US 6304216

I'll search for this specific patent number across USPTO and litigation sources.

Let me search for the specific claims of US6304216 and any litigation/CAFC docket activity.

Let me retrieve the full claim set to identify the independent claims precisely.

Search basis and scope

I searched for the literal number 6304216 / US 6,304,216 B1. Results returned only this patent — no similarly-numbered patents were mixed in. Sources used: the Google Patents full text (authoritative, provided in your message), the USPTO/ITC record, RPX Insight, uspto.report, and Federal Register/ITC notices.

US 6,304,216 B1 — Bibliographic summary

Abstract (per the patent): "A signal detector is provided in which correlation analyses of multiple segments of a received signal from separate and distinct periods of time are combined to improve the detection capability of the signal detector. In one embodiment, the signal detector is part of a GPS receiver."

Plain-language overview of the independent claims

Caveat on sourcing: the Google Patents text supplied for US 6,304,216 breaks off before the claims, so the claim summaries below are reconstructed from RPX Insight, uspto.report and the ITC's identification of asserted claims. Summaries for claims 1, 10, 13, 17 and 22 are well corroborated; claims 43 and 84 are confirmed as independent but I could not retrieve their full verbatim text. Verify against USPTO Patent Center before relying on exact language.

Claim 1 — Signal detector (apparatus, broadest). A receiver captures a first and a second segment of a signal taken at separate, distinct time periods; the signal is a signal‑of‑interest buried in noise/pseudo‑noise. A correlator produces first and second correlation data (how well each segment matches a hypothesis). A combiner measures a parameter difference between the two correlation results, adjusts one of them to account for that difference, and adds the adjusted result to the other to build "cumulative correlation data" that helps detect the signal or a parameter of it. (This is the core "capture separate snips, fix them up, and stack them" idea.)

Claim 10 — Signal detector, PN‑code specific. Same architecture, but the signal is a plurality of signals each being a repeating PN code modulated on a carrier. The correlator tests a combined PN code + code phase hypothesis, and the combiner determines a code phase difference between corresponding portions of the two correlation results, adjusts one portion, and combines. Dependent claim 11 recites the correlator is a matched filter.

Claim 13 — Method, broad. Parallel method counterpart of claim 1: receive two segments from separate, distinct periods; derive first and second correlation data against a hypothesis; determine a parameter difference; adjust a selected one of the two; combine adjusted with unadjusted to get cumulative correlation data. Dependents (14–16) add refinement.

Claim 17 — Method, PN‑code specific. Parallel method counterpart of claim 10: generate a code phase hypothesis, correlate each segment against a combined PN‑code/code‑phase hypothesis, determine a code phase difference between corresponding portions, adjust one, and combine. Dependents 18–21.

Claim 22 — Signal detector, means‑plus‑function. Written as a "means for" claim covering: means for receiving two segments from distinct periods; means for correlating each segment against a hypothesis; means for determining a parameter difference; means for adjusting a selected correlation result; and means for combining to produce cumulative correlation data. Because it is means‑plus‑function, scope is tied to the corresponding structure disclosed in the specification (receiver, matched filter/correlator, GPS processor/combiner).

Claim 43 — Independent signal detector claim (exact text not retrieved). Confirmed independent by its dependent claim 63 ("The signal detector of claim 43, wherein the combiner successively combines the correlation data for a plurality of successive segments to detect a parameter of the signal of interest"). Its likely focus is the successive/iterative accumulation aspect, but I flag this as unverified.

Claim 64 — Method, "select-and-adjust" form. Receive a first and a second segment of a signal of interest; derive first and second correlation data against a hypothesis; determine a parameter difference; select one of the two correlation data sets; adjust the selected one in response to the parameter difference; combine the adjusted data with the non‑selected data to form cumulative correlation data. Dependents add: non‑overlapping periods (65), plural signals of interest (66), PN‑modulated carrier (67), combined PN‑code/code‑phase hypothesis (68), and — notably — claim 86, which recites the code‑phase‑difference equation ΔCP = [(F_PN + D) × ΔT] modulo T_PN, and claim 87, which recites the offset equation ΔT = (T_n − T_2 + OS₁ − OS₂) × S. Those two dependent claims tie the claim set directly to the specification's Equations (1) and (2).

Claim 84 — Computer‑readable medium. A program/instructions claim covering: receiving first and second segments representing distinct periods of time, deriving correlation data for each, and combining. Dependent claims 85–87 add code‑phase‑difference determination between PN‑code/Doppler‑shift/code‑phase portions and the same ΔCP and ΔT equations.

Technical core across all independents: the specification's key insight is Equation (1), ΔCP = [(F_PN + D) × ΔT] modulo 1 mS, with ΔT = (T_n − T_2 + OS₁ − OS₂) × S (Equation 2) — i.e., you can circularly shift a correlation row by the computed code‑phase difference and add it to the accumulated array even though the actual code phases CP₁ and CP₂ are unknown. This is what lets a GPS receiver in a cellular phone harvest many short, irregular idle‑slot windows and stack them to raise effective SNR.

Litigation and docket check

ITC Investigation No. 337‑TA‑596 — Certain GPS Chips, Associated Software and Systems, and Products Containing Same (confirmed).

• Complaint filed by SiRF Technology, Inc. (San Jose, CA) on Feb. 8, 2007 (supplemented Feb. 22, 2007); investigation instituted March 13, 2007, 72 Fed. Reg. 11378.
• The '216 patent was one of four asserted (with US 7,043,363; 7,091,904; 7,132,980). Respondent: Global Locate, Inc., later joined by Broadcom, Inc.
• Asserted claims of the '216 patent: 1, 6, 10–12, 17–19, 64, 65, 69, 70, 72, 73.
• ALJ final Initial Determination June 13, 2008: Global Locate's products did not infringe the asserted '216 claims; the asserted claims were not invalid as anticipated; and SiRF failed to prove a domestic industry for the '216 patent. Commission declined review (73 Fed. Reg. 49219, Aug. 20, 2008) — final determination of no violation of Section 337.
• The Google Patents litigation link (Unified Patents) also points to ITC case 337‑TA‑596, consistent with the above.
• Note: this litigation is the inverse posture from the parallel Broadcom‑v‑SiRF dispute (ITC 337‑TA‑658 and D. Del./C.D. Cal. actions), where SiRF was the respondent on different Broadcom patents. Do not conflate the two.

CAFC 2026 dockets — no match found. I found no Federal Circuit 2026 appeal docket involving US 6,304,216. The only CAFC 2026 result returned was Clear Imaging Research LLC v. Google LLC, No. 2026‑1485 (S.D. Cal. 3:25‑cv‑00221, Judge Curiel), dismissed by agreed order on June 22, 2026 — unrelated to this patent. Given the '216 patent expired March 30, 2019 and its ITC assertion ended in 2008 with no appeal, absence of 2026 CAFC activity is expected. I state this as "no evidence found," not as a guarantee, since I could not run an exhaustive docket‑by‑number query across all CAFC 2026 filings.

Uncertainty disclosure

1. Claim text for independent claims 43 and 84 was not retrieved verbatim; only their independence and the content of their dependents is confirmed.
2. The claim summaries are drawn from secondary aggregators (RPX Insight, uspto.report) plus the ITC's asserted‑claim list, because the patent copy supplied to me is truncated before the claims section. Exact wording should be confirmed at USPTO Patent Center.
3. Google Patents lists "Current Assignee: Qualcomm Inc" and "Original Assignee: Conexant Systems LLC," while the assignment records say Conexant Systems, Inc.; I have reported the assignment‑record version and noted both.
4. No CAFC 2026 docket was found, but this is a negative search result rather than an affirmative clearance.

I'll research litigation involving US Patent 6,304,216 (note: the patent number is literally "6304216", which as granted is US 6,304,216 B1).

The general searches returned mostly unrelated Unified Patents materials. Let me search the specific litigation leads from the patent record (ITC 337-TA-596) and SiRF/CSR enforcement campaigns.

