US 6449070

I'll search for this specific patent number, including any CAFC/litigation activity.

Let me run a couple of narrower searches to check for any litigation or CAFC activity tied to this number specifically.

Search results for US Patent No. 6,449,070

Scope/limits of this search. I searched for the specific seven-digit number 6449070 (US 6,449,070 B1). I relied on web-accessible mirrors that index USPTO/EPO data (Google Patents, FreePatentsOnline, Justia) plus open web search for any Federal Circuit activity. I could not log into USPTO PatentCenter/PAIR or the CAFC docket system directly from this session, so the litigation check is a negative-result search, not a certified docket clearance. Everything below matches the authoritative full text you supplied, which I treat as controlling over any search snippet.

Bibliographic data (high confidence — consistent across sources)

Abstract (as published)

An optical transmission device applicable to a wavelength-multiplexed optical transmission system includes a first part creating a supervisory control signal transmitted optically together with an optical main signal, the supervisory control signal comprising a vector of elements respectively corresponding to wavelengths, and a second part converting a vector in a supervisory control signal received from a remote device into another vector using a path matrix in which the received vector forms the diagonal elements; the resulting vector is passed to a next stage of the system.

CAFC 2026 / litigation check — no matching record found

• No Federal Circuit 2026 docket, appeal, IPR, or district-court assertion tied to US 6,449,070 was surfaced. Given the patent expired on 2019-01-27 (past its 20-year term), active 2026 appellate litigation would be unusual, and none appeared.
• ⚠️ Do not confuse this with "JP 6449070." A Japanese patent number JP 6449070 B2 exists but is a completely unrelated Fukubi Chemical Industries patent (insulated under-floor structure). The bare digit string "6449070" resolves to different documents depending on jurisdiction; only the US document is the Fujitsu optical patent.
• Search snippets for "6449070" typically returned the Justia/"patent 7899331" prior-art citation listing "6449070 | September 10, 2002 | Izumi," which is a reference to this same US patent, not separate litigation.

Plain-language overview of the independent claims

The patent has six independent claims — 1, 14, 18, 20, 22, and 24. Claims 20–25 read like later-added claims (apparatus/means-plus-function/method variants of claim 1); the authoritative text I was given presents them within the 25-claim set, and I flag mild uncertainty about whether they were in the original grant or added by a post-issuance certificate/reexamination.

• Claim 1 — Optical transmission device (core invention). A device for a WDM system has (a) a transmitter that builds a "supervisory control signal" — an optical overhead signal carrying a vector whose bits each correspond to a wavelength — and sends it alongside the main optical signal; and (b) a receiver that takes a vector from an incoming supervisory signal from a remote device and transforms it into a new vector using a path matrix in which the received vector sits on the diagonal, then forwards that new vector downstream. In plain terms: each node recomputes the "which wavelengths are in use" map as it hands traffic to the next node, using a matrix multiplication.

• Claim 14 — Optical amplifier. An amplifier that receives the WDM signal plus the supervisory signal containing a wavelength-information vector, optically amplifies the WDM signal, and runs automatic level control (ALC) based on the count of wavelengths indicated by that vector — so gain is set per how many channels are actually present, without per-wavelength monitors or software.

• Claim 18 — Wavelength-multiplexed optical transmission system. A whole system of multiple such optical transmission devices, each with the claim-1 create-and-path-matrix-convert functionality, chained together.

• Claim 20 — Apparatus (transmitter/receiver form). An apparatus whose optical transmission device has a transmitter that sends the vector-bearing supervisory signal on the WDM line and a receiver that performs the diagonal-path-matrix conversion on an incoming remote supervisory signal and forwards the result downstream. Substantively the claim-1 subject matter recast in transmitter/receiver terms.

• Claim 22 — Apparatus (means-plus-function form). Same invention expressed as means for creating/transmitting, means for receiving, means for converting via the diagonal path matrix, and means for transmitting downstream — a §112(f)-style counterpart to claim 1.

• Claim 24 — Method. The corresponding method: create the supervisory control signal; transmit it with the WDM optical signal; receive a remote supervisory signal containing a vector; convert that vector via a path matrix with the received vector as diagonal elements; and transmit the converted vector downstream.

Dependent-claim themes worth noting (not independent, but they define the commercial scope)

Claims 2–13 add: zero-wavelength notification on main-signal loss (2); collision detection across routes (3); a returned received-state vector (4); OR/AND combination of received-state vectors across branched routes (5–6); fault identification by comparing wavelength-information vs. received-state vectors (7); wavelength-retrieval signaling with device ID and counter (8–9); stored receive-direction memory and its reset after route switching (10–11); route-switching instruction in the supervisory signal (12); and network/terminal-side role assignment by comparing device IDs (13). Claims 15–17 add the amplifier behaviors: hold last-known wavelength count on supervisory-signal-only failure (15); treat wavelength count as zero when both main and supervisory signals fail (16); and ramp-then-rollback on saturation/fault (17). Claim 19 adds the master/slave station scheme with master handover on fault.

