US 6981065

I'll search for this patent number to check for any litigation or additional records, then provide the summary based on the authoritative patent text.

US Patent 6,981,065 — Verified Summary

A note on the search: I searched for this exact number. I found no USPTO litigation docket, no PTAB proceeding, and no CAFC appeal (2026 or otherwise) involving U.S. Patent No. 6,981,065. I am stating that as a negative finding from the searches I ran, not as a positive assertion that none exists. I did not retrieve a USPTO PatentCenter/Assignment record directly; the bibliographic data below comes from the authoritative full patent text you supplied (Google Patents rendering, fetched 2026-09-27).

Bibliographic data

Assignment chain (per the record in the patent text): inventor Xiang Lu → Nortel Networks Limited (2000-08-30) → Ciena Luxembourg S.A.R.L. (2010-04-09, eff. 2010-03-19) → Ciena Corporation (2010-04-19). Subsequent events are security interests and releases: Deutsche Bank AG New York Branch (2014) released 2019; Bank of America, N.A. as administrative/collateral agent (2014, 2019), released 2023-11-20.

Abstract (as issued)

"A method of provisioning a connection across a SONET/SDU [sic — OCR of SONET/SDH] network formed from multiple sub-networks comprises: determining a model of the entire network, the model indicating the connections between all sub-networks, and representing each sub-network as a single unit; calculating a route between the sub-networks containing the first and second nodes using the model; passing the signals along the determined route, and determining the route between nodes within each sub-network locally within the respective sub-network. This method enables a simplified model of the network to be stored in each node of the network, so that a route calculation can take place at the sending node. The specific route through the network is determined as the signal passes through the network, as the specific path through individual sub-networks is determined locally. Thus, each node only needs to be capable of calculating a general route using the simplified model of the whole network and also capable of calculating a specific route through the sub-network of the node itself."

(Note: the "SONET/SDU" spelling is a transcription/OCR artifact in the source; the specification and claims consistently use "SONET/SDH." I am flagging it rather than silently correcting it.)

Plain-language overview of the independent claims

Claim 1 — method of provisioning a connection (the core claim).
A method for setting up a connection across a SONET/SDH network between two nodes, where:

• the network is hierarchical — at least a top-level sub-network plus one or more further levels of sub-networks, and at least one of those further levels contains two or more separate sub-networks;
• each sub-network contains two or more network nodes, and levels are joined by shared network nodes;
• the two endpoint nodes are not in the same sub-network.

The method has three substantive steps: (a) build a model of the whole network in which each sub-network is collapsed to a single node/unit and only the inter-sub-network connections are shown; (b) compute a route using that reduced model, i.e., decide which sub-networks/path to traverse; and (c) forward the signals along that route, while the route inside each sub-network is decided locally by that sub-network. The key insight is the split: a coarse, global routing decision from a stored simplified topology, plus fine-grained node-by-node routing resolved locally hop-by-hop as traffic traverses each sub-network — so no single node needs the full detailed topology of the whole network.

Claim 8 — computer-readable medium (the software/CRSM claim).
The same invention framed as a program product: a computer-readable medium whose instructions, when used to define a route between two nodes in a SONET/SDH network of the same hierarchical structure described in claim 1, (a) generate a data structure that is the simplified whole-network model (each sub-network as a single unit), (b) compute a route between the sub-networks containing the endpoint nodes using that model, and (c) generate a label to attach to a signal being transmitted, the label defining the determined route. In effect: the routing software plus the label-generation output that carries the route decision with the traffic.

Claim 11 — the network itself (apparatus/system claim).
The invention framed as a SONET/SDH network: hierarchical (top level plus one or more further levels, at least one level having two or more separate sub-networks, each sub-network having two or more nodes, adjacent levels joined by shared nodes), in which every network node stores the simplified whole-network model, every node has means to compute the inter-sub-network route between two nodes not in the same sub-network, and each node also has means to compute locally the route between nodes within its own sub-network. Written in "means for" form, so its scope turns on the corresponding structure disclosed and equivalents (see the specification's statement that nodes "require memory for storing the model of the entire network and require processing power").

Dependent claims, briefly: 2 (the recursive bottom-up tree-building and top-down distribution procedure for the model), 3 (each sub-network self-discovers its own topology via a routing protocol), 4 (stacked routing label with one level per network level), 5 (labels via a Label Distribution Protocol), 6 (label-stack level added/removed when crossing sub-network levels), 7 (label carried in the SONET/SDH datacomm channels); 9–10 (medium at a node that also computes the intra-sub-network route, and modifies an existing label to add/remove the intra-sub-network portion); 12 (node with means to generate a level of a stacked routing label for its own sub-network).

Points worth flagging / uncertainties

• Terminology artifacts in the source text. The specification's OCR contains several typos I have not corrected: "Open Shortcut Path First" for Open Shortest Path First (OSPF), "SONET/SDR DCC," "he SONET/SDH datacomm channels," "singe unit," and "been each first sub-network" in claim 2. These appear to be transcription errors, but per your rule I am reporting them as they appear rather than silently fixing them.
• The claim language is broader than the "summary" narrative. Claims 1/8/11 were issued with added limitations (at least one further level having two or more separate sub-networks; each sub-network having two or more nodes; endpoint nodes in different sub-networks) that do not appear in the SUMMARY OF THE INVENTION section. For any infringement or validity analysis, the issued claims control.
• Cited prior art of record (11–12 references) includes US 6,021,673, US 6,160,651, US 6,271,946 (all Telcordia, optical tag/label switching), US 6,022,954 and US 6,347,336 (network topology/auditing/discovery), US 6,020,733 (Cisco, efficient path determination in a routed network), US 6,052,722 (MCI, distributed intelligence and state management), US 6,556,593 (IBM, SDH/SONET add/drop multiplexing), and others.
• The patent is expired, so it cannot be asserted now; the Ciena ownership trail is primarily relevant historically or to licenses/settlement agreements tied to the 2000-era Nortel optical portfolio.
• No litigation confirmed. If you need a definitive litigation check, I would need to run a dedicated docket search (e.g., Docket Navigator, PACER, or the CAFC docket by patent number) — my web searches returned nothing specific to 6981065, and several "6981065" hits were unrelated parts-catalog numbers.

I'll run targeted searches across litigation-tracking sources for this specific patent number.

Let me run additional targeted searches on docket-specific sources and the patent's Google Patents "litigation" section.

Litigation search: US Patent 6,981,065

Bottom line: I found no litigation involving US Patent 6,981,065. As of my searches, there is no known district court case, no CAFC appeal, and no PTAB proceeding in which US 6,981,065 is the patent-in-suit or the challenged patent.

