US 7554995

I'll search for authoritative information on this patent number.

Search results for US 7,554,995 — status of my search

I ran searches targeting the exact number (including quoted "7554995", "7,554,995", Nomadix, and Federal Circuit 2026 dockets). Results below. Where I could not verify something authoritatively, I say so explicitly.

1. Patent identification (authoritative, from the full text)

Continuity (from the cross-reference section): continuation of Ser. No. 09/684,937 (now US 7,088,727) → CIP of Ser. No. 09/041,534 (now US 6,130,892) → CIP of Ser. No. 08/816,174 (abandoned). Later family members recorded on the same page include US 8,027,339, US 8,594,107 and US 2014/0079055.

2. Abstract (verbatim from the patent)

"A system and method for connecting a user device to a network where the user device settings, the network settings, or both are unknown include intercepting packets transmitted by the user device and modifying the packets to be compatible with the network. The system and method are particularly suited for use by mobile computers, such as laptop computers, which are connected to various foreign networks. Depending upon the particular application, a device may be carried with the mobile computer, or attached as a node on the network. The device automatically determines the network settings of the user device and/or the network and modifies packets appropriately so that the user device can communicate over the network without having to reconfigure the user device with appropriate settings for each network it may encounter. Communication settings such as network address, gateway, proxy address, etc. are automatically determined using various techniques."

3. Independent-claim overview — with an explicit caveat

Uncertainty note: the authoritative text I retrieved for US 7,554,995 covers the bibliographic data, abstract, background, summary, and the body of the detailed description, but it was truncated before the "Claims" section. I therefore do not have the verbatim claim language, and I cannot confirm the exact number or scope of the independent claims (independent claims in this family are, by convention, a method claim and a corresponding system/apparatus claim, but I have not verified this for the '995 patent). The following is a plain-language description of the invention's core operative steps as disclosed, which is the scope such independent claims would most likely track. It should be treated as inference, not as a quotation of claim language.

Plain-language elements recurring throughout the disclosure:

1. Interception without being addressed. The device ("nomadic router") receives packets from a user host even though the host is not configured to use it — via Proxy ARP (responding to the host's ARP request with the router's own MAC address, and/or reconfiguring the host's cached MAC) or via promiscuous mode (accepting all packets on the link).
2. Learning the host configuration by inspecting outbound packet content (source IP address, gateway, DNS/proxy settings), so the host's "home" configuration is discovered rather than manually entered.
3. Learning the network configuration by DHCP request, listening to router-information packets, passive promiscuous-mode observation (electing an unused IP address and switching if it is taken), or manual/SNMP/web configuration.
4. Packet translation at multiple protocol-stack layers — link, network, transport and application. Outbound: replace the host's source address with the router's address, translate location-dependent parameters (e.g., DNS server name, FTP-embedded addresses, proxy server), recompute checksums, and (where multiple hosts are attached) remap port numbers via a stored table. Inbound: perform the reverse translation so the host believes replies came directly to it.
5. Transparency to both ends. The router "pretends to be" the host's home gateway toward the host and "pretends to be" the host toward the network, so neither the host nor the network needs reconfiguration; optionally it provides DHCP service to the host when no DHCP server exists on the foreign network.
6. Device independence / multiple substrates (in the broader specification) — selecting among and simultaneously using multiple communication interfaces.

4. Litigation results — no CAFC 2026 docket found for this patent

• USPTO/Google Patent record (Google Patents litigation links): five C.D. Cal. cases are associated with this patent family: 2:19-cv-10202, 2:14-cv-00322, 2:11-cv-02604, 2:10-cv-00381, and 2:09-cv-08441.
• Stanford NPE Litigation Database (npe.law.stanford.edu/patent/7554995) lists exactly one case naming this patent: Nomadix, Inc. v. Solutioninc Technologies Limited, 2:10-cv-00381 (C.D. Cal.), with Nomadix as asserter and Solutioninc as alleged infringer. The lead/related C.D. Cal. cases are 2:09-cv-08441 (Nomadix v. Hewlett-Packard et al., Judge Pregerson) and 2:07-cv-01946.
• Federal Circuit 2026 dockets: I found no appeal involving US 7,554,995. My "7554995 + Federal Circuit 2026" search returned only unrelated 2026 decisions (e.g., Netlist v. Micron, involving the '918 patent on memory-module power). Given that the '995 patent's recorded legal status is Expired – Fee Related with an adjusted expiration of Dec. 3, 2018, it is unlikely to be the subject of an active 2026 Federal Circuit appeal, but I cannot represent my search as an exhaustive, authoritative check of the CAFC docket — treat "no 2026 CAFC activity found" as a negative search result, not as a certified absence.

5. Things I am explicitly uncertain about

• Independent claim text and count — not in my retrieved material (see caveat above).
• Current ownership — Google Patents lists "Gate Worldwide Holdings LLC" as current assignee while also recording a Nov. 22, 2019 quitclaim from Gate Worldwide Holdings back to Nomadix, Inc., plus earlier NTT DOCOMO security interests and a 2019 Knobbe Martens security interest (released May 28, 2019). The chain is inconsistent-looking and Google itself disclaims accuracy; I would not state a definitive current owner without an assignment-database pull.
• Exact event history of the five C.D. Cal. cases (outcomes, which patents were asserted in each) — I only confirmed case numbers and the one NPE-database listing.
• Whether an IPR/CBM was filed against the '995 patent — not established by my searches.