Confidence statement

The bibliographic data, abstract, and claim text are high confidence (they come from the authoritative full text you provided, corroborated by Google Patents, FreePatentsOnline, and Justia). The litigation/CAFC result is a negative finding with moderate confidence — no 2026 Federal Circuit docket for this number appeared, and the 2019 expiration makes such activity unlikely, but I was unable to run a certified CAFC/PAIR docket query. I found no authoritative basis for any assertion, IPR, or appeal involving US 6,449,070.

I'll research litigation involving US patent 6449070 across patent litigation databases.

Let me search more specific litigation databases.

Let me check a few more specialized sources for litigation records.

Litigation Search Results: US Patent 6,449,070

Bottom line

No litigation involving US Patent 6,449,070 was found. Despite multiple targeted searches of patent-litigation sources (including Unified Patents' litigation portal, CourtListener/CAFC-adjacent records, and general docket-style queries), I could not identify any district court action, ITC investigation, PTAB proceeding, or Federal Circuit appeal in which US 6,449,070 was asserted, challenged, or otherwise at issue.

What the patent is (context)

The patent covers a WDM optical transmission device that creates a supervisory control signal carrying a wavelength vector and converts received vectors via a "path matrix," plus related optical-amplifier automatic level control. Family members include EP 0936769 B1, JP 3551407 B2, and DE 69930985 T2.

Important clarifications on the search

Several results contained the number "6449070" but are not litigation against this US patent, and I want to be explicit about the false positives so the record is clean:

• JP 6449070 B2 is an unrelated Japanese patent (Fukubi Chemical Industry K.K.) for a "heat-insulated under-floor structure and construction method for a heat-insulated under-floor person-passage opening." Different jurisdiction, different owner, different technology — not this patent.
• FCC, SEC, and municipal filings surfaced "6449070" only as incidental order/accession IDs, not patent matters.
• Google Patents "Cited By" / "Citations" lists (e.g., OL2/OnLive streaming patents, Cisco, Ciena, NEC references) are patent-document citations within prosecution, not litigation involving the '070 patent.
• References to a "'907 patent" (e.g., the Netlist ITC case) are to different patents (e.g., US 7,619,912 / haptics patents), again not this patent.

Limitations

• Google Patents shows the patent as "Expired – Lifetime," which is consistent with a patent that was never actively asserted. A granted patent that lapses to expiration in 2019 rather than being litigated is consistent with my finding, but the absence of a record is not affirmative proof of zero activity.
• I could not directly render the Unified Patents District Court litigation caselist or PACER for the exact number within the available steps; the searches returned no case matching US 6,449,070.

Recommendation

To confirm conclusively, I would recommend running the exact number 6,449,070 through:

• PACER Case Locator (all federal district courts),
• Unified Patents Portal → Litigation → search patent number 6449070,
• USPTO Patent Trial and Appeal Board (PTAB) API / Patent Center for the application number 09/237,904,
• Docket Navigator or Lex Machina for asserted-patent queries.

Based on the searches performed, the honest answer is: no known litigation involving US 6,449,070.

PTAB / AIA Trial Landscape — U.S. Patent No. 6,449,070 B1

Proceedings overview

Total AIA trial proceedings on file: 0 (zero). The USPTO Open Data Portal returns no IPR, PGR, or CBM proceedings for U.S. Patent No. 6,449,070. There are therefore no claims invalidated, none sustained, none settled, and no institution decisions to report. The defensive posture this gives a defendant is unusual and, on balance, favorable — but not for the reason a defendant might assume. This is not a "hardened" patent that survived attack; it is an expired patent that the PTAB market never bothered to challenge:** the record shows an "Anticipated expiration" of 2019-01-27, and the patent's legal status is "Expired – Lifetime."** The absence of IPR activity here is best read as a signal that the patent was never a prolific assertion target during its enforceable life, not that it is a validated, litigation-tested asset.

Verification note: I ran targeted web searches against PTAB dockets, Docket Alarm, Unified Patents-style aggregator filings, and general "IPR/litigation 6449070" queries. Nothing surfaced. I found no proceeding number that I can honestly report, and per the operating constraints I am not inventing one. If a recently-filed petition exists outside ODP's ingest, it is not discoverable through the sources I searched as of 2026-10-01.

Key dates and status (from the structured record)

Why there are no proceeding sections

The output format calls for a ### {PROCEEDING_NUMBER} — {Petitioner} v. {Patent Owner} block per proceeding, ordered by impact. With a zero-count list, there is no petitioner, no panel, no FWD, and no CAFC appeal to describe. Populating that template would require fabricating proceeding numbers, which is expressly prohibited. I am leaving it empty and telling you plainly: the canonical list is empty.

For completeness, here is what I affirmatively checked and did not find:

• IPR / PGR / CBM petitions naming 6,449,070 — none indexed in ODP.
• Federal Circuit appeals originating from an FWD on this patent — none found (no Board decision exists to appeal).
• Defensive-aggregator activity (e.g., Unified Patents) against this patent — none found.
• Reissue / reexamination — the record shows ordinary prosecution leading to grant; no post-grant reexamination is noted in the structured data.

One caution for the record: the assignee Fujitsu is a frequent petitioner, not a frequent target — e.g., Fujitsu entities filed IPRs on Core Optical's U.S. 6,782,211 (see IPR2016-01618) and appeared across multiple optical-networking IPRs. That is the reverse posture from this patent, and it is a useful pattern signal: Fujitsu acts as an IPR petitioner in the optical space, but appears not to have had 6,449,070 challenged by others. Do not conflate the two.