Results table

What I searched and what came back

• Google Patents record for US6981065B1 (the authoritative text you supplied): the "Cited By," "Similar Documents," and "Patent Citations" sections list only other patents. There is no litigation, PTAB, or "Family" litigation entry attached to this patent — a meaningful negative signal, since Google Patents usually surfaces US litigation and PTAB events where they exist for a patent of this vintage and profile.
• General web searches for the number string "6981065" / "6,981,065": returned no patent litigation. The hits were unrelated — a stainless-steel funnel catalog item ("6981065 — Entonnoir en inox de 100 ml"), a Sartorius lab-equipment parts list, a Russian equipment catalog, and mining drill-hole data ("6981065 | 750495 | 540 | -90 | 340"). None involve this patent.
• Unified Patents / litigation-tracking queries: returned Unified Patents' own IPR activity (Velos Media, Bridge and Post, MasterObjects, etc.) and unrelated matters, but nothing referencing 6,981,065.
• Searches combining the patent with its owner and title ("Ciena patent 6981065 infringement"; "Provisioning of connection through a SONET/SDH network" + litigation): returned the Ciena/Nortel Capella Photonics, Oyster Optics, Fiber, LLC, and Cheetah Omni litigations, and the Corvis/Ciena dispute — none of which assert 6,981,065.

Important caution — do not confuse this with other "…065" patents

Several superficially similar numbers appear in litigation and are not US 6,981,065. I am flagging them only so they don't get merged into your answer (per your instruction not to return results for similar numbers):

• US 6,098,065 (Nortel) — one of the seven "associative search engine" patents asserted by the Rockstar consortium (Microsoft/Apple et al.) against Google and Android OEMs in the Eastern District of Texas (case 2:13-cv-00893, 2013). This is a different patent, different subject matter (search, not SONET/SDH), even though it arose from the same Nortel portfolio family of headlines.
• US 7,631,065 — asserted in Amdocs (Israel) Ltd. v. Openet Telecom, Inc. (E.D. Va.).
• US 6,226,065 — asserted in the Guardian v. InnoLux matter (D. Del., 1:04-cv-01338).
• US 6,665,500 / US 6,594,055, RE42,678 / RE47,905 / RE47,906 — the Capella Photonics ROADM patents litigated against Ciena, Cisco, Fujitsu, Infinera, Coriant.
• US 6,981,487 — a Toyota engine patent cited as prior art in an MIT IPR; another accidental string match.

None of these is US 6,981,065.

Why the absence of litigation is unsurprising here

• The patent expired before it could be asserted in the modern NPE era. The record shows a maintenance-fee lapse effective 2017-12-27 (LAPS/STCH/FP events in early 2018), with anticipated expiration 2020-05-18. An expired patent cannot support a claim for ongoing infringement, which sharply limits its value as a litigation asset.
• It left Nortel's estate via Ciena, not Rockstar. The assignment chain is Nortel Networks Limited → Ciena Luxembourg S.A.R.L. (2010-04-09, eff. 2010-03-19) → Ciena Corporation (2010-04-19). So this patent was not part of the 2011 Nortel/Rockstar auction that generated the high-profile 2013–2014 Google litigation. (This is consistent with, and reinforces, the prior section's ownership analysis.)
• Ciena's litigation posture in this area was overwhelmingly as a defendant (Capella, Oyster Optics, Fiber, Cheetah Omni) or as the patentee in the Corvis dispute over different, bit-rate-transparent patents — not as an enforcer of 6,981,065.

Confidence and caveats — read before relying on this

I want to be precise about what I did and did not verify:

1. This is a negative finding from the searches I ran, not a certified "no litigation exists" statement. I did not directly retrieve a PACER docket report by patent number, a CAFC docket listing, or the Unified Patents litigation caselist page for this number; my results came from open web searches that surfaced secondary sources. I also reached my search-step limit before completing a dedicated docket-database query on courtlistener/PACER.
2. No PTAB record found, but I likewise did not query the USPTO PTAB API/PTAB Center directly by this number.
3. Pre-2010 or non-US proceedings (e.g., foreign counterparts, ITC Section 337 investigations, or early-2000s Nortel cross-license disputes where this patent may have been listed in a portfolio but not individually litigated) would not necessarily surface in these searches. Nothing I saw suggests any such proceeding, but I cannot rule it out with certainty.
4. Because the patent is expired and fee-lapsed, any "outcome/status" column would in all events be historical.

If you need a definitive check

To convert this negative finding into a verified one, run in this order:

1. USPTO Patent Center / Patent Public Search for application 09/573,239 — confirm the maintenance-fee lapse and check for any reissue or certificate of correction.
2. PTAB Center / Patent Public Search search by patent number for any IPR/PGR/CBM.
3. PACER (or Docket Navigator) nationwide search on "6,981,065"; CourtListener/RECAP as a free cross-check.
4. CAFC docket search by patent number.
5. Unified Patents litigation caselist — https://portal.unifiedpatents.com/litigation/caselist — filter on the patent number.

I can run those steps if you'd like the negative finding hardened into a docket-verified one — tell me which source to prioritize and I'll proceed.

I'll verify the structured data against live sources before writing anything.

Let me run a few more targeted checks before concluding.

One more verification pass on any PTAB docket or ex parte reexam.

PTAB / AIA Trial Proceedings — U.S. Patent No. 6,981,065

Verification note: I treated the structured "PTAB proceedings on file" block as the canonical source (USPTO Open Data Portal, AIA trial feed). I separately ran live searches against PTAB decision text, PTAB E2E / PTACTS documents, CourtListener, and CAFC opinion text. All returned zero hits tying any IPR, PGR, or CBM proceeding number to U.S. 6,981,065. The numerous "6981065" web hits are unrelated: a Sartorius 100 ml filter funnel catalog number, a Seattle construction permit, an FDA MAUDE report key, and an Australian mining tenement — i.e., exactly the false-positive class the prior summary warned about. Do not treat any of those as a proceeding number.

Proceedings overview

Total AIA trial proceedings on U.S. 6,981,065: 0. Breakdown: 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 institution-denied. The bottom line for a defendant is stronger than any IPR outcome could deliver: the patent issued 2005-12-27, lapsed 2017-12-27 for non-payment of maintenance fees (docketed 2018-01-22), and reached its anticipated expiration 2020-05-18. It is expired and unenforceable against new conduct as of 2020-05-18. There is no infringement-cause-of-action value left in it, which is almost certainly why no petitioner ever paid to file an IPR.

Why there are no proceedings — the three structural reasons

This absence is not an accident and it is not a sign the claims are bulletproof. In this patent's particular timeline, an AIA trial was either legally unavailable or economically pointless:

1. PGR was never available. Post-Grant Review under 35 U.S.C. §§ 321–329 applies only to patents with an effective filing date on or after 2013-03-16. This patent has a 2000-05-18 priority date and is pre-AIA. PGR was off the table by statute for its entire life.
2. CBM was never available. The transitional Covered Business Method program (AIA § 18) reached only patents claiming a financial product or service — "a method or corresponding apparatus for performing data processing or other operations used in the practice, administration, or management of a financial product or service." U.S. 6,981,065 claims SONET/SDH transport-network provisioning (a multi-layer routing/topology-discovery method). It is a network-infrastructure patent, not a finance patent, so it was categorically outside CBM. (The CBM program has in any event been closed to new petitions since 2020-09-16.)
3. IPR was the only tool, and it was never economically attractive. IPRs have been available since 2012-09-16 — but by then this patent was 12 years old and only ~5 years from lapse. An IPR petition cost (in that era, all-in) roughly $300K–$500K through FWD. Meanwhile the patent was never asserted in a way that would have forced anyone's hand: the only Ciena-era assertion I could positively confirm in the 2000s was the 2005–2006 Ciena v. Nortel clash in the Eastern District of Texas, where the Markman order I located (Ciena Corp. v. Nortel Networks, N.D. Cal., 2006-04-25) construed other Nortel patents ('076, '176, '115, '392, '535, '324, '363, '142, '760, '519) — I found no document tying 6,981,065 to that or any other asserted set. So the last realistic window in which an IPR would have been filed against this patent (roughly 2012–2017, while it was still unexpired and assertable) saw no assertion worth responding to, and no petitioner spent the money.

