US 11095387

I'll search for the specific patent number and any related 2026 litigation activity.

Let me search more specifically for litigation involving this patent and its current owner.

Let me look specifically at whether the '387 patent itself appears in any 2026 docket, distinct from its continuations.

US Patent 11,095,387 (US11095387B2) — Analyst Summary

Confidence note up front: The bibliographic and claim details below come from the patent's own full text (authoritative). The litigation/ownership findings come from web search and are reported as found, with an explicit flag where I could not confirm a match. I did not locate any CAFC 2026 docket naming US 11,095,387; treat the litigation portion as partially unverified.

Bibliographic data (literal, as listed)

Ownership chain (as recorded): NEC Corp → NEC Asia Pacific Pte Ltd. (2024-03-21) → IP WAVE PTE. LTD. (2024-10-03) → PLS XLIV LLC (2025-03-20) → RADIANT PATENTS LLC (2025-10-01, "change of name"). A subsequent 2025-11-13 corrective assignment lists PLS XLIV LLC again. Note also a Google Patents "Other versions" / earliest-family entry: PCT/JP2016/000346 (WO2016121361A1) and the parent US 15/545,588 (US 10,374,740 B2). The '387 is a continuation of 15/545,588; 15/545,588 was itself a National Stage of PCT/JP2016/000346, claiming priority from JP 2015-013258 (2015-01-27).

Abstract (as published)

Provided is a wavelength path communication node device with no collision of wavelengths and routes, capable of outputting arbitrary wavelengths, and capable of outputting them to arbitrary routes. An add/drop multiplexer (11) includes a communication unit (101) that communicates an optical signal with at least one client device and at least one network and a control unit (102) that indicates a transfer destination of the optical signal according to an attribute of the received optical signal to the communication unit (101). The control unit (102) indicates an attenuation amount of the optical signal to the communication unit (101) for each connected device. When a connected device is changed, the control unit (102) instructs the communication unit (101) to change the attenuation amount. The communication unit (101) attenuates the optical signal with the attenuation amount indicated by the control unit (102) and transfers the attenuated optical signal to a transfer destination.

Plain-language overview of the independent claims

The patent has three independent claims: claim 1 (apparatus), claim 5 (system), and claim 9 (method). They are essentially the same invention expressed in three statutory categories.

Claim 1 — Optical communication apparatus

• A demultiplexer splits a wavelength-multiplexed signal into multiple optical signals (one per wavelength).
• An attenuator attenuates one of those optical signals by an amount that corresponds to a first optical device, and the attenuated signal is sent to that first optical device.
• A controller sends "attenuation information" for the first optical device to the attenuator.
• The key limitation: when the destination is switched from the first optical device to a second optical device, the controller sends the attenuation information corresponding to the second optical device, and the attenuator then attenuates with the amount matching the second device.
• In short: per-wavelength, per-destination adjustable attenuation that is re-programmed when you repoint the output.

Claim 5 — Optical communication system

• Same inventive core, but claimed as a system comprising a network management apparatus that controls the optical communication apparatus, plus the apparatus itself with the same demultiplexer / attenuator / controller and the same "switch destination → change attenuation information" behavior.
• The management device aspect ties to the fourth embodiment (an NMS/EMS that registers "connection information" and pushes connected-device changes to the add/drop multiplexer).

Claim 9 — Optical communication method

• The method counterpart: demultiplex the WDM signal; attenuate an optical signal with an amount corresponding to a first device; send the attenuated signal to the first device; and attenuate with an amount corresponding to a second device when the destination is switched from the first to the second device.
• Notably, claim 9 as written does not explicitly recite the "adding/inserting" step or a controller — it is the attenuation-switching method.

Selected dependent claims of interest

• Claims 2 & 10: controller receives connection information from a management device and identifies the attenuation information based on it.
• Claims 3 & 11: same, but specifically where the connection information indicates the switch from first to second device.
• Claims 4, 8, 12: the destination is a client device or a network.
• Claims 13, 14, 15: the attenuation amount is changed depending on whether the optical device is a cross-connect or a client device — this is the claim set that most directly maps to FIGS. 6 and 8 (the "XC" vs "DROP"/"ADD" port-repurposing examples).

Technical context (from the specification)

The stated problem: conventional ROADM-style add/drop multiplexers use a fixed attenuation amount, so the device connected to a given port cannot be changed. The invention makes attenuation variable and destination-dependent, so an arbitrary device (cross-connect, network, client/transponder, terminal station) can be hung off an arbitrary port. Embodiments cover: (1) a general add/drop multiplexer with communication + control units; (2) a drop unit / add unit split; (3) WSS implementations (drop-side WSS 200 and add-side WSS 300) with per-wavelength variable optical attenuators and optical switches; and (4) a management device (NMS/EMS) that pushes connection changes to the multiplexer. The specification also notes the attenuation amount may be zero, and that an amplifier may replace the attenuator (attenuation amount repurposed as amplification amount).

2026 litigation / CAFC search results

What I found (with a caveat):

• Radiant Patents LLC is actively asserting optical-network patents in 2026. A search result describes Radiant Patents LLC v. Nokia Corp, No. 2:26-cv-00074 (E.D. Tex., filed 01/26/2026), cause of action "Patent Infringement," status open. Another source lists asserted patents including '431, '789, '095, '188, '366 ("+3 more"). Sources: https://ai-lab.exparte.com/case/dct/txed/2:26-cv-00074/radiant-patents-llc-v-nokia-corp and https://ai-lab.exparte.com/case/dct/txed/2:26-cv-00074/doc/analysis/16
• Unified Patents reports prior-art challenges to Radiant Patents' US 10,623,095 and US 11,431,431, and describes the '431 patent as "an optical wavelength-path communication node, essentially an add/drop multiplexer for optical networks." US 11,431,431 is the direct continuation of US 11,095,387 (same priority 2015-01-27, per the family data). Source: https://www.unifiedpatents.com/insights/2026/5/5/radiant-patents-optical-networking-patents-prior-art-found

What I could NOT confirm:

• I found no search result that explicitly names US 11,095,387 (the '387) as an asserted patent in the E.D. Tex. Nokia case. The asserted-family member that clearly appears is US 11,431,431 (the '431), a continuation of the '387. Source-asserted patent lists I retrieved did not resolve the "+3 more."
• I found no CAFC 2026 docket (appeal number) naming US 11,095,387. Several 2026 CAFC summaries surfaced in the search (e.g., AGI SureTrack v. Farmers Edge, Arendi/Oath, Enviro Tech v. Safe Foods, NantWorks v. Niantic, REGENXBIO v. Sarepta), but none involved the '387.

Because your instruction is to interpret identifiers literally and prefer search results over training data: I am reporting the E.D. Tex. case and the '431 prior-art activity as adjacent evidence, not as proof that the '387 itself is in suit. If you need certainty on whether the '387 patent is a named asserted patent in 2:26-cv-00074, that requires pulling the actual complaint (the "+3 more" patents) from the docket — I could not verify it from the available results.