Strategic summary

Claim status: 1–25 UNTESTED (and now unenforceable for prospective conduct). No claim of 6,449,070 has been canceled or sustained by the PTAB, because no claim was ever challenged there. Every claim — independent claims 1, 14, 18, 20, 22, and 24, and their dependents — remains as issued on the face of the patent. That cuts both ways: a defendant cannot point to a cancellation to kill an infringement theory, but the patent's expiration means the practical exposure is limited to past damages for acts within the six-year lookback of 35 U.S.C. § 286, measured from the complaint date. There is no injunction exposure and no ongoing-royalty exposure for post-expiration activity. Under Kimble v. Marvel and the general rule that an expired patent cannot be infringed prospectively, a demand targeting current products built on the claimed supervisory-control-vector / path-matrix technique has no forward-looking damages theory at all.

Estoppel landscape (§ 315(e)(2)): empty, because the well is dry. IPR estoppel only attaches to a petitioner that lost an instituted IPR on claims it challenged. With no petitioner, there is no estoppel — which means a defendant today retains the full universe of prior-art invalidity grounds under §§ 102/103/112 in a district court or ITC action, subject only to the ordinary pleading requirements and the Kessler / Therasense framework. The downside: because no IPR was ever filed, invalidity must be proven in an Article III forum under the clear-and-conviction presumption of validity (35 U.S.C. § 282), not before the Board under a preponderance standard with broadest-reasonable-interpretation claim construction. Given the patent's 1999 filing date, the prior-art base is deep (the prosecution file itself cites Fujitsu's JPS58-200642, NEC's JPH2-270036 and JPH7-154368, NTT's JPH6-153248, Fujitsu's US 5,841,558, and NEC's US 6,055,082), so a § 102/§ 103 case is theoretically available — but it is a litigation-grade effort, not an IPR shortcut.

Pattern signals. (1) No serial-petitioner activity — nobody filed even one IPR, let alone multiple. (2) No patent-owner PTAB appeals — Fujitsu never had an FWD to appeal on this patent. (3) No defensive aggregator in the chain. The combined picture is of a dormant, expired patent with no adversarial history, which is genuinely unusual for a Fujitsu optical-networking asset from this era and reinforces that the patent was not a commercial assertion vehicle.

Recommended next steps

If you are a defendant: Confirm the expiration date (2019-01-27 per Google Patents; verify against the USPTO Patent Center "Expiration Date" field, accounting for any § 154(b) PTA and any terminal disclaimer) and pin the last day of the enforceable term. Then run the § 286 six-year lookback from the date you were actually served. Any demand built on future sales of an accused product should be rejected outright — an expired patent cannot be infringed prospectively, so a demand letter citing ongoing royalties is not supported by law.

• Because no claim has been invalidated, there is no FWD to link or quote. Do not represent otherwise to a court or opposing counsel.
• If the patent owner nonetheless asserts claims 1–13 (device), 14–17 (amplifier), 18–19 (system), 20–23 (apparatus), or 24–25 (method) for past conduct, your invalidity case must be built in district court. Start from the prosecution-file art listed above and the family citations (GB 9401488 D0; JPH09-321701), and expect a § 282 clear-and-conviction burden.
• Consider a laches / § 286 damages cutoff attack and an expiration-based non-infringement defense as your primary, cheapest dispositive theories before investing in a full invalidity case.

If you are evaluating this patent as an IPR candidate: You cannot file an IPR against it — the statutory window and the "patent" eligibility presuppose a live patent, and regardless, an expired patent offers no commercial upside to challenge. There is no trial-stage milestone to track because no trial exists.

Bottom line: No PTAB activity on file. The absence is itself the signal — this was never a heavily asserted patent, and it is now expired. A demand letter that cites 6,449,070 as a basis for prospective infringement or ongoing royalties is legally hollow; the only remaining theory is past damages inside the six-year window, and even that must overcome a presumption of validity that no IPR has ever disturbed.

I'll reconstruct the assignment chain for US 6,449,070, starting with the USPTO assignment record.

Let me search more specifically for the reel/frame and any post-issuance transfers.

Ownership & Assignment Analysis — US 6,449,070 B1

"Optical transmission device and wavelength-multiplexed optical transmission system" — Fujitsu Limited — App. 09/237,904

Source note / search limitation (stated up front). I attempted to pull the recorded assignment directly from USPTO Assignment Center (assignmentcenter.uspto.gov) and its legacy index (assignment.uspto.gov / legacy-assignments.uspto.gov). I was able to reach the USPTO assignment document corpus through search but did not obtain a reel/frame number for the Fujitsu recording; the public indices I could reach render the record without the reel/frame field. Per your instruction not to fabricate reels/frames, every reel/frame below is marked as not retrieved rather than guessed. The Google Patents legal-events register (the authoritative full text you supplied) shows exactly one reassignment record for this patent, and that is the basis for the timeline. Readers should verify at the Assignment Center and Google Patents legal events (links at the end).

