US 7792775

I'll verify against live sources before summarizing.

Let me check for the claim text and any litigation/docket records.

US 7,792,775 — Verification Summary

Bottom line: I could confirm the bibliographic and claim data for US 7,792,775 from the patent record and the published application. I found no litigation or Federal Circuit docket activity for this patent number, including nothing in 2026 — see the "Uncertainty" section, which is important here.

1. Identification (verified)

Sources: https://patents.google.com/patent/US7792775/en ; https://patents.google.com/patent/US20080215518A1 ; https://www.patentsencyclopedia.com/app/20080215518

2. Abstract (verbatim)

"A filtering rule analysis system for analyzing the rules of a packet filtering process that is set in network devices includes: a rule storage for storing sets of rules for which an order of priority has been established; a matrix generator for generating matrix spatial data that contains information of the order of priority and that indicates the correspondence relation between a minimum region identified by starting points and end points of ranges of packet attributes described in each rule as the condition of application of that rule and those rules; and an overlap analyzer for referring to the matrix spatial data to analyze overlap between the rules."

3. Plain-language overview

The patent addresses a practical firewall-administration problem: real policies contain tens to hundreds of overlapping IF-THEN packet-filter rules, so a human cannot tell which rules are actually doing any work. The disclosed approach is to convert the whole rule set into a multi-dimensional "matrix space" keyed on packet attributes (source/destination address, source/destination port, protocol), in which the "minimum regions" ("matrices") are the atomic cells carved out by all rules' boundary values, and each cell is annotated with the ordered list of rules that cover it. Analyses then run on those cells rather than on rule text.

4. Independent claims — plain language

Important caveat on numbering: The pre-grant publication shows claims 1–38 as canceled and a substitute set numbered 39–76. So the numbered independent claims in the issued patent are very likely 1, 9, 30, 36, 37, and 38, corresponding to published claims 39, 47, 68, 74, 75, 76. That renumbering is my inference, not something I verified against the granted claim text — the granted claim text was not in the materials available to me. Treat the claim numbers below as published-application numbers.

• Claim 39 (method): A filtering-rule analysis method with two steps — (a) storing a rule set having a defined order of priority, for use in packet filtering; and (b) a spatial-data generation step producing data that (i) carries the priority-order information and (ii) maps each rule to a "minimum region" defined by the start and end points of the packet-attribute ranges stated in that rule's condition. This is the base, broadest disclosure.
• Claim 47 (system): The same subject matter in apparatus form — "rule storage means" plus "spatial data generation means" producing the priority-bearing spatial data that maps minimum regions to rules.
• Claim 68 (recording medium): A storage medium holding a program that causes a computer (one having a rule-storage area for a priority-ordered rule set) to execute the process of generating the described spatial data.
• Claim 74 (broader method): Same as claim 39 but omitting the priority-information and start/end-point limitations — simply storing a priority-ordered rule set and generating spatial data indicating a correspondence between a minimum region "identified by a range of packet attributes" and the rules.
• Claim 75 (broader system): The claim-74 subject matter in means-plus-function apparatus form.
• Claim 76 (broader medium): The claim-74 subject matter as a program-bearing recording medium.

Dependent subject matter (grouped): integratable-rule identification and rule integration (claims 40–42, 63–65, 69–71); deletion of minimum regions corresponding only to the default rule (43, 44, 66, 72); boundary points computed by subtracting/adding a prescribed value to start/end points (45, 46, 67, 73); overlap analysis and detection of rules concealed by higher-priority rules (48, 49); pass testing of input packet attributes against the minimum region and its corresponding rule (50, 51); multiple rule sets with sameness analysis, including outputting the differing minimum regions (52, 53); rule deletion and rule amendment/reduction (54–57); and frequency analysis of actual packet history with rule creation/reordering (58–62).