I found the key ITC investigation. Let me dig into the parallel district court case and any further assertions/assignments.

Summary of findings — US Patent 6,304,216 ("6304216")

Interpretation note: The identifier "6304216" corresponds literally to U.S. Patent No. 6,304,216 B1, titled "Signal detector employing correlation analysis of non-uniform and disjoint sample segments" (inventor: Steven A. Gronemeyer; original assignee Conexant Systems, Inc.; filed and priority date March 30, 1999; granted October 16, 2001). All results below relate to that patent. Note the patent expired March 30, 2019 (anticipated expiration), and current assignee of record is Qualcomm Incorporated (2024 assignment from CSR Technology Inc.).

I found one (1) known litigation matter in which the '216 patent was actually asserted — a Section 337 investigation at the U.S. International Trade Commission. I found no district court or appellate case in which the '216 patent was a patent-in-suit.

Case 1 — ITC Investigation No. 337-TA-596

In the Matter of Certain GPS Chips, Associated Software and Systems, and Products Containing Same

Outcome detail specific to the '216 patent (from the Commission's Notice, 73 FR 49219):

• The ALJ found that Global Locate's products do not infringe asserted claims 1, 6, 10–12, 17–19, 64, 65, 69, 70, 72, and 73 of the '216 patent.
• The ALJ found the asserted claims of the '216 patent were not invalid as anticipated by the prior art.
• The ALJ found that SiRF failed to prove that a domestic industry exists for articles protected by the '216 patent.
• The investigation had earlier been terminated as to the '904 and '980 patents and certain claims of the '216 and '363 patents. SiRF petitioned for Commission review on the '216 patent (June 27, 2008); after briefing, the Commission declined review, so the "no violation" determination became final. (The companion '363 patent was found invalid as anticipated by the "GPS Builder System" and the "First GPS system.")

Sources: Federal Register notice of investigation, 72 FR 11378 (Mar. 13, 2007), https://www.federalregister.gov/api/v1/documents/E7-4555; Federal Register notice of Commission decision, 73 FR 49219 (Aug. 20, 2008), https://www.govinfo.gov/content/pkg/FR-2008-08-20/html/E8-19209.htm; USITC Pub. 4133 (final ID, "IV. '216 Patent"), https://www.usitc.gov/intellectual_property/documents/pub4133.pdf.

Related proceedings that did NOT assert the '216 patent (flagged to avoid confusion)

These arose from the same SiRF/Global Locate/Broadcom GPS dispute but involve different patents:

1. SiRF Technology, Inc. v. Global Locate, Inc., et al. — patent infringement complaint filed in the U.S. District Court for the Central District of California (announced mid-December 2006; SiRF's 10-K states the complaint was filed December 15, 2006). The patents asserted were 7,091,904; 7,043,363; 6,850,557; and 6,636,178 — not the '216 patent.
2. Global Locate's ITC complaint, Inv. No. 337-TA-602 (instituted May 7, 2007; respondents SiRF, E-TEN, Pharos, MiTAC, Mio) — asserted Global Locate's patents (6,417,801; 6,606,346; 6,651,000; 6,704,651; 6,937,187; 7,158,080), not the '216 patent. Final determination Jan. 15, 2009; affirmed by the Federal Circuit in SiRF Tech., Inc. v. U.S. Int'l Trade Comm'n, 601 F.3d 1319 (Fed. Cir. 2010) (the Federal Circuit opinion at 09-1262 concerns these A-GPS patents, not the '216 patent).
3. Broadcom Corp. v. SiRF Technology, Inc. and CSR plc, No. SACV08-546 JVS (MLGx) (C.D. Cal., filed May 14, 2008) — asserted Broadcom's own patents (e.g., 7,310,104), not the '216 patent.
4. SiRF Technology, Inc. v. Orrick, Herrington & Sutcliffe LLP, No. CV 09-4013 MHP (N.D. Cal.) — a legal-malpractice suit referencing ITC Investigations 337-TA-596 and 337-TA-602; the '216 patent was not asserted there.

Caveats and limits of this report

• I performed targeted web searches of Federal Register/ITC records, SEC filings, Federal Circuit opinions, and third-party litigation aggregators. I hit the tool-call limit, so this should not be treated as an exhaustive PACER, Docket Alarm, or Darts-IP search.
• I found no evidence of any post-2011 assertion of the '216 patent (i.e., during the CSR/Qualcomm ownership period), and no NPE-style litigation campaign.
• I did not locate a Federal Circuit appeal specifically addressing the '216 patent. Because SiRF's petition for Commission review was denied and the Commission found no violation, any appeal avenue would have run from the August 2008 final determination; I could not confirm whether SiRF filed one, and I am not asserting that it did or did not.
• The Google Patents family/litigation data for this patent points to the same matter: Unified Patents' litigation record lists ITC case 337-TA-596 (https://portal.unifiedpatents.com/litigation/International%20Trade%20Commission/case/337-TA-596).

Bottom line: US 6,304,216 was asserted once, by SiRF Technology against Global Locate (and Broadcom) in ITC Investigation 337-TA-596, instituted March 13, 2007. SiRF lost — the ALJ found non-infringement and no domestic industry as to the '216 patent, and the Commission declined review on August 14, 2008, ending the investigation with a finding of no violation of Section 337.

I'll verify the "no PTAB activity" default against public sources before finalizing.

I hit the tool-call limit while verifying, so this reflects the canonical structured data plus the targeted searches I completed. No contradiction surfaced between them.

Proceedings overview

Total AIA trial proceedings on US 6,304,216: 0 — no IPR, no PGR, no CBM, and no institution denials, settlements, or claim cancellations, because no petition was ever filed; the canonical USPTO ODP list in the structured block returns an empty set, and my independent web searches surfaced no AIA proceeding naming this patent. Bottom line for a defendant: you are not facing a PTAB-hardened patent or a PTAB-killed one — you are facing an expired patent (expired 2019-03-30) that no one ever bothered to attack at the Board, so any § 102/§ 103 defense must be built from scratch in district court or in a § 282 trial, not imported from an FWD.

(No proceedings to rank)

There are no proceeding entries to list. Per the task instructions I will not invent proceeding numbers, so this section is intentionally empty of the ### {PROCEEDING_NUMBER} blocks.

What I searched for, and what I found

Because "no PTAB activity" is an affirmative claim, I tried to falsify it:

Number-collision warning (important, and the most likely source of a wrong answer to this question). Several unrelated patents are habitually called "the '216 patent" in PTAB and Federal Circuit papers. None is US 6,304,216:

• US 5,490,216 (Uniloc) — Uniloc v. Sega, IPR2017-…, cert. petition 17-1018 (seventh-Amendment / expired-patent arguments). Different patent, different owner.
• US 8,195,216 (Philips/UMTS power control) — IPR2021-00563 (Banner & Witcoff copy at https://bannerwitcoff.com/wp-content/uploads/2025/04/IPR2021-00563.pdf), FWD 2022-09-13 invalidating claim 9, Fed. Cir. No. 23-1896 (nonprecedential, 2024-08-29) affirming-in-part/vacating-in-part and remanding; opinion at http://cafc.uscourts.gov/opinions-orders/23-1896.OPINION.8-29-2024_2376156.pdf.
• A third "216 patent" (case 6:20-cv-01707, D. Del.) appears in a Fintiv/Advanced Bionics institution brief against Quectel — that is a different '216 patent in a 2020 Delaware campaign, not this GPS patent.

If a docket aggregator reports "the '216 patent survived/fell in IPR," check the patent number digit-for-digit before relying on it. Nothing in those proceedings is estoppel or persuasive authority against US 6,304,216.

Why zero is unsurprising (structural, not just empirical).

