US 9473336

I'll search for the specific patent number and any associated litigation activity.

Let me check for the specific litigation history (ITC and district court) and any Federal Circuit appellate activity.

US Patent 9,473,336 B2 — Analyst Summary

Scope note on sources and the CAFC check: I searched for the specific number "9473336" and did not auto-correct or normalize it. I could not query USPTO PatentCenter or PACER directly (no authenticated access), so bibliographic data below is taken from the Google Patents and Justia mirrors of the official record, plus primary litigation documents. I found no active or docketed 2026 Federal Circuit appeal involving US 9,473,336. I will not invent one — see the caveat at the end.

Bibliographic data

Classifications: H04B 1/005; H04B 1/06; H04B 1/10; H04B 7/26; H04L 27/06 (receiver front ends / multi-band receivers / noise suppression).
Source: https://patents.google.com/patent/US9473336/en

Abstract (verbatim)

"A radio frequency (RF) front end having multiple low noise amplifiers modules is disclosed. In an exemplary embodiment, an apparatus includes at least one first stage amplifier configured to amplify received carrier signals to generate at least one first stage carrier group. Each first stage carrier group includes a respective portion of the carrier signals. The apparatus also includes second stage amplifiers configured to amplify the first stage carrier groups. Each second stage amplifier configured to amplify a respective first stage carrier group to generate two second stage output signals that may be output to different demodulation stages where each demodulation stage demodulates a selected carrier signal."

Plain-language overview of the independent claims

Claim 1 (two-stage amplification split across demodulators).
Imagine a receiver that hears several cellular "carriers" at once. A first-stage amplifier block amplifies the incoming carriers and sorts them into one or more bundles ("first stage carrier groups"), each bundle holding a subset of the carriers. A set of second-stage amplifiers then re-amplifies each bundle. The point of novelty is the output wiring: a single second-stage amplifier puts out two signals, and those two outputs go to two different demodulators (rather than feeding one demodulator or being combined). This lets one amplified bundle serve two separate demodulation paths.

Claim 4 (internal architecture of the first-stage amplifier).
This claim is narrower and defines what the first stage is built from: two (or more) LNAs — a first LNA that produces a first and second amplified output (each a portion of the carriers), and a second LNA that produces a third and fourth amplified output. A routing module (implemented in the spec as a bank of controlled switches, FIGS. 4–5) can steer any of those amplified outputs to either of two output ports. Second-stage amplifiers then amplify the resulting carrier groups. In effect: same two-stage concept, but with an explicit "reconfigurable switch fabric" between LNA outputs and the two output ports, so a given carrier or band group can be routed to either port on the fly. This is the claim Qualcomm later asserted in litigation.

Claim 16 (means-plus-function version).
This restates claim 1 in functional language: "means for amplifying received carrier signals" to produce first-stage carrier groups, plus "means for amplifying" those groups to produce second-stage outputs, with a first output of the second means coupled to a "means for demodulating" a first output signal and a second output coupled to another "means for demodulating" a second output signal. Under § 112(f), the corresponding structures are the first-stage amplification (302), second-stage amplification (304), and the demodulator/mixer stages described in the specification.

Notable dependent claims worth flagging (not independent, but they carry much of the commercial weight):

• Claim 3: the received carriers include primary and diversity carriers.
• Claim 6–10: second stage = "interface amplifiers" (iLNAs); matching circuits at the input; transformer modules converting to differential signals; double-balanced mixers demodulating to baseband using a local oscillator.
• Claims 11–14: first stage on one or more ICs, second stage on a separate single IC, both on a PCB, connected by PCB signal traces and interface pins.
• Claim 15: the "carrier-centric" (CA-centric) pin arrangement — pins for a second-stage amplifier's output are placed proximate to the corresponding demodulator, and traces to different demodulators cross orthogonally, to reduce coupling.

Litigation context (grounded in primary documents)

The Google Patents record lists this family as having litigation, including ITC Investigation 337-TA-1093 and California Southern District Court case 3:17-cv-02398 (Qualcomm v. Apple, filed Nov. 2017). The '336 patent was one of the patents Qualcomm asserted against Apple devices that did not use a Qualcomm-brand baseband modem (e.g., iPhone 7/8/X with Intel modems).

Two points from the ITC Initial Determination are directly relevant to claim scope:

• The asserted claim of the '336 patent was claim 4 ("first stage carrier group" and "first portion / second portion of the carrier signals" were among the disputed claim terms in the Joint Claim Construction Chart, Docket No. 337-TA-1093).
• The ALJ found Apple did not infringe claim 4 of the '336 patent, while finding claim 4 valid. The ALJ also found that none of Qualcomm's domestic industry products satisfied the technical prong of the domestic industry requirement for the '336 patent (though the economic prong was satisfied).
  Sources: https://patents.google.com/patent/US9473336/en ; https://www.docketalarm.com/cases/[PTAB](/ptab)/IPR2019-00186/.../Exhibit-1038-29-Joint_Claim_Construction_Chart... ; 337-TA-1093 Initial Determination summary.

The Nov. 30, 2017 SDCA complaint (3:17-cv-02398) asserted the '336 patent alongside US 9,154,356; US 8,063,674; US 7,693,002; and US 9,552,633. That case was later dismissed with prejudice as part of the 2019 Qualcomm–Apple global settlement/license, which is what gave rise to the Federal Circuit standing decisions in Apple Inc. v. Qualcomm (2021) — but those appeals concerned other Qualcomm patents / IPRs, not the '336 patent.

Explicit uncertainty / limits

1. No CAFC 2026 docket for US 9,473,336 found. Searches for "9473336" in combination with Federal Circuit/2026 returned only unrelated Qualcomm appeals (e.g., Federal Express Corp. v. Qualcomm Inc., Nos. 2024-1235/1236/1237, decided Apr. 29, 2026; ParkerVision v. Qualcomm; Centripetal v. Cisco). Those concern different patents (e.g., 7,623,033; 8,766,797; 9,182,231). I cannot confirm any 2026 appellate activity on the '336 patent, and I am stating that as an absence of evidence rather than a verified negative — PACER/Westlaw docket checks would be needed to be conclusive.
2. I could not independently verify from USPTO PatentCenter's own fee/status records (only from the Google/Justia mirrors), so the "Active" legal status and the 2035-03-27 anticipated expiration are as reported by those mirrors, not a first-hand USPTO status read.
3. The ITC record I reviewed is at the ALJ Initial Determination stage; I did not verify the Commission's final determination on the '336 patent, so the ultimate outcome (including any Commission reversal of the non-infringement finding) should be confirmed against the final Commission opinion before being relied on.
4. I found no reissue, reexamination, or IPR proceeding invalidating the '336 patent in the sources checked; the cited-by/citation lists show only routine IPR activity on other Qualcomm patents.

I'll research litigation involving US patent 9473336, starting with the leads already present in the patent record (the ITC investigation and the Southern District of California case).

Good — the patent record and news reports both point to Qualcomm v. Apple. Let me pin down the exact case numbers, dates, and outcomes.

Let me verify the outcome of the ITC investigation and check for any PTAB challenges or additional suits involving this specific patent.

Litigation Involving U.S. Patent No. 9,473,336

I searched patent-litigation resources (including the Unified Patents litigation portal caselist, the ITC's Federal Register notices/EDIS-derived reporting, and CourtListener/ PACER-derived dockets) for the literal identifier 9,473,336 / 9473336. Two enforcement proceedings against Apple are documented; both were part of the Qualcomm–Apple dispute and both named the '336 patent. No other litigation asserting this specific patent was found.

1. Qualcomm Incorporated v. Apple Incorporated — District Court

Notable procedural milestones: claim construction hearing Nov. 14, 2018; Order Construing Claims Dec. 13, 2018 (ECF 163); Apple's invalidity contentions and Qualcomm's motions for leave to amend infringement contentions were litigated through March 2019. A jury trial had been set for Oct. 21, 2019, but never occurred.

2. In the Matter of Certain Mobile Electronic Devices and Radio Frequency and Processing Components Thereof (II) — ITC Section 337

The investigation had narrowed during the proceeding (due to withdrawal of asserted claims) to claims 1 and 17 of the '356 patent; claim 4 of the '336 patent; and claims 1 and 8 of the '674 patent.