5. Uncertainty and data-quality flags

• No litigation or CAFC 2026 docket found. My searches returned no infringement suit, no IPR/PGR, and no Federal Circuit appeal involving 7,792,775. I cannot confirm a negative — my tooling queried general web sources, not the USPTO PatentCenter/PTAB APIs or the CAFC docket system directly. Do not treat "no litigation" as established.
• Do not conflate with a similarly-numbered patent. Search results surfaced extensive IPR petition material for U.S. Pat. No. 7,979,277 (speech recognition — Jiang/Baumgartner art, calculating circuits and feature vectors). That is an entirely different patent and entirely different technology. Per your instruction to interpret identifiers literally, I have excluded it from this summary. It is a likely source of false positives if anyone searches "7792775"-adjacent strings.
• Priority-date discrepancy. The Google Patents record gives 2005-02-24; a third-party aggregator entry (portal.unifiedpatents.com) lists 2005-02-23 for US-7792775-B2. I have used the patent record's 2005-02-24 but cannot resolve the one-day difference.
• Legal status is an assumption, not a legal conclusion — that caveat is stated on the Google Patents page itself. "Expired – Fee Related" and the 2027-09-27 adjusted expiration are assertions by the database, not verified USPTO maintenance-fee records.
• Date discrepancy in my source materials. Your prompt states the current date is April 26, 2026, but the patent document supplied to me carries a fetch timestamp of 2026-09-17T11:47:29Z. I flag this rather than silently reconciling it; it means the bibliographic snapshot I relied on postdates your stated "today."
• Truncated primary text. The authoritative full text provided to me ends mid-sentence in the Fourth Exemplary Embodiment ("...stored in, for example, a program storage de") and does not include the granted claims. The claim content above therefore comes from the published application (US 2008/0215518 A1) via patentsencyclopedia.com, not from the granted patent's claim set. This is the single largest gap in my answer.

I'll research litigation involving US patent 7792775 across patent litigation databases.

I'll verify the structured data with targeted searches before writing the analysis.

Proceedings overview

Total AIA trial proceedings on US 7,792,775: 0. Breakdown by status: 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials. The USPTO Open Data Portal returns no AIA trial proceedings for this patent, and my independent web searches (Google Patents/FPO patent pages, PTAB petition text indexed at ptacts.uspto.gov, PTAB docket aggregators, and CourtListener-adjacent results) surfaced no IPR, PGR, or CBM petition ever naming US 7,792,775 as the challenged patent. Bottom-line defensive posture: the patent has never been tested at the Board — no claim has been canceled, but no claim has been validated either. A defendant gets no ready-made invalidity kill shot from a prior FWD, but also faces zero § 315(e)(2) estoppel, meaning the entire prior-art field remains open for a first-filed IPR.

⚠️ Verification caveat (read before relying on the above): I could not directly query PTAB E2E / P-TACTS or the ODP proceedings endpoint from this session; the "no proceedings" finding rests on the structured block you supplied plus negative results from the searches I could run. Confirm with a party-name and patent-number search at PTAB E2E before you treat the record as empty.

No proceedings to report

Because the structured block lists zero AIA trials, there is no proceeding number, petitioner, panel, institution decision, FWD, settlement, or appeal to profile for US 7,792,775. I will not manufacture proceeding numbers or attributes. For completeness, the four zero-count buckets are:

• Claims invalidated: none. No FWD has ever canceled a claim of this patent.
• Claims sustained: none. No FWD has ever held a claim of this patent patentable over a PTAB challenge.
• Settled / terminated pre-FWD: none identified.
• Institution denied: none identified.

One adjacent data point, clearly not this patent: NEC Corporation does appear as a patent owner in PTAB proceedings on other patents — e.g., Peloton Interactive, Inc. v. NEC Corporation, IPR2023-01240, filed 2023-07-27, panel of APJs Charles Boudreau, Christopher Ogden, and Michael Strauss, FWD dated 2025-02-05, patent owner notice of appeal 2025-04-09, and a Board order dismissing the appeal 2025-07-14 (docket aggregator). I could not confirm which NEC patent that case involves from the material retrieved, and it should not be assumed to be the '775 patent. Likewise, a 2025 NEC assertion against Dell (referenced in an indexed petition document discussing an "ONF pledge" and a "'177 Patent") involves a different NEC patent. These cases establish only that NEC litigates and tolerates IPRs — not that anyone has challenged '775.

Strategic summary

Claim status: everything is UNTESTED. No claim of US 7,792,775 — independent or dependent — has been canceled, confirmed, or even construed by the Board, because no petition has ever been filed. Do not tell a court or a counterparty that "claims 1–5 are dead"; there is no FWD to cite, and no claim-level record exists (US 7,792,775 B2). The patent's own prosecution history is the only claim-scope record: application 11/885,050, filed 2006-02-23 as a national-stage entry, priority 2005-02-24, granted 2010-09-07, inventor Katsushi Matsuda, assignee NEC Corp.