Inventors

• Single-inventor patent. No co-inventor team to track, so the classic "all inventors departed within 12 months" fire-sale tell cannot be evaluated from the inventor list.
• No inventor-departure evidence surfaced. Nothing in the assignment record or in the citable search results indicates Izumi left Fujitsu near filing. I have no data on his later employment — I will not infer a departure from silence, since a single-inventor, corporate-assigned Japanese filing is the ordinary pattern and produces no observable signal.
• ⚠️ Do not conflate with the patent full text's own typography. The specification's OCR occasionally garbles inventor-ish strings, and one search hit returned an unrelated Fujitsu oath/declaration PDF (different application, OCR-mangled inventor names such as "Tanayuki ANAHOSI"). That document is not an assignment for 09/237,904 and is excluded.

Original assignee

Family: EP 0936769 B1, DE 69930985 T2, JP 3551407 B2 — a normal Japanese-origin corporate family, consistent with a single multinational applicant prosecuting in the US/EP/JP/DE.

Assignment timeline

Chronological record (one entry — the original corporate assignment; no post-issuance assignments found):

• 1999-01-27 (executed, date per Google Patents legal-events rendering) / recorded 1999-01-27 — Reel not retrieved
  • Conveyance: Assignment of Assignors' Interest ("ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignor: IZUMI, FUTOSHI (sole inventor)
  • Assignee: FUJITSU LIMITED
  • Correspondent: not retrieved — the indices reached did not expose the recording correspondent (attorney/firm) for this entry, so I cannot state the correspondent of record. No recurrence analysis is therefore possible.
  • Context: Ordinary in-house corporate assignment executed contemporaneously with the US filing — inventor assigns to his employer, the applicant of record. Not a fire-sale, reorg, securitization, or transfer-to-asserter.

Post-issuance (2002-09-10) to expiration (2019-01-27): no recorded assignments, no security interests, no change-of-name, no merger, no license recordings surfaced. Google Patents' legal-events chain for this patent contains only the single 1999 reassignment line plus the filing/grant/expiration entries; nothing between issuance and expiry.

What this means: Under your own rule, this is the "original assignee still owns it" case. The chain never left Fujitsu. Because I could not positively render the Assignment Center caselist for app. 09/237,904, I cannot certify that no later instrument exists — but the two independent renders I did reach (Google Patents legal events; FPO/Justia bibliographic pages, both listing "Fujitsu Limited" as current assignee) show no second link.

Timeline diagram

timeline
    title Ownership of US 6449070
    1998 : JP priority filing 10-032954
    1999 : US app 09/237,904 filed
         : Izumi assigns to Fujitsu Limited
    2002 : US patent granted
    2019 : Term expiry no further assignments

NPE / troll-pattern signals

Aggregate: zero NPE signals present; one "unclear" (correspondent unretrieved); the chain is a one-link corporate assignment that never moved.

Verdict

Insufficient data (only the original assignment).

Per the defined categories, "no records, or only the original assignment" is the exact fit: the sole recorded instrument is the inventor-to-employer Assignment of Assignors' Interest executed/recorded 1999-01-27, assignor IZUMI, FUTOSHI → assignee FUJITSU LIMITED (reel/frame not retrieved; correspondent not retrieved). Fujitsu Limited is the current assignee of record and held the patent uninterrupted to its natural term expiry on 2019-01-27, so there is no shell-entity transfer, no known asserter, no cascading chained-LLC transfers, no pre-litigation transfer, no bankruptcy sale, and no privateering — likewise no affirmative defensive-aggregator acquisition. Substantively the patent reads as a never-transferred, ever-expired operating-company asset, but because I could not certify the Assignment Center caselist and could not retrieve the reel/frame or correspondent fields, I am recording it as insufficient data rather than over-claiming a clean bill of health.

Verify at:

• USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search patent 6449070 or app 09/237,904)
• Google Patents legal events — https://patents.google.com/patent/US6449070/en (see "Reassignment" / legal-events timeline)
• Family status for the EP/DE/JP siblings: EP 0936769 B1, DE 69930985 T2, JP 3551407 B2

Items I could not resolve (labeled, not guessed): the reel/frame of the 1999 assignment; the correspondent of record; whether any unrecorded license exists; and any post-2019 reassignment of the expired right (expired patents are occasionally still pooled, and the Assignment Center would be the check).

I'll research the prior-art citations for US 6,449,070, starting with the patent itself and each cited reference.

Let me research the remaining cited references — the Japanese prior art patents and the family citations.

Prior-Art Analysis — US 6,449,070 B1

Task scope note. I searched targeted mirrors of USPTO/EPO data (Google Patents, FreePatentsOnline, Justia) for the specific number 6,449,070 and for each reference cited on its face. I confirmed the target document is the Fujitsu/Izumi optical patent (not the unrelated JP 6449070 B2 of Fukubi Chemical, which is a different jurisdiction and technology — flagged earlier and again here). Below I list every cited reference exactly as it appears in the patent's citation tables and give a §102-oriented read. Where I could not retrieve the full text of a Japanese-language reference within this session, I say so explicitly rather than assert what it discloses.