1. Timing. US 6,304,216 issued 2001-10-16; AIA trials became available 2012-09-16, and the PGR window (9 months post-grant) closed long before. Its only serious assertion window — ITC Inv. No. 337-TA-596, instituted 2007-03-13 — predates IPR entirely. Global Locate/Broadcom therefore litigated validity the old-fashioned way, in the ITC.
2. The ITC already adjudicated validity, and SiRF lost. Per the Commission's notice of decision (73 FR 49219, 2008-08-20, https://www.govinfo.gov/content/pkg/FR-2008-08-20/html/E8-19209.htm) and USITC Pub. 4133 ("IV. '216 Patent", https://www.usitc.gov/intellectual_property/documents/pub4133.pdf), on 2008-06-13 the ALJ found no violation: Global Locate's products do not infringe asserted claims 1, 6, 10–12, 17–19, 64, 65, 69, 70, 72, 73; the asserted claims were not invalid; and SiRF failed to prove domestic industry. The Commission declined review (determined 2008-08-14). So the only governmental validity adjudication on this patent is a defense-favorable non-infringement/no-DI outcome — the claims were expressly not held invalid, which is exactly the opposite of a "claims canceled" posture.
3. Fintiv-era discretionary denial would have been near-certain anyway. Had anyone filed post-2019, the patent was seven-plus years old and expired; the "settled expectations" line of Director discretion now used to deny IPRs rests squarely on patents six years or older. This is a hypothetical, but it reinforces that a 2026 attack would have to be a district-court § 282 or § 112/101 defense or an ex parte reexamination request (which several aggregators have pursued where IPR denial is likely — cf. Unified's 2025 practice notes).

Strategic summary

Claim status: 100% UNTESTED at the PTAB. No claim of US 6,304,216 has been canceled, confirmed, or even challenged in an AIA trial. Claim 1, claim 10 (the second independent signal-detector claim), claim 12 (the GPS-receiver claim), claim 17 and claim 64 (independent method claims), and the asserted dependents (6, 11, 18–19, 65, 69, 70, 72, 73) all stand as issued — but the entire patent expired 2019-03-30 (anticipated expiration, confirmed in the structured data; current assignee Qualcomm Incorporated via a 2024 assignment from CSR Technology Inc.). The operative practical limit is damages: infringement after 2019-03-30 is legally impossible, so recovery is confined to pre-expiration acts within the § 286 six-year lookback, meaning a 2026 filing can capture at most post-2013-09-29 conduct, and realistically only conduct up to expiration.

Estoppel landscape: essentially none, which cuts both ways. Because there is no IPR/PGR, § 315(e)(2) estoppel is a non-issue — no petitioner and no privy is estopped from anything. No IPR petitioner had to stipulate away § 102/§ 103 grounds. Conversely, there is no FWD to borrow: you cannot point to a PTAB claim construction, an instituted ground, or a cancellation to short-circuit a plaintiff's case. A defendant's invalidity case must be independently built and, if you elect to try it, you must be prepared to prove it by clear and convincing evidence under § 282 — the ITC's 2008 "not invalid" finding, while not issue-preclusive against a newcomer (different parties, different forum, and SiRF/Global Locate are both long gone), is a bad-fact for your narrative and should be scoped early with a motion in limine if you intend to relitigate the same primary references. Conversely, the ITC record (USITC Pub. 4133, including the stipulated technology overview and the '216 claim constructions) is a ready-made roadmap of how the claims were mapped and what the respondents argued — read it before you choose a § 102/§ 103 theory.

Pattern signals. No serial petitioner, because there is no petitioner. No patent owner appeal of an FWD, because there was no FWD. No defensive aggregator participation at the PTAB on this patent — Unified Patents' only recorded involvement is as a litigation-data source pointing at ITC 337-TA-596, not as a petitioner. The enforcement pattern for this patent is single-shot and historic: one ITC Section 337 complaint by SiRF Technology against Global Locate/Broadcom, instituted 2007-03-13, ending in a 2008-08-14 final determination of no violation. The family's sibling patents were the ones that kept fighting — US 6,636,178 was asserted in the C.D. Cal. SiRF v. Global Locate action and reissued as US 7,002,516 lineage, and the '363/'904/'980 patents featured in ITC 337-TA-596/602 and the subsequent § 101/prior-art battles. Anyone asserting 6,304,216 today is therefore doing so on a patent with no modern litigation or PTAB history at all, which suggests either a stale portfolio assertion, a licensing/damages-only theory against pre-2019 devices, or a mistake about which patent they own.

Recommended next steps

1. Confirm the obvious first: is the patent still enforceable on your facts? US 6,304,216 expired 2019-03-30. If the demand letter or complaint accuses conduct after that date, the claim fails as a matter of law on the face of the pleading — check maintenance fees, term, and any terminal disclaimer, and move to dismiss or for summary judgment on quantum/accrual rather than spending money on validity.
2. Do not import "the '216 patent IPR" from any docket report. Have counsel pull the number from the PTAB Center/PTAB E2E (https://ptacts.uspto.gov/ptacts/) and PATENTCENTER, and independently confirm zero trial proceedings and zero reexamination control numbers for 6,304,216. My searches found none, but I hit the tool-call limit and did not complete an exhaustive Docket Alarm/PACER/Darts-IP sweep — treat the absence as very well supported, not as certified.
3. Leapfrog the Board if you need an administrative kill. With no IPR available in practice on a 2019-expired patent and discretionary denial now near-automatic for patents over six years old, a § 302 ex parte reexamination request on the strongest single-reference § 102 art (plus a § 112 written-description/claim-support attack on the "parameter/partial data" claims, which an examiner can consider) is the realistic administrative route. It has no § 315(b) clock and cannot be estopped.
4. Mine the ITC record before choosing grounds. USITC Pub. 4133 (https://www.usitc.gov/intellectual_property/documents/pub4133.pdf) contains the stipulated '216 technology overview, the claim constructions, and the respondents' invalidity theories. It also tells you what the ALJ expressly rejected — meaning those references were considered and survived, and are unlikely to carry a reexamination on their own.
5. If a plaintiff does allege infringement, extract admissions on the expired-claim scope. The asserted claim set from 337-TA-596 (1, 6, 10–12, 17–19, 64, 65, 69, 70, 72, 73) is your starting universe for § 282 — pin the plaintiff to those claims and to pre-2019 accused products, then attack the "combiner"/"parameter difference" limitations under § 112 (the specification's support for "parameter difference" beyond code phase is thin) rather than fighting only the prior art.
6. Set no PTAB calendar items. There is no institution deadline, no oral hearing, and no FWD due date to track, because no trial exists. Any litigation stay you seek must rest on the expired-patent and damages issues, not on a pending Board proceeding.

Caveat on completeness: I completed targeted searches of USPTO PTAB materials, Federal Register/ITC records, and secondary PTAB commentary, but I hit the tool-call limit before running a full PTAB E2E docket lookup and a CourtListener sweep for any Federal Circuit appeal touching this patent. The absence of AIA proceedings is consistent across every source I did consult and with the canonical structured data, but I flag the residual search gap rather than claim an exhaustive PACER/Docket Alarm review.

I'll research the assignment chain for US 6,304,216 across USPTO records, SEC filings, and corporate history.

Good — I found the corporate deal context. Now let me try to locate the actual reel/frame records and correspondents.

Assignment & Ownership Reconstruction — US 6,304,216 ("6304216")

Verified baseline (authoritative for dates/parties): the Google Patents legal-events table for US6304216B1 (https://patents.google.com/patent/US6304216/en), which mirrors the USPTO Assignment record. Important limitation up front: I was unable to retrieve USPTO reel/frame numbers or correspondent fields for this patent within this research session — the Assignment Center / Assignment Search interfaces are form- and session-driven and did not surface machine-readable per-patent records to my searches. Everything below is tagged by confidence. I have not invented reel/frame numbers.