Bottom line

US 11,095,387 is a NEC-originated (inventor Yurie Matsuyama), post-AIA continuation patent on variable per-wavelength attenuation that is reconfigured when the optical destination changes, claimed as apparatus (cl. 1), system (cl. 5), and method (cl. 9), with dependent claims distinguishing cross-connect vs. client device attenuation. It was owned by Radiant Patents LLC (an MVRE LLC NPE entity, f/k/a PLS XLIV LLC) at last record, and its direct continuation US 11,431,431 is a confirmed 2026 assertion/prior-art target against Nokia in E.D. Tex. I have no confirmed record of the '387 patent itself in a 2026 CAFC docket or as a named asserted patent, and I flag that as an open item rather than asserting it.

I'll search specifically for litigation naming US 11,095,387.

Let me run more targeted searches on the specific patent number and its litigation portals.

Let me run additional targeted searches, including the Unified Patents litigation portal and docket-level terms.

Litigation Search Results — US Patent 11,095,387 (US11095387B2)

Bottom line: I found no litigation that names US 11,095,387 as an asserted patent. Interpreting the identifier literally, I could not confirm a single case, complaint, IPR, or CAFC docket in which US 11,095,387 itself is the patent-in-suit. I want to be explicit about that rather than pad the answer with the closely related family activity, which is what dominates the search results.

What the searches returned for the specific number

No confirmed hits for "11095387" / "11,095,387." Searches keyed to the exact number returned:

• A UK Companies House gazette entry containing the string "11095387" as a company registration number ("ONCE UPON A TIME ILLUSTRATIONS 11095387 LIMITED") — a false positive, not the patent.
• Various unrelated patent complaints ('617, '248, '439, '243 patents, Realtime Data, AuthenTec, Mullen, etc.) that surfaced only because the search engine matched on docket/patent-number formatting, not the identifier.
• No PACER, CourtListener, Docket Alarm, Darts-IP, or unifiedpatents.com result naming US 11,095,387 as an asserted patent.

The Google Patents page for the '387 carries a "Family has litigation" / Darts-IP banner linking to https://patents.darts-ip.com/?family=56542984, i.e., the litigation flag is attached to family ID 56542984, not necessarily to the '387 member itself. A family-level flag does not establish that the '387 was named.

Adjacent activity — clearly labeled as NOT the '387

So you are not misled into treating these as '387 cases, here is the nearby activity I did find, with the patent numbers as literally listed:

The relevant caveat: US 11,431,431 (the '431) is a continuation of US 11,095,387 (same priority 2015-01-27, same family 56542984). The '431 — not the '387 — is the family member described by Unified Patents as "an optical wavelength-path communication node, essentially an add/drop multiplexer for optical networks," and it is a confirmed 2026 assertion/prior-art target. That is cousin activity, not '387 activity.

CAFC

I found no 2026 (or earlier) Federal Circuit docket naming US 11,095,387. Several 2026 CAFC matters surfaced in searching (e.g., AGI SureTrack v. Farmers Edge, NandWorks v. Niantic, REGENXBIO v. Sarepta, Enviro Tech v. Safe Foods), none of which involve the '387.

Classification of the requested fields

Open items / where certainty would require a primary source

I could not verify the "+3 more" patents in the Radiant Patents v. Nokia complaint. If the '387 is among them, it would be a named plaintiff-side assertion by Radiant Patents LLC (f/k/a PLS XLIV LLC, an MVRE LLC entity) against Nokia Corp in E.D. Tex. — but I have no source that states this, so I am recording it as an unverified possibility, not a finding. Confirming it requires pulling the actual 2:26-cv-00074 complaint (¶ references) or the Darts-IP family-56542984 litigation record at https://patents.darts-ip.com/?family=56542984.

Final answer, stated plainly: No litigation is known to me that specifically names US 11,095,387. The 2026 assertion activity I located involves sibling family member US 11,431,431 (asserted against Nokia; prior art charted by Unified Patents) and other Radiant Patents optical patents ('095, '188, '789) — none of which is US 11,095,387.

I'll verify the "no PTAB activity" default against live sources before reporting.

Let me run a few more targeted checks to confirm the absence of any AIA trial on this patent or its family.

Proceedings overview

Total AIA trial proceedings on US 11,095,387: 0. The structured "PTAB proceedings on file" block drawn from the USPTO Open Data Portal returns no IPR, PGR, or CBM proceeding for this patent, and my independent web checks surfaced none either. So the breakdown is: active — 0; claims invalidated — 0; claims sustained — 0; settled — 0; institution denied — 0. Bottom-line defensive posture: the '387 has never been tested at the PTAB, and all 15 claims — including independent claims 1 (apparatus), 5 (system), and 9 (method) — remain live and unchallenged. Unlike a "hardened" patent that has survived IPRs, or a "dead" one whose claims were canceled, this is an untested patent: the claims are intact because nobody has attacked them, not because they were attacked and survived. For a defendant, that cuts both ways — there is no favorable FWD to point to, but there is also no § 315(e)(2) estoppel, no adverse claim construction, and no patent-owner-established record shrinking the claim scope.

Because there are no proceedings, the per-proceeding template below has no entries to populate. I set out instead what I verified, what I could not verify, and what the absence means.

---

No proceedings located — evidence and caveats

• Canonical source (structured block): "The USPTO ODP API returns no AIA trial proceedings for this patent as of the most recent ingest." I treat this as the authoritative answer for US 11,095,387.
• Independent web checks: Searches for the '387 in combination with "IPR," "PTAB," and "Radiant Patents" returned no petition, institution decision, FWD, or appeal naming this patent. The hits that did surface were uninvolved third-party matters (Ameranth, Capella Photonics, CIG USA/Applied Optoelectronics, ASUS/Nokia video-coding, etc.) and are not proceedings on this patent. I did not auto-attribute any of them.
• Adjacent-but-distinct activity I did find (do not confuse with a proceeding on the '387):
  • Unified Patents is running PATROLL prior-art bounty contests — these are crowdsourced prior-art searches, not AIA trials — against sibling/related Radiant Patents (f/k/a PLS XLIV LLC) assets: US 10,623,095, US 10,404,366, US 10,924,188, and US 11,431,431. The '431 is the direct continuation of the '387. Source: https://www.unifiedpatents.com/insights/2026/5/5/radiant-patents-optical-networking-patents-prior-art-found and https://patroll.unifiedpatents.com/contests/fJpp6n7SYsQis8RvA.
  • Radiant Patents is asserting an optical patent portfolio against Nokia in Radiant Patents LLC v. Nokia Corp., No. 2:26-cv-00074 (E.D. Tex., filed 2026-01-26). Consistent with the earlier section of this analysis, I could not confirm the '387 itself is a named asserted patent in that complaint.
• What a PATROLL contest does and does not mean: a bounty on the '431 signals that a defensive aggregator believes there is real prior art in this family and is stockpiling it. It is not a filed IPR and creates no estoppel and no PTAB record. But it is a leading indicator that a petition on a family member — possibly the '387 — may follow once the Nokia litigation sets the § 315(b) clock.