0. Target document (control)

§102 framework used below: anticipation requires a single reference disclosing every element of a claim. A reference that discloses only the preamble (e.g., "a WDM optical signal with a supervisory signal") cannot anticipate an independent claim that also requires the per-wavelength vector and the diagonal path-matrix conversion. I therefore distinguish (i) references relevant only to preambles/background, (ii) references relevant to specific dependent claims, and (iii) any reference that could plausibly reach an independent claim.

1. The six references cited on the face of the patent ("Citations (6)")

Source: Google Patents citation table for US6449070B1 (also visible in the EP 0936769 counterpart).

C1 — JPS58200642A (Fujitsu), "Wavelength division multiple type optical bus system"

• Dates: JP priority 1982-05-18; published 1983-11-22.
• §102 status: printed publication >1 year before the 1998 priority → §102(b) art.
• Description (from title/classification; full JP text not retrieved): an early WDM "optical bus" — multiplexing multiple wavelength channels onto a shared bus. It establishes the environment of wavelength-division multiplexing but is not a supervisory-control/path-matrix teaching.
• Potential §102 mapping: touches only the preamble of claims 1, 18, 20, 22, 24 ("wavelength-multiplexed optical transmission system"). Cannot anticipate any claim — it lacks the supervisory control signal, the per-wavelength vector, and the diagonal path-matrix conversion. Low relevance; likely cited as background.

C2 — JPH02270036A (NEC), "Redundant switch protecting circuit"

• Dates: JP priority 1989-04-12; published 1990-11-05.
• §102 status: §102(b) art.
• Description (title-based): a protection-switching circuit for redundant (work/protection) switch paths. Not retrieved in full.
• Potential §102 mapping: relevant at most to the redundancy/branching concepts underlying claims 5–6 (OR/AND combination of received-state vectors across branched work/protection routes) and claim 19 (master/slave failover). But a "redundant switch protecting circuit" does not itself disclose a vector of per-wavelength elements, a received-state vector, or the OR/AND vector combination recited in claims 5–6. Does not anticipate those claims as written; best characterized as §103-type background for redundancy handling.

C3 — JPH06153248A (NTT), "Wavelength cross connect circuit"

• Dates: JP priority 1992-11-12; published 1994-05-31.
• §102 status: §102(b) art.
• Description (abstract retrieved): A wavelength cross-connect circuit comprising (a) optical-signal separating means that demultiplexes a WDM signal per wavelength, (b) wavelength conversion means that converts demultiplexed signals to predetermined wavelengths, (c) optical-path determining means that decides the output port from the input port and the wavelength, and (d) optical-signal combining means that combines multiple signals onto one output port. Its stated purpose is to reduce component types, circuit scale, and power attenuation from splitting.
• Potential §102 mapping: This is the most structurally on-point of the Japanese references for the optical cross-connect environment of the specification (Fig. 7/12) — i.e., the "second part"/routing context of claim 1 and the cross-connect subject matter behind claims 10–12 (retrieval/receive-direction and route switching). However, it discloses a per-wavelength switching/assignment decision, not a supervisory control signal carrying a vector of elements respectively corresponding to wavelengths that is transformed by a path matrix in which the received vector forms the diagonal elements. Cannot anticipate claims 1, 18, 20, 22, 24; relevant only to background/§103 combination.

C4 — JPH07154368A (NEC), "Optical amplifier"

• Dates: JP priority 1993-11-30; published 1995-06-16.
• §102 status: §102(b) art.
• Description: An optical amplifier (full JP text not retrieved in this session).
• Potential §102 mapping: the only cited reference whose title maps to the independent claim 14 ("An optical amplifier … automatic level control based on the number of wavelengths indicated by the received wavelength information vector"). ⚠️ Important caveat: I could not retrieve this document's disclosure, so I cannot state that it teaches automatic level control based on a count of wavelengths decoded from a supervisory vector. A generic "optical amplifier" reference would anticipate claim 14 only if it disclosed the ALC-on-wavelength-count feature — which is the novelty point of claim 14. Treat as relevant art for claim 14, not demonstrated anticipation.

C5 — US5841558A (Fujitsu; Tamura et al.), "Auxiliary-signal transmission system"

• Dates: U.S. filed 1995-02-14 (parent Ser. No. 08/388,357, now U.S. 5,625,481; the '558 is a division filed 1997-01-22); JP priority 1994-06-21; granted 1998-11-24.
• §102 status: effective U.S. filing pre-dates the '070 priority (1998-02-16) → §102(e) art (and §102(a)/(b) via its 1994 JP priority/1995 filing).
• Description (full text retrieved): Transmits a supervisory/auxiliary signal together with a main optical signal through an optically amplified link. Supervisory signaling is carried by modulating spontaneous emission of an erbium-doped fiber amplifier — e.g., via a variable optical band-pass filter whose pass band is varied by the supervisory signal, then WDM-combined with the main signal. It covers transmitting terminals, receiving terminals, and optical repeaters, and covers response information from repeaters.
• Potential §102 mapping: Strongly relevant to the concept of co-propagating a supervisory control signal with the WDM main signal — the shared premise of claim 1's first part and the claim 14 preamble ("receiving a WDM optical signal and a supervisory signal"). It does not, however, disclose (i) a supervisory signal whose payload is a vector of elements respectively corresponding to wavelengths, (ii) transformation by a diagonal path matrix, or (iii) ALC based on the number of wavelengths in that vector. Cannot anticipate claims 1, 14, 18, 20, 22, 24; it is the natural §103 base for the supervisory-channel aspect.