Inventors

• Sole inventor — no joint inventors. The patent face (75) reads "Steven A. Gronemeyer, Cedar Rapids, IA (US)"; assignee (73) is "Conexant Systems, Inc., Newport Beach, CA (US)" (US6304216B1 front page, https://patentimages.storage.googleapis.com/fb/0a/39/4a67f3eb6e70f5/US6304216.pdf).
• Employer determination: Conexant's GPS design center was located in Cedar Rapids, Iowa, and Newport Beach, CA — corroborated by the SiRF/Conexant deal coverage: "SiRF will establish offices in Newport Beach and Cedar Rapids, Iowa, where Conexant's GPS employees are currently located" (EE Times, "Conexant sells GPS chip business to SiRF," 2001‑07‑02, https://www.eetimes.com/conexant-sells-gps-chip-business-to-sirf/). Gronemeyer's Iowa residence matches the Conexant GPS group.
• Departure pattern — signal NOT present. Gronemeyer did not leave the assignee within 12 months of filing. He is named inventor on later patents flowing from the same business unit, e.g. US 7,999,733 (SiRF Technology Inc.), US 8,446,223 (CSR Technology, Inc.), and US 2017/0242130 A1 (CSR/Qualcomm era). He also appears as a conveying party on a SiRF confirmatory assignment recorded at Reel 019103/0984 (executed 12/14/2006; app. 10/570,833) — i.e., he was still inside the SiRF-side chain seven years after this filing. There is no inventor-exodus signal preceding a portfolio sale.

Original assignee

Conexant Systems, Inc., Newport Beach, CA (Delaware corporation; formerly the semiconductor spin-off of Rockwell International, spun off in 1999).

• Primary line of business: communications semiconductors. Its GPS line traced back to Rockwell's GPS division (the "Jupiter"/"Zodiac" GPS modules and chip sets).
• Did it ship a product embodying the claims? Yes, in the practical sense: Conexant sold a GPS chip set / OEM receiver-module line (Zodiac chip set, Jupiter modules). The '216 patent's own specification names Conexant hardware: the matched-filter chip code-named "Magna", the processor chip code-named "Scorpio," Part No. 11577‑11, and the RF chip code-named "Gemini/Pices Monopack," Part No. R6732‑13, all "available from the assignee of the subject application" (Conexant). Commercialization is documented in the specification itself.
• Current status: Operating, but no longer in GPS. Conexant exited GPS entirely — its GPS assets (including IP) went to SiRF; the module business then moved to Navman (EE Times, 2002‑04). Conexant later sold its broadband-access/Wi‑Fi lines (e.g., to Ikanos, Xocyst Transfer AG / IPValue-managed nCube-type vehicles, and eventually Synaptics acquired Conexant in 2017, with remaining assets going to Sckipio/Espressif-era divestitures). Conexant is not a current owner of this patent.

Assignment timeline

Confidence key: Dates + conveyance types + parties = verified (Google Patents legal events, which reproduce the USPTO Assignment record). Reel/frame + correspondent = not retrieved this session (see caveat at top).

• 1999-03-30 (executed on/at filing) / recorded 1999-03-30 — Reel unverified

  • Conveyance: Assignment of assignors' interest
  • Assignor: Steven A. Gronemeyer (inventor)
  • Assignee: Conexant Systems, Inc., Newport Beach, CA
  • Correspondent: not retrieved. Lead (unconfirmed for this patent): Conexant's recorded correspondent across its assignment family is Keith Kind, Conexant Systems, Inc., Office of the General Counsel, 4000 MacArthur Blvd., Newport Beach, CA 92660 — e.g., the Conexant/MINDSPEED records reproduced in the PTAB Exhibit 1012 (IPR2017‑00276) assignment tables, Reel 014546/0305 and Reel 018847/0871. Treat as a probable correspondent, not a confirmed one for the '216 record.
  • Context: Standard employee invention assignment to the operating employer; filed the same day as the application.

• 2000-01-03 — Reel unverified

  • Conveyance: Security Interest / Security Agreement
  • Assignor: Conexant Systems, Inc.
  • Assignee: Credit Suisse First Boston
  • Correspondent: not retrieved
  • Context: Securitization — collateral grant under Conexant's corporate credit facility; this is a lien, not a title transfer.

• 2001-11-05 — Reel unverified

  • Conveyance: Release of Security Interest
  • Assignor: Credit Suisse First Boston
  • Assignee of record entry: Conexant Systems Worldwide, Inc.; Brooktree Worldwide Sales Corp.; Brooktree Corp.; Conexant Systems, Inc.
  • Correspondent: not retrieved
  • Context: Securitization release — the 2000 lien discharged (coincides with the GPS business sale closing in this window).

• 2002-06-17 — Reel unverified

  • Conveyance: Assignment of assignors' interest
  • Assignor: Conexant Systems, Inc.
  • Assignee: SiRF Technology Holdings, Inc.
  • Correspondent: not retrieved
  • Context: Asset acquisition. This recordation is the IP leg of the Conexant→SiRF GPS deal. Independent corroboration: SiRF's Form S‑1/A states "In connection with the purchase of the GPS semiconductor product line of Conexant, we entered into an agreement with Conexant," and "On September 21, 2001, the Registrant issued 6,704,483 shares of series G preferred stock as part of an Agreement and Plan of Contribution and Merger with Conexant Systems, Inc." (SEC EDGAR, https://www.sec.gov/Archives/edgar/data/[1163943](/patent/1163943)/000119312504043279/ds1a.htm). See contradiction flag below.

• 2002-10-29 — Reel unverified

  • Conveyance: Assignment of assignors' interest
  • Assignor: SiRF Technology Holdings, Inc.
  • Assignee: SiRF Technology, Inc.
  • Correspondent: not retrieved. Lead (unconfirmed): SiRF's IP-recording correspondent in the 2006–2007 window was a Granada Hills, CA firm at 10605 Balboa Blvd., Suite 300 (contact e-mail domain eclipsegrp.com), submitter first name "Jeffrey C." — per the SiRF confirmatory assignment at Reel 019103/0984 (a different SiRF property). Same caveat: probable, not confirmed here.
  • Context: Internal reorganization — movement from the holding company to the operating subsidiary, ~4.5 months after the acquisition recordation.

• 2011-12-22 — Reel unverified

  • Conveyance: Change of Name (not an assignment)
  • Assignor: SiRF Technology, Inc.
  • Assignee: CSR Technology Inc.
  • Correspondent: not retrieved
  • Context: Change of name only — follows CSR plc's acquisition of SiRF (announced 2009) and the rebranding of SiRF's US entity; no change in ultimate beneficial ownership.

• 2024-10-14 — Reel unverified

  • Conveyance: Assignment of assignors' interest
  • Assignor: CSR Technology Inc.
  • Assignee: Qualcomm Incorporated
  • Correspondent: not retrieved
  • Context: Post-acquisition cleanup/recordation. Qualcomm completed its acquisition of CSR plc in August 2015 (≈$2.4B); this 2024 recording is a ~9-year-late confirmatory/clean-up recordation typical of large-cap portfolio integration, not a fresh arm's-length sale. This matches the patent page's "Current Assignee: Qualcomm Inc" and the 2024‑10‑14 reassignment entry.

No further recorded events. The patent reached its anticipated expiration on 2019-03-30 (Google Patents legal status: "Expired – Lifetime"), i.e., it was already expired when the Qualcomm recordation was filed.

⚠️ Contradiction flag (source conflict, not a fabrication)

The deal-announcement/close dates conflict with the recordation date, and I am not resolving this by assumption:

Reconciliation hypothesis (flagged as hypothesis): the deal was structured as a contribution + stock issuance (July 2001 agreement → Sept 21, 2001 stock issuance), with the formal IP assignment documentation executed/recorded at USPTO in June 2002. The "2003" figure in the KIPO piece appears to be an error, and I note it as such. Anyone relying on the ownership date between Sept 2001 and June 2002 should pull the actual recorded instrument.

