams et al. 2003 (N2) and Engen et al. (N1) as the line-standard baseline

N1 and N2 establish (i) that a thru/line standard connecting two feed lines is the canonical first calibration standard on planar substrates and (ii) that line-based standards can be self-calibrated with unknown propagation constant. Claim 1's "transmission line segment calibrator (two transmission lines + a segment between them)" is the definitional thru/line standard of N1/N2. These references reinforce, but are not necessary to, Grounds 1–3.

Ground 5 (secondary): US 2011/0151596 (P3) for the "identical error box" feature of claim 2

Claim 2's equal-transmission-line-length requirement exists to guarantee a common error box / common reference plane across standards and the DUT. P3 teaches exactly this design principle by a different mechanism: forming an auxiliary structure that "replicates" a portion of the primary structure so that the shared ABCD matrices cancel in the de-embedding algebra. Equal feed-line lengths across standards and DUT site is the most direct, routine way to achieve a common reference plane in a probe-contacted layout, and would be an obvious implementation choice to a POSITA designing a calibration kit that must share one probe interface. P1 reinforces this: its standards must be connected "between the two test ports" at the same reference planes.

7. Claim-by-claim conclusion

8. Anticipation overlay (§ 102) — worth flagging

Because the claims are structural and the mapping in § 5 is close to one-to-one, a § 102 challenge should also be considered, not just § 103:

• N4 (Huang et al. 2010) alone. If, as its abstract suggests, the paper's on-wafer LSS calibration kit comprises (i) a line/thru structure, (ii) a resistive series standard, and (iii) a resistive shunt standard, each on CPW, together with a PHEMT DUT site, then every element of claim 1 may be disclosed in that single document. Caveat: I have only the abstract and index metadata, not the full text or figures, so I cannot certify that N4 shows the specific "two transmission lines + offset transmission line + resistor" topology per calibrator. This is the single most important document to obtain in full for a validity challenge, and I flag that I have not verified it.
• Section 102(b) timing of N4. N4 published online 2010-08-12; the application was filed 2011-09-01 — approximately 385 days, i.e., more than one year. Under pre-AIA § 102(b), a printed publication more than one year before the U.S. filing date is prior art even against the inventor's own work (pre-AIA has no general grace period of the AIA § 102(b)(1) type outside the statutory bars). If accurate, N4 is available as § 102(b) art and, by extension, for § 103. Caveat: this is my own date arithmetic from the records I retrieved; the § 102(b) critical date should be independently verified against the operative filing/priority date and the publication's actual distribution date.

9. Counterarguments the patentee would raise, and their likely force

1. "No reference discloses the identical error boxes achieved by equal-length transmission lines." Likely weak. This appears in the abstract/specification and in dependent claim 2, and the underlying design principle (common reference plane / replica structure) is taught by P3 and is inherent in any calibration kit that must share one probe interface (P1, N3). Equal feed-line lengths is also the simplest conceivable way to get there.
2. "The examiner cited these references and the claims still issued." True — the reference record shows the Engen (N1), Huang & Lin (N3), and Huang et al. (N4) citations in the patent's citation list, meaning they were of record and considered. But consideration during prosecution does not foreclose an obviousness challenge; it only means the applicant was able to persuade the examiner. Without the prosecution history, I cannot tell whether a rejection was made and overcome, or never made — I flag this as an unverified point that materially affects the analysis.
3. "The claims are directed to a structure, not the algorithm, so the algorithm's novelty cannot be read out of the claims." This cuts both ways. If the patentee resists importing the algorithm, the claims are broad structural recitations of known calibration-standard topologies — which strengthens the obviousness case. If the patentee imports the algorithm to save validity, it risks a § 112 written-description/enablement fight and the claim-construction question of whether the structure claims can be limited by unclaimed method steps. The patent's central dilemma: the only arguably novel content was left unclaimed.
4. "The claim requires an offset transmission line in the series and shunt calibrators." Weak as a distinction. The specification's own Figure 2 equivalent circuit treats the offset line as the physical length between the probe/calibration reference plane and the lumped element — which is inherent in any planar series/shunt resistor standard, since a resistor cannot be placed exactly at the probe tip. N3/N4's series/shunt kits and P1's concentrated-component standards, once implemented in planar form, necessarily have such an offset.
5. Intrinsic-record contradiction to press on the patentee (from the prior summary, restated for use here): the specification states the offset transmission lines of the series and shunt calibrators "differ," while claim 3 recites "equal to or different from." This inconsistency (a) confirms claim 3 adds no patentable weight, and (b) is a claim-construction vulnerability if the patentee later argues the "differing offset lengths" is the point of novelty.