What is CANCELED vs. SUSTAINED vs. UNTESTED

Because there is no FWD, there is no claim-level disposition to report. I will not manufacture one.

• CANCELED: none. No AIA trial, no certificate cancelling any claim. (I also checked for an ex parte reexamination of this patent and found none. The Ciena financial-disclosure references to a reexamination and a "Notice of Intent to Issue a Reexamination Certificate" concern a different patent — the '016 patent — and I could not tie them to 6,981,065. Stating that as a negative finding, not a positive assertion.)
• SUSTAINED: none through PTAB. Note carefully: "never adjudicated" is not "held valid." All 12 claims stand exactly as issued in 2005, with only whatever presumption their unexpired status once carried.
• UNTESTED: all 12 claims. Independent claims 1 (method), 8 (computer-readable medium), and 11 (network/§ 112 ¶ 6 "means for") remain as issued, along with dependents 2–7, 9–10, and 12. No claim of this patent has ever been construed by the PTAB or cancelled by certificate.

Estoppel landscape

• § 315(e)(2) IPR estoppel: inapplicable. There are no petitioners, no instituted trials, and no FWDs, so no party is estopped by § 315(e)(1) or (e)(2). Any prior-art ground that could ever be run has never been run.
• The real bar is Article III standing / justiciability, not estoppel. With the patent expired since 2018 (lapsed) / 2020 (anticipated), a defendant cannot be liable for post-expiration conduct. Pre-expiration damages claims would be barred by the 6-year limitation in 35 U.S.C. § 286 for anything accruing before roughly 2020-09-27. That leaves a vanishing rhetorical window, not an estoppel problem.
• If a demand letter does somehow invoke this patent, the defensive posture is straightforward: (a) expired/unenforceable for any ongoing conduct; (b) § 286 time-barred for anything old; (c) no claims have been cancelled, so a validity challenge would have to be built from scratch — but nobody needs one, because infringement can't be established in the first place.
• Prior-art notes for completeness (not estoppel-relevant). The 11–12 references of record remain the best starting set if validity ever mattered: US 6,021,673 / US 6,160,651 / US 6,271,946 (Telcordia, optical tag/label switching — directly on point for the MPLS-style stacked-label aspect of claims 4–7 and 12) and US 6,022,954 / US 6,347,336 (topology auditing / auto-discovery). Those relate to dependent claims 4–7 and 12 more than to independent claims 1/8/11.

Pattern signals

• Repeat petitioners: N/A. No petitioner has ever filed against this patent, so there is no serial-filer pattern to read.
• Patent-owner appeals: N/A. Ciena has litigated aggressively around other Nortel and Capella patents, but there is no 6,981,065 PTAB appeal to trace to the Federal Circuit. I found no CAFC docket number tied to a 6,981,065 FWD, because no FWD exists.
• Defensive aggregators: none. No Unified Patents / RPX-style filing was found. Unified's CAMB (formerly CBM) and IPR activity is well-documented and this patent does not appear in it — again, consistent with a patent no operating company was asserting.

Recommended next steps

• Say it plainly in any internal memorandum: there is no PTAB activity of record on U.S. 6,981,065. Zero proceedings. Cite the USPTO ODP structured feed (the source block) plus a PTAB E2E search at https://ptacts.uspto.gov/ and a CourtListener docket search at https://www.courtlistener.com/ for independent verification.
• Lead with expiry, not with PTAB. The dispositive defense is that the patent lapsed on 2017-12-27 and expired 2020-05-18 — see the USPTO PatentCenter legal-events record and the Google Patents page at https://patents.google.com/patent/[US6981065](/patent/US6981065)/en. Any demand that cites claims 1–12 of this patent can be answered on enforceability and § 286 grounds before any validity work is warranted.
• Calibrate expectations about what "no IPRs" means. The prior summary is right that "well-asserted patents eventually attract IPRs." Here the correct inference is the weaker one: this patent was never meaningfully asserted in the IPR era (2012–2017), and it lapsed before anyone had a reason to challenge it. That is a fact about assertion history and timing, not a testimonial to the strength of claims 1, 8, and 11.
• If you need a definitive docket check, the searches I ran are strong negative evidence but not a certified clearance. A PACER/§ 286 damages audit and a full Docket Navigator or Lex Machina pull on both "6,981,065" and the 09/573,239 application number would close the loop on whether it ever appeared in an infringement complaint, ITC Section 337 investigation, or license-demand campaign.

Constraints honored: no proceeding numbers invented; no FWD attributed where none exists; no claim-level disposition asserted for unadjudicated claims; all dates in YYYY-MM-DD.

Cross-reference check against the prior sections: consistent — the earlier "Patent summary" flagged the litigation search as a negative finding and the "PTAB proceedings on file" block defaulted to no activity. Nothing in my searches contradicts those. One item to add, not correct: the prior summary's claim-count and status data are confirmed, and the lapsed/expired dates I use above (lapse effective 2017-12-27; docketed 2018-01-22; anticipated expiration 2020-05-18) come directly from the authoritative legal-events record in the patent text and should be treated as controlling over any secondary source.

I'll attempt to retrieve the recorded assignment details, including correspondents, before building the analysis.

Two follow-ups to try to pin the correspondents of record.

Assignment-chain reconstruction — US 6,981,065 B1

Sourcing note (read first). Reel/frame numbers below come from the REEL/FRAME: strings embedded in the legal-events block of the authoritative patent text (Google Patents, https://patents.google.com/patent/US6981065/en, fetched 2026-09-27). Google Patents does not expose the correspondent of record for these records, and my web searches did not retrieve the corresponding assignment PDFs from legacy-assignments.uspto.gov for reels 011074, 024213/0653, 024252/0060, 033329, 033347, 050938, 050969, or 065630. The Correspondent fields below are therefore marked "not retrieved" rather than guessed. Signal #3 (repeat correspondent) is consequently unclear — this is a data-availability gap, not a negative finding. To close it, pull the eight recordation PDFs by reel/frame at https://assignmentcenter.uspto.gov/ (query "6981065", or "573239").

Inventors

Unusual-pattern check: with a single inventor there is no multi-inventor "everyone leaves within 12 months" pattern to detect. What is notable is the opposite of a fire-sale tell: Lu assigned to Nortel on 2000-06-19 and there is no record of any inventor-side transfer, lien, or second assignment — the entire downstream chain runs through corporate entities, not individuals. No inventor-related anomaly.