Timeline diagram

timeline
    title Ownership of US 6304216
    1999 : Inventor Gronemeyer assigns to Conexant
         : Application filed 30 Mar 1999
    2000 : Conexant grants security interest to Credit Suisse
    2001 : Credit Suisse releases security interest
    2002 : Conexant GPS IP assigned to SiRF Technology Holdings
         : Moved to SiRF Technology Inc
    2007 : Asserted by SiRF in ITC 337-TA-596
    2008 : SiRF loses - no violation found
    2011 : Name change to CSR Technology Inc
    2019 : Patent expires 30 Mar 2019
    2024 : Assigned to Qualcomm Incorporated

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. Every recorded assignee is an identified operating entity with a real principal place of business: Conexant Systems, Credit Suisse First Boston (lien only), SiRF Technology Holdings, SiRF Technology Inc., CSR Technology Inc., Qualcomm Incorporated. No "IP/Holdings/Ventures/Licensing" LLC, no registered-agent-service address, no single-member Delaware/Texas vehicle appears anywhere in the chain.

2. Known asserter in the chain — NOT PRESENT. None of the recorded owners matches the NPE directories named in the brief (Acacia, Marathon, IV, IPNav, Wi‑LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation, Spangenberg entities). Conexant, SiRF, CSR and Qualcomm are all product companies. Note carefully: SiRF was a plaintiff — but as an operating company suing a competitor (Global Locate/Broadcom) in ITC 337‑TA‑596, which is operating-company assertion, not NPE conduct. SiRF's public position was that its portfolio comprised "more than 250 patents awarded worldwide and more than 370 additional patent applications pending," used defensively/offensively in a two-way A‑GPS chip war (KIPO dispute bulletin, 2008‑10‑21).

3. Repeat correspondent across the chain — UNCLEAR / NOT ESTABLISHED. I could not retrieve the correspondent-of-record field for any '216 entry. Two unconfirmed leads surfaced and are flagged precisely as leads: (a) Keith Kind, Conexant Systems, Inc. (Newport Beach) recurs as correspondent across Conexant's recorded assignment family — but on other Conexant patents (e.g., Reel 014546/0305; Reel 018847/0871, reproduced in PTAB Exhibit 1012, IPR2017‑00276), not verified on this patent's reel/frame; (b) a Granada Hills, CA firm with the eclipsegrp.com domain, submitter "Jeffrey C.," appears as SiRF's recording correspondent on Reel 019103/0984, again a different SiRF property. Under the brief's own rule — recurrence is the finding, a single appearance is not — neither rises to a finding, and neither is an NPE attorney. Corroborating one against the other is not evidence.

4. Cascading transfers — WEAK / NOT PRESENT. There are exactly two SiRF-side transfers, 2002‑06‑17 and 2002‑10‑29 (~4.5 months apart). They are a parent→subsidiary reorganization into a named corporate family, not chained LLCs, and there is no shared-secret correspondent address or common principal pattern visible. Subsequent links are separated by 9 years (2002→2011) and 13 years (2011→2024). No cascade.

5. Pre-litigation transfer — NOT PRESENT. The nearest transfer to the only suit (ITC 337‑TA‑596, instituted 2007‑03‑13) is the 2002‑10‑29 SiRF Holding→SiRF Inc. recordation — over four years earlier. The 2011 and 2024 transfers post-date the litigation by 3 and 16 years.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 sale appears in the chain. Conexant divested GPS as a strategic exit (EE Times: "In a move to cut costs…"), for SiRF stock, not via insolvency. SiRF's 2009 combination with CSR plc was a merger/acquisition, not a bankruptcy sale; CSR's 2015 acquisition by Qualcomm was a cash tender. Nothing here resembles Kodak/Nortel/Polaroid.

7. Privateering — NOT PRESENT. Conexant did not hand the patent to an NPE to sue its rivals. It sold the GPS business and the employees to a direct acquirer/operator ("SiRF plans to retain all of the employees from Conexant's GPS business," EE Times 2001‑07‑02). Conexant's retained rights were back-licenses for its own mobile-communications products — the opposite of a privateering structure. One adjacent, ownership-neutral data point worth recording: SiRF and Qualcomm signed a mutual patent non-assertion agreement in October 2008 covering each party's portfolio — a defensive arrangement between the then-owner and the eventual owner's parent, which tends to reduce NPE risk for this asset.

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT, Unified Patents or OIN. It terminates at Qualcomm Incorporated — a large operating semiconductor/telco-R&D company with every incentive to hold rather than neutralize.

Verdict

Operating-company assertion.

The chain is a clean, single-direction corporate lineage — inventor → Conexant (1999) → SiRF Technology Holdings → SiRF Technology, Inc. (2002 recordations) → CSR Technology Inc. as a name change only (2011) → Qualcomm Incorporated (2024 cleanup) — with the only non-transfer event being the 2000-01-03 Credit Suisse First Boston security-interest grant discharged by the 2001-11-05 release. The patent was asserted exactly once, by the operating GPS chip vendor that owned it (SiRF), against a direct competitor (Global Locate/Broadcom) in ITC 337-TA-596 — and SiRF lost (ALJ final ID 2008‑06‑13 finding non-infringement and no domestic industry; Commission declined review 2008‑08‑14). There is no shell-entity transfer, no known asserter, no pre-litigation transfer, no bankruptcy sale, and no defensive-aggregator endpoint, so no NPE verdict is supportable.

Caveat on verdict strength: this conclusion rests on verified dates/conveyances/parties but on unverified reel/frame and correspondent data. Because the primary evidence base is the Google Patents legal-events table rather than the Assignment Center's reel/frame and correspondent fields, I have marked item 3 (repeat correspondent) as unclear rather than not present. If a full reel/frame pull later shows a single repeat correspondent running the 2002 SiRF-side filings and the 2011 name change, that would not by itself change the verdict — recurrence of counsel is only probative when combined with a shell-entity assignee, and none exists here.

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (alternate interface: https://assignment.uspto.gov/patent/index.html), search by patent number 6304216; corresponding application 09/281,741. Family members to pull alongside it: US 6,636,178 B2 (app. 09/971,293, filed 2001-10-04) and US 7,002,516 B2 (app. 10/644,311, filed 2003-08-19) — both same priority date, both should show the same SiRF-side chain and are useful to cross-check the 2002 recordation dates and correspondents.

Sources used to ground the corporate narrative (non-USPTO):

• EE Times, "Conexant sells GPS chip business to SiRF," 2001-07-02 — https://www.eetimes.com/conexant-sells-gps-chip-business-to-sirf/
• EE Times, "Conexant completes GPS exit as it transfers module business to customer," 2002-04-05 — https://www.eetimes.com/conexant-completes-gps-exit-as-it-transfers-module-business-to-customer/
• EE Times, "Former Rockwell GPS modules move one more time," 2002-04-11 — https://www.eetimes.com/former-rockwell-gps-modules-move-one-more-time/
• SiRF Technology Holdings, Form S-1/A (2004), SEC EDGAR — https://www.sec.gov/Archives/edgar/data/1163943/000119312504043279/ds1a.htm
• SiRF confirmatory assignment, Reel 019103/0984 (app. 10/570,833), legacy USPT

I'll search for the specific patent and its cited references.

Let me pull the full "References Cited" list from the patent front page and Google Patents.

Let me try dedicated patent databases for the cited references list.

Prior-art analysis — US 6,304,216 B1 ("6304216")

0. Identifier handling and search scope

Per the strict rule, I interpreted "6304216" literally as U.S. Patent No. 6,304,216 B1 — "Signal detector employing correlation analysis of non-uniform and disjoint sample segments," Steven A. Gronemeyer, filed/priority 1999-03-30, granted 2001-10-16, app. 09/281,741. I did not substitute any similarly-numbered patent (e.g., US 5,043,736 or US 5,043,216-style neighbors) and screened out the many "similar number" hits that USPTO/Google-family pages throw off.

Two evidence limitations you must know before relying on this section:

1. The USPTO/Google full text supplied for this patent is truncated. It ends mid-sentence in the Detailed Description ("…In one implemen"). It therefore contains no "References Cited" table, no claims, and no "Other References" (NPL) list. Everything I say about the cited references comes from the face-page image text of the printed patent and from family records — not from the truncated narrative.
2. I hit the tool-call/search limit while pulling the continuation page of the front-page reference list. The face of the patent reads "(List continued on next page.)" and I could retrieve only part of that continuation. The list below is therefore complete only for the portion I retrieved. I flag every place where the record is a fragment.