Estoppel landscape: wide open. Because no petitioner has ever instituted an IPR on this patent, there is no § 315(e)(2) estoppel binding anyone, no § 325(e)(2) estoppel, and no IPR-driven narrowing via a certificate of correction or disclaimer. A defendant served today can raise § 102 and § 103 over any prior art, any combination, and any theory — nothing is foreclosed. Two practical constraints on vehicle choice, though: PGR is unavailable (the patent's priority date of 2005-02-24 predates the AIA's 2013-03-16 PGR window), and CBM review is unavailable both because the claims are directed to network-security/firewall rule analysis rather than a "financial product or service" and because the § 18 CBM program sunset on 2020-09-16. IPR is the only AIA trial vehicle, and the § 315(b) one-year bar from service of a complaint is the clock that matters.

Pattern signals and a potentially decisive non-PTAB fact. No repeat petitioner, no defensive aggregator (no Unified Patents proceeding on this patent anywhere in the results), and no patent-owner PTAB appeal history for this patent. More important: Google Patents reports the patent's legal status as "Expired - Fee Related" with an adjusted expiration of 2027-09-27, which on its face indicates lapse for failure to pay a maintenance fee rather than expiration by term. Google's own page carries the disclaimer that "the legal status is an assumption and is not a legal conclusion." If that lapse is confirmed at USPTO Patent Center, it is worth more to a defendant than any IPR — but note that lapse is not necessarily permanent: 35 U.S.C. § 41(c) permits petitions to revive (unintentional or unavoidable delay), and a revived patent comes back with intervening-rights protections under § 41(c)(1) for parties who made substantial preparations during the lapsed period. Check the fee record and any revival petition before building a defense on lapse.

Recommended next steps

1. Confirm the empty PTAB record by patent number at PTAB E2E / P-TACTS and the ODP proceedings API. Absence of PTAB activity is itself a signal: well-asserted, high-value patents almost always attract at least one IPR, so a clean record suggests this patent has not been a significant assertion vehicle.
2. Pull the maintenance-fee and legal-status record at USPTO Patent Center for application 11/885,050 / patent 7,792,775. If the patent lapsed for non-payment and no revival petition is pending, unenforceability is a cheaper and faster defense than an IPR — and it is not subject to the Board's claim-by-claim framing.
3. Check litigation history before doing invalidity work. Search "7792775" on CourtListener and in district-court dockets; if a complaint has been served, diary the § 315(b) one-year deadline, the § 315(a)(1) bar if you filed a DJ action first, and the real-party-in-interest/privity rules under § 315(e)(2) that would capture affiliates and co-defendants.
4. If you file an IPR, expect no PGR or CBM fallback, and build the petition around § 102/§ 103 on the firewall-policy-analysis art (rule-conflict detection, minimum-region/Cartesian decomposition, rule-containment and redundancy elimination). Notably, the '775 patent is itself cited as prior art against later patentees' rule-redundancy-elimination claims (e.g., the family published as CN 103827853 B discusses US 7,792,775 as background art), which tells you the field is dense — good hunting grounds for an obviousness combination.
5. Preserve a § 101 / § 112 posture in parallel. A 2005-priority patent claiming computer-implemented analysis of packet-filtering rule sets is exposed to Alice § 101 arguments and to written-description/enablement attacks on the broad "matrix spatial data" and rule-integration functional language. These are not PTAB-available grounds (IPR is limited to § 102/§ 103 on patents and printed publications), so keep them in the district court and ITC lanes.
6. Do not cite an FWD. There is none. If opposing counsel or a demand letter asserts that claim 1 (or any claim) "has been invalidated," that assertion is false on the current record.

I'll research the assignment record for US patent 7792775 across the specified sources.

Let me search for litigation and any post-issuance transfers.

Assignment Record Reconstruction — US 7792775 (Filtering rule analysis method and system)

Verification caveat: I was able to confirm the existence and basic shape of the assignment chain from Google Patents legal events and general patent databases, but I was not able to retrieve the exact reel/frame numbers, correspondent-of-record details, or execution dates directly from the USPTO Assignment Center interface during this session. Where a field is unavailable, I say so rather than fill it in. The single-event reconstruction below is consistent across every source consulted.

Inventors

Unusual-pattern check: Not present. There is no evidence of inventors departing NEC around filing, and no inventor-side assignment activity beyond the routine employer assignment. The inventor and the assignee are aligned (standard corporate-employee invention).

Original assignee

NEC Corporation (Japan) — named assignee on the issued patent and listed as current assignee by Google Patents.