C6 — US6055082A (NEC), "Optical transmission system"

• Dates: JP priority 1996-11-25; U.S. filing ~1997; granted 2000-04-25.
• §102 status: U.S. filing pre-dates the '070 priority → §102(e) art (despite the post-1999 grant date).
• Description (full text retrieved): An automatic dispersion-compensation optical transmission system. A transmitting unit outputs an optical main signal (a) and a monitoring-control optical signal (b); a low-frequency multiplexing circuit superposes low-frequency pulse signals synchronized with each of them; a WDM coupler combines them. A linear repeater branches the line, WDM-separates main and monitoring signals, opto-electrically converts them, detects a phase difference between the multiplexed low-frequency tones, and selects one of several DCFs (dispersion-compensating fibers) via optical branch/confluence switches under a control circuit to optimize dispersion compensation.
• Potential §102 mapping: Relevant to the main signal + supervisory/monitoring channel co-transmission premise (claim 1 first part / claim 14 preamble) and to the general notion of a repeater acting on decoded control information. It is not a wavelength-count vector or path-matrix teaching, and its control target is dispersion, not amplifier gain vs. wavelength count. Cannot anticipate any claim of the '070 patent.

2. Family-cited references ("Family Cites Families (2)")

These appear as citations in the prosecution of the family (EP 0936769 / JP 3551407), not necessarily as §102 art against the U.S. claims, but they are the remaining cited documents:

• GB9401488D0 — a UK application publication ("Optical communications"), pre-dating the priority by ~4 years. General optical-communications background; not a WDM supervisory-vector/path-matrix disclosure. Does not anticipate.
• JPH09321701A (Fujitsu, "Optical communication system and optical amplifier") — priority 1996-05-31, published 1997-12-12 (≈13.5 months before the '070 U.S. filing) → potential §102(b) art. ⚠️ Full JP text not retrieved here. By title alone it is the most directly relevant piece for independent claim 14 (optical amplifier in an optical communication system) of all the cited art, and a natural candidate for a §103 combination with C5/C6. I cannot confirm, without its text, whether it discloses ALC keyed to a decoded wavelength count — the specific claim-14 feature.

Supplementary observation: the EP counterpart's citation list (EP 0936769 A2/B1) also surfaces JPH088835A (Fujitsu, "Optical transmission system," 1994-06-21 → 1996-01-12), which is the apparent Japanese counterpart of US 5,841,558 (C5). If you want the JP-language disclosure mapped claim-by-claim, JPH088835A and JPH09321701A are the two documents to pull in full.

3. Bottom-line §102 assessment

No cited reference anticipates any independent claim (1, 14, 18, 20, 22, 24). The reason is that every independent claim of the '070 patent requires at least one of two features that the cited art does not show in a single disclosure:

1. a supervisory control signal whose payload is a vector of elements respectively corresponding to wavelengths, and/or
2. conversion of that received vector into another vector via a path matrix in which the received vector forms the diagonal elements, forwarded to the next stage — and, for claim 14, ALC based on the number of wavelengths indicated by the vector.

Consistent with the patent having issued over this art, the six references function as follows:

Most probative art for the two commercial axes:

• Supervisory channel over an amplified WDM link → US 5,841,558 (C5) and US 6,055,082 (C6) (plus JPH088835A as the JP counterpart of C5).
• Optical amplifier / ALC subject matter of claim 14 → JPH07154368A (C4) and JPH09321701A (family cite), with C5 and C6 as §103 scaffolding.

Confidence / limitations.

• Dates, titles, assignees, and the two U.S. references' disclosures are high confidence (retrieved and corroborated).
• The JPS58200642A, JPH02270036A, JPH07154368A, JPH09321701A, and GB9401488D0 are analyzed from their citation-table metadata only — I reached the session's search limit before retrieving their full texts. Any statement that one of them "anticipates" a claim would therefore be unsupported; I have flagged those as not demonstrated rather than asserting anticipation. To complete the §102 mapping rigorously, pull the full texts of JPH07154368A and JPH09321701A (claim 14) and JPH088835A (supervisory-channel aspect), and the machine translations of JPS58200642A, JPH02270036A, and JPH06153248A.
• The 28 "Cited By" documents (OnLive/OL2, Cisco, Ciena, NEC, etc.) are forward citations and are not prior art against the '070 patent.

I'll ground this §103 analysis in the actual prior-art references listed on the patent page. Let me pull what's available on the six examiner-cited references and the two family-cited references.

Let me pull the remaining prior-art references — the two family-cited ones (Fujitsu JPH09321701A, BT GB9401488D0) and the remaining examiner citations.