I'll research litigation involving US patent 7554995 using multiple sources.

Let me gather more comprehensive data from the NPE litigation database and other sources.

Let me look at the specific case numbers and the litigation portals directly.

Litigation Involving U.S. Patent No. 7,554,995

Important scope note: I searched for the specific number 7,554,995 (and the literal string 7554995). Because this patent sits in a large Nomadix patent family (which also includes U.S. Pat. Nos. 6,130,892; 7,088,727; 7,194,554; 8,156,246; 6,636,894; etc.), many search results are family-level and mention sister patents rather than the '995 patent itself. Below I separate cases where I could confirm the '995 patent was actually asserted/at issue from family-level litigation that may or may not reach the '995 patent. Where I could not verify an item with confidence, I say so.

A. Cases where the '995 patent was specifically asserted or at issue

1. Nomadix, Inc. v. Hewlett-Packard Company et al.

• Plaintiff: Nomadix, Inc.
• Defendants: Hewlett-Packard Company and iBAHN Corp. (additional co-defendants in the same action; the reported claim-construction decision identifies iBAHN as counterclaimant)
• Jurisdiction: U.S. District Court, Central District of California (No. CV-09-08441 DDP (VBKx))
• Case No.: 2:09-cv-08441
• Filing date: November 2009 (the complaint for infringement of, inter alia, U.S. Pat. No. 7,554,995 is dated Nov. 17, 2009; Law360 reported the suit Nov. 23, 2009)
• Status/outcome: The court issued a Markman/first claim-construction order (Aug. 31, 2011, and a further order reported at 2011 Markman 5037162), construing terms across the asserted Nomadix patents (the '554, '399, '009, '894, '754, '073, '376 and related patents, with the '995 listed among the patents-in-suit in the complaint). I could not confirm a final judgment or settlement on the docket from the sources retrieved; treat the ultimate disposition as unverified.
• Source: Google Patents litigation link (https://portal.unifiedpatents.com/litigation/California%20Central%20District%20Court/case/2%3A09-cv-08441); docket/claim-construction records at https://docs.justia.com/cases/federal/district-courts/california/cacdce/2:2009cv08441/[459068/458](https://assignmentcenter.uspto.gov/search/patent/reelFrameDetail?reelFrame=459068-0458) ; Law360 case page https://www.law360.com/cases/4d2ed0eb5cdfbf6a07000035

2. Nomadix, Inc. v. SIA Mikrotikls et al.

• Plaintiff: Nomadix, Inc.
• Defendants: SIA Mikrotikls; Global IT Communications, Inc.; Creative Wireless, Inc.; Microcom Technologies, Inc.
• Jurisdiction: U.S. District Court, Central District of California (2:19-cv-10202-DDP-E)
• Case No.: 2:19-cv-10202
• Filing date: 2019 (docket last retrieved Mar. 8, 2023)
• Status/outcome: The '995 patent ("System and method for establishing network connection with unknown network and/or user device") was expressly asserted in the Amended Complaint, alongside U.S. Pat. Nos. 7,194,554; 6,636,894; 8,156,246; and 7,088,727, with willful/induced/contributory infringement allegations. The court issued claim-construction/indefiniteness rulings (document filed Mar. 8, 2023, addressing '995 claim 24, among others). Final disposition not confirmed from the retrieved sources — unverified.
• Sources: https://dockets.justia.com/docket/california/cacdce/2:2019cv10202/[766115](/patent/766115) ; https://www.courtlistener.com/docket/16536918/33/nomadix-inc-v-sia-mikrotikls/ ; https://storage.courtlistener.com/recap/gov.uscourts.cacd.766115/gov.uscourts.cacd.766115.147.0.pdf

3. Nomadix, Inc. v. Solutioninc Technologies Limited

• Plaintiff: Nomadix, Inc.
• Defendant: Solutioninc Technologies Limited
• Jurisdiction: U.S. District Court, Central District of California (C.D. Cal., Western Division; No. 2:10-CV-00381 DDP (VBKx))
• Case No.: 2:10-cv-00381
• Filing date: 2010 (exact date not confirmed)
• Status/outcome: The '995 patent was at issue — it appears in Defendants' Supplemental Joint Invalidity Contentions as "Exhibit C3, Claim Comparison for U.S. Patent No. 7,554,995" (dated Jul. 22, 2011). This is also the one case listed for patent 7554995 in the Stanford NPE Litigation Database. Final disposition not confirmed — unverified.
• Sources: https://npe.law.stanford.edu/patent/7554995 ; https://patents.google.com/patent/[US8516083B2](/patent/US8516083B2) ; Google Patents litigation link case/2%3A10-cv-00381

B. Family-level litigation flagged on the patent record (specific '995 assertion NOT confirmed)

The Google Patents "Family has litigation" flag for this file lists these additional C.D. Cal. cases, but I could not verify that the '995 patent itself (as opposed to sister-family patents such as the '554, '246, '727, or '892) was asserted:

• 2:14-cv-00322 (C.D. Cal., 2014) — filing date/details unverified
• 2:11-cv-02604 (C.D. Cal., 2011) — filing date/details unverified
• Source (family litigation links): https://patents.google.com/patent/[US7554995](/patent/US7554995)/en

Also related to the Nomadix campaign but not confirmed as asserting the '995 patent:

• Nomadix, Inc. v. Hospitality Core Services LLC d/b/a Blueprint RF, No. 2:14-cv-08256 DDP (VBKx) (C.D. Cal.) — asserted patents centered on the '246 and '554 patents; the '995 patent does not appear to be the asserted patent in the retrieved filings.
• Nomadix, Inc. v. Guest-Tek Interactive Entertainment Ltd., No. 2:19-cv-04980-AB-FFM (C.D. Cal.), aff'd/discussed on appeal at 9th Cir. No. 20-55439 — this was a license/forum-selection-clause dispute (and parallel PTAB IPR challenges to several Nomadix patents), not a confirmed '995 infringement suit.
• Earlier Nomadix assertions (e.g., Nomadix v. Second Rule LLC, 2:07-cv-01946; IP3 Networks v. Nomadix, 3:04-cv-01485) predate the '995 patent's issuance (June 30, 2009) and therefore do not involve the '995 patent.

Summary

• Confirmed '995 assertions: Nomadix, Inc. v. Hewlett-Packard Co. et al. (2:09-cv-08441, C.D. Cal.); Nomadix, Inc. v. SIA Mikrotikls et al. (2:19-cv-10202, C.D. Cal.); Nomadix, Inc. v. Solutioninc Technologies Ltd. (2:10-cv-00381, C.D. Cal.).
• All confirmed cases are in the Central District of California, consistent with Nomadix's patent-enforcement campaign.
• Outcomes: I could not verify final dispositions (settlement, judgment, or dismissal) for these cases from the sources reached; treat all outcome/status statements above as unverified except the interim rulings (Markman orders in 2:09-cv-08441 and 2:10-cv-00381, and the 2023 claim-construction/indefiniteness order in 2:19-cv-10202).
• Note on the patent itself: Per the Google Patents record, U.S. Pat. No. 7,554,995 is now Expired – Fee Related, with an adjusted expiration of 2018-12-03, and the file carries a "Family has litigation" flag plus a "First worldwide family litigation" entry via Darts-IP (family 25219872). That Darts-IP family record was not accessible in the sources retrieved, so I could not enumerate any foreign/non-U.S. litigation.

Caveat on completeness: Because my searches surfaced family-level records and the searches were limited, this list may not be exhaustive. To be certain of every case specifically asserting 7,554,995 — especially filing dates and final outcomes — a direct docket check on PACER/CourtListener for each case number above, plus the Darts-IP family record, would be required. I did not auto-correct any patent or case number; all identifiers are reported exactly as found.

I'll verify the structured "no proceedings on file" default against live sources before concluding.

Nomadix faced several IPRs — but so far all on other family members. Let me check specifically whether the '995 patent itself was ever challenged.

No hits yet on the '995 itself. Let me probe from a few other angles before concluding.

Proceedings overview

Total AIA trial proceedings on US 7,554,995: zero. The USPTO Open Data Portal structured record (the canonical source supplied in this analysis) returns no IPR, PGR, or CBM proceeding naming this patent, and my independent searches — Google Patents litigation/PTAB links, Stanford NPE Litigation Database (npe.law.stanford.edu/patent/7554995), Docket Alarm PTAB dockets, and keyword probes pairing "7554995"/"7,554,995" with IPR/PTAB terminology — surfaced no petition, no institution decision, no Final Written Decision, and no appeal directed to this patent. Breakdown: active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0.

Bottom-line defensive posture: the '995 is untested at the PTAB — not hardened, not broken. No claim has been canceled, so there is no "the troll has no case" argument to make; but equally, no petitioner has ever forced Nomadix (now Gate Worldwide Holdings / Nomadix) to defend these claims in an AIA trial, so there is no § 315(e)(2) estoppel and no Board precedent construing the claims. The most important practical fact is different: the '995's recorded legal status is Expired – Fee Related, with an adjusted expiration of 2018-12-03. An expired patent cannot support injunctive relief, and damages are limited to the six-year lookback before filing — which substantially reduces both the value of the assertion and the incentive to spend $500k+ on an IPR.

Important negative finding, stated plainly: the absence of PTAB activity is not an artifact of the patent being obscure — it is conspicuous. This patent was asserted in at least two C.D. Cal. campaigns (Nomadix v. Hewlett-Packard, 2:09-cv-08441, where the '995 was one of seven patents in the November 17, 2009 complaint; and Nomadix v. Solutioninc, 2:10-cv-00381). Nomadix's other family members attracted roughly ten IPR petitions from Hospitality Core Services/Blueprint RF and Guest-Tek. Yet the '995 itself was never petitioned. The likely explanations are timing (most IPR activity post-dates the '995's 2018 expiration) rather than any defect in the claims.