Original assignee

• Entity named on the issued patent: Nortel Networks Ltd. (Nortel Networks Limited, Canada). The grant event (2005-12-27) sits on the Nortel-continuity side of the chain; the 2000-08-30 assignment to Nortel Networks Limited was effective 2000-06-19.
• Primary line of business: telecommunications network equipment — including the optical/SONET-SDH transport line (long-haul OC-192 and metro/regional OC-48/OC-12 shelves) that this patent's provisioning method is written for. The patent's own FIG. 2 architecture (OC-192 long haul / OC-48 regional / OC-12 metro) is a Nortel product-line diagram in prose.
• Did they ship a product embodying the claims? In substance, yes — Nortel was a top-tier optical transport vendor and this disclosure reads as an internal provisioning/routing feature of that gear. I have not found a specific Nortel product manual or release note tying a named shelf to claims 1/8/11, so treat "shipped an embodying product" as highly likely but not documentary in this record.
• Current status: Insolvent and wound down. Nortel and affiliates filed under the CCAA in Ontario and (for Nortel Networks Inc.) Chapter 11 in the District of Delaware on 2009-01-14 (confirmed in the Ciena IP License Agreement filed as Ex. 10.3 to Ciena's 10-Q, dated 2010-03-19: http://getfilings.com/sec-filings/[100610](/patent/100610)/CIENA-CORP_10-Q/w78484exv10w3.htm). The Metro Ethernet Networks (MEN) business, including optical networking and carrier Ethernet switching, was sold to Ciena; the remaining Nortel estates were liquidated and the Nortel patent portfolio was sold in tranches to multiple buyers (Ericsson, Rockstar/Bell Canada consortium, Ciena, etc.).

Assignment timeline

⚠️ Terminology caution. The patent text labels the 2010-03-19 effective date on both the Nortel→Ciena Luxembourg and the Ciena Luxembourg→Ciena Corporation links. That is a common two-step acquisition structure (target IP lands in an intermediate holdco on closing, then is pushed up), and the dates bear it out. Do not read it as two independent acquisitions.

1. 2000-06-19 (executed) / recorded 2000-08-30 — Reel 011074/0443

• Conveyance: Assignment
• Assignor: Lu, Xiang (individual inventor)
• Assignee: Nortel Networks Limited (Canada)
• Correspondent: not retrieved from the Assignment Center (Google Patents does not surface correspondents; the reel-011074 PDF was not retrievable in my searches). Flag: no recurrence determination possible for this link.
• Context: Ordinary inventor-to-employer assignment, executed ~1 month post-filing; the root of title.

2. 2010-03-19 (executed) / recorded 2010-04-09 — Reel 024213/0653

• Conveyance: Assignment
• Assignor: Nortel Networks Limited
• Assignee: Ciena Luxembourg S.A.R.L. (560A rue de Neudorf, L-2220 Luxembourg)
• Correspondent: not retrieved. Flag: this reel is close to but distinct from reel 024213/0620 (a 2010 Panasonic/Hasegawa recording, correspondent Michael E. Fogarty, Reel 024213 Frame 0620) — do not merge the two; same reel, different frames.
• Context: Bankruptcy fire-sale / §363-style asset sale. Nortel's MEN business sold to Ciena under an asset sale agreement dated 2009-11-24 (as amended), with Ciena Luxembourg designated as the IP-acquiring entity; executed on the 2010-03-19 closing and confirmed in the SEC-filed IP License Agreement of the same date. This is a distressed-seller divestiture, not a voluntary portfolio sale.

3. 2010-03-19 (executed) / recorded 2010-04-19 — Reel 024252/0060

• Conveyance: Assignment
• Assignor: Ciena Luxembourg S.A.R.L.
• Assignee: Ciena Corporation (Delaware; 1201 Winterson Road, Linthicum, MD 21090)
• Correspondent: not retrieved. Flag: if the same correspondent filed reel 024213/0653 and reel 024252/0060 (likely, same closing, same counsel), that would be a two-link recurrence — but I cannot confirm it from the data I obtained.
• Context: Internal group restructuring. Push-up of the acquired IP from the Luxembourg acquisition subsidiary to the listed U.S. operating parent, 31 days after the upstream transfer. Benign intra-group step; Ciena Corporation expressly guaranteed the Luxembourg entity's obligations under the 2010-03-19 IP License Agreement (Section 4.14), which is inconsistent with a licensing-only shell.

4. 2014-07-15 (executed) / recorded 2014-07-15 — Reel 033329/0417

• Conveyance: Security Interest (grant of security)
• Assignor: Ciena Corporation
• Assignee: Deutsche Bank AG New York Branch
• Correspondent: not retrieved. Flag: security-interest recordings of this type are typically filed by lender-side or borrower IP counsel; a recurrence check against reels 033347, 050969 is warranted.
• Context: Securitization / credit support. Collateral grant in favor of the term-loan agent. The patent is not conveyed — it is pledged.

5. 2014-07-15 (executed) / recorded 2014-07-16 — Reel 033347/0260

• Conveyance: Patent Security Agreement
• Assignor: Ciena Corporation
• Assignee: Bank of America, N.A., as Administrative Agent (North Carolina)
• Correspondent: not retrieved.
• Context: Securitization. Parallel collateral grant to the ABL/administrative agent on the same day as the Deutsche Bank pledge — a standard two-agent intercreditor collateral package, not a conveyance.

6. 2019-10-28 (executed) / recorded 2019-10-30 — Reel 050938/0389

• Conveyance: Release by Secured Party
• Assignor: Deutsche Bank AG New York Branch
• Assignee: Ciena Corporation
• Correspondent: not retrieved.
• Context: Release of security interest following refinancing/repayment; Deutsche Bank's lien cleared.

7. 2019-10-28 (executed) / recorded 2019-11-08 — Reel 050969/0001

• Conveyance: Patent Security Agreement
• Assignor: Ciena Corporation
• Assignee: Bank of America, N.A., as Collateral Agent (Illinois)
• Correspondent: not retrieved.
• Context: Securitization, re-papered. Replacement collateral package for the restated credit facility — the same role Bank of America held in 2014, now as collateral agent. Note this reel number (050969/0001) is not related to SEC accession 0001104659-24-050969, which surfaced in search and is unrelated.
• Cross-check: Ciena's 10-K exhibit index lists the "Omnibus Amendment to Security Agreement and Pledge Agreement, dated September 28, 2018" among Ciena Corporation, Ciena Communications, Inc., Ciena Government Solutions, Inc. and Bank of America, N.A. as Administrative Agent — consistent with a single standing IP collateral program, filed by one set of counsel over time.

8. 2023-10-24 (executed) / recorded 2023-11-20 — Reel 065630/0232

• Conveyance: Release by Secured Party
• Assignor: Bank of America, N.A.
• Assignee: Ciena Corporation
• Correspondent: not retrieved.
• Context: Release of security interest. Last recorded event. Note the execute→record gap of ~27 days is unusually long for a routine release and may indicate a housekeeping/cleanup filing rather than a contemporaneous closing document.

Net effect: title runs Lu → Nortel Networks Limited → Ciena Luxembourg S.A.R.L. → Ciena Corporation, where it has sat since 2010-04-19, encumbered twice and released twice, with no conveyance to any third party, ever.