10. Secondary considerations (objective indicia)

No evidence of secondary considerations appears in the record I retrieved. The assignee is Yuan Ze University, a Taiwanese academic institution, and (per the litigation section) no enforcement activity is known. There is no evidence of long-felt but unmet need, unexpected results, industry praise, licensing, or commercial success with a nexus to the claimed structure. The patent's stated advantages (accuracy improvement, wideband operation, low-cost fabrication, self-calibration) are the same advantages already reported by N3 and N4, which undercuts any "unexpected results" argument. Absent such evidence, secondary considerations do not rebut the § 103 case.

11. Bottom line

On the prior art listed on this patent's own page, claims 1–5 of US 8,798,953 are, in my assessment, vulnerable to a § 103 obviousness attack — and claim 1 is additionally vulnerable under § 102 — subject to the verification caveats below.

• The strongest single ground is Huang & Lin 2008 (N3) in view of Huang et al. 2010 (N4), both already on the face of the patent: N3 supplies the line/series/shunt standard set on planar CPW, and N4 supplies the on-wafer implementation, resistive-standard parasitics, and reference-impedance transformation — i.e., essentially the entire disclosure, with an express improvement relationship providing the motivation to combine.
• The strongest structural ground is US 5,608,330 (P1) in view of N3/N4: P1 affirmatively teaches a three-standard, seven-term self-calibration using a line plus concentrated-component series and shunt two-ports, states the standards are "simple and economical," and declares the method "especially suited for on-wafer measurements" — supplying both content and motivation.
• Dependent claims 2–5 add little: claim 2's equal lengths follow from the common-reference-plane principle (taught by P3); claim 3 excludes nothing on its face; claims 4 and 5 recite conventional planar media and standard commercial probe footprints, respectively.
• The patent's core problem is structural to how it was drafted: the only arguably novel subject matter (the self-calibration mathematics) was never claimed, leaving the claims to cover a physical arrangement of calibration standards that the art already described or would obviously have arrived at.

Uncertainty disclosure / verification needed:

• I retrieved abstracts and metadata, not the full texts of N3, N4, N2 and N5. The § 102 analysis of N4 in particular turns on the figures and the description of the physical kit, which I have not inspected.
• The § 102(b) date arithmetic for N4 (published 2010-08-12; filed 2011-09-01; ~385 days) is my own computation and should be verified against the operative priority date.
• I did not retrieve the prosecution history. Whether any of these references was the basis of an actual rejection, and how it was overcome, is unknown to me and is material.
• I did not verify whether the Taiwanese family members (TWI426289B, "Radio frequency scattering parameter correction method with three correctors"; TWI463147B) published before 2011-09-01, which would bear on § 102(a)/(e). The prior summary correctly treats these as related family members, not as this patent.
• The reference-type labeling on the patent page (citation vs. examiner-cited) is reproduced as it appears; I have not independently confirmed which references the examiner affirmatively relied upon versus merely listed.

Sources: the patent text you supplied (https://patents.google.com/patent/[US8798953B2](/patent/US8798953B2)/en); US 5,608,330 (https://patents.google.com/patent/US5608330A/en, https://www.everypatent.com/comp/pat5608330.html); US 2010/0176789 A1 (https://patents.google.com/patent/US20100176789A1/en); US 2011/0151596 A1 / US 8,436,626 B2 (https://patents.google.com/patent/US20110151596A1); Huang & Lin 2008, DOI 10.1109/TIM.2007.915121 (https://ieeexplore.ieee.org/document/4436225/); Huang et al. 2010, DOI 10.1109/TMTT.2010.2058570; the patent-page citation record at https://patents.google.com/patent/US20130060501A1/en.