No proceedings to report on the '{PROCEEDING_NUMBER}' template

Per the instructions, I do not invent proceeding numbers. There is no proceeding to render in the required per-proceeding format. I am not able to complete the "for each proceeding" sections because the count is zero. Rather than fabricate, below is the adjacent-proceeding context a defendant actually needs — clearly labeled.

Adjacent IPR activity on sibling Nomadix patents (NOT US 7,554,995)

These are different patents in the same family (the '995 is a continuation of Ser. No. 09/684,937 → US 7,088,727 → US 6,130,892). None of them cancels or narrows the '995's claims. I list them only to show the shape of the challenge pattern and to flag which panels/art theory were in play. Where I could not verify an outcome, I say so.

Two additional data points worth recording:

• Nomadix's own public framing (2018-07-02, GlobeNewswire): "Like previous challenges to Nomadix's patents by others, these challenges by Guest-Tek failed… In the decisions denying Guest-Tek's petitions, the PTAB explained that Guest-Tek had failed to demonstrate a reasonable likelihood that it could show unpatentability of any of the claims." The denials issued 2018-06-29; Guest-Tek had (incorrectly) announced institution on 2018-06-21. (press release)
• The one Nomadix PTAB reversal of note cuts against the patent owner: in Guest-Tek's separate contract litigation, Guest-Tek argued — and the C.D. Cal. record reflects — that Nomadix sought a permanent injunction only "after Guest-Tek began winning in the PTAB," and that the injunction forced Guest-Tek to withdraw its IPRs on the eve of the Board's decisions. That dispute went up on the contract/injunction issue, not on any patent: Ninth Circuit No. 20-55439 (appeal from 2:19-cv-04980-AB-FFM). This is not a PTAB appeal and involves no '995 claim.

Explicit uncertainty: I could not verify the institution/termination outcomes for IPR2016-00052, -00073, -00077, -00092, IPR2019-00211, or IPR2019-00253. I also did not find any CBM petition, PGR petition, or ex parte reexamination naming the '995. (The '894 patent did have a reexamination, Control No. 90/007,220 — again, a different patent.)

Strategic summary

Claim status. For US 7,554,995: no claims are canceled, no claims are sustained-in-IPR (because no IPR issued), and all claims are untested at the PTAB. I do not have the verbatim claim text — the earlier-generated section flagged that the authoritative text available was truncated before the Claims section — so I cannot list surviving claim numbers, and I will not guess at them. Any statement that "claims X–Y of the '995 survive" would be fabrication. What is verifiable is the legal status field: Expired – Fee Related, adjusted expiration 2018-12-03. Practically, the entire claim set is presumptively valid and unadjudicated, but the enforceable damages window is largely closed.

Estoppel landscape. Because no IPR was ever instituted on the '995, § 315(e)(2) estoppel is empty for every party. No petitioner, real party in interest, or privy is barred from raising any § 102/§ 103 ground against these claims. Guest-Tek's institution denials on the sibling '899, '922, '857, and '917 patents produced no estoppel on the '995 either (and, more importantly, denials produce no appealable FWD — see St. Jude v. Volcano, ZOLL Lifecor v. Philips). A defendant today therefore has a clean sheet: every prior-art ground is available, subject only to the § 315(b) one-year clock running from service of a complaint alleging infringement of this patent.

Pattern signals. Nomadix's IPR exposure was driven by two repeat petitioners — Hospitality Core Services/Blueprint RF (five petitions, 2015–2016) and Guest-Tek (four-plus petitions, 2017–2019) — and the Board denied institution across the board on the petitions I could verify a status for. Nomadix litigated PTAB issues aggressively, including obtaining (and then defending on appeal) a permanent injunction that forced Guest-Tek to withdraw pending IPRs. No defensive aggregator appears in the chain on the '995 — no Unified Patents filing, no IPR2016/2017 "open market" challenge. Unified Patents' litigation link on the patent page is a C.D. Cal. case portal, not an evidence of a Unified-filed IPR. The patent's assignment chain is also a red flag for due diligence — Google Patents lists Gate Worldwide Holdings LLC as current assignee while recording a 2019-11-22 quitclaim from Gate Worldwide Holdings back to Nomadix, Inc., plus NTT DOCOMO security interests and a 2019 Knobbe Martens security interest released 2019-05-28. Confirm standing/ownership via the USPTO Assignment database before responding to any demand.