Obviousness Analysis — US 6,449,070 B1 under 35 U.S.C. § 103

Note on scope and a flagged inconsistency. Two things before the analysis:

1. Date inconsistency (per the cross-reference rule). The prior sections were generated against a fetch date of 2026-10-01 (Google Patents fetch timestamp), while this task states the current date as April 26, 2026. Nothing in the §103 analysis turns on which is correct — no intervening-art question exists because every reference below predates the 1998-02-16 priority date. I proceed under the task's stated date and flag the mismatch rather than silently reconciling it.
2. Which references count as "Prior Art." Per the page's Prior Art section, that means the Citations (6) examiner references and the Family Cites Families (2). The Cited By (28) documents (Cisco, Ciena, NEC, the OnLive/OL2 cluster, etc.) cite the '070 patent and therefore post-date it — they are not §102/§103 prior art and I exclude them. This is consistent with the false-positive cleanup already done in the litigation section.

1. Governing law and the art-date gate

This is a pre-AIA case (priority 1998-02-16; filed 1999-01-27), so pre-AIA §102/§103 apply, with the KSR Int'l v. Teleflex (2007) standard governing motivation to combine. All references clear the 1998-02-16 priority gate, but by different statutory routes:

Common-ownership caveat (important for a validity opinion). Three references — US5841558A, JPS58200642A, and JPH09321701A — are Fujitsu's own art, commonly owned with the '070 patent. For an application filed in 1999, pre-AIA §103(c) as then in force excluded only §102(f)/(g) art from §103 combinations; the CREATE Act expansion to §102(e) art did not take effect until 2004. So commonly-owned §102(e) art (US5841558A) remains available for a §103 combination here. (Hedge: this is a legal-characterization point a litigator may contest; I state it as the statutory position for a 1999 filing, not as a certified holding.)

Evidentiary limitation (stated up front). I have the NTT abstract and the two US references' disclosures, but the four JP publications were retrieved only as titles/abstracts (largely in Japanese). The §103 case below is therefore a reference-level, element-mapping analysis; a full invalidity contention would require certified translations and page-and-line pin cites. Where I am inferring content from a title/abstract, I say so.

2. The invention's own admissions (useful against the claims)

The '070 specification's Background concedes that the elements of the amplifier claims were already proposed:

"A method has been proposed in which … the optical amplifier 205 is informed of the presence of the above wavelength by the control information. Hence, the optical amplifier 205 receives the control information and performs the automatic level control based on a change of the number of wavelengths." (col. describing FIG. 1)

That is an applicant admission of a prior-art ALC-by-wavelength-count method. It is powerful §103 ammunition against claim 14, and it anchors the "design incentive" rationale for claims 15–17.

3. Closest prior art

Two references carry the load, and neither alone anticipates:

• Primary A — JPH06153248A (NTT; https://patents.google.com/patent/JPH06153248A/en). A wavelength cross-connect receives WDM signals, demultiplexes by wavelength, wavelength-converts, and — critically for claim 1 — "the optical signal separating means determines the output port from the wavelength … the optical path determining means determines the output port from the input port and the wavelength of the optical signal." That is a routing map indexed by (input port, wavelength) → output port. A "path matrix with the wavelength vector as its diagonal elements" is, mathematically, a diagonal-matrix encoding of that same map. Primary A supplies the conversion/routing function and the network topology.
• Primary B — US5841558A (Fujitsu; https://patents.google.com/patent/[US5841558](/patent/US5841558)). Expressly transmits a supervisory information signal ("auxiliary signal") together with the main optical signal, through optical repeaters, and the repeater "extract[s] the auxiliary signal … and the amplified main signal is then transmitted … together with its response information corresponding to the supervisory information signal as an auxiliary signal." Primary B supplies the supervisory channel + return-path/response architecture, and it is same-assignee Fujitsu art (a strong "same field, same problem" combination signal — KSR).
• Secondary C — US6055082A (NEC; http://www.everypatent.com/comp/pat6055082.html). Transmits the optical main signal a and a separate "monitoring control optical signal b" on one WDM line through a directly-amplifying repeater. Confirms the separate supervisory-wavelength channel is a known, conventional design choice.
• Secondary D — JPH07154368A (NEC; family EP0732786A1) and JPH09321701A (Fujitsu). Optical-amplifier gain/level control art — the level-control side of claims 14–17.
• Secondary E — JPH02270036A (NEC, "Redundant switch protecting circuit"). Protection/redundancy switching art — the redundancy, branch-OR/AND, route-switch, reset, and master-handover side of claims 5–6, 11–12, 19.

4. Claim 1 — element chart and the obviousness case

Motivation to combine (KSR rationales). (i) Same field / same problem — A, B, C all address managing optical channels along a WDM link; combining the cross-connect's routing intelligence with the existing in-band supervisory channel is the natural, predictable improvement. (ii) Predictable result — encoding "which λ is in use on which port" as a bit vector and re-deriving the downstream occupancy from the node's own routing map has no unexpected mechanism; the specification itself admits the diagonal path-matrix is just the routing information ("the diagonal elements … correspond to the input terminals λ1–λN"). (iii) Design incentive / known technique — a supervisory channel that already carries control data (B, C) invites carrying wavelength-occupancy data.

Honest counter-point (weakest link). I do not see a reference that explicitly teaches propagating a per-wavelength occupancy vector downstream by diagonal-matrix multiplication. That specific formulation appears novel-ish. The §103 challenge therefore rests on characterizing the "path matrix" as a functional/mathematical expression of a routing table already known from A plus the conventional supervisory channel of B/C — a strong but not airtight combination. Expect the patentee to argue the matrix formalism adds a specific, non-routine computation (collision detection, downstream vector derivation).