Cross-references and practical notes

• The Google Patents "Family has litigation" flag for US 9,473,336B2 lists exactly these two proceedings — the ITC case (portal.unifiedpatents.com/litigation/International%20Trade%20Commission/case/337-TA-1093) and the S.D. Cal. case (portal.unifiedpatents.com/litigation/California%20Southern%20District%20Court/case/3%3A17-cv-02398) — which is consistent with my findings.
• Both proceedings arose from the same November 2017 infringement pleading package: Qualcomm filed the district court complaint (Nov. 29, 2017) and the ITC complaint (Nov. 30, 2017) asserting the same five patents, including the '336 patent.
• Sister S.D. Cal. actions between the same parties (e.g., 3:17-cv-01375, 3:17-cv-02402, 3:17-cv-02403) appear in related-case notices, but I found no evidence that the '336 patent was asserted in those actions. In the ITC, the parallel investigation was 337-TA-1065 (the "Mobile Electronic Devices (I)" case), which involved a different patent set.
• IPR/PTAB: Apple filed numerous IPRs against Qualcomm patents in this campaign, but I was unable to confirm a petition specifically directed to U.S. 9,473,336 within the available results. I therefore do not assert that any IPR on the '336 patent exists. (Note that a PTAB proceeding would in any event not be "litigation" in the district-court/ITC sense.)
• Appellate: No CAFC appeal specific to the '336 patent was identified — consistent with the district case being dismissed with prejudice and the ITC investigation terminating on settlement before a Commission final determination.

Caveats

• The complaint-filing date for the district court case is Nov. 29, 2017 (per the docket/complaint PageID), while contemporaneous press accounts describe the suit as filed "Wednesday" Nov. 29, 2017 — reported in some outlets as Nov. 30 due to time-zone/syndication differences.
• The ITC termination notice (84 FR 44330) cites "claim 1 of the '674 patent" as the violation basis, while ALJ McNamara's remedy recommendation refers to "claim 8 of U.S. Patent No. 8,063,674." That discrepancy concerns the '674 patent, not the '336 patent, and does not affect the '336 outcome.
• Docket details were compiled from public aggregators (CourtListener/PACER-derived, Docket Alarm, PlainSite) and ITC Federal Register notices; the underlying filings remain the authoritative source.

Bottom line: The only known litigation involving U.S. Patent No. 9,473,336 is (1) Qualcomm Inc. v. Apple Inc., No. 3:17-cv-02398-DMS-MDD (S.D. Cal., filed Nov. 29, 2017; dismissed with prejudice Apr. 23, 2019), and (2) ITC Inv. No. 337-TA-1093 (Certain Mobile Electronic Devices and Radio Frequency and Processing Components Thereof (II), complaint filed Nov. 30, 2017; no violation found as to the '336 patent; investigation terminated on settlement). Both plaintiff: Qualcomm Incorporated; both defendant/respondent: Apple Inc.

PTAB / AIA Trial Proceedings — US 9,473,336 B2

Patent: US 9,473,336 B2 — "Radio frequency (RF) front end having multiple low noise amplifier modules"
Patent owner (current assignee): Qualcomm Incorporated
Application: 14/671,939 · filed 2015-03-27 · granted 2016-10-18 · priority (provisional 61/994,158) 2014-05-16 · AIA patent (effective filing date after 2013-03-16)
Anticipated expiration: 2035-03-27

Proceedings overview

There are zero AIA trial proceedings on file for US 9,473,336 — 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 institution denials — so all 20 claims (1–20) are PTAB-untested and the patent stands intact, which means a defendant gets neither the windfall of a cancellation nor the roadmap of a tested claim construction: any IPR-based defense must be built from scratch on a patent that the Patent Owner already elected to drop from its own infringement case.

Sourcing and confidence:

• The structured "PTAB proceedings on file" block (USPTO Open Data Portal) returns no AIA trial proceedings for this patent as of the most recent ingest. That is the canonical list and it is the basis for the count above.
• My independent web checks did not surface any IPR, PGR, or CBM naming U.S. Patent No. 9,473,336 / application 14/671,939. I also found no PTAB "Notices of Filing Date Accorded" or FWDs referencing it in the Apple–Qualcomm IPR wave.
• Caveat (stated plainly because it matters): ODP ingest lags, and absence from my search results is evidence of absence, not proof of it. The authoritative confirmation is a manual PTAB E2E / Patent Trial and Appeal Board End-to-End docket search on application number 14/671,939. Treat the "zero" as the strong working assumption and verify before you rely on it in a brief.

Because there are no proceedings to report on this patent, the required per-proceeding template is not populated below. Instead, the sections that follow explain the nearby proceedings that a defendant will inevitably be shown by opposing counsel, and why none of them touch this patent.

No proceeding on US 9,473,336 — nothing to report

• Type: N/A
• Filed: N/A
• Status: No AIA trial proceeding on file (USPTO ODP; corroborated by web search, not conclusively confirmed)
• Judge panel: N/A
• Petition grounds: N/A — no petition has been filed against this patent
• Institution decision: N/A
• Final Written Decision: N/A — no claim of US 9,473,336 has ever been construed, canceled, or sustained in an AIA trial
• Settlement / termination: N/A
• Appeal: N/A — no FWD exists to appeal
• Defensive value: There is no PTAB record to leverage. You cannot point to a cancellation, and you cannot borrow the Board's claim constructions. Conversely, nobody has spent their § 315(e)(2) estoppel on this patent, so the field is wide open for a first-filer IPR.

Related proceedings that do not involve US 9,473,336 (context only — do not cite as this patent's history)

Opposing counsel may gesture at the Qualcomm–Apple PTAB war. These are real proceedings, but they concern different patents. Listing them so you are not blindsided:

• IPR2018-01315 and IPR2018-01316 — Apple Inc. v. Qualcomm Incorporated (U.S. Patent No. 8,063,674, not the '336 patent). Filed 2018-06-29; instituted 2019-01-18 (per RPX). FWD 2020-01-03 (Board Op., Apple Inc. v. Qualcomm Inc., IPR2018-01315, -01316, 2020 Pat. App. LEXIS 5250) found the challenged claims unpatentable over applicant-admitted prior art (AAPA) in view of Majcherczak. The Federal Circuit vacated and remanded on 2022-02-01 on § 311(b) grounds (Qualcomm Inc. v. Apple Inc., 124 F.4th 1367 line of cases), and in April 2025 issued a second precedential decision again reversing the Board and reinstating the patent (Qualcomm Inc. v. Apple Inc., 134 F.4th 1355 (Fed. Cir. 2025)). Docket: CAFC No. 23-1208. Opinion: https://www.courtlistener.com/opinion/[10382636](/patent/10382636)/qualcomm-incorporated-v-apple-inc/
• Apple's broader IPR campaign. Qualcomm's SEC disclosures state Apple filed roughly 50 IPRs against Qualcomm patents in 2017–2018, with the Board instituting on 13. Docket Alarm's Apple-v.-Qualcomm PTAB list shows petitions on Qualcomm patents such as 8,683,362; 7,844,037; 9,203,940; 8,497,928; 8,768,865; and 9,608,675 — none of which is 9,473,336. See the Docket Alarm PTAB search at https://www.docketalarm.com/search/PTAB/dockets/?f=party_obj.name_exact-Apple
• Litigation (not PTAB) touching the '336 patent:
  • ITC Inv. No. 337-TA-1093 — Qualcomm's 2017-11-30 complaint against Apple asserted the '336 patent along with the '356 and '674 patents; the ITC staff recommended a finding of no infringement on the three patents, and the ALJ/Commission ultimately found a violation on only one patent. https://portal.unifiedpatents.com/litigation/International%20Trade%20Commission/case/337-TA-1093
  • Qualcomm Inc. v. Apple Inc., No. 3:17-cv-02398-DMS-MDD (S.D. Cal.) — Qualcomm's 2017-11-29 complaint asserted 46 claims of five patents, including the '336 patent. In its 2018 motion to assert additional claims, Qualcomm stated it "agreed not to assert any claims of the '336 Patent" and dropped it from the case. See https://storage.courtlistener.com/recap/gov.uscourts.casd.[522828](/patent/522828).108.2.pdf
  • The worldwide Apple–Qualcomm litigation settled in April 2019 (six-year global license), which ended the district court case with prejudice and mooted the parties' PTAB appeals (Apple's appeals were dismissed for lack of Article III standing: Apple Inc. v. Qualcomm Inc., Nos. 2020-1683 et al., decided 2021-11-10, https://caselaw.findlaw.com/court/us-federal-circuit/[2151412](/patent/2151412).html).