Recommended next steps

1. Do not build a defense around PTAB outcomes that don't exist. There is no FWD to cite, no canceled claim to point to. Any letter or motion asserting that "the PTAB invalidated claims of the '995" would be false.
2. Lead with expiration. The '995 is recorded as Expired – Fee Related, adjusted expiration 2018-12-03. Confirm the maintenance-fee/expiration status in Patent Center at USPTO PatentCenter and the recorded expiry date, then build the damages analysis around the six-year lookback from any complaint. An expired patent supports neither injunctive relief nor ongoing royalties.
3. If an IPR is still strategically useful, it remains fully available. No § 315(e)(2) estoppel applies, no § 325(d) "same or substantially the same art" issue has been exhausted by prior petitions on this patent, and there is no General Plastic follow-on problem (there is no earlier petition on the '995 by anyone). File within one year of service of the first complaint alleging infringement of the '995 (§ 315(b)) or the door closes.
4. CBM is off the table. The CBM transitional program sunset on 2018-09-16 and the '995 is not a financial-services patent in any event.
5. Trial-stage tracking is moot — but the sibling track is not. If the demand letter cites other Nomadix patents (the '899, '917, '922, '857, '266, '269, '806, '246, '894), pull the IPR dockets directly from PTAB E2E / PTACTS and Docket Alarm; several of those proceedings' institutions were denied, so the same clean-sheet estoppel analysis likely applies to them as well.
6. Verify my negative finding before relying on it in a filing. My searches returned no '995 proceeding, and the structured ODP block is empty — but I did not query PTAB E2E by patent number directly, and ODP ingest gaps are possible. A one-click confirmation is available at PTAB E2E (search "Application Number: 11/097,925" and "Patent Number: 7554995", plus a party-name search for "Nomadix"). Treat "no PTAB activity" as a strongly supported negative, not a certified absence.
7. Resolve the ownership chain before any substantive response, given the contradictory Gate Worldwide Holdings / Nomadix quitclaim records noted above.

I'll research the assignment record for this patent using targeted searches.

Let me dig deeper into the assignment records and the corporate entities involved.

Ownership Reconnaissance — US 7,554,995 B2

Sourcing caveat up front (read before the timeline): I was unable to retrieve the USPTO Assignment Center records directly. The Assignment Center UI (https://assignmentcenter.uspto.gov/) and the legacy index (https://assignment.uspto.gov/patent/index.html) are JavaScript/query-driven and did not return machine-readable records to my searches. What follows is reconstructed from the Google Patents legal-events feed for this patent, which mirrors a subset of recorded-assignment data.

Consequence: I have recorded dates, conveyance types, assignors and assignees, but I do not have reel/frame numbers, execution dates, or correspondent-of-record for any entry. The task's Signal 3 (repeat correspondent) is therefore not assessable from my data — I will not guess at it. Every "Reel" field below is marked as not retrieved rather than invented. Verification should be done by hand at the Assignment Center search page, and I recommend doing so before relying on this for standing analysis.

Two additional caveats worth carrying forward:

• Assignee data on Google Patents is self-flagged as possibly inaccurate, and its "current assignee" field appears stale (see contradiction flag below).
• These recordings may have been made against the parent application or the portfolio as a whole, not per-patent. In a family with US 6,130,892 → US 7,088,727 → US 7,554,995 → US 8,027,339 → US 8,594,107, a single reel/frame commonly covers dozens of patents. Do not assume a reel/frame retrieved for one family member is exclusive to '995.

Flagged contradiction with the prior section: the prior section records Google Patents listing Gate Worldwide Holdings LLC as current assignee, while the same feed records a 2019-11-22 quitclaim from Gate Worldwide Holdings LLC back to Nomadix, Inc. These cannot both be true propositions about present ownership. Google's field is likely stale (last refresh before the quitclaim, or it is reading the 2018 transfer as the last assignment and ignoring the quitclaim). I could not resolve this; treat current ownership as unconfirmed pending a manual Assignment Center pull. A minor secondary discrepancy: Unified Patents' portal renders the priority date as 1997-03-11 where Google Patents renders 1997-03-12. I note it but cannot adjudicate it.

Inventors

Key structural finding — the assignee did not exist when the invention was filed. The '995 patent claims priority to 1997-03-12 (application 08/816,174). Nomadix, Inc. was founded in 1998 in Santa Monica, California. So neither inventor was an employee of the original assignee at the time of filing, and the inventors' interest had to be assigned to a company formed later.

Corroborating fact in the patent text: the patent carries a government-interest statement — "The U.S. government may have rights in this invention as provided for by the terms of Contract No. DAAH01-97-C-R179 awarded by DARPA." The 1997 contract predates Nomadix's incorporation, consistent with university/contractor-funded work. I could not determine the named contracting institution from the DARPA contract number, so I will not assert UCLA held rights.

No inventor-departure pattern. The "all inventors leave within 12 months → fire-sale precursor" heuristic is inapplicable: Kleinrock and Short moved into the assignee (chairman; CTO) rather than out of it. The relevant anomaly here is the reverse and is discussed under the timeline — an 11-year delay before the inventors' assignment was recorded.

Original assignee

Nomadix, Inc. (name on the issued patent), Agoura Hills / Westlake Village / Woodland Hills, California.

• Line of business: network access gateway hardware and embedded software for public-access and hospitality networks (Visitor-Based Networks). The product line is real and shipped: the Nomadix Universal Subscriber Gateway shipped September 1999, followed by the USG II (2003), the AG 5800 access gateway (2012), and the Nomadix Service Engine (NSE) software licensed to third parties including Agere Systems (2002) and packaged with Boingo Wireless (2002). This directly undermines any "never practiced the invention" theory — the '995 family's packet-interception and translation functionality is the core of the USG product.
• Ownership history: acquired by Singapore-based MagiNet (announced December 2006); MagiNet then acquired by DOCOMO interTouch Pte. Ltd., a subsidiary of NTT DOCOMO, announced December 2007 (~$150M). Nomadix operated as a DOCOMO interTouch subsidiary through the 2009–2014 litigation campaign.
• Current status: operating. Nomadix, Inc. is part of Assa Abloy, which acquired it in March 2024. It continued to file new patents as recently as 2024 (e.g., Passpoint/loyalty-integration filings per Nomadix's own 25-year history post).
• Note on the corporate family: a 2019 court filing by Nomadix identified GWHSPV, LLC as Nomadix's parent and Gate Worldwide Holdings LLC as the parent of GWHSPV. That is the single most probative piece of evidence in this whole record — it establishes that the 2018 transferee was an affiliate holding company inside the Nomadix family, not an unrelated third-party shell.