1. What "patent citation" means here, and the two distinct lists

2. U.S. Patent Documents on the face of the '216 (as retrieved)

Retrieved verbatim from the printed patent's (56) section; inventor surnames are as printed (OCR-level spelling preserved, not corrected):

Foreign Patent Documents (as retrieved):

"Other References" (NPL): not retrieved for the '216. (By contrast, the continuation US 6,636,178's record does list NPL — D.J.R. van Nee & A.J.R.M. Coenen, "New Fast GPS Code-Acquisition Technique Using FFT," Electronics Letters, vol. 27, no. 2, Jan. 17, 1991, per Justia. Do not attribute that NPL to the '216 without verifying its face page.)

3. The family/PCT record's short citation list

Google Patents' record for the WO counterpart (WO 2000/058746 A1) shows a "Patent Citations (4)" set: US 3,604,911; US 5,018,088; US 5,293,398; US 5,640,429. Three of those four cross-check exactly against the '216 face page I retrieved; US 5,640,429 falls in the un-retrieved continuation portion. Treat this as corroboration, not as the '216's citation table.

4. §102 analysis — does any cited reference anticipate?

Legal frame: The '216 is pre-AIA (filed 1999-03-30; no earlier priority). Critical date = 1999-03-30. The §102 universe from the cited art is §102(a)/(b) printed publications (everything issued before 3/30/1999) and §102(e) for US patents whose US filing date precedes 3/30/1999 (candidates: US 6,075,809, US 6,133,873, and any other post-1999-issued patents on the continuation page).

The independent-claim element set that any anticipating reference must disclose in a single embodiment:

Result of the mapping (based on the references retrieved):

Why the element set matters: the '216's only real point of novelty over every cited reference is the conjunction of (C) and (D) — deriving ΔCP = [(F_PN + D) × ΔT] modulo 1 mS (Eq. 1), with ΔT = (T_n − T_2 + OS₁ − OS₂) × S (Eq. 2), and using it to circularly shift a row of one correlation array before summing it into the cumulative array — even though the actual code phases CP₁ and CP₂ are unknown. That limitation is not present in the correlator art, the GPS tracking art, or (on the retrieved record) the Krasner art. Element (E) alone (adding correlation results) is old; (C)+(D) applied to disjoint, non-uniform segments is the differentiator.

Dependent claims worth specific §102 attention:

• Claims 6, 73 (parameter difference is a code phase difference) — broader than Eq. 1; a reference teaching code-phase alignment of two accumulated correlation results would reach these. None retrieved does.
• Claim 79 ("combining step is implemented with a matched filter") — the correlator art comes closest to this single limitation, but claim 79 still carries all of claim 64.
• Claims 86 and 87 (Eq. 1 and Eq. 2, in haiku form) — anticipation requires a reference that literally discloses the modulo-code-phase-difference formula; none of the cited art is known to. These are the most defensible claims against a §102 attack on formula grounds.
• Claim 12 ("A GPS receiver including either of the signal detectors of claims 1 and 10") — a GPS-receiver reference that also had disjoint-segment accumulation would be needed; none retrieved.

Corroboration from the contested record: In ITC Inv. No. 337-TA-596, the ALJ's final Initial Determination (June 13, 2008) found the asserted claims of the '216 patent were not invalid as anticipated, and the Commission declined review (73 Fed. Reg. 49219, Aug. 20, 2008). Asserted claims were 1, 6, 10–12, 17–19, 64, 65, 69, 70, 72, 73. That is an adjudicated confirmation of the analysis above: the cited art, as presented in that forum, did not §102-anticipate these claims.

Bottom line on §102: On the evidence retrieved, no single cited reference, alone, discloses every element of any independent claim — each lacks at least the parameter-difference-adjustment/cumulative-combination limitation (C)/(D)/(E). The cited art supports, at most, §103 obviousness combinations (digital correlator + GPS receiver + adaptive/segment processing), not anticipation.

5. Forward citations ("Cited By") — context only, not prior art

Google Patents lists five citing documents for the '216: US 2008/0151813 A1 (Adaptix, fast initial acquisition in mobile WiMAX); US 2009/0303067 A1 (CSIRO, position tracking); US 2010/0048203 A1 (data reduction in a mobile network); CN 103270701 A (EnModus, signal detection in noise); US 2022/0179098 A1 (u-blox, GNSS signal processing). These are later documents and cannot be §102 art against the '216. Note the interesting one: the '216's data-reduction/disjoint-segment concept was cited into GNSS receivers (u-blox) two decades later.

6. Corrections and contradictions to flag against the earlier-generated sections

Per the cross-reference instruction, here are explicit flags:

1. CORRECTION — Claim 43 is no longer "unverified." The earlier section recorded claim 43's verbatim text as not retrieved. It is now available (uspto.report / RPX): "A signal detector, comprising: a receiver configured to detect a first segment of a signal of interest and a second segment …; a correlator …; a parameter determiner, coupled to the correlator, configured to determine a parameter difference between the first and the second correlation data; a processor, coupled to the parameter determiner, configured to adjust a selected one …; and a combiner … to generate cumulative correlation data." It is a third apparatus variant (parameter determiner + processor + combiner), with dependents 44–63 — including claim 63 ("successively combines … for a plurality of successive segments"), which confirms the earlier section's inference about the iterative-accumulation focus.

2. CORRECTION — claims 86 and 87 belong to the claim 84 (computer-readable-medium) tree, not the claim 64 tree. The earlier section's phrasing attached "claim 86 … the code-phase-difference equation" and "claim 87 … the offset equation" to claim 64's dependents. Per uspto.report, claim 64's dependents run 65–83, and claims 86 and 87 depend from claims 85 and 86 respectively, under claim 84: 84 (medium) → 85 (medium, code-phase difference for PN/Doppler/code-phase portions) → 86 (ΔCP = [(F_PN + D) × ΔT] modulo T_PN) → 87 (ΔT = (T_n − T_2 + OS₁ − OS₂) × S). The substance of what the earlier section said is right (the equations are claimed); the parent claim is wrong.

3. CONFIRMED — claim 1's verbatim text matches the earlier summary exactly (RPX/ITC wording), including "signal of interest perturbed by noise or pseudo-noise." Claim 10, 13, 17, 22, 64, 84 independent-claim structure also confirms.

4. NEW — the earlier sections said nothing about prior art (the source was truncated and no citation table was reached). This section fills that gap, with the incompleteness caveats above.

7. What to pull next to close the gaps (ranked)

1. Face pages 1–2 of the printed patent — https://patentimages.storage.googleapis.com/fb/0a/39/4a67f3eb6e70f5/US6304216.pdf — to capture the entire continuation-page list (the missing block between US 5,323,164 and US 6,075,809) and the complete "Other References" (NPL) list.
2. USPTO Patent Center / IFW for app. 09/281,741 — the examiner's cited references and any §102/§103 rejections, which is the authoritative citation list and reveals which reference was applied against which claims.
3. File histories of US 6,075,809 and US 6,133,873 — to fix their US filing dates and confirm §102(e) status relative to 3/30/1999 (the only two retrieved references whose §102(e) eligibility is date-sensitive).
4. Full texts of US 5,018,088 (Higbie), US 5,640,429 (USAF), US 3,604,911 (Schmitt), US 5,293,398 (Hamao) — to test the closest correlator art against claims 6/73/79 and the Eq. 1/Eq. 2 claims 86–87.
5. USITC Pub. 4133, §IV ("'216 Patent") — https://www.usitc.gov/intellectual_property/documents/pub4133.pdf — for the tribunal's own prior-art/claim-construction analysis, which is the best available proxy for a rigorous anticipation study.