Why this matters: the party with the greatest incentive to IPR the '336 patent (Apple) never appears to have filed against it, and Qualcomm withdrew the patent from its own district court case. That combination is the most likely explanation for the empty PTAB docket on this patent.

Strategic summary

Claim status. All 20 claims of US 9,473,336 are UNTESTED before the PTAB. Independent claims 1 (apparatus with first-stage amplifier + second-stage amplifiers, first output to a first demodulator, second output to a second demodulator) and 16 (means-plus-function counterpart) plus dependents 2–15 and 17–20 have never been held canceled or patentable in an AIA trial. There is no narrowed claim set to work from and no FWD quote to borrow. If a demand letter cites claims 1–20, all of them are presumptively valid and enforceable — you cannot say "the claims are dead," because they are not.

Estoppel landscape. Because no IPR/PGR/CBM was ever instituted on this patent, no petitioner (or privy) is subject to § 315(e)(2) estoppel with respect to it. The prior-art field is completely open. Practical constraints to map before filing:

• § 315(b) one-year bar. Any IPR petition is time-barred if filed more than one year after the petitioner (or a real party in interest/privy) was served with a complaint alleging infringement of the '336 patent. Parties served during the 2017–2019 Apple litigation window may be barred; a newly asserted defendant is not — and the clock starts on service.
• Available fora. The patent's effective filing date is 2014-05-16, so it is an AIA patent. PGR is unavailable (window closed ~2017-07-18, nine months after the 2016-10-18 grant), CBM is unavailable (not a financial-services business method, and the program sunset on 2020-09-16). IPR is the only AIA vehicle.
• Ground scope. Under Qualcomm Inc. v. Apple Inc., 134 F.4th 1355 (Fed. Cir. 2025) (precedential), which restated and reinforced the 2022 decision, AAPA cannot be "the basis" of an IPR ground under § 311(b) — grounds must rest on prior-art patents or printed publications. AAPA, expert testimony, and general-knowledge evidence may still inform the level of ordinary skill and motivation to combine, but build your grounds on actual prior-art documents. (This is also now embodied in the USPTO's July/August 2025 guidance for petitions filed on or after 2025-09-01: https://www.ptablaw.com/2025/08/07/july-uspto-guidance-sets-stricter-standards-for-evidence-in-ipr-petitions/)

Pattern signals. No petitioner has filed more than one IPR against this patent because no petitioner has filed one at all. The patent owner (Qualcomm) is a sophisticated, well-resourced litigant that litigates aggressively and has repeatedly appealed adverse Board decisions to the Federal Circuit — most notably winning two precedential reversals in the '674-patent appeals. There is no defensive aggregator (e.g., Unified Patents) IPR on file for this patent either. The realistic signal is not "hardened through IPR" but "never tested" — the patent was asserted in the ITC and in the S.D. Cal. action and then quietly withdrawn from the district court case as part of Qualcomm's case narrowing, before the April 2019 global settlement ended the dispute.

Recommended next steps

1. Verify the negative directly. Run a manual PTAB E2E search (https://ptab.uspto.gov) and a PatentCenter "Patent Trial" lookup on application 14/671,939 to confirm there is no petition, no notice of filing date, and no FWD. The ODP block is the canonical source, but ODP ingest lags; do not represent "no PTAB proceedings" to a court or an adversary without this confirmation.
2. If you are a defendant now being asserted against: note that there is no FWD to quote and no canceled claim to point to. Your invalidity story has to stand on its own. The most productive angles given the specification are (a) § 103 combinations of carrier-aggregation two-stage receiver/LNA-bank art (Qualcomm's own related disclosures such as US 2013/0231080 and US 2013/0231064 are cited references on the face of the patent and are fair game as printed publications), and (b) § 112 written-description/enablement and means-plus-function definiteness attacks on independent claim 16 ("means for amplifying," "means for routing," "means for demodulating"), which has never been construed.
3. Statutory clock if you file an IPR. Institution decision is due within six months of the petition filing date (§ 314(b)); the Final Written Decision is due within twelve months of institution (§ 316(a)(11)), extendable up to six months for good cause. Budget the § 315(b) one-year service bar first.
4. Check your own service history. If you (or a privy) were served with an '336 infringement complaint in the 2017–2019 window, the § 315(b) bar may already have run. That is the single most consequential fact to establish before deciding whether an IPR is even available to you.
5. Commercial/contract posture. The patent expires 2035-03-27, so there is no "wait it out" strategy. If you take a license, insist that the agreement state whether the validity of this specific patent affects your payment obligations — the Federal Circuit's Apple–Qualcomm standing decisions (Apple I, 992 F.3d 1378; the 2021 consolidated dismissals) make clear that a broad portfolio license can strip a licensee of standing to challenge validity later.

Bottom line: no PTAB proceedings, no invalidated claims, no estoppel, no roadmap — US 9,473,336 is a live, untested Qualcomm patent, and the absence of IPRs against it (despite two years of ITC and district court assertion) is itself the most important defensive fact on the docket.

I'll verify the assignment record, litigation context, and correspondent details before reconstructing the chain.

Let me attempt to pull the specific assignment record and check for any post-issuance transfers.

Ownership Analysis — US Patent 9,473,336 B2

"Radio frequency (RF) front end having multiple low noise amplifier modules"
App. No. 14/671,939 · Filed 2015‑03‑27 · Issued 2016‑10‑18 · Priority 2014‑05‑16 (Prov. 61/994,158)

Bottom line up front: This patent has exactly one recorded assignment — the routine inventor-to-employer assignment to Qualcomm. There is no NPE chain, no shell LLC, no cascading transfer, and no post-issuance conveyance of any kind. The asserted-plaintiff in the litigation record is Qualcomm itself. The NPE signal scan comes back essentially empty.

Inventors

Basis for employer: All six are listed as assignors on the single recorded assignment (Reel 035922/0033) conveying their entire right, title and interest to Qualcomm Incorporated, with an executed/recorded window of late May – mid June 2015, i.e. ~2 months after the 2015‑03‑27 non-provisional filing. That is the standard employee-invention assignment pattern at a large corporate filer.

Unusual-pattern check: No anomalous signal. There is no evidence in the sources reviewed that all (or any) inventors departed Qualcomm within 12 months of filing. The invention is a core San Diego RF/analog front-end design — the assignee is the R&D shop of record, and the specification is written in Qualcomm's house style with a domestic-industry/PCB-mounting narrative. The "mass inventor departure preceding a fire-sale" pattern does not appear here.

Original assignee

Qualcomm Incorporated (Delaware corporation; 5775 Morehouse Drive, San Diego, CA 92121‑1714) — the entity named on the face of the issued patent and still the record owner.

• Primary line of business: fabless semiconductor design and wireless R&D (QCT — Snapdragon modems, RF front-end modules), plus the QTL patent-licensing business. Publicly traded (NASDAQ: QCOM); SEC filings regularly describe its patent portfolio as "tens of thousands of patents."
• Did they ship a product embodying the claims? Qualcomm unquestionably ships carrier-aggregation RF front-end / modem products in this technology space, and it litigated the patent as a domestic-industry complainant. But note the nuanced record: in ITC Inv. No. 337‑TA‑1093 the ALJ found that "none of Qualcomm's domestic industry products satisfy the technical prong of the domestic industry requirement for the '336 patent" (Initial Determination, 2019‑03‑26). Qualcomm satisfied only the economic prong for '336. So there is no adjudicated product embodying this patent's asserted claims — the assertion rested on economic-prong licensing investment plus the patent-in-suit.
• Current status: Operating. No bankruptcy, no dissolution, no acquisition. The patent itself is listed Active on Google Patents with an anticipated expiration of 2035‑03‑27 (i.e., no maintenance-fee lapse and, critically, no recordal of transfer).

Assignment timeline

The Assignment Center / Google Patents legal-events record for US 9,473,336 contains one and only one recorded assignment. There is no second link — no re-recorded change of name, no security interest, no license recordal, no release, no transfer to any third party.