Assignment timeline

All entries: execution date not separately retrieved; reel/frame not retrieved; correspondent not retrieved. Dates shown are recording dates from the Google Patents legal-events feed. I could not verify whether each recording is patent-specific or portfolio-wide.

1. recorded 2008-09-29 — Reel not retrieved

• Conveyance: Assignment of Assignors' Interest (initial inventor assignment)
• Assignor: Kleinrock, Leonard; Short, Joel E.
• Assignee: Nomadix, Inc.
• Correspondent: not retrieved — cannot assess recurrence
• Context: Curative / confirmatory inventor assignment recorded ~11 years after the 1997 priority filing, landing within roughly nine months of the December 2007 DOCOMO interTouch acquisition. This is textbook acquisition due-diligence cleanup — a chain-of-title gap being closed so the buyer's counsel can deliver a clean opinion. Analyst note: the delay itself is the finding. An 11-year gap between invention and recorded assignment is exactly the kind of defect that later surfaces in a § 281 standing challenge.

2. recorded 2015-09-30 — Reel not retrieved

• Conveyance: Security Interest (patent security agreement)
• Assignor: Nomadix, Inc.
• Assignee: NTT DOCOMO, Inc.
• Correspondent: not retrieved
• Context: Securitization — parent-company lien over the subsidiary's patent portfolio. Financing/credit-support documentation, not a change in beneficial ownership.

3. recorded 2017-08-29 — Reel not retrieved

• Conveyance: Modification to Patent Security Agreement
• Assignor: Nomadix, Inc.
• Assignee: NTT DOCOMO, Inc.
• Correspondent: not retrieved
• Context: Securitization — amendment of the 2015 lien, typically a re-advance, covenant reset, or extension. Two consecutive DOCOMO security filings two years apart is a financing signal.

4. recorded 2018-01-23 — Reel not retrieved

• Conveyance: Assignment of Assignors' Interest
• Assignor: Nomadix, Inc.
• Assignee: Gate Worldwide Holdings LLC
• Correspondent: not retrieved
• Context: Internal reorganization / transfer to affiliate holding company (the "Holdings" step). Supported by the 2019 court disclosure that Gate Worldwide Holdings sits above Nomadix through GWHSPV, LLC. On its face this is the shape of a shell-entity transfer; the affiliate relationship is what defeats that reading.

5. recorded 2018-01-23 — Reel not retrieved

• Conveyance: Assignment of Patent Security Agreement
• Assignor: NTT DOCOMO, Inc.
• Assignee: Gate Worldwide Holdings LLC
• Correspondent: not retrieved
• Context: Securitization housekeeping — the existing DOCOMO lien is conveyed to the new holding entity so the security and the collateral sit in the same hands. Recorded same day as entry 4, which is the tell that entries 4 and 5 are two halves of one transaction.

6. recorded 2019-03-20 — Reel not retrieved

• Conveyance: Security Interest
• Assignor: Nomadix, Inc.
• Assignee: Knobbe, Martens, Olson & Bear, LLP
• Correspondent: not retrieved
• Context: Securitization in favor of outside counsel. A law firm taking a security interest in the client's patents is normally an unpaid-fee / fee-deferral arrangement, and Knobbe Martens is Nomadix's long-running patent prosecution and litigation firm. This is the strongest distress signal in the chain — but note it is a lien, not an assignment, so it is not a "known asserter" event. A law firm is not an NPE.

7. recorded 2019-05-28 — Reel not retrieved

• Conveyance: Release by Secured Party
• Assignor: Knobbe, Martens, Olson & Bear, LLP (as secured party, releasing)
• Assignee: Nomadix, Inc.
• Correspondent: not retrieved
• Context: Lien discharge — the March 2019 Knobbe security interest is extinguished roughly two months later. Consistent with fees being paid or the arrangement being unwound.

8. recorded 2019-11-22 — Reel not retrieved

• Conveyance: Quitclaim
• Assignor: Gate Worldwide Holdings LLC
• Assignee: Nomadix, Inc.
• Correspondent: not retrieved
• Context: Reversal of the 2018 holding-company transfer — title pulled back down into the operating company. Timing is conspicuous: recorded ten days before the December 2, 2019 filing of Nomadix, Inc. v. SIA Mikrotikls et al., 2:19-cv-10202 (C.D. Cal.). Pulling title into the plaintiff immediately before filing is the standard move for establishing a clean standing record.