---

Strategic summary

Claim status. All claims are UNTESTED. Nothing is canceled; nothing is confirmed. Claim 1 (demultiplexer + destination-corresponding attenuator + controller sending attenuation information that is re-sent when the destination switches from a first to a second optical device), claim 5 (same core plus a network management apparatus), claim 9 (method counterpart), and dependents 2–4, 6–8, 10–12, and especially 13–15 (attenuation amount changed depending on whether the device is a cross-connect or a client device) remain as issued. For any defendant, all 15 claims are fair game and none has been narrowed by prosecution-style amendment or PTAB cancellation.

Estoppel landscape. Because no IPR has reached an FWD, there is no § 315(e)(2) estoppel against anyone on the '387. A defendant asserting invalidity in district court faces no IPR-based estoppel bar, and any third party is free to file a first-ever IPR on the '387 provided it satisfies § 315(b) (one year from service of an infringement complaint on that petitioner or its privy) and § 315(a) (not filed after the petitioner's own DJ action). Practically, the entire universe of prior-art patents and printed publications is available — none has been "used up" by a prior petitioner. That is the single most defendant-favorable fact about this patent.

Pattern signals. No petitioner has filed on this patent — not once. The recurring actor in this family is Unified Patents (a defensive aggregator/NPE-opposition organization) via PATROLL, but only at the prior-art-hunting stage, and only on siblings ('095, '366, '188, '431), not the '387. The patent owner (Radiant Patents LLC, an MVRE LLC entity) has no PTAB appeal history on the '387, because there is no PTAB history to appeal. There is a live 2026 district-court campaign (Nokia, E.D. Tex.), which is the typical precursor to IPR filings — so the "zero proceedings" state is likely a snapshot of a patent that is asserted but not yet challenged, rather than a patent that challengers have examined and declined to attack.

Two honest flags. (1) The ODP block is the authority; if the ODP ingest lags, a very recently filed petition (e.g., an IPR2026-xxxxx) might not yet appear — I found no such filing in web sources, but I cannot rule out a paper filed within the last few days. (2) The '431 ('387's continuation) is the family member confirmed in the 2026 assertion/prior-art crosshairs; do not let anyone tell you a proceeding on the '431 is a proceeding on the '387 — they are separate patents with separate claims and separate estoppel effects (estoppel runs patent-by-patent, claim-by-claim).

---

Recommended next steps

If you are a defendant facing assertion of US 11,095,387:

• There is no FWD to cite. Do not represent to a court or adversary that any claim of the '387 has been canceled or held unpatentable — it has not. The prior-art charts discussed above are for the '095, '366, '188, and '431, not the '387; the '431 charts may be technically adjacent (same family, same specification) but they do not establish anything about the '387's claims.
• Evaluate the § 315(b) clock now. If you have been served with a complaint alleging infringement of the '387, you have one year to file an IPR. That deadline, not the ODP listing, governs your window.
• Exploit the clean slate. With no prior petitioner and therefore no § 315(e)(2) estoppel, you (or a defensive aggregator on your behalf) can challenge the '387 on any § 102/§ 103 ground supported by patents and printed publications. The '387's own prosecution history and the four references cited on its face (JP 2011-040997; JP 2013-123205; JP 2013-531909; WO 2009/145118) are the natural starting point, and the prior art that the family's PATROLL charts target on the '431 may be adaptable to the '387's apparatus/system/method claims.
• Watch for a first petition. Given the active 2026 Nokia litigation and the Unified PATROLL activity on siblings, the probability of a first IPR on the '387 (or on the '431) is non-trivial. Monitor the PTAB E2E docket and Unified's filings; if an IPR on the '431 is filed, read its grounds closely — and, if you are a defendant, weigh whether SIPR/joinder or your own parallel petition makes sense before the '387's § 315(b) year runs.
• If you want certainty, pull the complaint. The open item from the earlier section stands: to confirm whether the '387 is a named asserted patent in Radiant Patents LLC v. Nokia Corp., 2:26-cv-00074, the complaint's patent list must be read directly. I could not verify it from available search results and will not assert it.

Bottom line for a defendant: No PTAB proceedings exist on US 11,095,387. All 15 claims are untested, uncanceled, and unconstrained by any prior PTAB record — and you face no § 315(e)(2) estoppel. The absence of IPRs is not evidence the patent is strong; it is evidence no one has yet fired. Given the active 2026 assertion campaign and sibling-level prior-art activity, treat the "zero" as a temporary, pre-challenge state and move on your § 315(b) timeline accordingly.

I'll search for the assignment records and correspondent details before reconstructing the chain.

Excellent — I found a related assignment record with reel/frame and correspondent data. Let me dig deeper on the specific '387 records and the PLS XLIV / Radiant legs.

Assignment & Ownership Analysis — US 11,095,387 B2

⚠️ Retrieval limitation stated plainly up front. I was able to retrieve the structure and dates of every recorded reassignment in this chain (via Google Patents legal events, which are derived from USPTO Assignment records), but I was only able to retrieve actual reel/frame numbers for the parallel records on a sibling family patent (US 9,482,632), which went through the same assignor chain but a different LLC branch (Cloud Byte LLC rather than PLS XLIV LLC). I did not retrieve the reel/frame numbers for the '387's own recordings. I am not going to invent them. Every reel/frame below is labeled with the record it actually belongs to. The '387's own reel/frame entries remain an open verification item at https://assignment.uspto.gov/patent/index.html (search patent 11095387).

---

Inventors

Pattern assessment: No unusual inventor pattern detected. There is a single inventor, so the "all inventors departed within 12 months of filing" tell is inapplicable. I found no evidence of inventor departure, no inventor-to-assignee reversion, and no inventor-named assignment record (the chain begins with NEC Corporation as assignor, which implies an omnibus employee invention assignment pre-dating the 2015-01-27 priority filing). The priority document is JP 2015-013258 (JP 6395219 B2), and the PCT was filed by NEC Corp as applicant (PCT/JP2016/000346). Inventor signal: not present.

---

Original assignee

NEC Corporation (Tokyo, Japan) — named on the issued patent as original assignee.

• Primary line of business: diversified ICT — telecom network infrastructure (including WDM/optical transport systems), IT services, semiconductor and display in earlier eras.
• Did they ship a product embodying the claims? Likely yes, in the directly-relevant sense: the specification's FIG. 9 wavelength path communication node device (WXC 401/402, WSS 401A–402B, add/drop, transponder aggregator) is the architecture of a commercial CDC-ROADM line card. NEC sold WDM/optical transport systems using reconfigurable optical add/drop multiplexing. Caveat: I did not locate a specific NEC product datasheet or a § 287(a) marking record tying a SKU to these claims; treat "shipped a product" as strongly plausible but not documentary-verified here.
• Current status: Operating company, but no longer owns this patent. NEC Corporation remains an active global vendor. It divested this patent family through NEC Asia Pacific Pte Ltd. Note that NEC's optical transport business has been through repeated restructuring; I found no bankruptcy for NEC Corporation — the divestiture was a corporate portfolio monetization, not an insolvency sale.
• Family relationship to patent: The '387 is a continuation of US 15/545,588 (US 10,374,740 B2), which is itself the National Stage of PCT/JP2016/000346 (WO2016121361A1). Family ID 56542984 also contains US 11,431,431 ('431) and US 11,973,579 ('789).