• 2015‑05‑29 → 2015‑06‑08 (executed; per the assignment's own signing-date range) / recorded 2015‑06‑16 — Reel 035922/0033
  • Conveyance: Assignment — "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)"
  • Assignor: PAN, Dongling; TASIC, Aleksandar Miodrag; RANGARAJAN, Rajagopalan; LEUNG, Lai Kan; NARATHONG, Chiewcharn; TANG, Yiwu (all six joint inventors)
  • Assignee: Qualcomm Incorporated (California)
  • Correspondent: Not retrievable for this specific reel from the sources available to me. I want to be precise here rather than guess: I could not pull the cover sheet for Reel 035922/0033. What I can confirm is that contemporaneous Qualcomm assignment cover sheets handled through the same channel list Michael Drapkin, P.O. Box 11583, Salt Lake City, UT 84147, as correspondent / submitter (see e.g. Qualcomm Reel 037883/0116, recorded 2016‑02‑23, and later Reel 054979/0905). That is Qualcomm's assignment-recordal convention, not an NPE tell — it recurs across Qualcomm's ordinary employee-assignment filings. I am flagging it as likely but unverified for this reel.
  • Context: Internal / routine inventor-to-employer assignment. Not a fire-sale, not a reorg, not a securitization, not a transfer-to-asserter. This is the initial vesting step, executed within ~9 weeks of filing.

No later link exists. The Google Patents legal-events table shows only the 2015‑06‑16 assignment and the 2016‑10‑18 grant; the Assignment Center surface reflects the same single record. Per the search instructions: when the only record is the original assignment, the operative finding is that the original assignee still owns the patent — which is exactly the case here.

Timeline diagram

timeline
    title Ownership of US 9473336
    2014 : Provisional filed 16 May
    2015 : Utility application filed 27 Mar
         : Inventors assign to Qualcomm 16 Jun
    2016 : Patent US 9473336 issues 18 Oct
    2017 : Qualcomm sues Apple in SD Cal
         : Qualcomm files ITC complaint
    2018 : ITC investigation instituted
    2019 : ITC finds claim 4 valid but not infringed
         : Investigation terminated on settlement

NPE / troll-pattern signals

Cross-reference note on the litigation (relevant context, not an NPE signal):

• ITC 337‑TA‑1093 (Certain Mobile Electronic Devices and RF and Processing Components Thereof II) — instituted 2018‑01‑18 on Qualcomm's complaint; claim 4 of the '336 patent was the asserted claim. ALJ McNamara's Initial Determination (2019‑03‑26) found Apple did not infringe claim 4 and that claim 4 is valid, and that Qualcomm's DI products did not meet the technical prong for '336. The Commission terminated the investigation on settlement (notice dated 2019‑08‑20).
• Qualcomm Inc. v. Apple Inc., 3:17‑cv‑02398‑DMS‑MDD (S.D. Cal.) — filed 2017‑11‑29 asserting the '336 patent among five; Qualcomm later agreed to drop the '336 claims to narrow the case (see Doc. 109). Case terminated 2019‑04‑23.
• Sources: ITC notice of termination · ALJ ID via Docket Alarm · Qualcomm S.D. Cal. brief

Verdict

Operating-company assertion.

The chain of title is a single link — inventors → Qualcomm Incorporated, recorded 2015‑06‑16 at Reel 035922/0033 — and it has never moved since. Qualcomm is the original assignee, remains the record owner per the Assignment Center, is a publicly traded operating semiconductor company, and asserted the '336 patent in its own name against a direct competitor, Apple, in both the ITC (337‑TA‑1093) and the Southern District of California (3:17‑cv‑02398). Every one of the eight NPE-pattern signals is not present; there is no shell entity, no repeat asserter-lawyer, no cascade, and no defensive-aggregator termination.

One honest caveat for the file: the ITC found claim 4 valid but not infringed, and found no Qualcomm DI product satisfied the technical prong for '336 — so the "ships a product embodying the claims" prong is weaker for this specific patent than the ownership pattern would suggest. That does not change the verdict (Qualcomm is plainly an operating-company plaintiff, not an NPE), but it is the kind of detail that matters if the patent is ever re-asserted.

Verify at: USPTO Assignment Center · Patent Assignment Search — search patent number 9473336; the sole hit should be Reel 035922/0033.

Limitations to note (per my operating rules): I could not retrieve the actual cover-sheet PDF for Reel 035922/0033, so the correspondent field for that specific reel is unverified — I have flagged the Drapkin / Salt Lake City name only as a corroborated Qualcomm-wide convention from adjacent reels, not as a confirmed entry on this one. Everything else above (single assignment, assignee, dates, reel/frame, litigation posture) is grounded in the sources cited.

I'll research the specific patent and its cited prior art references. Let me start with targeted searches.

Let me research the remaining cited references for more detail.

Let me research the remaining cited references.

Prior-Art Analysis — US 9,473,336 B2

0. Scope, method, and caveats

• Patent under analysis: US 9,473,336 B2, "Radio frequency (RF) front end having multiple low noise amplifier modules." Application US 14/671,939, filed 2015‑03‑27; priority date 2014‑05‑16 (provisional US 61/994,158); granted 2016‑10‑18; assignee Qualcomm Incorporated. Source: https://patents.google.com/patent/US9473336/en
• Databases checked: I searched Google Patents' mirror of USPTO bibliographic/front‑page data and the EPO/IST records surfaced through it. Google Patents reproduces the USPTO "References Cited" front page for this patent, so the citation list below is the authoritative examiner‑cited set (10 US patent documents) rather than a re‑derived set. The single non‑patent citation is the PCT ISR/WO (PCT/US2015/029966, ISA/EPO, 20 Jul 2015).
• Critical date for § 102: Because the application was filed after 2013‑03‑16 but claims a 2014‑05‑16 priority date, prior art must predate 2014‑05‑16 (with the usual AIA § 102(a)(1)/(a)(2) carve‑outs). Note this makes two of the cited references extraordinarily close to the line (both published 2014‑05‑15 — one day before the priority date), discussed below.
• Honest limitation: I retrieved the full front matter, abstract, and (for several references) claims and specification text for US 20130230080, US 20130231064, US 20140134960, US 8532588, US 9300420, US 9774485 and US 9603187. For US 4054842, US 20060068746, US 7349488, US 20090088124, US 20140134959 and US 9136811 I retrieved bibliographic data and titles but not full claim/specification text; my anticipation remarks for those are therefore provisional and flagged as such.

Important framing: The examiner cited these ten documents as the references of record. The overwhelming majority are background/§ 103 obviousness art, not § 102 anticipatory art. As explained in § 3, no single cited reference appears to disclose every element of independent claims 1, 4, or 16; the realistic challenge to the '336 patent is obviousness (single reference + POSITA, or combinations), not strict anticipation.

Constructed claims of the '336 patent (for reference in the analysis):

• Claim 1 — at least one first‑stage amplifier amplifying received carrier signals into a first‑stage carrier group; second‑stage amplifiers each amplifying a respective first‑stage carrier group; first output of a first second‑stage amplifier → input of a first demodulator, second output → input of a second demodulator.
• Claim 4 — first‑stage amplifier having a first LNA generating first and second amplified outputs (different carrier portions), a second LNA generating third and fourth amplified outputs, and a routing module routing selected outputs to first and second output ports; plus second‑stage amplifiers.
• Claims 5–15 — single‑ended first‑stage groups; interface (i)LNA configurable as first/second selected carriers; matching circuits; transformer modules producing differential signals; mixers/LNAs using an LO and double‑balanced mixers; first stage on one or more ICs and second stage on a separate single IC; both on a PCB; PCB traces; IC interface pins; carrier‑centric pin arrangement with orthogonal trace crossing (claim 15).
• Claims 16–20 — means‑plus‑function counterparts of claims 1, 4 and the differential/mixing features.

1. The ten examiner‑cited references (full citations, dates, description, § 102 exposure)

R1. US 4,054,842 A

• Citation: Rockwell International Corporation, "Channel gain imbalance compensation for FSK demodulator." Priority 1977‑01‑19; published 1977‑10‑18. Google Patents: https://patents.google.com/patent/[US4054842A](/patent/US4054842A)/en
• Description: A demodulator for frequency‑shift‑keyed (FSK) signals that compensates for channel gain imbalance between I and Q channels.
• § 102 exposure: Essentially none for the '336 independent claims. It dates from 1977 and concerns FSK demodulation gain balancing, not multi‑band/multi‑carrier RF front ends. It cannot anticipate any claim (no carrier aggregation, no first/second‑stage LNA, no routing module, no carrier‑centric pins). Its only role is as generic evidence that demodulator/mixer signal‑conditioning techniques were long known. Not anticipatory of any of claims 1–20.