Post-chain: no assignment recorded after 2019-11-22. Ownership per the recorded record terminates at Nomadix, Inc., which was acquired by Assa Abloy in March 2024. No record of a further assignment of the portfolio to Assa Abloy appears in this feed — expected if the acquisition was a stock purchase (the subsidiary keeps record title).

Timeline diagram

timeline
    title Ownership of US 7554995
    1997 : Priority application filed by Short and Kleinrock
    1998 : Nomadix Inc founded
    1999 : Universal Subscriber Gateway ships
    2006 : Nomadix acquired by MagiNet
    2007 : MagiNet acquired by DOCOMO interTouch
    2008 : Inventor assignment recorded to Nomadix
    2009 : First infringement campaign launched
    2015 : NTT DOCOMO security interest recorded
    2017 : Security agreement modified
    2018 : Patents transferred to Gate Worldwide Holdings
         : DOCOMO security interest conveyed same day
    2019 : Knobbe Martens security interest recorded
         : Knobbe Martens release recorded
         : Quitclaim back to Nomadix
         : New suit filed against SIA Mikrotikls
    2024 : Nomadix acquired by Assa Abloy

NPE / troll-pattern signals

1. Shell-entity transfer — unclear.
The 2018-01-23 transfer (entries 4–5) moved the portfolio to an entity bearing the "Holdings" suffix — a named tell. But the same-day conveyance of the DOCOMO security agreement, and Nomadix's own 2019 court disclosure identifying "GWHSPV, LLC, Parent Company of Nomadix, Inc.; and Gate Worldwide Holdings LLC, Parent Company of GWHSPV, LLC," establish an affiliate holding structure inside the Nomadix family, not a third-party licensing shell. The transfer was also reversed by the 2019-11-22 quitclaim. I could not verify Gate Worldwide Holdings' address, formation state, or whether it has any products in commerce — so I cannot call this present. It is not "not present" either. Directionally, the affiliate relationship is the dominant evidence.

2. Known asserter in the chain — not present.
No entity in the recorded chain — Kleinrock/Short, Nomadix, Inc., NTT DOCOMO, Inc., Gate Worldwide Holdings LLC, or Knobbe Martens — matches the enumerated lists (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities). Important nuance that is not a substitute for this signal: Nomadix, Inc. is itself a high-volume patent plaintiff — it sued eight companies in November 2009 including HP, Wayport, iBAHN, LodgeNet, Guest-Tek, Aruba Networks and Superclick; obtained a $3.2M judgment and permanent injunction against Second Rule (March 2009); sued Blueprint RF (October 2014); and monetized through licenses to HP (Nov 2012), AT&T/Wayport/Superclick (Mar 2013) and Aruba (Sep 2013). But it did so as an operating company selling the accused product category, which is the distinction that matters and pushes this toward the operating-company verdict rather than an NPE one.

3. Repeat correspondent across the chain — not assessable.
I could not retrieve any correspondent-of-record data. I am explicitly declining to name an attorney or firm on this point rather than infer one from Knobbe Martens' role as litigation/prosecution counsel — being named as secured party on reel entries 6–7 is not the same as being correspondent of record, and conflating the two would fabricate a finding. This is the single largest hole in the analysis and should be the first thing checked manually. If a single recording attorney recurs across entries 4, 5 and 8, that materially changes the weight of signals 1 and 4.

4. Cascading transfers — present, but non-probative.
Four recordings within roughly 22 months (2008-01-23 ×2, 2019-03-20, 2019-05-28, 2019-11-22 — i.e., entries 4, 5, 6, 7, 8). The cadence matches the signal. The substance does not: the parties are an affiliate holding company (entries 4–5, 8), the parent (DOCOMO, entry 5), and a law firm taking and releasing a lien (entries 6–7). There is no chained sequence of unrelated single-purpose LLCs, and no evidence of shared correspondent addresses or common principals beyond the disclosed Nomadix/GWHSPV/Gate affiliation. I mark the timing present and the pattern not present.

5. Pre-litigation transfer — not present for the first suit; present on timing for the 2019 re-filing.
The first infringement suit naming this patent, per the Stanford NPE database, is Nomadix, Inc. v. Solutioninc Technologies Ltd., 2:10-cv-00381 (C.D. Cal.), in the campaign that traces to the November 2009 filings. The nearest preceding assignment is the 2008-09-29 inventor assignment — roughly 14 months earlier, outside the 6-month window. So the signal fails for the first suit.
However, for the 2019 action, the 2019-11-22 quitclaim recording precedes the 2019-12-02 complaint in 2:19-cv-10202 by ten days. That is a clean standing-record sequence and I flag it as notable. Caveats: (i) I could not confirm that the '995 patent specifically was among the patents asserted in 2:19-cv-10202 (Google Patents associates the case with the family; the docket does not enumerate patents); and (ii) recorded dates can lag execution dates. I therefore weight this as an observation, not a finding.

6. Bankruptcy fire-sale — not present.
No Chapter 7 or Chapter 11 proceeding for Nomadix appears in any source I retrieved. The company's transitions were M&A events (MagiNet 2006, DOCOMO interTouch 2007, Assa Abloy 2024), not insolvency sales. Note the distinction: the 2019 Knobbe Martens security interest (entry 6) and the 2018 holding-company step are consistent with financial stress, but stress is not bankruptcy, and I have no filing to cite.