---

Assignment timeline

The '387's recorded chain (dates are recorded dates as surfaced on Google Patents legal events; execution dates for the '387's own instruments were not retrievable):

• Executed date not retrieved / recorded 2024-03-21 — Reel not retrieved (parallel record on US 9,482,632 is 066124/0752, exec. 2023-12-13, rec. 2023-12-22, 90 properties)

  • Conveyance: Assignment (Assignment of Assignors Interest, see document for details)
  • Assignor: NEC CORPORATION
  • Assignee: NEC ASIA PACIFIC PTE LTD.
  • Correspondent: Not retrieved for the '387. On the parallel US 9,482,632 record this leg shows the assignee address 80 Bendemeer Road #05-01/02, Singapore 339949; no correspondent attorney name was retrieved for this leg.
  • Context: Internal corporate reorganization / intra-group transfer to NEC's Singapore regional holding entity — the first step of a staged divestiture.

• Executed date not retrieved / recorded 2024-10-03 — Reel not retrieved (parallel record on US 9,482,632 is 066376/0276, exec. 2024-01-18, rec. 2024-01-27)

  • Conveyance: Assignment
  • Assignor: NEC ASIA PACIFIC PTE LTD.
  • Assignee: IP WAVE PTE. LTD. (recorded address per the parallel record: 20 Collyer Quay, #23-01, Collayer Quay, Singapore 049319)
  • Correspondent: SUGHRUE MION, PLLC, 2000 Pennsylvania Avenue NW, Suite 9000, Washington, DC 20006 (appears on the parallel US 9,482,632 recording — see signal 3)
  • Context: Sale/monetization to a patent-holding intermediary — first transfer-to-non-practicing-entity step.

• Executed date not retrieved / recorded 2025-03-20 — Reel not retrieved

  • Conveyance: Assignment
  • Assignor: IP WAVE PTE. LTD.
  • Assignee: PLS XLIV LLC (Delaware)
  • Correspondent: Not retrieved for the '387. On the parallel US 9,482,632 record, the equivalent IP-WAVE-to-LLC leg (067944/0332, exec. 2024-03-05, rec. 2024-06-27) names correspondent MR. DEREK M. RABY, 2000 Pennsylvania Avenue NW, Suite 9000, Washington, DC 20006 — i.e., the same street address as Sughrue Mion, PLLC.
  • Context: Transfer-to-asserter. IP WAVE PTE LTD disgorged NEC's optical portfolio into multiple sibling Delaware LLCs in this same window (see signal 4).

• Executed date not retrieved / recorded 2025-10-01 — Reel not retrieved

  • Conveyance: Change of Name (per the record)
  • Assignor: PLS XLIV LLC
  • Assignee: RADIANT PATENTS LLC
  • Correspondent: Not retrieved.
  • Context: Change of name only — the same legal entity rebranded. This is the entity Unified Patents publicly identifies as "Radiant Patents LLC (f/k/a PLS XLIV LLC), an NPE and entity of MVRE LLC."

• Corrective instrument / recorded 2025-11-13 — Reel 070575/0077 cited within the corrective text (the corrective document itself references the reel it corrects)

  • Conveyance: Corrective Assignment ("Corrective assignment to correct the application and patent numbers for patent numbers 10924188, 10581518, 10200117 and 12009865, which were erroneously recorded as application numbers")
  • Assignor: IP WAVE PTE. LTD.
  • Assignee: PLS XLIV LLC (per the listing)
  • Correspondent: Not retrieved.
  • Context: Housekeeping correction of erroneous serial/patent-number listings in an IP WAVE recording — itself a tell that these transfers were bulk-recorded in volume (see signal 4).

If you need the '387's own reel/frame: it is directly retrievable by patent number at the Assignment Center; the dates above should be used to disambiguate the correct instruments from the ~90-property bulk recordings.

---

Timeline diagram

timeline
    title Ownership of US 11095387
    2015 : Priority filed by NEC Corp
    2016 : PCT filed by NEC Corp
    2019 : US continuation filed
    2021 : Patent issued to NEC Corp
    2024 : Assigned to NEC Asia Pacific Pte Ltd
         : Assigned to IP WAVE PTE LTD
    2025 : Assigned to PLS XLIV LLC
         : Renamed Radiant Patents LLC
         : Asserted against Nokia
    2026 : Nokia suit in E D Tex

---

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT (strong).
NEC Corporation → NEC Asia Pacific Pte Ltd → IP WAVE PTE. LTD. → PLS XLIV LLC → Radiant Patents LLC (recorded 2024-03-21, 2024-10-03, 2025-03-20, 2025-10-01 respectively). The terminal assignee carries the classic suffix ("Patents LLC"), Unified Patents confirms Radiant Patents LLC is "an NPE and entity of MVRE LLC" (https://www.unifiedpatents.com/insights/2026/4/29/4000-awarded-for-radiant-patents-optical-patents-prior-art), and the IP-WAVE-to-LLC legs on the parallel sibling chain land at bulk-registered Delaware addresses (CLOUD BYTE LLC, 850 Library Ave., Suite 204, Newark, Delaware 19711 per reel 067944/0332). Multiple differently-named LLCs were seeded from the same IP WAVE source, which is inconsistent with an operating business and consistent with single-purpose assertion vehicles.

2. Known asserter in the chain — PRESENT (strong).
Radiant Patents LLC (f/k/a PLS XLIV LLC) appears as an active plaintiff: Radiant Patents has asserted optical-networking patents against Nokia, and Unified Patents ran PATROLL prior-art contests on its US 10,623,095 and US 11,431,431 (the '431 being the '387's direct continuation). Sources: https://www.unifiedpatents.com/insights/2026/5/5/radiant-patents-optical-networking-patents-prior-art-found and https://www.unifiedpatents.com/insights/tag/New+Contests. Note: the '387 itself was not confirmed as a named asserted patent in the 2026 E.D. Tex. Radiant Patents v. Nokia docket — only its continuation was confirmed. This signal therefore rests on the assignee entity, not on the '387 being in suit.