R2. US 2006/0068746 A1

• Citation: Nokia Corporation, "Direct conversion receiver radio frequency integrated circuit." Filed/priority 2004‑09‑30; published 2006‑03‑30. Google Patents: https://patents.google.com/patent/US20060068746A1/en
• Description: A direct‑conversion (zero‑IF) receiver RFIC architecture — LNA, downconversion mixers and baseband chain arranged for direct conversion.
• § 102 exposure: Low. It predates the priority date and is relevant background to the "LNA → mixer → baseband" element and to single‑ended‑to‑differential handling, but it does not disclose (i) a two‑stage amplified front end with distributed first‑stage carrier groups, (ii) a routing module steering first/second portions to two output ports, or (iii) one second‑stage amplifier output feeding two different demodulators. Does not anticipate claims 1, 4, or 16. (Full text/claims not retrieved — provisional.)

R3. US 7,349,488 B1

• Citation: Xilinx, Inc., "Frequency shift keying signaling for integrated circuits." Priority 2003‑04‑10; published 2008‑03‑25. Google Patents: https://patents.google.com/patent/[US7349488B1](/patent/US7349488B1)/en
• Description: FSK signaling schemes for on‑chip/IC communication, including demodulation of FSK signals.
• § 102 exposure: Essentially none for the '336 claims. Like R1, this is FSK signal‑handling art with no bearing on carrier‑aggregation RF front ends, multi‑LNA routing, or second‑stage distribution. Does not anticipate any claim. (Full text not retrieved — provisional.)

R4. US 2009/0088124 A1

• Citation: Nanoamp Solutions, Inc. (Cayman), "Radio Frequency Receiver Architecture." Filed/priority 2007‑09‑27; published 2009‑04‑02. Google Patents: https://patents.google.com/patent/US20090088124A1/en
• Description: An RF receiver architecture (LNA/mixer chain, gain and impedance handling) for multi‑standard receivers.
• § 102 exposure: Low–moderate. Relevant background for LNA and mixer arrangement and possibly for matching, but it is a single‑chain receiver architecture; it does not disclose the two‑stage distributed front end with carrier‑group routing or the one‑amplifier‑to‑two‑demodulator coupling of claim 1. Does not anticipate claims 1, 4, or 16. (Full text not retrieved — provisional.)

R5. US 2013/0230080 A1 (granted as US 9,172,402 B2)

• Citation: Qualcomm Incorporated, "Multiple‑input and multiple‑output carrier aggregation receiver reuse architecture." Priority 2012‑03‑02; published 2013‑09‑05. Google Patents: https://patents.google.com/patent/US20130230080A1/en ; grant: https://patents.justia.com/patent/[9172402](/patent/9172402)
• Description: A MIMO/carrier‑aggregation receiver that reuses carrier‑aggregation receiver paths across a transceiver chip and a separate receiver chip (first/second PRx and SRx), with routing from four antennas through respective receivers, multiple LNAs per receiver, and shared synthesizers. Expressly states the device "may not require a power splitter, an external low noise amplifier or die‑to‑die signal routing."
• § 102 exposure: Moderate for the multi‑receiver/CA context, low for the specific claim limitations. It is squarely in the same field (carrier aggregation, multiple LNAs, separate transceiver/die arrangement — cf. '336 claims 11–13), and its "multiple LNAs / multiple receivers on chips" disclosure is relevant to the general architecture. However, it discloses reuse of parallel receiver paths, not a first‑stage carrier‑group amplifier feeding a second‑stage amplifier whose two outputs go to two different demodulators (claim 1), nor the claim‑4 first/second‑LNA + routing‑module trio. Does not anticipate claims 1, 4, or 16; it is a strong § 103 candidate when combined with, e.g., R6 or R7.

R6. US 2013/0231064 A1 (granted as US 9,362,958 B2) — among the two most relevant citations

• Citation: Qualcomm Incorporated, "Single‑chip signal splitting carrier aggregation receiver architecture." Priority 2012‑03‑02; published 2013‑09‑05. Google Patents: https://patents.google.com/patent/US20130231064A1/en ; grant: https://patents.google.com/patent/[US9362958B2](/patent/US9362958B2)/en
• Description: A single‑chip signal‑splitting CA receiver. Each LNA has a first‑stage amplifier (transconductance/Gm) and a second‑stage amplifier (cascode/Cas). Signal splitting stages route a signal from a first‑stage amplifier of one receiver into a second‑stage amplifier (or a mixer) of another receiver to reuse a simultaneous‑hybrid‑dual‑receiver path. Explicit inter‑/intra‑band configurations with low/mid/high bands, multiple LNAs for a first band and a second band, and passive mixers producing I/Q.
• § 102 exposure: Highest of the group for the two‑stage concept and for claim 4's topology. It discloses a two‑stage LNA (Gm + cascode), multi‑band LNAs, splitting a first‑stage output and routing it to a second‑stage amplifier/mixer, and differential/passive mixing to baseband. It does not, on the retrieved disclosure, teach a first‑stage amplifier that groups portions of received carriers and a second‑stage amplifier whose two outputs are coupled to two different demodulators as claimed in claim 1, nor the specific claim‑4 pairing of two LNAs each producing two carrier‑portion outputs plus a routing module feeding two output ports. So it is not a clean anticipation of claim 1 or 4, but it is the closest structural reference and a very strong § 103 basis, alone (applying the artisan's ordinary skill) or in combination with R5/R7. If one reads the '336 claim 1 "second stage amplifier … first output → demodulator 1, second output → demodulator 2" as a bare two‑output splitting function, R6's splitting stages come closest to anticipating — but the mapping is imperfect, so I would not assert anticipation with high confidence.

R7. US 8,532,588 B1

• Citation: The Connectivity Patent Trust (inventors Weber, Terrovitis, Su), "Apparatus for signal power loss reduction in RF communication systems." Priority 2002‑08‑13 (CIP of 11/559,382; continuation of 10/217,699); published 2013‑09‑10. Google Patents: https://patents.google.com/patent/[US8532588B1](/patent/US8532588B1)/en
• Description: Redundant front‑end LNAs/PAs to avoid diversity‑switch losses. An RF receive front end includes a plurality of LNAs, each receiving an RF signal from a dedicated antenna, and a "signal designator" that receives the amplified LNA signals and selects designated output signals; each of several mixer sets receives a designated output and an LO (I/Q); adders produce baseband I/Q.
• § 102 exposure: Moderate, and specifically relevant to claims 1 and 16. The "plurality of LNAs → signal designator (a routing/selection function) → designated outputs → mixers → baseband" chain maps onto several '336 elements (multiple LNAs, routing/selection of amplified outputs, distribution to mixers/demodulators). What it does not show is the '336 two‑stage architecture where a second‑stage amplifier is interposed after routing and its two outputs fan out to two different demodulators, nor the carrier‑grouping (by band group) of claim 1. Does not, standing alone, anticipate claims 1, 4, or 16, but it is a meaningful § 103 combination partner for R6, particularly against claim 16 (means‑for‑amplifying / means‑for‑routing / means‑for‑demodulating).

R8. US 2014/0134959 A1 (granted as US 9,048,928 B2)

• Citation: Qualcomm Incorporated, "Expandable transceivers and receivers." Priority 2012‑11‑13; published 2014‑05‑15. Google Patents: https://patents.google.com/patent/US20140134959A1/en ; grant: https://patents.google.com/patent/[US9048928B2](/patent/US9048928B2)/en
• Description: Expandable transceiver/receiver architecture (scalable receiver front‑end blocks).
• § 102 exposure: Legally notable because of timing, technically modest. It published 2014‑05‑15, exactly one day before the '336 priority date (2014‑05‑16), so it is § 102(a)(1) prior art only by that one‑day margin (and would not be § 102(a)(2) art, being a pre‑filing publication). Its subject matter (expandable multi‑receiver front ends) is background to the "scalable/expandable to more than 4 DL carriers" theme of the '336 patent, but the retrieved material does not show the claim‑1 second‑stage two‑demodulator fan‑out or the claim‑4 routing module. Does not anticipate claims 1, 4, or 16. Flag for re‑verification: the one‑day publication margin makes the exact publication date worth confirming from the USPTO/EPO record. (Full claims not retrieved — provisional.)