Timeline diagram

timeline
    title Ownership of US 6981065
    2000 : Inventor Lu assigns to Nortel Networks
         : Application filed 18 May 2000
    2005 : Patent issues 27 Dec 2005
    2009 : Nortel files CCAA and Chapter 11
    2010 : Nortel assigns to Ciena Luxembourg
         : Ciena Luxembourg assigns to Ciena Corporation
    2014 : Ciena pledges to Deutsche Bank
         : Ciena pledges to Bank of America
    2019 : Deutsche Bank pledge released
         : Ciena re pledges to Bank of America
    2023 : Final Bank of America pledge released

NPE / troll-pattern signals

Verdict

Operating-company assertion.

The complete 8-record chain (reel 011074/0443 through reel 065630/0232) shows a single distressed-sale acquisition of an operating vendor's IP rather than any NPE structuring: Xiang Lu → Nortel (011074/0443, 2000) → Ciena Luxembourg (024213/0653, executed 2010-03-19, out of the Nortel CCAA/Chapter 11 proceedings) → Ciena Corporation (024252/0060, recorded 2010-04-19), which has held title ever since through nothing but secured-lien churn (033329/0417 and 033347/0260 in 2014; 050938/0389 and 050969/0001 in 2019; 065630/0232 in 2023). Signals 1, 2, 5, 7, and 8 are all absent, and the signal that the rubric weights most heavily — bankruptcy fire-sale (#6) — points at an operating buyer, not an aggregator.

Two caveats on this verdict, stated plainly. First, the bucket's second prong ("suing actual competitors") is not satisfied on this record — I found no suit, PTAB filing, or CAFC appeal naming US 6,981,065, so the accurate description is operating-company-held, and never asserted. Second, the patent is expired for fee non-payment: maintenance lapsed with an effective date of 2017-12-27 (lapse recorded 2018-01-22), ahead of its 2020-05-18 anticipated expiration, which means Ciena abandoned it rather than monetizing it. So the practical NPE risk from this particular patent today is nil; the ownership trail matters only as portfolio context or for license/settlement records covering the broader Nortel→Ciena 2010 transfer.

Verify at: https://assignmentcenter.uspto.gov/ (search 6981065 or application 09/573,239) — and specifically request the recordation PDFs for the eight reels above to fill the Correspondent field and complete signal #3.

I'll verify the patent on USPTO/Google Patents and pull details on each cited reference. Let me run targeted searches.

Let me pull details on the remaining cited references.

Prior-Art Analysis — US Patent 6,981,065 B1

Analyst note on sources and step limits: I searched for the patent and its cited references. I confirmed the patent's identity and full text (authoritative Google Patents rendering, fetched 2026-09-27, as supplied). I retrieved substantive full-text/abstract detail for four references (US 6,023,733; US 6,111,673; US 6,556,593; US 6,052,722) and title/bibliographic data for the rest. I hit my search-step limit before retrieving full text for US 6,229,540, US 6,347,336, US 6,643,286, US 6,523,064, US 6,496,862, US 6,229,538, US 6,160,651, and US 6,271,946. For those, descriptions below are drawn from their titles/abstracts and search snippets and are flagged accordingly. Where I could not verify a reference's internal disclosure, I say so rather than assert anticipation.

1. The patent under analysis (verified)

This is not US 6,098,065 (Rockstar associative-search patent), not US 7,631,065, and not any similarly numbered patent. All analysis below is confined to 6,981,065.

2. The §102 standard I applied

A reference anticipates under 35 U.S.C. § 102 only if a single reference discloses every limitation of the claim, arranged as in the claim (Net MoneyIN, Inc. v. VeriSign, Inc., 545 F.3d 1359 (Fed. Cir. 2008)); In re Gleave, 560 F.3d 1331 (Fed. Cir. 2009)). A reference that discloses only some limitations is at most a § 103 (obviousness) reference, and a bare citation "of record" in a patent is not a finding of anticipation or even of materiality.

The independent claims are broad but structurally specific. Claim 1 requires, in combination: (a) a SONET/SDH network with a top-level sub-network plus one or more further levels, at least one further level containing two or more separate sub-networks, each sub-network containing two or more nodes; (b) levels joined by shared nodes; (c) endpoint nodes in different sub-networks; (d) a model of the entire network with each sub-network represented as a single unit; (e) route calculation between sub-networks using that model; and (f) local determination of the intra-sub-network route. Claim 8 mirrors (a)–(e) plus label generation; claim 11 mirrors the structure plus "means" for (d)–(f).

No cited reference I reviewed discloses element (d) combined with (f) — i.e., the hierarchical two-tier routing split (collapsed global model + locally resolved intra-sub-network path) in a SONET/SDH multi-level network. That is the point of novelty, and it is why I find no single-reference § 102 anticipation below; the references are best characterized as § 103 combinations.

3. Prior art of record — the 12 "Patent Citations"

(Order as listed in the patent record. "*" = cited by examiner; US 6,023,733 is the one not examiner-flagged in the record, i.e., appears applicant-cited.)

4. Reference-by-reference analysis

(1) US 6,023,733 — Cisco, "Efficient path determination in a routed network"

• Full citation: US 6,023,733 A; filed 1997-10-30; granted 2000-02-08; Cisco Technology, Inc.
• Description (verified): Route/path determination in a routed (packet) network. Its specification text is strikingly close to the '065 background section — the OSI seven-layer protocol-stack discussion (stack 125/175, layers 112/114/116/118/120/122/124, source station 110, destination station 150, channel 180) appears essentially verbatim. It concerns router/bridge path determination and route caching in a connectionless packet network.
• §102 assessment: Relevant to the general routing concepts of claims 1 and 8, and it is the evident source of the '065 background. But it is not a SONET/SDH network, has no hierarchical multi-level sub-network model, no "each sub-network as a single unit," and no global-model/local-route split. Not anticipatory. Best characterized as § 103 background/teaching for the routing step.

(2) US 6,052,722 — MCI Communications, "System and method for managing network resources using distributed intelligence and state management"

• Full citation: US 6,052,722 A; filed 1997-03-07; granted 2000-04-18; MCI Communications Corp.; inventor Mehdi Taghadoos.
• Description (verified): Defines an entity/object for each managed network resource; organizes entities in one or more levels of a network management system based on association/containment; propagates state changes between entities; entities proactively initiate network management and protection actions. Claims recite layered objects with "affected object list" and "supported by object list," proactively triggering protection (including SDH multiplex-section protection units).
• §102 assessment: Directly relevant to the "distributed intelligence" rationale of claim 1 and to the '065 statement that "provisioning may be inhibited during line protection switching." It discloses hierarchical, distributed-intelligence network management over SONET/SDH resources. It does not disclose a collapsed whole-network routing model or label-based intra-sub-network routing. Not anticipatory; a strong § 103 reference against claim 1's distributed-intelligence framing and possibly against claim 11's distributed "means."