3. Repeat correspondent across the chain — PRESENT (moderate; same-address recurrence).
On the parallel sibling recordings, the same street address — 2000 Pennsylvania Avenue NW, Suite 9000, Washington, DC 20006 — carries two different recorded correspondent names: SUGHRUE MION, PLLC on reel 066376/0276 (NEC Asia Pacific → IP WAVE) and MR. DEREK M. RABY on reel 067944/0332 (IP WAVE → Cloud Byte LLC). A recording firm and an individual filer sharing an identical suite is a recurrence tell rather than two independent filers. Caveat I will not paper over: these correspondents are documented on the sibling patent's records, not on the '387's own records, and I could not verify the '387's correspondent of record. A single appearance is not a finding — what I have is a same-address recurrence on a parallel chain, which I flag as corroborating rather than dispositive.

4. Cascading transfers — PRESENT (strong).
Four recorded transfers of the same asset within roughly 18 months (2024-03-21 → 2025-10-01), through four successively less substantial entities (operating parent → regional holding co → Singapore patent intermediary → Delaware LLC → renamed LLC). Corroborating volume indicators: the parallel NEC→NEC Asia Pacific recording covered 90 properties (reel 066124/0752), other NEC recordings covered 179 properties, and the 2025-11-13 instrument is a corrective assignment fixing erroneous patent/application numbers in an IP WAVE recording — the signature of bulk, assembly-line recordation. Multiple sibling LLCs (Cloud Byte LLC; PLS XLIV LLC/Radiant Patents) were populated from the same IP WAVE source.

5. Pre-litigation transfer — PRESENT (moderate).
IP WAVE → PLS XLIV LLC recorded 2025-03-20, followed within months by Radiant Patents' assertion activity against Nokia (Unified's PATROLL contests on the '095 and '431 patents closed 2025-09-30) and, per prior analysis, the E.D. Tex. Nokia complaint filed 2026-01-26. The '387's chain terminated at the asserting entity roughly 10 months before the 2026 filing and before the 2025 assertion activity. This is timing-consistent with assertion-enablement (clean standing record + venue setup), though for the '387 specifically it is inference from dates, not from a complaint naming the '387.

6. Bankruptcy fire-sale — NOT PRESENT.
I found no Chapter 7/11 filing by NEC Corporation or NEC Asia Pacific Pte Ltd, and no § 363 sale. This divestiture was a solvent portfolio monetization, not an insolvency liquidation. Explicitly distinguishing this from the Kodak/Nortel/Polaroid pattern.

7. Privateering — UNCLEAR (leaning present).
NEC transferred to IP WAVE PTE LTD, which seeded multiple assertion LLCs, and those entities are asserting (against Nokia — a direct NEC competitor in telecom infrastructure). That is structurally the privateering pattern. However, I found no SEC filing, no license-back, and no public statement by NEC confirming a retained economic interest or a "license on behalf of NEC" arrangement. The behavior fits; the documentary proof of NEC's retained stake does not exist in what I retrieved. Unclear.

8. Defensive aggregator (anti-NPE) — NOT PRESENT.
The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. Unified Patents is on the opposing side, running PATROLL contests against this portfolio. The patent is actively weaponized, not neutralized.

Cross-reference flag (contradiction check vs. prior section): The previously generated section correctly noted that the '387 is not confirmed in the 2026 E.D. Tex. Nokia case, and that the confirmed asserted family member is US 11,431,431. My findings agree and add the reason: the '387 and '431 sit in the same family (ID 56542984) under the same assignee, so the asserting entity is the same — Radiant Patents LLC — even though the specific patent-in-suit is the '431. There is no contradiction, but there is one date tension worth logging: Unified's PATROLL activity on this portfolio closed 2025-09-30, while the reported Nokia case was filed 2026-01-26 — consistent with two waves of assertion (a 2025 wave preceding the 2026 complaint), which I could not fully resolve from available sources.

---

Verdict

NPE — high confidence

Justification (2–3 sentences, evidence-cited): The recorded chain moves the '387 from an operating vendor (NEC Corporation) through an intra-group holding entity (NEC Asia Pacific Pte Ltd, recorded 2024-03-21) and a Singapore patent intermediary (IP WAVE PTE. LTD., recorded 2024-10-03) into a Delaware assertion vehicle (PLS XLIV LLC, recorded 2025-03-20), which was then simply renamed Radiant Patents LLC (recorded 2025-10-01) — four transfers in ~18 months with no evidence of any product-bearing assignee at any post-NEC step. The terminal assignee is publicly identified by Unified Patents as "an NPE and entity of MVRE LLC" and is actively asserting (Nokia; PATROLL contests on the sibling '431 and the '095), satisfying signals 1, 2, 4, and 5; the only signals not present are bankruptcy fire-sale and defensive-aggregator neutralization. The single residual uncertainty is that I retrieved reel/frame for the parallel sibling recordings rather than the '387's own (the '387's legs ran 2024-03-21 / 2024-10-03 / 2025-03-20 / 2025-10-01, and the sibling's parallel legs were reels 066124/0752, 066376/0276, and 067944/0332) — the chain shape and assignee identities are solid; the exact '387 reel/frame should be confirmed at the link below.

Verify at: https://assignment.uspto.gov/patent/index.html (search patent number 11095387) — or the newer front end at https://assignmentcenter.uspto.gov/

Open items I did not fabricate and could not confirm: (a) the '387's own reel/frame numbers; (b) the '387's own correspondent of record; (c) the '387's execution dates (vs. recordation dates); (d) whether NEC retained a license-back or economic interest in the IP WAVE/PLS XLIV transfer; (e) whether the '387 itself (not just its continuation '431) is a named patent in Radiant Patents LLC v. Nokia Corp, No. 2:26-cv-00074 (E.D. Tex.).

I'll research US 11,095,387 in the USPTO and analyze its cited prior art.

Prior-Art Analysis — US 11,095,387 B2

Search status / honesty note up front. I confirmed the patent identity and pulled the full reference list of record from the patent text you provided (authoritative). I completed targeted verification searches on the highest-relevance citations. I hit my tool-call limit before I could independently open every one of the 51 cited references, so for the lower-relevance references the "description" below is drawn from the patent's own face-of-record titles/assignees plus general knowledge, and I flag where I could not independently verify content. All identifiers are reproduced literally from the patent — I have not corrected any numbers.

---

1. Patent under analysis (confirmed)

Claim architecture (the target of any § 102 analysis): independent claim 1 (apparatus: demux + attenuator whose attenuation corresponds to a first device + controller that re-sends attenuation info for a second device when the destination is switched), claim 5 (same, as a system with a network management apparatus), claim 9 (same, as a method). Dependents: 2/3 & 10/11 (management-device connection information), 4/8/12 (destination = client device or network), 13/14/15 (attenuation changed depending on whether the device is a cross-connect or a client device).

---

2. Methodology caveat on "anticipates under § 102"

This is a granted patent. Every reference below was already before the examiner (or is family art), and the claims issued over them. So "potentially anticipates" here means: would this reference, alone, disclose every element of a given claim as construed — i.e., a reexam/invalidity threat assessment, not a statement that the reference destroyed novelty during prosecution. I mark the four strongest § 102 candidates explicitly. § 103 combinations are noted only where a single reference is plainly insufficient.