5. Claim 14 (optical amplifier) — the strongest obviousness case

This claim is squarely obvious. The specification concedes that informing the amplifier of the number of wavelengths via control information and performing ALC on that count was already proposed. Claim 14 adds only the routine step of deriving the count from a bit vector — which the '070 specification itself describes as merely summing the "1" bits in hardware. Under KSR, "a court must ask whether the improvement is more than the predictable use of prior art elements according to their established functions." Counting set bits to drive a level-control loop is the paradigm of a predictable, design-choice implementation. Strong §103 rejection: B (or C) in view of D and the applicant's admission.

Claims 15–17 (hold last-known count on supervisory-only loss; zero-count when main+supervisory lost; ramp-and-rollback on saturation):

• 15 — fail-safe "hold last value" is a classic control-engineering measure; combine B/C/D with JPH02270036A (protection circuit) → predictable.
• 16 — shutting the pump/level to zero on total signal loss to protect the amplifier is standard EDFA surge/loss-of-signal protection (confirmable from the D-family amplifier art). Predictable.
• 17 — gradual level ramp with rollback on fault is ordinary rate-limiting + fault reversion; JPH02270036A supplies the protection/rollback motivation.

None of 15–17 recite a structural or algorithmic detail that escapes "predictable use of known elements."

6. Dependent claims 2–13 and independent claims 18–25

Claims 18–25 get no separate patentable weight. Claim 18 is a system claim aggregating claim-1 devices (A + B/C in a network — explicit in A). Claims 20, 22, 24 are the same subject matter recast as apparatus (transmitter/receiver), means-plus-function, and method forms; claims 21/23/25 just add the "wavelength-information vector" limitation. Recasting a device claim into apparatus/means/method form does not create patentability. (Consistent with the earlier section's flag that 20–25 may be later-added claims — either way, obviousness is unaffected.)

7. Motivation-to-combine summary (why a POSITA would combine)

A person having ordinary skill in the art of WDM transport (circa 1997–98, EDFA-based long-haul + emerging OXC/ADM networks) would combine these references because:

1. Same field, same problem — A, B, C, D, and JPH09321701A all target WDM channel management and amplifier level control; there is no field-crossing.
2. Known technique to solve the identified problem — the applicant's own background identifies the problem as getting wavelength-count information to amplifiers/cross-connects without per-λ monitors or slow software, and admits the control-information approach was already proposed. KSR holds that "if a technique has been used to improve one device … a person of ordinary skill would recognize that it would improve similar devices in the same way."
3. Predictable result — bit-vector encoding + set-bit counting + a routing-map (matrix) transform produces no unexpected behavior; every step is a known mathematical/engineering operation with a known outcome.
4. Design incentives — reusing the already-present supervisory channel (B/C) avoids new hardware; deriving downstream occupancy from the node's own routing map (A) avoids an external network-management round trip.

8. Countervailing considerations (what a defense would say)

• No explicit "diagonal path matrix" teaching — the closest single reference (A) discloses a routing map but not its diagonal-matrix, propagated-vector formulation; a defense will argue claim 1's specific mechanism is not taught, only the environment is.
• B's "response information" ≠ a per-wavelength received-state vector — B's auxiliary signaling is generic; mapping it to claims 4/7 requires a motivation that a defense may call hindsight.
• No objective indicia of record — per the earlier sections, no litigation/PTAB activity for US 6,449,070 was found and the patent expired 2019-01-27. There is consequently no commercial-success, licensing, or long-felt-need evidence to rebut obviousness — but absence of secondary considerations is not affirmative proof of obviousness, and I do not overstate it.
• Enablement of the JP art — the four JP references were reviewed only at title/abstract level; a real invalidity opinion must confirm they are §102(b) publications with enabling disclosures.

9. Bottom line and confidence

• Claim 14 is the most clearly obvious — defeated by US5841558A or US6055082A in view of the amplifier-control art (JPH07154368A / JPH09321701A) and, decisively, the applicant's own admitted prior art of ALC-by-wavelength-count. High confidence.
• Claims 15–17 — obvious as routine fail-safe/rollback engineering over the claim-14 combination plus JPH02270036A. Moderate-to-high confidence.
• Claim 1 (and its 20/22/24 recasts, and claim 18) — prima facie obvious over JPH06153248A + US5841558A (or US6055082A), with the "path matrix" characterized as a mathematical restatement of the known (input-port, λ)→output-port routing map. Moderate confidence — the specific diagonal-matrix, downstream-propagated-vector formulation is the vulnerable point and is where the patentee would most plausibly distinguish.
• Claims 2–13, 19 — each maps to a routine combination of A/B/C/D with the NEC protection-circuit teaching (JPH02270036A). Moderate confidence, with claims 8–13 (retrieval/counter/ID/memory) the most dependent on a "conventional signaling" rationale.

Key evidentiary caveat: this is a reference-level analysis built on the page's Prior Art section (Citations (6) + Family Cites (2)). Full-text translations of JPS58200642A, JPH02270036A, and JPH09321701A should be obtained before any contention is filed, since I relied on titles/abstracts for those three.