(3) US 6,111,673 — Telcordia, "High-throughput, low-latency next generation internet networks using optical tag switching"

• Full citation: US 6,111,673 A; filed 1998-07-17; granted 2000-08-29; Telcordia Technologies, Inc.; inventors Gee-Kung Chang, Sung Joo Yoo.
• Description (verified in detail): Optical (WDM) tag switching. Claim 1 recites "generating and storing a local look-up table in each of the network elements, each local look-up table determining a local route through the associated one of the network elements," adding an optical header indicating the route, and optically determining the header to select the local route. The spec expressly compares to Cisco electronic "Tag Switching" and to IETF MPLS, and describes a Tag Distribution Protocol for distributing tag information.
• §102 assessment: This is the most substantive single reference and is highly material to claim 4 (routing label/stack), claim 5 (Label Distribution Protocol), and — importantly — the "local route through each network element" idea that maps onto claim 1's "determining the route between nodes within each sub-network locally." But it is an optical/WDM, single-tier tag-switched network: no multi-level SONET/SDH hierarchy, no top-level + further sub-network levels, and no collapsed whole-network model (each sub-network as one unit). It therefore does not anticipate claims 1/8/11, but it is a strong § 103 reference for claims 4–7 and for the "local routing" element.

(4) US 6,160,651 — Telcordia, "Optical layer survivability and security system using optical label switching and high-speed optical header reinsertion"

• Full citation: US 6,160,651 A; filed 1999-01-25; granted 2000-12-12; Telcordia Technologies, Inc.
• Description: Optical label switching with high-speed optical header reinsertion for survivability/security. (Companion to US 6,271,946 and US 6,657,757.)
• §102 assessment: Relevant to label/header handling — detection and reinsertion maps onto claim 6 (a label-stack level added/removed when crossing levels) and claim 4. No SONET/SDH hierarchical provisioning. Not anticipatory; § 103 reference for claims 4/6.

(5) US 6,229,538 B1 — Compaq, "Port-centric graphic representations of network controllers"

• Full citation: US 6,229,538 B1; filed 1998-09-11; granted 2001-05-08; Compaq Computer Corp.
• Description: Graphical/port-centric representations of network controllers (a GUI/visualization reference).
• §102 assessment: Not anticipatory of any claim. At most, background on representing network elements/ports graphically. No meaningful § 102 or § 103 mapping to claims 1, 8, or 11.

(6) US 6,229,540 B1 — Visionael, "Auditing networks"

• Full citation: US 6,229,540 B1; filed 1996-02-23; granted 2001-05-08; Visionael Corporation.
• Description: Network auditing — discovering/validating the actual composition and topology of a network and building a representation ("model") of it.
• §102 assessment (§102 caveat — full text not retrieved): Relevant to claim 1 step (a) ("determining a model of the entire network") and claim 2 discovery/distribution concepts. It does not disclose the SONET/SDH multi-level "each sub-network as a single unit" model combined with local sub-network routing. Not anticipatory; potential § 103 reference for the model-building step.

(7) US 6,347,336 B1 — Samsung, "Automatic discovery and positioning method of the network elements in the network management system…"

• Full citation: US 6,347,336 B1; filed 1998-04-06; granted 2002-02-12; Samsung Electronics Co., Ltd.
• Description: Automatic discovery and positioning of network elements when network topology is configured (auto-discovery of topology).
• §102 assessment (§102 caveat — full text not retrieved): Material to claim 3 ("each sub-network determines its own topology using a routing protocol") and to claim 2's auto-discovery/distribution. It does not disclose per-sub-network collapsed modeling with hierarchical SONET/SDH levels. Not anticipatory; § 103 reference for the topology-discovery limitations.

(8) US 6,496,862 B1 — Mitsubishi, "Remote monitoring and control of devices connected to an IEEE 1394 bus via a gateway device"

• Full citation: US 6,496,862 B1; filed 1998-08-25; granted 2002-12-17; Mitsubishi Electric Research Laboratories, Inc.
• Description: Gateway device bridging an IEEE 1394 bus to a broader network for remote monitoring/control.
• §102 assessment: Marginal. At most, background on gateway/inter-network interconnection. Not anticipatory of any claim; weak § 103 relevance at best.

(9) US 6,523,064 B1 — Mitsubishi, "Network gateway for collecting geographic data information"

• Full citation: US 6,523,064 B1; filed 1999-04-29; granted 2003-02-18; Mitsubishi Electric Research Laboratories, Inc.
• Description: A network gateway collecting geographic data.
• §102 assessment: Marginal, same category as reference (8). Not anticipatory; negligible § 103 relevance to the routing/provisioning claims.

(10) US 6,556,593 B1 — IBM, "Digital cross connect and add/drop multiplexing device for SDH or SONET signals"

• Full citation: US 6,556,593 B1; filed 1996-12-11; granted 2003-04-29; International Business Machines Corp.; inventors Herkersdorf, Lemppenau, van As.
• Description (verified): SONET/SDH digital cross-connect and add/drop multiplexing using parallel processing modules, operating on tributary signals carried in higher-rate STM-N/STS-N frames; discusses the SDH/SONET hierarchy (STM-1…STM-64; VC/AU/AUG containers) and cross-connecting lower-rate tributaries without fully demultiplexing frames.
• §102 assessment: Material background to claim 1's premise that a SONET/SDH network is hierarchical, with lower-rate sub-networks (tributaries) attached at shared nodes of higher-rate networks — the '065 spec describes exactly this OC-192/OC-48/OC-12 topology and "tributary shelves." But it is a node/hardware-level cross-connect, with no whole-network model and no route calculation/provisioning. Not anticipatory; § 103 reference supplying the SONET/SDH tributary-hierarchy teaching.

(11) US 6,643,286 B1 — Dunti Corporation, "Modular switches interconnected across a communication network…"

• Full citation: US 6,643,286 B1; filed 1999-05-14; granted 2003-11-04; Dunti Corporation.
• Description (§102 caveat — full text not retrieved): Modular switches interconnected across a communication network to achieve minimal address mapping/translation.
• §102 assessment: Potentially relevant to hierarchical/modular interconnection and address mapping (tangentially to claim 1/11's modular network structure and to labeling/address translation). It does not disclose the collapsed whole-network model plus local sub-network routing in SONET/SDH. Not anticipatory; weak § 103 relevance.

(12) US 6,271,946 B1 — Telcordia, "Optical layer survivability and security system using optical label switching and high-speed optical header generation and detection"

• Full citation: US 6,271,946 B1; filed 1999-01-25; granted 2001-08-07; Telcordia Technologies, Inc.
• Description: Companion to US 6,160,651, addressing optical label/header generation and detection (and, per its family, label-swapping/header replacement).
• §102 assessment: Relevant to labels on transmitted signals (claim 4) and label add/remove (claim 6). No SONET/SDH provisioning hierarchy. Not anticipatory; § 103 reference for claims 4/6.

5. §102 bottom line — per-claim mapping

Conclusion: None of the twelve references of record appears to anticipate any claim of US 6,981,065 under § 102, individually. The likely prosecution posture is that the examiner used these references in § 103 combinations — principally the Telcordia optical-tag-switching family (US 6,111,673 / 6,160,651 / 6,271,946) for the label and local-routing elements, MCI's US 6,052,722 for distributed/layered intelligence, Samsung's US 6,347,336 for auto-discovery, IBM's US 6,556,593 for the SONET/SDH tributary hierarchy, and Cisco's US 6,023,733 for path determination.