• Primary line of business: Diversified information technology and electronics — computing, networking, telecommunications, and information-security products/solutions.
• Product embodying the claims: NEC has a long-standing network-security product line (the "UNIVERGE" / "Secured Firmware"/firewall-management and security-policy tooling families) directly in the field of this invention (firewall policy analysis and management). Whether a specific shipped product reads on the claims is not verifiable from the sources consulted; the patent's subject matter (packet-filtering rule analysis for firewalls/routers) aligns squarely with NEC's network-security offerings, so there is no evidence the patent is off-product for NEC.
• Current status: Operating. NEC Corporation remains an active, publicly traded Japanese company (TSE: 6701). No bankruptcy, dissolution, or fire-sale event is associated with this patent.

Assignment timeline

Only one recorded assignment appears in the public record for this patent, and it is a routine inventor → employer assignment executed and recorded during prosecution, before issuance — not a post-issuance transfer.

• Approx. 2006–2007 (executed) / recorded 2007-08-24 — Reel/Frame: not retrievable in this session (see verification caveat above; verify at https://assignmentcenter.uspto.gov/)
  • Conveyance: Assignment of Assignors' Interest (see document for details)
  • Assignor: Katsushi Matsuda (individual inventor)
  • Assignee: NEC Corporation
  • Correspondent: Not captured in the sources consulted. This is the field to pull directly from Assignment Center on verification — for a single NEC-inventor assignment the correspondent is typically NEC's outside IP counsel of record, and its recurrence (or non-recurrence) across the file is the meaningful datum.
  • Context: Internal corporate assignment — inventor assigns rights to his employer; standard employee-invention paperwork filed during prosecution.

Post-issuance assignments: none found. Google Patents lists NEC Corp as the current assignee, and no transfer to any IP-holding LLC, aggregator, or asserter appears in the record. Per the task's own rule, the absence of post-issuance records is itself the finding: the original assignee (NEC) appears to still own the patent.

I also found no litigation naming this patent (search for infringement suits / defendants returned nothing relevant).

Timeline diagram

timeline
    title Ownership of US 7792775
    2005 : Priority date filed JP
    2006 : US application filed
    2007 : Inventor assigns to NEC Corp
    2008 : Application published
    2010 : Patent issued to NEC Corp
    2027 : Adjusted expiration

NPE / troll-pattern signals

Because the record contains only the original inventor-to-employer assignment, every NPE signal is either "not present" or "unclear — insufficient record," and the honest call is driven by the absence of downstream transfers.

1. Shell-entity transfer — Not present. No operating-company → licensing-LLC transfer is recorded. Current assignee is NEC Corporation (operating company).

2. Known asserter in the chain — Not present. No assignee in the chain matches Acacia, Marathon, IV, IPNav, Wi-LAN/Mosaid, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, or any Spangenberg entity. Only NEC Corporation appears.

3. Repeat correspondent across the chain — Unclear / not assessable. The chain has a single link, so recurrence cannot exist by definition. The correspondent name was not captured this session — pull it from Assignment Center to complete the field.

4. Cascading transfers — Not present. No chained LLC transfers; no multi-hop sequence within any window.

5. Pre-litigation transfer — Not present. There is no litigation and no assignment dated near a suit. The only assignment (2007) predates issuance (2010) by three years and is a prosecution-stage employer assignment, not a litigation-enabling transfer.

6. Bankruptcy fire-sale — Not present. NEC has not been in bankruptcy; no proceeding-associated sale of this patent is recorded.

7. Privateering — Not present. No transfer from NEC to an NPE that sues on NEC's behalf.

8. Defensive aggregator (anti-NPE) — Not present (but see note). The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. It terminates at NEC itself. Note that the patent expired (fee-related) with an adjusted expiration of 2027-09-27 per Google Patents — a lapsed, fee-expired patent is effectively neutralized regardless of owner of record, though this is a maintenance-fee lapse, not a defensive-aggregator acquisition.

Verdict

Insufficient data / non-asserting original owner — no NPE pattern.

The only assignment on record is the inventor-to-employer assignment of Katsushi Matsuda's interest to NEC Corporation, executed during prosecution and recorded 2007-08-24 (reel/frame not retrievable this session — verify at https://assignmentcenter.uspto.gov/). There is no post-issuance transfer to any IP-holding or licensing entity, NEC remains the assignee of record, and no litigation naming the patent was found. This does not meet any NPE threshold: with a single internal-assignment link and an operating-company owner, signals 1–8 are all "not present" or non-assessable.