Explicit uncertainty: I did not obtain the full text of the cited references, the complete face-page citation list, the NPL list, or the examiner's rejections. Descriptions marked "not verified" are exactly that. My §102 conclusions are a documentary-gap-limited judgment grounded in (i) the partial face-page list, (ii) the claim element sets as retrieved, and (iii) the ITC's non-invalidity finding — not an exhaustive novelty search.

I'll search for the prior art references cited on the patent's Google Patents page, since the supplied full text is truncated before the citations section.

Let me retrieve the complete prior art citation list from the patent front page and the related family member.

I have the front-page U.S. reference list. Let me retrieve the remainder of the "References Cited" section (the later U.S. patents and foreign references).

Obviousness Analysis — U.S. Patent No. 6,304,216 B1

Under 35 U.S.C. § 103; priority date March 30, 1999

1. Scope, sourcing, and a flag on the record

Source note / discrepancy flag. The task says to "use the results from the Prior Art section of this page." The Google Patents text supplied to me is truncated before the claims and the citation section, so I retrieved the front-page "(56) References Cited" list from the patent PDF itself (patentimages, US6304216.pdf) and cross-checked it against Justia and RPX. This creates a minor inconsistency with the earlier-generated sections of this analysis: those sections noted the claim text for independent claims 43 and 84 was not retrieved verbatim, yet the ITC publication (pub4133.pdf) and uspto.report have now surfaced the dependent claim language for both (e.g., claim 63 depends from 43; claims 85–87 depend from 84). I flag this because the obviousness analysis below leans on that newly-surfaced language. Claim 43's own full text remains unverified.

Also flagging: the earlier sections state the priority/filing date as 1999-03-30 (Google Patents), while the Unified Patents record renders the priority date as 1999-03-29. I use March 30, 1999 per the patent face, but the one-day difference matters for prior-art cutoff only if a reference published between those dates. None identified here does.

2. Framework

Under § 103 as construed per Graham v. John Deere and KSR Int'l v. Teleflex, I must (a) determine the scope and content of the prior art, (b) ascertain the differences between the prior art and each claim, (c) resolve the level of ordinary skill in the art, and (d) weigh objective indicia. KSR further provides that a combination is obvious where a known technique is applied to a known device to yield predictable results, where there is a design incentive or market pressure, or where the reference combination is "obvious to try" with a finite number of identified, predictable solutions.

Level of ordinary skill (assumed, none recited in the patent): a B.S. in EE plus ~3–5 years in spread-spectrum/GPS receiver design, or equivalent. The art at the 1999 date was mature: matched-filter and serial correlator acquisition, Doppler-bin search, non-coherent (post-detection) integration over multiple dwells, and GPS/cellular co-design were all well-worn.

3. The prior art of record on the patent face

(56) U.S. Patent Documents (as listed on US6304216B1): 3,604,911 (Schmitt); 3,975,628 (Graves); 4,426,712 (Gorski-Popiel); 4,445,118 (Taylor et al.); 4,463,357 (MacDoran); 4,578,678 (Hurd); 4,602,023 (Counselman III); 4,701,934 (Jasper); 4,754,465 (Trimble); 4,785,463 (Janc et al.); 4,809,005 (Counselman III); 4,821,294 (Thomas Jr.); 4,890,233 (Ando); 4,894,662 (Counselman); 4,894,842 (Broekhoven); 4,992,720 (Hata); 4,998,111 (Ma et al.); 5,014,066 (Counselman III); 5,018,088 (Higbie); 5,043,736 (Darnell); 5,048,329 (Ando); 5,108,334 (Eschenbach et al.); 5,148,042 (Nakazoe); 5,153,591 (Clark); 5,179,724 (Lindoff); 5,202,829 (Geier); 5,225,842 (Brown et al.); 5,253,268 (Omura et al.); 5,276,765 (Freeman); 5,293,170 (Lorenz et al.); 5,293,398 (Hamao); 5,297,097 (Etoh); 5,311,195 (Mathis et al.); 5,323,164 (Endo); and further entries continuing to 6,075,809 (Naruse, 12/2000).

(56) Foreign documents: EP 0 511 741 A1 (11/1992) / B1 (11/1997); EP 0 639 901 A2 (2/1995) / A3 (11/1998); and a JP publication fragment ("080065205 A").

Incorporated-by-reference companions (from the "Related Applications" section):

• U.S. 6,044,105 — "Doppler Corrected Spread Spectrum Matched Filter" (Ser. No. 09/145,055, filed Sep. 1, 1998) — expressly incorporated "as though set forth in full."
• U.S. 6,494,145 / 6,917,331 — "Signal Detector Employing Coherent Integration" (Ser. No. 09/281,566, filed even date herewith) — expressly incorporated.

Confidence caveat. I have verified the identity, inventor, and date of these references from the patent face, but I have not obtained and read the full disclosures of most of them (tool limit reached). My characterizations of what each teaches are therefore based on their titles/known subject matter plus general knowledge of the 1999 GPS art, and are marked accordingly. A production § 103 analysis requires reading each reference and pin-citing columns/lines.

4. The claims grouped for analysis

The crux of every independent claim is the same two-step idea: (i) compute a relative parameter difference between two independently-captured correlation results, and (ii) shift/align one and add it to the other, so that data captured at different, non-overlapping times can be accumulated even though the absolute code phases are unknown.

5. Proposed § 103 combinations

Combination 1 — Primary: matched-filter GPS correlator + secondary: multi-dwell accumulation + tertiary: GPS/cellular idle-window constraint

Primary (G1, G2, G4 element — the correlator): Janc et al., US 4,785,463 (assigned to Trimble, on the patent face) and/or Gorski-Popiel, US 4,426,712, and/or Ma et al., US 4,998,111. These are GPS digital-signal-processor references that generate a locally-replicated PN code and correlate it against digitized received samples across a two-dimensional (code-phase × Doppler) hypothesis space, i.e., they teach deriving "correlation data representative of the correlation between [a] segment and a hypothesis" and, for G2, a "combined PN code and code phase hypothesis." Janc in particular is a matched-filter/correlator implementation, which maps to claim 11's "matched filter."

Secondary (the accumulating-combiner element): Lorenz et al., US 5,293,170 (Motorola) and/or Eschenbach et al., US 5,108,334, and the express incorporation by reference of the coherent-integration companion, US 6,494,145 / 6,917,331. Multi-dwell signal search — accumulating correlation magnitudes across several successive dwells to raise effective SNR before declaring detection — was standard textbook technique in 1999 (non-coherent/post-detection integration; M-of-N sequential detection). This teaches "cumulative correlation data" and, in Lorenz/5,293,170 (a cellular-integrated GPS context), the motivation to accumulate across discontinuous windows.

Tertiary (the "separate and distinct periods" element + motivation): The patent itself concedes the constraint at column 1: "in the case of a mobile wireless phone integrated with a GPS receiver, the sampling window should be limited to those periods in which the phone is not transmitting." A POSITA seeking to add GPS to a handset would necessarily be driven to capture only during idle (non-transmit) slots, which are by definition disjoint and of non-uniform length. Any GPS/cellular-integration reference of record (Brown et al., US 5,225,842; Darnell, US 5,043,736; or the carrier-integration references) supplies this.

Motivation to combine: KSR design-incentive. The well-documented 1990s problem was that a handset's GPS dwell is far too short to detect a weak (~−130 dBm) spread-spectrum signal. Integrating across many short dwells is the predictable engineering response. There is no teaching away.

Assessment: This combination would, in my view, support a strong prima facie § 103 challenge to G1, G4, and G5, but is materially weaker on G2/G3(G2 portion)/G6 — see § 6.

Combination 2 — The two incorporated companion applications, read together

Both US 6,044,105 (Doppler-corrected matched filter) and US 6,494,145 (coherent integration) are expressly incorporated by reference into the '216 specification. Because § 103 prior art includes material incorporated by reference into the challenged patent, these two applications can be combined as if their disclosures were physically in the '216 specification.