R9. US 2014/0134960 A1 (granted as US 9,603,187 B2) — the other most relevant citation

• Citation: Qualcomm Incorporated (inventors Tasic, Davierwalla, Narathong, van Zalinge), "Omni‑band amplifiers." Priority 2012‑11‑14; published 2014‑05‑15. Google Patents: https://patents.google.com/patent/US20140134960A1/en ; grant: https://patents.justia.com/patent/[9603187](/patent/9603187)
• Description: An omni‑band LNA having at least one gain transistor and a plurality of cascode transistors for a plurality of band groups (low/mid/high band). A selected gain‑transistor/cascode‑transistor pair is enabled to amplify the input RF signal and provide an output for the selected band group; gain transistors may couple to different taps of a shared source‑degeneration inductor or to separate inductors; load circuits include transformers; a single input feeds multiple band‑group outputs; a tunable matching circuit is disclosed.
• § 102 exposure: High for the amplifier‑level limitations; low for the system‑level claims. This is the clearest disclosure among the citations of a multi‑band‑group LNA with stacked gain+cascode devices and switchable, band‑group‑selectable outputs feeding transformers/downconverters — the structural core of the '336 first‑stage LNAs (see '336 FIG. 5, transistors 512/516 PMOS + 514/518 NMOS and switches). It is very close to claim 4's "first LNA … second LNA … generating first/second portions of carriers" in the sense of enabling one of several amplifier circuits per band group and providing corresponding outputs. However, R9 does not disclose the claim‑4 routing module that routes first/second/third/fourth amplified outputs to first and second output ports, nor the claim‑1 second‑stage two‑demodulator fan‑out. Like R8, it published 2014‑05‑15 (one day before the '336 priority date), so it is § 102 art only by that margin. Does not anticipate claims 1, 4, or 16 as a whole, but is a powerful § 103 reference (and, because of its overlapping inventors, also bears on common‑ownership/§ 103(c)‑type considerations).

R10. US 9,136,811 B2

• Citation: Edgewater Wireless Systems Inc., "Automatic gain control (AGC) for multichannel/wideband communications system." Priority/filing 2004‑07‑26; published 2015‑09‑15. Google Patents: https://patents.google.com/patent/[US9136811B2](/patent/US9136811B2)/en
• Description: AGC for multichannel/wideband communications.
• § 102 exposure: Low. The effective prior‑art date is its 2004 filing (pre‑priority), notwithstanding its 2015 grant date. Its subject (AGC across wideband/multichannel) is tangential to the '336 two‑stage LNA routing/pin architecture and does not disclose claims 1/4/16. Does not anticipate any claim. (Full text not retrieved — provisional.)

2. Ranking by relevance to the '336 claims

3. § 102 anticipation assessment — bottom line

No single examiner‑cited reference appears to anticipate any of independent claims 1, 4, or 16. For strict anticipation, one reference must disclose every element, including:

• (claim 1) a first‑stage amplifier producing a carrier group and a second‑stage amplifier whose two outputs are coupled to two different demodulators; and
• (claim 4) two LNAs each producing two carrier‑portion outputs plus a routing module feeding two output ports.

The closest references fail on these specific elements:

• R6 supplies the two‑stage LNA and splitting, but not the claim‑1 second‑stage→two‑demodulator fan‑out or the claim‑4 dual‑LNA + routing‑module trio.
• R9 supplies the multi‑band‑group gain/cascode LNA, but not the routing module to two output ports nor the second‑stage fan‑out.
• R7 supplies multiple LNAs plus a selection ("designator") function, but not the second‑stage amplifier interposed between routing and demodulators.

Consequently, the defensible prior‑art theory against the '336 patent is obviousness under § 103 — most plausibly R6 + R9 (two‑stage/multi‑band LNAs) and/or R6 + R7 or R5 + R6 — rather than § 102 anticipation. I would not represent that any cited reference anticipates a specific claim.

4. Additional context worth noting

• Same‑family/companion filings: The '336 family includes EP 3143700 B1, EP 3573245 B1, JP 6178025 B1, KR 101804529 B1, CN 106464277 B, CN 110581713 B, ES 2729865 T3, HU E044947 T2, and WO 2015/175349 A1 (https://patents.google.com/patent/US9473336/en). These are not prior art to the '336 patent (common priority) but are relevant to any global validity assessment.
• References citing the '336 patent (post‑dating, not prior art) include US 10,439,858 B2 ("Multiplex modules for carrier aggregation receivers," Qualcomm, 2019) and US 10,390,343 B2 ("Carrier aggregation diversity antenna module with integrated LNA banks," Qualcomm, 2019) — useful for claim‑family mapping.
• Litigation/family note (from the patent record): an ITC § 337 proceeding (337‑TA‑1093) and a California S.D. case (3:17‑cv‑02398) are listed; if validity is at issue there, the R6/R9 combination and the one‑day‑margin publications R8/R9 would be the priority references to verify against full text.

Verification recommendations (before relying on this): (1) pull the full claim sets of R2, R4, R8, R10 and the specifications of R1/R3 from USPTO PatentCenter/Google Patents to confirm the "no anticipation" conclusions; and (2) confirm the exact publication dates (2014‑05‑15) of US 2014/0134959 A1 and US 2014/0134960 A1, since their status as § 102(a)(1) art depends on a single day.

I'll research the key prior-art references cited on the patent page to ground the obviousness analysis.

Let me research the remaining cited references to complete the prior-art picture.

Obviousness Analysis — US 9,473,336 B2 under 35 U.S.C. § 103

Analyst note / flag before starting: The task header states "Current Date: April 26, 2026," while the fetched source header for this patent page reads "Fetched: 2026-09-27." That is a minor date inconsistency, not a substantive contradiction, and it does not affect the analysis below.

Method note: Consistent with the rules, I take the prior-art section of the US 9,473,336 page as the authoritative universe of references. That section contains "Citations (10)" (the examiner-cited prior art) and "Family Cites Families (8)" (art cited in family members). The "Similar Documents" and "Cited By" lists are largely later art and I have used them only for context, clearly labeled. I did not auto-correct any patent numbers. Where I could not verify a reference's substance in the sources I retrieved, I say so rather than inventing content.

1. The § 103 framework applied here

• Priority date: 2014-05-16 (provisional 61/994,158); filed 2015-03-27. Prior art must predate the priority date. All ten examiner citations do; US 9,300,420 (filed 2012-09-11) does as well.
• Level of ordinary skill (PHOSITA): a designer of RF receiver front-ends for cellular carrier-aggregation (CA) transceivers — familiar with multi-band/multi-carrier LNAs, on-chip vs. off-chip partitioning, mixers/downconverters, LO distribution, and PCB/IC pin/routing isolation practice.
• The problem the patent itself names (Description/Background): path-to-path isolation and receiver desensitization when multiple receive paths run concurrently. That framing matters, because every primary reference below is directed at the same problem, which strengthens the motivation-to-combine showing.

I analyze the three independent claims first, then the dependents that carry commercial weight.

2. Independent claim 1 — strongest prima facie case

2.1 Claim 1 element breakdown

2.2 Primary combination A1: US 2013/0231064 A1 + US 2014/0134959 A1

US 2013/0231064 A1 (Qualcomm; "Single-chip signal splitting carrier aggregation receiver architecture"; granted as US 9,362,958 B2) — https://patents.google.com/patent/US20130231064A1/en

This reference is the closest to the core of claim 1. Its specification states directly that "signal splitting" involves:

"taking a signal from the output of a first stage amplifier (such as a transconductance stage (Gm)), splitting the signal and piping the signal into two separate second stage amplifiers (such as cascade stages (Cas)) and subsequent mixers for carrier aggregation. In another configuration … signal splitting refers to taking a signal from the output of a second stage amplifier (such as a cascade stage (Cas)), splitting the signal and piping the signal into two separate mixers for carrier aggregation." ([0035])

It also discloses that "each low noise amplifier may include a first stage amplifier and a second stage amplifier. The first stage amplifier may be a transconductance stage and the second stage amplifier may be a cascode stage," and characterizes a splitting stage as "a routing between a second stage amplifier in a low noise amplifier of the primary receiver and a mixer in the tertiary receiver" (claim 13; EP 2 820 761 B1 [0065]–[0068], routing 535a from Cas of LNA 348c into mixer 356c).