Why the claims survived: the independent claims as issued carry added limitations not present in the SUMMARY section — "at least one of the further levels comprising two or more separate sub-networks," "each of the sub-networks comprising two or more network nodes," and "the first and second nodes not being part of the same sub-network." These read as amendments introduced to distinguish the art. (This is inference from comparing the issued claims to the summary text; I did not retrieve the prosecution history to confirm the amendment reasons.)

6. Caveats you should weigh before relying on this

1. Four of the strongest references I read in full; eight I did not. Full text was verified for US 6,023,733, US 6,111,673, US 6,556,593, and US 6,052,722. For US 6,229,540, US 6,347,336, US 6,643,286, US 6,523,064, US 6,496,862, US 6,229,538, US 6,160,651, and US 6,271,946, I relied on titles/abstracts/snippets. A "no anticipation" conclusion for those eight is provisional.
2. This is a post-issuance prior-art-of-record review, not a validity opinion. Citations "of record" are not admissions of anticipation; only single-reference, all-elements disclosure is § 102. The real invalidity exposure, if any, is § 103 — a combination I have flagged but not fully briefed.
3. The patent is expired (fee lapse eff. 2017-12-27; anticipated expiration 2020-05-18), so this analysis is historical/defensive. I found no litigation, PTAB, or CAFC proceeding involving 6,981,065 (consistent with the prior section's negative finding; not a certified docket-verified negative).
4. Terminology artifacts in the source (e.g., "Open Shortcut Path First" for OSPF, "SONET/SDU," "SONET/SDR DCC") are reported as they appear and are not auto-corrected per your rule.
5. If you want this hardened, the next steps would be: pull the full prosecution history (file wrapper for 09/573,239) to identify the examiner's actual § 102/§103 rejections and the amendments that overcame them, and retrieve the complete text of the eight references I could not fully read.

I'll ground the obviousness analysis in the actual content of the cited references. Let me pull the key prior-art patents of record.

Let me pull the remaining key references — the SONET/SDH add-drop/cross-connect art and the Telcordia optical label-switching family.

Obviousness Analysis — U.S. Patent No. 6,981,065 B1

"Provisioning of connection through a SONET/SDH network"
Inventor: Xiang Lu · Original assignee: Nortel Networks Ltd. · Filed/priority: 2000-05-18 · Issued: 2005-12-27 · Expired (fee lapse eff. 2017-12-27)

Scope note / caveats. This analysis uses the prior art listed in the "Citations (11)" / "Patent Citations (12)" sections of the patent record you supplied (the Google Patents rendering, fetched 2026-09-27). I have grounded the reference-level descriptions below in retrieved text of the references themselves where available. Where I could only obtain a title/assignee/abstract, I say so explicitly rather than assuming full teaching. I reached my tool-call limit before retrieving full text for US 6,229,540 (Visionael) and US 6,271,946 (Telcordia), so those are characterized from their titles and family context. Note also that the source lists the cited-art set inconsistently (11 vs. 12 entries); US 6,271,946 B1 appears only in the 12-item list. This is an ex post invalidity-style analysis written from the perspective of a challenger; it is not a legal opinion, and the references were of record and were overcome during prosecution of the '065 patent, which then issued.

1. Legal standard applied

Under 35 U.S.C. § 103, a claim is obvious when the differences between the claimed subject matter and the prior art are such that the subject matter as a whole would have been obvious at the time of the invention to a person having ordinary skill in the art ("POSITA"). The Graham v. John Deere factors govern: (1) scope and content of the prior art; (2) differences between the prior art and the claims; (3) the level of ordinary skill; and (4) secondary considerations.

Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), a combination is obvious where: (a) the elements were known in the art; (b) there was a known problem for which the combination was an obvious solution; (c) the combination is of predictable, known elements performing their known functions; or (d) the combination was "obvious to try" because the prior art identifies a finite number of identified, predictable solutions with a reasonable expectation of success. A motivation to combine can come from the references themselves, from the knowledge of a POSITA, from the nature of the problem, or from market/design incentives — it need not be stated in the references. Conversely, the patentee may rebut with objective evidence of non-obviousness (teaching away, unexpected results, long-felt need, commercial success with nexus).

2. Scope and content of the prior art of record

All references have 1996–1999 priority dates and were cited of record — all qualify as prior art under § 102 (pre-AIA) to the '065 patent's 2000-05-18 priority date.

3. Level of ordinary skill in the art

The claims concern connection-oriented, layered transport-network provisioning — SONET/SDH framing and digital cross-connect, routing protocols, and label/tag-based forwarding. A POSITA in May 2000 would have had a bachelor's degree in electrical engineering or computer science and roughly 2–4 years of experience (or equivalent), and would be familiar with:

• SONET/SDH line rates, tributary mapping, and the SONET/SDH datacomm channels (DCC) as an in-band signaling channel;
• Link-state (OSPF) and hierarchical/inter-domain routing;
• MPLS / Cisco tag switching as of 1997–2000, including label distribution (TDP/LDP) and label stacks; and
• multi-layer/long-haul–regional–metro network architecture.

That skill level supports combining the transport-layer framing art with the label-switching and hierarchical-management art.

4. Key claim-construction points

• "Model of the entire network … representing each sub-network as a single unit" — an abstraction in which each sub-network collapses to a single node (the patent's FIG. 3). No unusual construction needed; this is a standard topology-aggregation concept.
• "Determining the route between nodes within each sub-network locally within the respective sub-network" — the intra-sub-network hop-by-hop path is computed by the sub-network itself, not centrally.
• "Routing label … comprising a stack having a level … for each network level" (claim 4) — an MPLS-style label stack, one entry per hierarchy level. This is the most distinctive limitation.
• The issued claims are narrower than the specification's SUMMARY: claims 1/8/11 add (a) at least one further level having two or more separate sub-networks, (b) each sub-network having two or more nodes, and (c) the endpoint nodes not in the same sub-network. (Consistent with the cross-reference note in the earlier sections. I am not re-arguing that here, but it is the boundary of any § 103 case.)

5. Ground A (primary): US 6,111,673 + US 6,052,722 + US 6,556,593 — claim 1

US 6,111,673 is a near-complete § 102/§ 103 match for the core of claim 1 once its optical-context language is mapped to SONET/SDH. Element by element:

Motivation to combine (KSR). (1) The references self-identify the problem — routing across many interconnected sub-networks while keeping per-node processing and signaling overhead low — which is exactly the problem '065 sets out to solve (its Background admits provisioning was "normally performed manually" and "labour and time intensive"). (2) US 6,111,673 expressly discusses interworking with SONET/ATM and the WDM/ATM/SONET testbed, making the transport-layer combination with US 6,556,593 natural. (3) US 6,052,722 supplies the hierarchical abstraction technique (multi-level entity containment + cross-level state propagation) that US 6,111,673 needs to make its "global knowledge" routing tractable in a multi-sub-network setting — a textbook "combine two known techniques to solve a known problem" scenario. (4) All three elements perform their known functions in the combination (framing, hierarchical abstraction, label-based local forwarding) with predictable results.