Recommended verification step: Pull the 2007-08-24 record directly from the USPTO Assignment Center (search patent 7792775) to capture the exact reel/frame and the correspondent of record, then confirm via Patent Center that no later assignment has been recorded since the Google Patents snapshot.

Sources:

• Google Patents, US7792775B2: https://patents.google.com/patent/US7792775B2/en
• USPTO Assignment Center: https://assignmentcenter.uspto.gov/ (search 7792775)
• Patent Leaderboard (inventor profile): https://www.patentleaderboard.com/nec/katsushi-matsuda/989148
• FreePatentsOnline full text: https://FreePatentsOnline.com/7792775.html

I'll research the prior art and citations for US patent 7792775.

Let me look specifically at the patent's citation records and prosecution history.

Prior art analysis for US 7,792,775 B2

Important scope/limitation note up front: My searches returned the patent's full text and family/bibliographic data, but I was not able to retrieve the USPTO front-page "References Cited" (examiner-cited) list — i.e., the U.S. patent documents, foreign patent documents, and "Other Publications" that the examiner (Michael Holmes) actually cited on Form PTO-892 during prosecution of application 11/885,050. I ran into a tool step limit before confirming that list. I am therefore not going to fabricate U.S. patent numbers the examiner may have cited. What I can ground firmly is the prior art cited on the face of the patent's own specification (the "Patent Document 1 / Patent Document 2 / Non-Patent Document 1" list), which is the reference set the applicant characterized and distinguished. Treat the examiner-specific §102 rejections as unconfirmed pending the PTO-892.

Interpretation is literal: patent number 7792775 = US 7,792,775 B2, application US 11/885,050, filed 2006-02-23, priority 2005-02-24, granted 2010-09-07, assignee NEC Corp, inventor Katsushi Matsuda. Family members: US 2008/0215518 A1, WO 2006/090781, PCT/JP2006/303277.

1. References cited within US 7,792,775 (the "Patent Documents" and "Non-Patent Document")

These are the only prior-art references I can confirm with high confidence as being cited by/against 7792775, because they appear verbatim in the patent text.

Key §102 observation grounded in the claim text: The independent claims of 7792775 require a spatial data generation step/means that produces spatial data "indicating the correspondence relation between: a minimum region identified by starting points and end points of ranges of packet attributes described in each rule … and the rules," plus (in the overlap embodiment) analysis of overlap via that spatial data. None of the three confirmed references (JP-A-2000-174808, JP-A-2003-333084, Okajo et al.) discloses generating such minimum-region / matrix spatial data. JP-A-2000-174808 and JP-A-2003-333084 both operate on pairwise rule comparisons (conflict table / containment collation) rather than an exhaustive minimum-region partition of the attribute space, and the SCCML paper is a language reference. On the record I can see, therefore, I would not identify any of these as a §102 anticipatory reference against the independent claims; they are better characterized as the §103 background the applicant distinguished.

2. What I could NOT confirm (and why this matters)

• Examiner-cited U.S. patent references (PTO-892): Not retrieved. I saw only that 7792775 is cited by later patents (e.g., US 9,667,596 "Firewall policy comparison"; US 2015/0358283 "Firewall Policy Converter"; US 2015/0237015 "Provisioning firewall rules"). Being cited by a later patent is not prior art against 7792775 and must not be confused with it.
• Foreign search-report references from PCT/JP2006/303277 / WO 2006/090781: not retrieved.
• CN 103827853 B: This Chinese document discusses US 7792775 as background art and criticizes it (asserting it is limited to interval-relation tests, assumes all rules match the same object type, and explodes combinatorially with many attributes). This is a later reference criticizing 7792775 — it is not prior art for §102 purposes.

3. Recommendation to complete the task

To give you a defensible §102 mapping, the decisive document is the USPTO PatentCenter file wrapper for application 11/885,050 (patent 7792775), specifically:

1. the front-page "References Cited" list, and
2. the PTO-892 / examiner's Non-Final and Final Office Actions, which identify exactly which cited reference was applied against which claim.

Because my searches hit a step limit before surfacing that form, I am flagging this as an unresolved gap rather than inventing citations. If you can provide the PatentCenter "References Cited" entries (or grant me another search pass), I will map each U.S./foreign patent and non-patent reference to the specific independent and dependent claims (minimum-region spatial data under the independent claim; overlap analysis, pass-test, sameness analysis, rule deletion/amendment, frequency analysis and rule creation, and rule-integration under the dependent claims) with the same citation-level rigor used above.