Mapping to claim 1:

• (a) the Gm stage amplifies received carriers and its output is split into carrier "portions" feeding downstream stages — a first stage amplifier generating a first stage carrier group;
• (b) the Cas stages re-amplify — second stage amplifiers generating second stage output signals; and
• (c) the Cas output is split to two different mixers (demodulators) — the "first output → first demodulator / second output → second demodulator" limitation, in substance.

US 2014/0134959 A1 (Qualcomm; "Expandable transceivers and receivers"; granted as US 9,048,928 B2) — https://patents.google.com/patent/US20140134959A1/en — supplies the two-IC, PCB-traced architecture the '336 relies on, and — critically — a single first-stage LNA driving two separate downconverters:

"LNA 540 also provides a second output RF signal via interface circuit 570 to buffer 582, load circuit 554, and downconverter circuit 564 … LNA 540 thus drives both downconverters 560 and 564 on two IC chips 510 and 512." ([0051]); see also [0052] ("Each enabled LNA may drive one or more on-chip downconverter circuits and/or an off-chip downconverter circuit") and [0059] ("An amplified RF signal for a given RX band may be sent between multiple IC chips…").

It also expressly frames the isolation problem the '336 patent addresses: a 100 dB TX/RX isolation requirement that "may be difficult or not possible to achieve since pin-to-pin isolation is worse than the isolation requirement" ([0036]), solved by splitting receivers across chips ([0037], [0082]).

Result: A1 discloses every element of claim 1. Even standing alone, US 2014/0134959 is arguably a single-reference § 102 case for much of claim 1 (one LNA, two outputs, two downconverters); combined with US 2013/0231064's explicit Gm/Cas split-to-two-mixers teaching, claim 1 is rendered obvious.

2.3 Alternative combination A2: US 2013/0230080 A1 + US 9,300,420 B2 + US 2014/0134959 A1

• US 2013/0230080 A1 (Qualcomm; MIMO CA receiver reuse) — https://patents.google.com/patent/US20130230080 — discloses a transceiver chip plus a receiver chip, with "each receiver … multiple low noise amplifiers" (claim 3) and mixer 450a coupled to the outputs of LNAs 446a–d, plus PLL/VCO and baseband filter (see FIG. 4 and [0047]–[0049]); routing of carriers across chips and across demodulator stages.
• US 9,300,420 B2 (Qualcomm; "Carrier aggregation receiver architecture") — https://patents.justia.com/patent/[9300420](/patent/9300420) — discloses "a plurality of low noise amplifiers (LNAs), a plurality of switches, and at least one downconverter … The switches are coupled to the outputs of the plurality of LNAs. The at least one downconverter is coupled to the plurality of switches," and that "the LNAs and the switches may be implemented on at least one front-end module or a back-end module. The downconverter(s) are implemented on the back-end module" (Abstract).

Together with US 2014/0134959's two-output LNA, A2 also reaches claim 1.

2.4 Motivation to combine (claims 1)

1. Same field, same problem, same assignee, overlapping inventors. US 2014/0134959/960 name Aleksandar Miodrag Tasic and Chiewcharn Narathong, who are also named inventors on the '336 patent itself. US 2013/0230080, US 2013/0231064 and US 9,300,420 are all Qualcomm carrier-aggregation front-end work. A PHOSITA would treat these as a single, mutually informing design line.
2. The references expressly seek the '336 patent's stated objective — reducing cross-path interference and meeting TX/RX isolation in CA. US 2014/0134959 states the goal and the reason ("stimulate ... spurious and isolation problems," "mitigate ... by splitting the receiver inputs and transmit outputs between multiple IC chips," [0037], [0082]). Substituting/combining known split-and-route structures is the predictable way to achieve it.
3. Predictable result / KSR. Routing an already-available carrier group from one LNA output to more than one downstream demodulator is (i) a known technique (splitter/routing between amplifier stages and mixers) and (ii) yields no more than the expected benefit (a carrier simultaneously visible to two demodulation paths). Adding a second-stage gain block to a routed RF carrier is equally routine.

3. Independent claim 4 — the claim Qualcomm asserted

3.1 Claim 4 element breakdown

3.2 Combination B: US 2014/0134959 A1 + US 2014/0134960 A1 + US 9,300,420 B2

US 2014/0134959 A1 maps the multi-amplifier-circuit LNA almost literally. Its dependent claims recite:

Claim 12: "the LNA comprising: a first amplifier circuit configured to receive an input RF signal and provide a first output RF signal to the first output of the LNA, and a second amplifier circuit configured to receive the input RF signal and provide a second output RF signal to the second output of the LNA."
Claim 13: "a third amplifier circuit configured to receive a second input RF signal and provide a third output RF signal to the first output of the LNA."

FIG. 9A/9B further show amplifier circuits 941/951 and 943a–953m, and FIG. 9B a "multi-input multi-output LNA 944" with M on-chip outputs plus off-chip output(s). That is the two-LNA, four-output structure of claim 4(a)–(b) in all but nomenclature (each amplifier circuit carries a different band-group portion of the received carriers). Specifically, US 2014/0134959 claim 9 recites a first LNA "for a first band group and the second LNA being for a second band group" — precisely the "first portion / second portion of the carrier signals" split.

US 2014/0134960 A1 (Qualcomm; "Omni-band amplifiers"; granted as US 9,603,187 B2) — https://patents.google.com/patent/US20140134960A1/en — adds both the multi-band-group output structure and a switching fabric between LNA outputs and downconverters:

"an omni-band LNA … has (i) a single input for all supported band groups and (ii) multiple outputs for the multiple band groups, e.g., one output for each band group" ([0029]); "each load circuit 470 includes a transformer … secondary coil providing a first differential amplified RF signal … to a downconverter"; and "multiple load circuits 470 may be coupled to a shared downconverter via switches. The switches may be controlled to pass an amplified RF signal from one load circuit to the shared downconverter at any given moment" ([0037]).

US 9,300,420 B2 supplies the "routing module" claim element explicitly: "a plurality of LNAs, a plurality of switches, and at least one downconverter … The switches are coupled to the outputs of the plurality of LNAs" (Abstract/claim 1), i.e., a controllable switch fabric that steers any LNA amplified output to a selected downstream port.

Result: claim 4(c)'s "routing module … route at least one of the first, second, third, and fourth amplified outputs to a first output port and … to a second output port" is met by the switch bank of US 9,300,420 and/or the shared-downconverter switches of US 2014/0134960; claim 4(a)–(b) is met by US 2014/0134959 FIGS. 9A–9B / claims 9, 12–13; claim 4(d) by the second-stage (Cas / iLNA) stages of US 2013/0231064 and US 2014/0134959.

3.3 Motivation to combine (claims 4)

• Design continuity + explicit "expandability" teaching. US 2014/0134959's stated purpose is a modular, expandable architecture — "the modular design of the expandable transceivers may allow any number of IC chips to be coupled together to support any number of carriers, any number of bands, and any number of band groups" ([0059]). A PHOSITA seeking to scale a CA receiver would naturally add the switch-steering fabric of US 9,300,420 and the multi-band-group output structure of US 2014/0134960.
• Reconfigurability rationale. Routing any amplified output to either of two output ports lets the receiver accommodate inter-band vs. intra-band CA configurations with the same silicon — the same flexibility rationale US 2014/0134960 articulates ("20 omni-band LNAs … configured to support … some other combination of receivers," [0071]).
• No unexpected result. The claimed combination produces band-group aggregation and port re-selection — exactly what the references promise.

Important litigation cross-reference (from the prior section): In ITC Inv. No. 337-TA-1093, claim 4 was the asserted claim; the ALJ found it valid but not infringed. The non-infringement finding is about the accused product's structure, not about whether the prior art renders claim 4 obvious — so this § 103 case stands or falls on the references, not the infringement record. I flag that the prior section's "no IPR/reexam invalidating the '336" statement and the separately referenced IPR2019-00186 exhibit are not reconciled in the material I reviewed; I could not confirm whether IPR2019-00186 targeted the '336 patent, so I do not assert that it did.