---

3. Complete reference list of record (51)

Dates are the priority/filing date and the publication date as listed for each citation.

Note: the specification separately names four references as "Patent Literature 1–4" (PL1 = JP2011040997A; PL2 = JP2013123205A; PL3 = JP2013531909A; PL4 = WO2009145118A1). PL1–PL4 are the references the applicant itself characterized as the closest ROADM background.

---

4. § 102 relevance — ranked

Tier A — strongest potential anticipatory references

A1. WO2009145118A1 — NEC, "Wavelength path communication node device, wavelength path communication control method, program, and recording medium" (filed 2008-05-26; pub. 2009-12-03).
This is the single most dangerous reference. It is same-assignee (NEC) as the '387, it is the specification's own Patent Literature 4, and it independently appears in the examiner's citation list. It describes a wavelength-path node in which optical signals at different wavelengths transmitted from a route are selectively output to a WSS, an aggregator, or an ODMUX/OMUX, "and then optimal ATT adjustment is performed based on the output/input port information" — language the '387 specification reuses almost verbatim (see the '387 discussion of FIG. 9). That maps directly onto the core of claim 1 (demultiplex → per-wavelength attenuate → route to a chosen destination) and the "change ATT when the connected device/port changes" idea behind claims 1/5/9.

• Potential § 102 target: claims 1, 5, 9; peripherally 13–15 (its node repurposes ports between WSS/XC and client/aggregator paths — the XC-vs-client distinction).
• Caveat: Whether it expressly discloses a controller that re-sends attenuation information when the destination is switched from a first device to a second (the precise claim-1 trigger) is the crux; the "output/input port information" concept is very close but I could not verify the exact switching-language from the summary I retrieved. If it does not, it is a § 103 (obviousness) reference rather than clean § 102 art.

A2. JP2011040997A — NTT, "Optical wavelength multiplex transmission system" (JP App. JP2009186602A, filed 2009-08-11; pub. 2011-02-24). (= Patent Literature 1)
Verified content: a colorless/directionless multi-degree ROADM using a WSS that multiplexes/demultiplexes a WDM signal, expressly addressing arbitrary setting/changing of the wavelength and path of Add/Drop signals. That disclosure covers the "selectively transfer a received wavelength-multiplexed signal to a device or network per wavelength" element of claim 1 and the routing attribute of claim 9.

• Potential § 102 target: claims 1, 9 (and 4/8/12 as to "destination is a client device or a network"). Weaker on the specific "attenuation amount keyed to the destination device" and "reprogram on switch."

A3. US20040179845A1 — NEC, "Optical node processor, optical network system and its control method" (filed 2003-03-10; pub. 2004-09-16).
Same-assignee NEC art predating the priority date by ~12 years. Title/field indicate an optical node with a control method for routing optical signals — the closest structural analogue to claim 1's "controller that indicates a transfer destination."

• Potential § 102 target: claims 1, 5, 9 structurally; requires the reference to disclose per-device attenuation. Treat as strong § 103 art unless the attenuation-as-function-of-destination element is expressly present.

A4. US20090297149A1 — Fujitsu, "Optical add/drop multiplexer" (filed 2008-05-28; pub. 2009-12-03).
An OADM with per-path optical handling; the type of reference most likely to disclose "attenuate one of the demultiplexed signals and send it to a device," plus "destination is a client device or a network."

• Potential § 102 target: claims 1, 4, 8, 9, 12. Low confidence on the destination-switch re-programming limitation.

Tier B — variable-attenuation / power-balancing art (strong § 103, weaker standalone § 102)

B1. US6959149B2 — Nortel (Bragg et al.), "Power balancing in DWDM optical networks" (filed 2002-03-29; pub. 2005-10-25). Verified. Discloses a DWDM network with variable optical attenuators (VOAs) that set per-channel power to a design value, controlled independently per wavelength by control loops ("span loss padding"). Directly supports the "attenuator … attenuation amount" element and per-wavelength control — but its inputs are power targets, not destination-device identity, so it does not reach claim 1's "attenuation corresponding to [the destination] device."

• Potential § 102 target: none cleanly on its own; § 103 against the attenuation mechanism of claims 1/9 when combined with an OADM/WSS reference (A2/A3).

B2. US7218858B1 — Ciena, "System and method for use of a variable optical attenuator in an optical network" (filed 2002-05-20; pub. 2007-05-15). Verified. Node with routing and add/drop multiplexing, VOA controlled by feedback in a "plural zone" method, and latching VOA that preserves its attenuation setting. Relevant to the attenuator element and to "attenuation amount" control, and its node-level provisioning/OAM discussion touches the management-device angle (claims 2/3/10/11).

• Potential § 102 target: claims 1/9 only in combination; strongest as § 103 support and as art against the "amount may be zero / latching" nuances.

B3. CN101222277A — Fiberhome, "Method for automatic equalization of optical channel power in WSS-type ROADM node equipment" (2008-02-02 / 2008-07-16). WSS-level automatic per-channel power equalization in a ROADM — squarely on the "variable optical attenuator per wavelength in a WSS" structure of the '387's third embodiment.

• Potential § 102 target: claims 1, 9 (attenuation element) — but not the destination-keyed reprogramming.

B4. JP2014007564A — Hitachi, "Optical cross-connect device and optical level control method" (2012-06-25 / 2014-01-16). An optical cross-connect with optical level control — directly relevant to dependent claims 13/14/15 ("attenuation amount is changed depending on whether the optical device is a cross-connect or a client device"), because it concerns level control at a cross-connect port.

• Potential § 102 target: claims 13, 14, 15 (dependent), and 1/9 if level is set per output port.

Tier C — management-device / control-plane art (relevant to claims 2, 3, 10, 11)

• JP2013531909A — NEC Labs America, "Transponder aggregator system and method of operation" (2010-04-21 / 2013-08-08) — the '387's Patent Literature 3; aggregator/optical-switch control, relevant to the FIG. 9 aggregator and to destination selection.
• JP2010103781A (NTT, 2008-10-23 / 2010-05-06, "Management control system and management device") and JP2012005088A (NTT, 2010-06-21 / 2012-01-05, "Network management apparatus and optical path designing method") — directly on the management device that provisions routes/attenuation, i.e., claims 2, 3, 5, 6, 7, 10, 11.
• US20120148236A1 (Kumar, "In-band control mechanism," 2010-12-13 / 2012-06-14) and US20110188851A1 (Fujitsu, "Optical network and control method therefor," 2010-01-29 / 2011-08-04) — control-plane signaling to reconfigure node behavior → claims 2/3/10/11.