Claim 8 (CRM). Same combination supplies every limitation: US 6,111,673's instructions/local look-up table generation and header/label generation map onto "instructions causing generation of a data structure … [and] generation of a label for attachment to a signal." Claim 8's "data structure defining the network … model … each sub-network as a single unit" is met by US 6,052,722's layered model data. No additional structural element distinguishes claim 8 beyond claim 1's subject matter recast as software — obvious for the same reasons.

Claim 11 (network/system). US 6,111,673's network of network elements, each storing a local routing table and each having header-detection + route-selection means, combined with the SONET/SDH multi-layer network of US 6,556,593 and US 6,052,722's hierarchical nodes, meets "each node … storing a model … means for calculating a route … and … means for determining a route between nodes within said sub-network locally." Under § 112(f), "means for" terms cover the disclosed structures (the references' NC&M processors, memory/storage, and per-element controllers) and equivalents — the patent's own specification concedes nodes need only "memory … and … processing power," which the references plainly disclose.

6. Dependent claims

7. Grounds B and C (parallel / alternative combinations)

Ground B — US 6,023,733 (Cisco) as primary for the routing-model limitations.
The '065 patent's Background is essentially a restatement of US 6,023,733's disclosure (protocol stacks, OSI, bridge/router distinction, source routing and cached paths, "all-paths explorer" broadcast). US 6,023,733 is titled and directed to "efficient path determination in a routed network" — path determination across interconnected subnetworks, including discovering and caching paths and having routing devices along a path cache the path. Combined with US 6,556,593 (to supply the SONET/SDH medium) and US 6,052,722 (hierarchical model), it supplies the "calculate a route … using the model" and "model of the entire network" limitations. Caveat: I retrieved only US 6,023,733's background, not its full claims, so this ground is a secondary/supplemental ground rather than the primary attack.

Ground C — Discovery-based grounds for claims 1–3.
US 6,347,336 + US 6,229,540 + US 6,052,722 together teach auto-discovery and maintenance of a network-topology model organized in hierarchical levels, plus per-level containment relationships. These supply the "determining a model of the entire network," "auto-discovery," and "each sub-network determines its own topology" limitations independently of US 6,111,673, reinforcing Ground A.

8. Motivation to combine — consolidated (KSR factors)

1. Known problem, known solution. The '065 Background itself frames the problem: multi-layer SONET/SDH provisioning was manual, labor-intensive, and needed to fit within limited in-band signaling bandwidth. US 6,111,673 targets exactly "keep routing tables/data-exchange small and the network scalable," and US 6,052,722 targets "manage network resources using distributed intelligence."
2. Predictable, known elements performing known functions. SONET/SDH framing + hierarchical topology abstraction + label-based local forwarding are each independently known and combine without changing any element's function — the hallmark of predictable combination.
3. Finite, identified solutions. By 2000, the label/tag approach, hierarchical routing, and topology aggregation were recognized, finite alternatives for automated transport provisioning — "obvious to try" with a reasonable expectation of success.
4. Market/design incentive. Reducing the per-node routing-table and signaling burden (the express rationale in both US 6,111,673 and US 6,052,722) is an economic/design incentive that KSR recognizes as a valid motivation.
5. Same field of endeavor. All primary references are in the transport/optical network control plane and would have been considered by a POSITA; the optical/electronic distinction in US 6,111,673 does not make it non-analogous to a SONET/SDH (a standard electronic-layer-label transport) provisioning problem — US 6,111,673 itself couples its optical layer to the electrical layer via ATM/SONET.

9. Rebuttals the patentee would raise, and how they fare

R1 — "US 6,111,673 is optical/WDM tag switching; '065 is SONET/SDH electronic provisioning."
Assessment: weak. US 6,111,673 expressly interworks with ATM/SONET and the ONTC WDM/ATM/SONET testbed, and describes its scheme as the optical analogue of Cisco/MPLS tag switching and distributed IP routing. A POSITA would treat it as analogous art. The physical/optical-vs-electrical distinction does not alter the claimed routing/abstraction steps.

R2 — "No reference teaches collapsing each sub-network to a single unit while also computing the intra-sub-network route locally."
Assessment: the patentee's best argument, but rebuttable. US 6,111,673 discloses the global-route-then-local-route split ("NC&M decides the initial outbound link … the rest of the routing decisions are taken at the individual NEs"), and US 6,052,722 discloses multi-level entity/containment modeling. The remaining step — representing each sub-network as one model unit — is the routine form of topology aggregation and is suggested by US 6,111,673's spanning-tree/global-knowledge approach and US 6,347,336/US 6,229,540's topology modeling. Expect the patentee to press this, but it is an aggregation-of-disclosures argument, not a missing-element argument.

R3 — "The label stack with one level per network level (claim 4) is absent from the art of record."
Assessment: modest strength on claim 4, but does not save claims 1, 8, 11. Later art expressly criticizes the Chang-family header for conveying only a single label-switch state (see US 2002/0141019's analysis of Chang). That absence may be meaningful for claim 4 — but MPLS label stacking and push/pop (TDP/LDP; RFC 3031) were known, so claim 4 is still vulnerable under KSR as a predictable application, and in any event claims 1/8/11 do not require a label stack at all.

R4 — Teaching away. Nothing in the record teaches away (e.g., no reference disparages hierarchical centralization plus local resolution or label stacking for SONET/SDH). This factor favors the challenger.

10. Secondary considerations

• No objective evidence of non-obviousness is present in the record I have. No unexpected results, copying, industry praise, or licensed-nexus evidence is in the patent or its prosecution materials as rendered.
• The patent lapsed for non-payment in 2017 (anticipated expiration 2020-05-18), which undercuts any inference of commercial significance tied to the claimed subject matter — an indirect negative signal on nexus. (Consistent with the expiration/ownership findings in the earlier sections; I am not re-stating them.)
• Persistence of the examiner's allowance is not itself evidence of non-obviousness — the references were of record, but obviousness under KSR can rest on combinations the examiner did not articulate.

11. Conclusion

Bottom line. A strong § 103 case exists against independent claims 1, 8, and 11 via the combination of US 6,111,673 (primary) + US 6,052,722 + US 6,556,593, with US 6,347,336 / US 6,229,540 reinforcing the model/auto-discovery limitations. The dependent claims fall largely to known MPLS label-distribution/push-pop mechanics and SONET/SDH overhead practice. The only materially contestable limitation is the label stack with one level per network level (claim 4) — and even that is vulnerable under KSR as a predictable application of known label-stacking, since claims 1/8/11 do not require it. The absence of any secondary-considerations evidence and the patent's fee-lapse further weaken any non-obviousness rebuttal.

Confidence and caveats. High confidence on the mapping of US 6,111,673, US 6,052,722, US 6,556,593, and US 6,347,336 (I retrieved substantive text). Lower confidence on US 6,023,733 (background only), and US 6,229,540 / US 6,271,946 / US 6,229,538 (title/abstract/record only) — those supporting references should be verified in full text before being relied on as primary. I did not retrieve every reference's full claim set, and this memo does not attempt anticipation (§ 102) analysis, which would require a full element-by-element comparison against a single reference. If you want, I can prioritize retrieving the full text of US 6,023,733 and US 6,271,946 to harden Grounds B and C.