4. Independent claim 16 — means-plus-function

Claim 16 is claim 1 in § 112(f) form. Its structures, per the '336 specification (FIG. 9), are: "means for amplifying received carrier signals" = first stage amplification 302; "means for amplifying the … first stage carrier group" = second stage amplification 304; the two "means for demodulating" = the demodulator stages dm0–dm3 / mixer modules 608(a)–(h).

Because § 112(f) permits a functional claim to be rendered obvious by a combination that discloses the corresponding structure performing the recited function, the A1, A2 and B combinations above fully cover claim 16:

• "means for amplifying received carrier signals to generate … first stage carrier group" ↔ LNA 540/944 (US 2014/0134959) or omni-band LNA 440 (US 2014/0134960);
• "means for amplifying the … first stage carrier group" ↔ Cas stages of US 2013/0231064 / downstream LNA+buffer of US 2014/0134959;
• "means for demodulating" ↔ mixers/mixer modules of US 2013/0230080, US 2013/0231064, US 2014/0134959.

Claims 17 and 19 are merely the functional mirror of claims 2 and 8 and fall with them.

5. Claims 2–3, 5–10 (dependent)

Verdict: claims 2–3 and 5–10 are obvious over Combination B with the noted secondary references. Claim 8's transformer limitation in particular is squarely met, since both US 2014/0134959 and US 2014/0134960 disclose transformer load circuits whose secondary coils provide differential output RF signals.

6. Claims 11–14 (IC partitioning, PCB, interface pins)

Verdict: claims 11–14 are obvious. The partitioning of an RF front-end into a front-end IC (LNA + switching) and a back-end IC (downconversion) was itself a known architectural choice for isolation (US 2014/0134959's core thesis), and pins/PCB traces are the necessary physical consequence.

7. Claim 15 — the "CA-centric pin / orthogonal routing" claim

Claim 15 recites the interface pins arranged "carrier-centric," selected pins "positioned proximate to a respective demodulator to isolate signal traces connected to the demodulator," and "signal traces connected to different demodulators cross orthogonally."

Prior-art support (weaker, but present):

• US 2014/0134960 FIG. 10 shows three band-group transformers with nested primary coils and ground guard rings between adjacent primaries that "provide[] isolation between these two primary coils" ([0067]) — i.e., deliberate layout to reduce mutual coupling between adjacent band-group paths.
• US 2013/0230080 and US 2013/0231064 both emphasize on-chip routing/coupling, and US 2014/0134959 states the isolation objective quantitatively (100 dB; pin-to-pin vs. chip-to-chip isolation, [0036]).
• The general body of IC layout knowledge (co-locating a source/destination, avoiding parallel runs, crossing at right angles to minimize capacitive coupling) is routine layout practice; the claim's benefit ("minimizing signal coupling") is a predictable result of a designer's choice.

Verdict and caution: This is the most vulnerable to a non-obviousness counterargument, for two reasons: (i) the "carrier-centric pin placement proximate to the corresponding demodulator with orthogonal crossing" is a layout optimization that prior art supports only by analogy, not element-by-element; and (ii) per the prior section, the ITC upheld validity of claim 4 (not claim 15), so there is no adjudicated validity holding on claim 15 either way. An obviousness challenge to claim 15 would need an additional layout/floor-planning reference whose pin/demodulator adjacency is expressly for coupling reduction. I did not verify such a reference in the prior-art section of record, so I will not manufacture one. (JP 5992649 B2, "Signal line wiring and shielding to improve insulation," appears only in the page's "Similar Documents" list — not the prior-art citations — and I could not confirm its date, so I do not rely on it.)

8. Unifying motivation-to-combine statement

For claims 1, 4 and 16 (and their dependents), a PHOSITA at the 2014 priority date had, from the cited art:

1. A recognized need — the '336 Background and US 2014/0134959/960 all identify CA path-to-path isolation and desensitization as the problem to solve.
2. A known two-stage LNA/split architecture — US 2013/0231064 (Gm → Cas, with splitting to two second-stage amplifiers and two mixers) and US 2014/0134959 (multi-output LNAs driving on-chip and off-chip downconverters).
3. A known multi-band-group routing fabric — US 2014/0134960 (one output per band group; switches to a shared downconverter) and US 9,300,420 (LNA outputs → switch bank → downconverter).
4. A known partitioned-IC / PCB-trace topology — US 2014/0134959, US 2013/0230080, US 9,300,420.
5. A reason to combine with a reasonable expectation of success — same assignee, overlapping inventors (Tasic, Narathong), same field, same objective, and each reference expressly designed for modular expansion/scaling across bands, carriers and chips.

That combination yields every element of independent claims 1, 4 and 16 without any allegedly new physical principle. On this record the prima facie obviousness case is strong for claims 1–14 and 16–20, and moderate-to-weak for claim 15.

9. Counterpoints a patent owner would raise (and my response)

• Secondary considerations. Qualcomm would likely argue commercial success and industry adoption of the two-stage/partitioned CA front-end; the page's own "Cited By" list shows Samsung (US 11,228,326; US 11,252,728), MediaTek (US 10,756,686) and others building CA receiver/LNA architectures — evidence of a crowded, commercially important field, which can support either side. I cannot quantify nexus to the claimed combination, so I treat this as an open factor rather than a rebuttal.
• The ITC validity holding on claim 4. Per the prior section, the ALJ found claim 4 valid. That is a § 103 sanity check pointing the other way, but an ALJ validity finding is not a holding that the specific reference combinations above were considered; the art-of-record and the asserted § 103 theory must be compared directly.
• Claim 15's specificity. The pin-adjacency + orthogonal-crossing limitation is the hardest to meet from the cited art; a strong § 103 case on claim 15 would require a floor-plan/pin-assignment reference that I did not verify here.

10. Explicit uncertainties and limits (do not over-read this memo)

1. Unverified references. I retrieved substantive text for US 2014/0134959 A1, US 2014/0134960 A1, US 2013/0230080 A1, US 2013/0231064 A1 and US 9,300,420 B2. I could not obtain reliable content for US 2009/0088124 A1 (Nanoamp) and US 8,532,588 B1 (Connectivity Patent Trust); my searches did not complete. I therefore did not rely on them for any element. Their relevance (RF receiver architecture; signal power-loss reduction) is inferred from their titles only.
2. US 4,054,842, US 7,349,488, US 9,136,811, US 2006/0068746. These are cited on the page but are peripheral (FSK channel-gain compensation, FSK signaling in ICs, AGC for multichannel/wideband systems, and a direct-conversion receiver RFIC). I used US 2006/0068746 only for the generic direct-downconversion/LO-to-baseband teaching relevant to dependent claim 9, and the others not at all.
3. Non-patent citation. The only NPL of record is "International Search Report and Written Opinion—PCT/US2015/029966—ISA/EPO—Jul. 20, 2015." I did not retrieve its content, so I cannot report its X/Y/A categorizations for the '336 claims.
4. Litigation/IPR status. I did not independently confirm the ITC final Commission determination, nor whether IPR2019-00186 challenged US 9,473,336. The prior section flags this; I do not resolve it and do not assert it.
5. Status/nature of this memo. This is a technical, prima facie § 103 analysis, not a legal opinion. Any invalidity challenge must be run against verified full-text references, the correct priority/§ 102 dates, and any applicable claim-construction rulings (e.g., the disputed "first stage carrier group" and "first/second portion of the carrier signals" terms noted in the ITC Joint Claim Construction Chart referenced in the prior section).

Primary sources used: US 9,473,336 page — https://patents.google.com/patent/US9473336/en ; US 2014/0134959 A1 — https://patents.google.com/patent/US20140134959A1/en (granted US 9,048,928 — https://patents.google.com/patent/US9048928) ; US 2014/0134960 A1 — https://patents.google.com/patent/US20140134960A1/en (granted US 9,603,187) ; US 2013/0230080 A1 — https://patents.google.com/patent/US20130230080 ; US 2013/0231064 A1 — https://patents.google.com/patent/US20130231064A1/en (granted US 9,362,958; EP 2 820 761 B1) ; US 9,300,420 B2 — https://patents.justia.com/patent/9300420.