Tier D — remaining references (component-, WSS-, or generic-node art)

These bear on individual elements (WSS construction, VOA devices, OADM structure, routing) but are individually distant from the destination-keyed attenuation limitation:
Nistica US7257288B1 / US7408639B1 (tunable optical routing); ETRI US20090154923A1 (WSS) and US20100329684A1 (channel express/add module); Sumitomo CN1666131A (optical variable attenuator); NTT JP2007148042A (wavelength selective switch/combiner); Ciena US20130315580A1 (SDN photonic routing); Boduch US20160156429A1 (wavelength equalizing arrays); Agilent JP2001230730A (optical switching node); Corvis US20020196490A1; Boertjes US20030123880A1 (programmable OADM); Thiele US20040017984A1; Barnard US20070025729A1; Hitachi Comm JP2007208591A (optical cross-connect); Fujitsu US7706690B2 (relay node); Fujitsu US20070147841A1 (optical add/drop device) and US20070138417A1 (submarine); Han Jin S US20060098983A1 (OADM); Mori US20060018658A1; ECI US20110176805A1; Fujitsu US20110182585A1; Fujitsu JP2011254309A; NTT JP2010124266A; NEC JP2010219729A and WO2011055724A1 and JP2014022865A and RU2505940C2; CN1466304A and CN102790653A and CN102959883A (ROADM/OADM variants); Wuhan/CN102790653A; JP2013005016A; JP2013201517A; US8917988B2 (AT&T); US20080232760A1 (Verizon). These are best viewed as § 103 secondary references establishing the state of OADM/WSS/VOA art.

---

5. Claim-by-claim § 102 exposure summary

---

6. Bottom line

• The single most relevant prior art to US 11,095,387 is WO2009145118A1 (NEC, 2009-12-03) — same-assignee, cited both by the applicant as Patent Literature 4 and by the examiner, and it teaches per-wavelength ATT adjustment based on output/input port information. It is the reference most likely to support an invalidity contention against claims 1, 5, 9, though I could not verify the exact "switch from first device to second device" trigger language, so I rate it § 103-leaning unless that trigger is expressly disclosed.
• The next tier — JP2011040997A (NTT), US20040179845A1 (NEC), US20090297149A1 (Fujitsu) — covers the demultiplex/route-to-arbitrary-destination elements (§ 102 on the narrow "destination = client or network" limitation; § 103 overall).
• US6959149B2 (Nortel), US7218858B1 (Ciena), and CN101222277A (Fiberhome) are the key variable-attenuator / power-balancing references and the backbone of any § 103 combination against the "attenuator" element.
• JP2014007564A (Hitachi, XC optical level control) is the most on-point reference for dependent claims 13–15.

Explicit uncertainty: because I reached my search-step limit, I could not open the full text of every one of the 51 references to confirm the exact disclosure of the "attenuation-info re-sent on destination switch" limitation. The claims apparently survived all 51 references during prosecution, which strongly implies none disclosed that combined limitation — meaning the realistic challenge is § 103, not § 102. If you need a definitive § 102 position on claims 1/5/9, the two references to pull in full are WO2009145118A1 and JP2011040997A.

I'll ground this obviousness analysis in the cited prior art. Let me pull details on the key references.

Let me confirm details on the remaining key references.

Obviousness Analysis — U.S. Patent 11,095,387 (US11095387B2) under 35 U.S.C. § 103

Scope note. This analysis uses the prior art identified in the patent's own "Prior Art"/Citations section (the four Patent Literature references recited in the specification, plus the 51 examiner/third-party citations listed on the face of the patent). It builds on — and does not repeat — the claim summary and bibliographic findings in the earlier sections. Where I could not independently verify a reference's full disclosure from the search results, I say so and rely on the reference's title/abstract as listed plus the specification's own characterization of it (which is itself an admission about the prior art).

---

I. Legal framework and the single point of novelty

Under Graham v. John Deere and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), the § 103 inquiry turns on: (1) scope and content of the prior art; (2) differences between the prior art and the claims; (3) level of ordinary skill; and (4) secondary considerations.

The '387 claims are narrow in the physical-layer arts but broad in their control logic. Every claimed element — a demultiplexer for a WDM signal, a per-wavelength variable attenuator, and a controller — is old and admits as much: the specification states that in the cited ROADMs "optical signals transmitted from a connection destination are attenuated, a difference in power between the optical signals is adjusted, and the adjusted optical signals are multiplexed." That is an admission that per-destination attenuation and power-adjustment-before-multiplexing was known.

The sole asserted advance is the re-programmability element: the controller "sends the attenuation information corresponding to a second optical device when a destination of the attenuated optical signals is switched from the first optical device to the second optical device" (claim 1). The specification frames this identically: "as the attenuation amount of the optical signals is fixed, there is a problem that a device to be connected to the add/drop multiplexer cannot be changed." The invention is, in substance, making a known fixed value into a known variable value and updating it on a reconfiguration event. That is a classic § 103 scenario.

---

II. Key claim terms (construed for this analysis)

---

III. Level of ordinary skill in the art (LOOS)

A bachelor's degree in EE/optical engineering plus 2–4 years in WDM/ROADM optical transport, or equivalent. The artisan is familiar with: WSS-based ROADMs, VOAs and their control, EDFA gain/power management, per-channel power equalization, transponder aggregators, CDC-ROADM architectures, and NMS/EMS-driven provisioning. Critically, this artisan knows that per-channel attenuation must be re-set when the number or routing of channels changes — a truism in the art long before 2015.

---

IV. The prior art (as listed on the '387)

(Several other citations — US 7,706,690; US 2007/0147841; US 2010/0329684; US 2006/018658; JP 2007-148042 — are secondary cumulatives showing per-channel VOA control and OADM attenuation were routine.)

Verification caveat: I confirmed the disclosures of US 6,959,149 and US 7,218,858 directly. For the Japanese/CN references I am relying on their listed titles/abstracts plus the specification's own description, not on a full-text read. This does not change the analysis, because the strongest combination rests on the references I did verify plus the applicant's own admissions.

---

V. Grounds of rejection

Ground 1 (primary) — WO 2009/145118 + US 6,959,149 + US 7,218,858 → claims 1, 4, 5, 8, 9, 12

Mapping to claim 1:

Motivation to combine (KSR rationales):

1. Known problem, known fix. US 7,218,858 states the exact deficiency the '387 purports to solve — a fixed attenuator "cannot adjust its attenuation" on a switching event "which may result in that network element not operating properly in accordance with the new signal power levels." One of ordinary skill, seeking to let a port serve different devices, would swap the fixed attenuator for the VOA of US 6,959,149 — a substitution of a known element (VOA for fixed pad) to obtain a predictable result (adjustable power).
2. Same field, same purpose. All three references are DWDM/ROADM optical-transport references addressing per-channel power management. KSR holds this weighs toward combinability.
3. Design incentive / obvious to try. WO 2009/145118 and US 6,959,149 already teach per-wavelength, independently controlled attenuation; nothing more is needed than to index the attenuation value to the destination device and update it when the destination changes — a "use of a known technique to improve a similar device in the same way."
4. Reasonable expectation of success. Per-channel VOA control loops were well known (US 6,959,149 claim 1; US 7,218,858). Setting a VOA to a stored, device-specific value is a routine control implementation with predictable results (correct power balance, no interference — the very benefit the '387 touts).