• '105 supplies the Doppler-correction front end and the matched filter that produces correlation arrays across PN code × Doppler shift × code phase hypotheses (this is verbatim the patent's own FIG. 8 architecture and the FIG. 5 array structure of claims 10/11 and 19).
• '145 supplies the coherent integration concept — accumulating correlation energy across dwells.

Motivation: The two applications share a common assignee (Conexant/SiRF), a common inventor lineage, and are cross-referenced to each other on their faces. A POSITA would combine them by design intent.

Caution — this is a double-edged combination. "'145 coherent integration" is arguably a different solution to the same SNR problem (coherent = phase-preserving within a continuous window), which the '216 patent distinguishes by handling phase-discontinuous, disjoint windows. If the '105+'145 combination were the only art, one could argue it does not teach the key limitation — computing a code-phase difference to realign segments whose absolute code phases are unknown — because coherent integration of contiguous samples never needs that realignment. The '216 patent may thus be patentably distinguishable over its own companion applications. I would not rest the § 103 case on Combination 2 standing alone.

Combination 3 — For the equation claims (G6: claims 86, 87) and the code-phase-difference element (G2/G3)

Claims 86–87 recite only algorithms: how the code-phase difference between two epochs is a function of the nominal PN rate, the Doppler hypothesis, and the elapsed time, and how the elapsed time is built from frame-mark timings and measured offsets OS₁/OS₂, scaled by the local-oscillator time-base error S.

Primary references: Any of Counselman (US 4,894,662 / 5,014,066), MacDoran (US 4,463,357), or Hurd (US 4,578,678) — the JPL/SRI-era literature on deriving pseudo-range/code phase from GPS satellite transmissions — teaches that code phase advances as a function of (chip rate + Doppler) × elapsed time, with periodicity modulo the code epoch. That relationship is a fundamental, well-known identity of GNSS, not an invention. Counselman's precision-ranging work and MacDoran's pseudo-range derivation both turn on it.

Secondary references: Any GPS time-transfer/synchronization reference teaching the use of epoch/frame marks plus a local oscillator to measure elapsed time and to correct for local-oscillator error (e.g., Trimble US 4,754,465; Higbie US 5,018,088 on phase calibration) supplies the ΔT = (T_n − T_2 + OS₁ − OS₂) × S structure.

Motivation to combine: These are arithmetic identities that a POSITA must apply — there is nothing to combine so much as to compute — making G6 highly vulnerable on § 103 if the underlying G1/G2/G5 claims fall. This is a KSR-type situation: "finite number of identified, predictable solutions."

Note on the ITC finding: The ALJ held the asserted claims (1, 6, 10–12, 17–19, 64, 65, 69, 70, 72, 73) not invalid as anticipated. That is an anticipation (§ 102) holding only. It does not immunize the claims from § 103. However, the fact that the ITC found no single reference anticipated these claims — and that claims 43/64/84's equation dependents (86, 87) map to well-known GNSS arithmetic — means a § 103 case would have to attack the specific alignment-of-disjoint-segments concept, not the formulas.

Combination 4 — Computer-readable-medium claims (G7: 84–87)

Dependent claim 84 is a CRM/instruction claim. Once the method of claim 64 is deemed obvious, G7 adds only the generic and conventional implementation choice of encoding the method as software on a medium (or as instructions in the embedded microprocessor that the specification itself describes — "an embedded microprocessor with an external bus"; "generic processor"). Under KSR and Fed. Cir. law on Beauregard-style claims (In re Beauregard), the CRM form adds no patentable weight beyond the underlying functionality. G7 is obvious per se if G5 is obvious.

6. Where the § 103 case is weakest — the genuine non-obviousness kernel

Two limitations resist the combinations above and (I believe) explain the ITC outcome:

1. "Even though the actual code phases are unknown" (spec, col. 8 / FIG. 7 discussion). All of the cited multi-dwell accumulation art works non-coherently — summing magnitudes, so phase alignment is irrelevant. The '216 advance is that it can combine in a way that exploits a computable relative code phase to align coherent correlation structure across disjoint windows. If the prior art's multi-dwell accumulation is non-coherent (as Lorenz/Eschenbach-type references generally are), then a § 102 rejection fails — consistent with the ALJ's finding — and a § 103 rejection must show a reason to compute and apply the phase-difference shift rather than to discard phase. That motivation is not obvious from the face of the cited references, and the specification's own admission that combining disjoint segments "is difficult" (col. 1) cuts toward non-obviousness under the "problem known, solution not obvious" line.

2. Offset measurement via frame marks (the OS₁/OS₂, T, S structure; claims 87). The '216 implements this with a dedicated offset-measurement circuit fed by locally-generated frame marks and corrects ΔT by the oscillator error S. That specific measurement architecture is not shown in any single reference of record retrieved.

My honest bottom line: The § 103 case is strong for claims 1, 13, 43, 64 (the generic "capture-two-segments-and-combine" concepts) and for the CRM/equation dependents (84–87), where the added features are conventional software on arithmetic identities. The case is weak-to-contestable for claims 10–12, 17–21 and their equation dependents if the reference set consists only of non-coherent multi-dwell art. A challenger would need a reference that teaches code-phase-corrected (quasi-coherent) realignment across dwells — which, pre-1999, I cannot identify from the record retrieved, and which the ITC apparently did not find.

7. Objective indicia and the real-world check

• ITC Inv. No. 337-TA-596 — the only matter in which the '216 patent was asserted (SiRF v. Global Locate/Broadcom, instituted March 13, 2007, 72 Fed. Reg. 11378). ALJ final ID June 13, 2008: no infringement and not invalid as anticipated; Commission declined review, 73 Fed. Reg. 49219 (Aug. 20, 2008), final no violation of § 337.
• The asserted claims (1, 6, 10–12, 17–19, 64, 65, 69, 70, 72, 73) thus survived a well-funded, adversarial § 102 attack — but no § 103 challenge was resolved, so the obviousness question is judicially untested.
• Secondary-consideration evidence is essentially absent from the record I retrieved: no evidence of commercial success nexus, licensing, or copying tied to the claims, and no evidence of unexpected results. If anything, the ~2-year commercial life before broad A-GPS deployment and the fact that the claims were never separately asserted after 2008 temper any nexus argument.
• The patent expired March 30, 2019, so § 103 has only retroactive/damages significance today.

8. Confidence and limitations

Explicit non-fabrications: I have not asserted column/line pin-cites or verbatim quotes from the cited references because I did not retrieve their full texts. The "Non-Patent Citations" (NPL) section of the '216 face was not retrieved and is likely to include Kaplan's Understanding GPS, the GPS Interface Control Document, and ION proceedings — I decline to list specifics I cannot verify. Verbatim claim text for independent claim 43 remains unretrieved.

One-line answer: On the record retrieved, the '216 claims are obvious under § 103 over a combination of a matched-filter GPS correlator (e.g., Janc '463 / Ma '111), multi-dwell correlation accumulation (e.g., Lorenz '170 / Eschenbach '108,334 / the incorporated '145), and the GPS-handset idle-window constraint — with KSR design-incentive supplying the motivation — for the broad "capture-two-disjoint-segments-and-combine" claims (1, 13, 43, 64, 84–87); but the specific code-phase-difference realignment of disjoint, unknown-phase segments (claims 10–12, 17–21 and dependent 86–87) is not clearly shown obvious by the non-coherent multi-dwell art of record, which is consistent with the ITC's 2008 finding of no invalidity.

Sources: Google Patents, https://patents.google.com/patent/[US6304216](/patent/US6304216)/en; patent PDF, https://patentimages.storage.googleapis.com/fb/0a/39/4a67f3eb6e70f5/US6304216.pdf; USITC Pub. 4133, https://www.usitc.gov/intellectual_property/documents/pub4133.pdf; RPX Insight, https://insight.rpxcorp.com/patent/[US6304216B1](/patent/US6304216B1); uspto.report, https://uspto.report/patent/grant/6304216; Federal Register 72 FR 11378 and 73 FR 49219.