Bottom line: On the evidence I can verify, the prior art cited by US 7,792,775 consists of JP-A-2000-174808, JP-A-2003-333084, and the Okajo/Matsuda/Ogawa SCCML paper (IPSJ SIG Notes Vol. 2004, No. 129, pp. 89-94). Each is a plausible §103 combination reference against the rule-overlap/rule-optimization aspects, but none appears to disclose the core "minimum-region matrix spatial data" limitation, so I do not identify a §102 anticipatory reference among them. The examiner's own PTO-892 citation list remains unverified in this analysis.

I'll ground this § 103 analysis in the verified record and prior art. Let me pull the claim text and reference details.

Obviousness Analysis — US 7,792,775 B2 under 35 U.S.C. § 103

Builds on the "Patent summary" and "Litigation summary" sections above. I do not repeat the bibliographic verification there; the caveats listed in that section (truncated specification, no granted claim text, date discrepancy) carry forward and materially limit the precision of this analysis.

0. Framework, effective law, and a scope caveat

Governing statute. Application US 11/885,050 was filed 2006-02-23 (international filing date; § 371 national stage), with a claimed priority of 2005-02-24. Because the application was filed before 2013-03-16, pre‑AIA § 103 applies, but the Graham/KSR obviousness framework is unchanged. The controlling rationales are:

• Graham v. John Deere — scope and content of the prior art; differences between the prior art and the claims; PHOSITA level; secondary considerations.
• KSR Int'l v. Teleflex (2007) — the TSM test is not the sole test; a combination is obvious if it flows from any of the recognized rationales (known elements combined per known methods with predictable results; known technique improving a similar device in the same way; design incentives/market forces; "obvious to try").
• In re Kahn — the combination must be supported by "articulated reasoning with a rational underpinning."

PHOSITA definition used here. A person with a bachelor's degree in CS or EE and ~2–3 years of experience in network security / firewall administration, including familiarity with (i) IF‑THEN packet-filter rule sets and priority ordering, and (ii) standard algorithmic data structures for interval/orthogonal-range partitioning of multidimensional coordinate spaces.

Two caveats that bound this opinion:

1. Granted claim text is not in the record (the authoritative text ends mid-way through the Fourth Exemplary Embodiment). I therefore analyze the claims as published in US 2008/0215518 A1 (claims 39–76), noting the probable granted renumbering to 1–38. A rigorous § 103 opinion must be run claim-by-claim on the granted text; treat the below as a strong prima facie case, not a final one.
2. This is a validity/defensive question only — the patent is listed Expired – Fee Related (adjusted expiration 2027-09-27) and no litigation was found (per the prior section). There is no live enforcement posture.

1. The invention, reduced to elements

2. The prior art on the record — what each reference actually teaches

All three references were verified against live sources; all three predate the one-year pre‑AIA § 102(b) critical date (2005-02-23), so none relies on the inventor's own grace period.

2.1 JP-A-2000-174808 — "Method of Operating a Data Packet Filter" (filed 1999-12-02; published 2000-06-23) — § 102(b) art

Verified summary (J-GLOBAL, https://jglobal.jst.go.jp/detail?JGLOBAL_ID=200903090566734367): a packet filter stores an ordered plurality of rules (Element A), automatically re-orders them so that rules that frequently match incoming packets move forward and rarely-matched rules move back, and computes a "match count" per rule. Critically, the applicant's own specification characterizes this reference as also performing pairwise conflict checking in which a conflict requires (i) differing actions and (ii) an intersection of the fields (protocol, source/dest address, source/dest port), with the results placed in a "conflict table." That characterization is an admission about the reference and may be used as evidence. → Discloses A, D (overlap/intersection detection), E-adjacent semantics, and the match-count basis for J.

2.2 JP-A-2003-333084 — "Packet Filtering Rule Setting Method" (filed 2002-05-09; published 2003-11-21) — § 102(b) art

Verified summary (J-GLOBAL, https://jglobal.jst.go.jp/detail?JGLOBAL_ID=200903044221690152): detects an existing rule that contains all or part of a new rule and inserts the new rule immediately before it; performs element-by-element collation (interface, protocol, chain, address, port) — i.e., orthogonal range containment/overlap testing between rules; and supports effective-time metadata with deletion of rules once the time passes. → Discloses containment/overlap analysis (D/E-adjacent), rule ordering manipulation (A-context), and rule deletion (H-adjacent). The applicant itself admits 333084 "uses a technique that is substantially equivalent to the step for determining intersections of fields … described in JP‑A‑2000‑174808."