Ground 1 also disposes of claims 4/8/12 (destination is "a client device or a network") as an express WO 2009/145118 / US 6,959,149 teaching, and of claim 5/6/7/10/11 to the extent the "management device" is any controller pushing configuration — further supported by JP 2010-103781 and US 2012/0224170.

Ground 2 — JP 2013-531909 + CN 101222277 + US 7,218,858 → claims 1, 5, 9

• JP 2013-531909 supplies the WSS-based aggregation node with selective switching between degrees and transponders (the '387 FIG. 9 architecture), i.e., the "selectively transferring … for each wavelength" selection function.
• CN 101222277 supplies automatic per-channel power equalization in a WSS ROADM — the "attenuation amount … for each wavelength" control.
• US 7,218,858 supplies the re-adjustment on a network switching event.

Combining a WSS ROADM with automatic per-channel equalization and re-adjustment on switchover yields claim 1 with a strong expectation of success, because every sub-combination was individually known and the result (balanced per-channel power after reconfiguration) is exactly what each reference sought.

Ground 3 — Claims 13–15 (attenuation differs for cross-connect vs. client device)

This is the claim set with the weakest margins. The '387's own FIGS. 6 and 8 show the XC-vs-DROP/ADD distinction, and the specification states attenuation is "specified and changed for each device connected" and reflects "connection destination information … relating to devices and networks." The prior art supplies device-class-dependent level control:

• JP 2014-007564 (Hitachi OXC "optical level control method") — level control at a cross-connect degree.
• US 2009/0297149 / US 2006/0098983 (Fujitsu / Han) — different attenuation for express/through vs. dropped/added channels.
• US 6,959,149 — the reference expressly assumes "the desired set-point for any particular wavelength is known in advance and is independent of the behaviour of all other wavelengths," which supports device-specific set-points.

Motivation: the known fact that an express/XC path and a drop-to-client path have different power targets (the '387 itself alleges that a fixed value "suitable" for one device is unsuitable for another) supplies the reason to store different attenuation values for XC vs. client destinations. This is "applying a known technique to a known device ready for improvement."

Ground 4 — US 2006/0098983 + US 6,959,149 (+ US 7,218,858) as an alternative primary

Where Ground 1/2 stress the WSS/aggregator architecture, this ground uses the OADM-level framing of the claims: an OADM (Han; Fujitsu) with per-channel VOAs (Nortel; Ciena) that are re-set on reconfiguration (Ciena). Because the claims use generic "demultiplexer / attenuator / controller" language rather than WSS-specific terms, this alternative mapping reaches the same result.

Ground 5 — claim 9 (method)

Claim 9 is the method counterpart and omits the controller (per the earlier section). Every recited step — demultiplex; attenuate with an amount for a first device; send to the first device; attenuate with an amount for a second device when the destination is switched — is performed by the structures of Grounds 1/2 when operated. A method claim covering the operation of an obvious apparatus is obvious for the same reasons.

---

VI. Why the "motivation" argument is unusually strong here

1. The specification concedes the premise. It admits the cited ROADMs attenuate per connection destination and adjust power. The only delta is variability + update-on-change.
2. The art expressly identifies the very failure mode the '387 claims to cure. US 7,218,858 literally says fixed attenuators fail on a switching event; CN 101222277 does automatic per-channel equalization in a WSS ROADM. A reference that names the problem and the fix is near-dispositive for motivation.
3. No unexpected result is asserted. The '387's stated benefit (avoid power imbalance/interference after a connection change) is the expected, predicted result of correct per-channel attenuation — not a new or surprising effect.
4. Per-wavelength independent control is old. US 6,959,149 claim 1 requires adjusting one channel's power "independently of the power levels of other optical signals," which is precisely the per-wavelength attenuation control the claims require.
5. The control layer is generic. "Controller send[ing] attenuation information" reads on any node controller/NMS — the sort of routine provisioning shown by JP 2010-103781, US 2012/0224170, and US 2013/0315580.

---

VII. Dependent-claim treatment

---

VIII. Honest weaknesses / where the patent owner will push back

I will not overstate the case. The following are the real counterarguments:

1. "Corresponding to a [device]" vs. "corresponding to measured power." US 6,959,149 and the Ciena references set attenuation based on measured power / target OSNR, not on a stored look-up keyed to the identity of the connected device. Patent owner will argue the '387 claims a device-indexed attenuation value, and that the power-balancing art is agnostic to what is connected. Counter: the '387 specification itself equates the two — it says the attenuation is "specified … for each device connected" and is derived from "connection destination information," and US 6,959,149 expressly notes the per-wavelength set-point is "known in advance," supporting a pre-stored device-specific value. The distinction, if any, is a design choice (KSR).
2. WO 2009/145118 / the JP references — I did not read full text. My WO 2009/145118 mapping rests on the '387's own description of it plus the family. A challenger must read WO 2009/145118 and JP 2013-531909 in full to confirm they teach re-setting attenuation on a destination change, not merely attenuating per wavelength. This is the top verification task for a real petition.
3. Claim 5's "network management apparatus" as a claimed element. If the owner argues the management-device-sends-connection-information limitation is a non-obvious architectural addition, the rebuttal is JP 2010-103781 / US 2013/0315580 (management plane pushing optical configuration).
4. Secondary considerations. None appear in the record (no unexpected results, no long-felt-but-unmet need shown in the spec). The '387 is untested (per the earlier PTAB section), so there is no presumption of validity beyond issuance and no prior PTAB claim-construction record.

---

IX. Bottom line

Every element of claims 1, 5, and 9 is disclosed or suggested by the art cited on the face of the '387. The strongest, cleanest grounds are:

• Ground 1 — WO 2009/145118 + US 6,959,149 + US 7,218,858 for claims 1, 5, 9 (and 2–4, 6–8, 10–12). US 7,218,858 supplies the explicit motivation (fixed attenuators fail on a switching event) and US 6,959,149 supplies per-wavelength, independent VOA control.
• Ground 2 — JP 2013-531909 + CN 101222277 + US 7,218,858, which matches the '387's own FIG. 9 architecture and adds automatic per-channel equalization.
• Ground 3 — add JP 2014-007564 / US 2009/0297149 for the XC-vs-client claims 13–15.

The claimed invention is best characterized as automating a known, manually-set per-channel attenuation value and updating it on reconfiguration — a predictable improvement over fixed-attenuation ROADMs that the cited art both teaches and motivates. I assess claims 1/5/9 and most dependents as reasonably likely to be found obvious, with the XC-vs-client claims 13–15 being the closest call and the device-indexed-attenuation point (item 1 in § VIII) being the principal dispute.

Two explicit caveats, consistent with the earlier sections: (1) I relied on the reference titles/abstracts as listed on the '387 for the Japanese/CN documents and did not fully read them; a petition must do so. (2) Nothing here should be read as a PTAB outcome — the '387 has zero AIA proceedings, so this is a defense-side invalidity theory, not an adjudicated result.