2.3 Non-Patent Document 1 — Okajo, Matsuda & Ogawa, "A Policy Description Language for Policy-Based Security Management" (SCCML), IPSJ SIG Notes Vol. 2004, No. 129, pp. 89–94, published 2004-12-20 — § 102(b) art

Verified (https://ipsj.ixsq.nii.ac.jp/records/44870): SCCML is a common, device-independent policy language capable of the unified expression of the configurations of multiple heterogeneous security devices (access-control and monitoring). Notably, the author list includes the patent's sole inventor, Katsushi Matsuda — this is the inventor's own publication. Under pre‑AIA § 102(b) there is no "by others" requirement, so a disclosure by the inventor more than one year before filing is a statutory bar; SCCML qualifies (published ~14 months before the 2006-02-23 filing date). This is worth flagging because the same research program's follow-on papers supply further corroboration (see § 3.4–3.5) — though those later papers should be date-checked before being relied on individually.

3. Claim-by-claim obviousness

3.1 Independent claims — Elements A + B (+C): claims 39, 47, 68 (probable granted 1, 9, 30)

Combination: JP‑2000‑174808 in view of JP‑2003‑333084, in further view of the ordinary knowledge of a PHOSITA.

• 174808 supplies A and, via its conflict-table/field-intersection analysis, the reason to model rules as multidimensional attribute ranges. 333084 supplies element-wise range containment/overlap testing between rules — i.e., the coordinate-by-coordinate comparison of start/end values.
• The only meaningful difference over that art is the representation: partitioning the attribute axes at the set of all rule boundary values to yield atomic "minimum regions" (a matrix), each annotated with the priority-ordered list of rules covering it.
• Why this is obvious: discretizing a multidimensional attribute space at the union of all interval endpoints is the standard, predictable way to enumerate every distinct combination of input conditions (orthogonal range decomposition / grid partitioning — the same principle behind grid files, k-d partitions, and segment/interval trees). Applied to a rule set whose overlap semantics the prior art already analyzed pairwise, this replaces an O(n²) conflict table with an atomic-cell decomposition in which the label of each cell is simply the highest-priority rule covering it. The patent itself concedes the longstanding unmet need is administrator comprehension of an entire policy — a recognized design objective, not an unexpected result.
• Motivation to combine: 174808 and 333084 are in the same field (packet-filter rule management), address the same problem (rule overlap/ordering errors), and 333084 was already recognized by the applicant as equivalent to 174808's intersection technique. The improvement is a known technique applied to a known device, ready for improvement, with predictable results.
• Claim 74/75/76 (the broader set omitting priority information and the start/end-point limitation) is a fortiori obvious over the same combination.

3.2 Overlap / concealment — claims 48, 49 (and 10, 11 if granted)

Combination: JP‑2000‑174808 + JP‑2003‑333084. 174808's conflict table is overlap detection, and its "differing action + field intersection" test is precisely the semantics that produces "a rule is concealed/overlapped by a higher-priority rule." 333084 contributes containment testing. Motivation: 174808 requires conflict analysis to guarantee that reordering preserves overall policy meaning — the same information directly answers "which rules are doing no work?" → Strong.

3.3 Pass test — claims 50, 51

Combination: JP‑2000‑174808 + JP‑2003‑333084 + SCCML. Once the cell representation exists, mapping input packet attributes to a cell and reading that cell's top-priority action is a trivial, predictable lookup. The SCCML program is expressly directed to unified analysis of firewall policy, and packet pass/deny simulation is the ordinary, expected function of a firewall-policy analyzer. Motivation: administrators need to answer "will this packet be passed?" → Strong.

3.4 Multiple rule sets + sameness analysis — claims 52, 53

Combination: SCCML (NPL 1) + the spatial representation of § 3.1. SCCML's stated purpose is to express the policies of multiple heterogeneous security devices in one language so their configurations can be combined/compared; the research program explicitly targets cross-device contradiction detection. Comparing whether two devices' filtering behaviors are identical is the direct, named objective of that art. Motivation: to find where two firewalls disagree (a security-hole source). → Strong (subject to confirming NPL 1's § 102(b) status, which is solid).

3.5 Rule deletion & amendment — claims 54–57

Combination: JP‑2003‑333084 + JP‑2000‑174808 + SCCML. 333084 already deletes rules (on effective time), and 174808's conflict analysis identifies semantically inert rules. Deleting a rule that "despite deletion exerts no influence" and narrowing an overlapped rule's range to remove inert sub-regions are the natural, expected optimization steps once overlaps are known. Motivation: the applicant's own specification states administrators wish to "delete rules that have become unnecessary to optimize rules … in typical packet filtering in which effective times are not set" — an articulated, pre-existing need. → Strong for 54/55; Moderate for the specific split-by-minimum-regions mechanic of 56/57 (it is a predictable decomposition, but the particular implementation is narrower).

3.6 Frequency analysis + rule creation — claims 58–62

Combination: JP‑2000‑174808 + the spatial representation. 174808 already counts how often each rule matches and moves high-match rules forward to speed filtering. Sorting minimum regions by measured packet counts and building new higher-priority rules from the busiest cells is the same incentive (filtering speed) applied to the finer-grained representation. Motivation: explicitly stated in 174808's purpose (performance improvement). → Strong, except the "join adjacent regions whose counts differ by less than a prescribed range" variant (62), which is a routine threshold heuristic → Moderate.

3.7 Integratable-rule identification + integration — claims 40–42, 63–65, 69–71

Combination: JP‑2003‑333084 + JP‑2000‑174808 + ordinary knowledge of logic minimization. Merging two same-action rules whose combined region is exactly their union is the classic cell-adjacency merge (Quine–McCluskey / Karnaugh adjacency). Motivation: the patent itself states the benefit — reducing rule count so administrators can comprehend the rules. → Strong for the general concept (40, 63, 69); Moderate-to-Weak for the specific tests in 41/42 etc. (the "sum of minimum-region counts equals the count of the union region" candidate test and the apex test are narrower algorithmic conditions that a defendant relying solely on these three references may not fully anticipate).

3.8 Boundary-point expansion — claims 45, 46 (and 67, 73)

Subtracting a prescribed value from start points and adding one to end points so that packets lying exactly on a boundary are captured. This is a predictable boundary-condition handling detail; integer/inclusive-range boundary handling is a routine implementation concern a PHOSITA would address. → Moderate (closer call, but within the skill level).

4. Synthesis — motivation to combine (KSR rationales)

5. Bottom-line assessment

Conclusion. On the substance, the independent claims and the bulk of the dependents are prima facie obvious under § 103 over JP‑A‑2000‑174808 in view of JP‑A‑2003‑333084, with SCCML (NPL 1) supplying the multi-device/sameness and policy-analysis motivations. The combination is supported by articulated reasoning: shared field and problem, express cross-references in the applicant's own specification, a named long-felt need, and predictable results. The residual non-obviousness risk sits in the narrow algorithmic dependents (the integration candidate/apex tests, the split-by-minimum-region mechanic, and the boundary-expansion rule) — none of which appears in the independent claims.

Rebuttal considerations (none evidenced): no secondary considerations (commercial success, long-felt-but-unresolved need, unexpected results, copying) are of record; the patent is expired with no litigation found, so there is no objective-indicia record to weigh against the prima facie case. A patentee would most plausibly argue that the "minimum-region with priority list" representation was a specific non-obvious design choice, but the discrete, coordinate-wise partition at rule boundary values is a predictable application of known techniques, which KSR treats as insufficient to confer patentability.

6. Reliability caveats specific to this analysis

• No granted claim text was available, so all claim citations use published-application numbering (39–76) with inferred granted mapping. If granted claim 1 includes additional limitations beyond published claim 39, the independent-claim conclusion must be re-run.
• JP‑A‑2000‑174808's conflict-table/field-intersection teaching rests partly on the applicant's own characterization (an admission usable as evidence) rather than the full Japanese text; the J-GLOBAL abstract confirms the ordering/match-count teaching but not every detail of the conflict analysis.
• SCCML is the inventor's own publication, which is still § 102(b) art under pre‑AIA law (no "by others" requirement, outside the one-year grace), but this should be double-checked against the exact publication date if it is ever contested.
• The FIT2005 / CSEC‑28 2005 companion papers (which describe deletable-rule detection and a firewall pass/deny simulator) are powerful corroboration for claims 54–57 and 50–51, but their publication dates appear to fall around/after the 2005-02-24 priority date and therefore may not be prior art; I have not relied on them as statutory art.
• The date discrepancy noted in the prior section (prompt says 2026-04-26; the fetched record is timestamped 2026-09-17) persists and is not resolved here.