7. Privateering — not present.
The relationship that superficially resembles privateering is DOCOMO interTouch (parent) ↔ Nomadix (subsidiary asserting patents). That is vertical integration within one corporate family, not an arm's-length operating company funding a separate NPE to attack competitors. I found no SEC filing, Patent Progress or EFF coverage suggesting an outsourced assertion vehicle. (Nomadix is a private subsidiary, so SEC coverage would be thin regardless — I note the evidentiary limitation.)

8. Defensive aggregator — not present.
The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at Nomadix, Inc., an operating company now inside Assa Abloy. Do not be misled by Unified Patents' portal listing US-7554995-B2 — Unified maintains a monitoring and prior-art database covering asserted patents; appearing there is a record of being tracked, not of being owned.

Verdict

Operating-company assertion.

Nomadix, Inc. — the assignee of record from the 2019-11-22 quitclaim and the entity that has held title continuously since — is an operating company whose Universal Subscriber Gateway is the commercial embodiment of the '995 family's interception-and-translation functionality, and the asserted lawsuits (Second Rule, HP, Wayport, iBAHN, LodgeNet, Guest-Tek, Aruba, Superclick, Blueprint RF) target actual competitors and product-category participants, not end users or unrelated industries. None of the eight NPE signals resolves affirmatively on the evidence I could retrieve; the two that show factual activity — the 2018 holding-company step and the 2019 quitclaim timing — are explained by an intra-family reorganization and a standing cleanup, and the only true "shell" tell is neutralized by Nomadix's own court disclosure that Gate Worldwide Holdings is its ultimate parent. Confidence is moderate, not high, for two reasons that a manual Assignment Center pull could resolve either way: (a) the 2018–2019 sequence of a Holdings-LLC transfer plus a law-firm security interest plus a quitclaim-10-days-before-filing is an unusual cluster that warrants monitoring; and (b) I retrieved no correspondent-of-record data, so I cannot rule out that Signal 3 — the strongest shell-family tell — is present.

Verification links:

• USPTO Assignment Center — search "7554995": https://assignmentcenter.uspto.gov/
• Legacy index (same underlying records): https://assignment.uspto.gov/patent/index.html
• Google Patents legal events (source of the dates above): https://patents.google.com/patent/US7554995/en
• Docket, Nomadix, Inc. v. SIA Mikrotikls et al. (interested-parties disclosure naming GWHSPV/Gate): https://dockets.justia.com/docket/california/cacdce/2:2019cv10202/[766115](/patent/766115)
• Stanford NPE Litigation Database entry: https://npe.law.stanford.edu/patent/7554995

Highest-value manual checks, in priority order: (1) correspondent-of-record on each reel/frame; (2) execution dates vs. the 2019-12-02 complaint; (3) Gate Worldwide Holdings LLC's formation state and registered-agent address; (4) whether the 2019-11-22 quitclaim covers '995 or only a subset of the portfolio.

I'll research the prior-art citations for US 7,554,995. Let me run targeted searches.

Let me pull the citation lists from patent databases directly.

Let me try to retrieve the front-page reference list and litigation invalidity contentions.

US 7,554,995 — Prior-art citation analysis

0. Bottom line up front

I was not able to retrieve the actual front-page "References Cited" list for US 7,554,995 through the searches available to me. The Google Patents full-text page I could reach renders the specifications/abstract/description, and the results returned to me repeatedly truncated before the "Patent Citations / Cited By" tables and before the Claims. I will not invent a list of cited references, because a fabricated "prior art" table for a specific patent is worse than an empty one.

Below I give you (a) exactly what I did verify, (b) the one specific reference I can tie to this patent family's prosecution with a citable source, (c) the § 102 framework that governs how the '995 citations must be analyzed (this is where the real analytical value is, and it is not trivial because of the CIP/continuation chain), and (d) the exact places to pull the authoritative list.

Consistency flag: the prior section's caveat still stands and I am not overriding it — the authoritative text available to me remains truncated before the Claims. Anything below about claim scope is analysis of effective priority dates, not a restatement of claim language.

1. What I verified about the '995 patent (unchanged from prior section, restated only where needed)

USPTO-database caveat: I did not query PatentCenter/Patent Public Search directly (those require an interactive session). Google Patents, Justia and the Stanford NPE Litigation Database were my reachable proxies. Treat the following as a verified-fragment report, not a certified USPTO extraction.

2. The one reference I can tie to this family's prosecution — with a caveat you must not ignore

US 6,385,653 — the "Sitaraman patent."

• Citation form: U.S. Patent No. 6,385,653 (Sitaraman).
• Where I found it: the claim-construction order in Nomadix, Inc. v. Hewlett-Packard Co., No. 2:09‑cv‑08441 (C.D. Cal., Judge D. Pregerson) — https://app.midpage.ai/document/nomadix-inc-v-hewlett-packard-1000000895881. The order states: "the examiner initially rejected the claims at issue in view of U.S. Patent No. 6,385,653 ('Sitaraman patent'), which involved a 'protocol gateway' that provided network access to client computers, which use application protocols that the gateway is configured to support."
• Description: a protocol gateway providing network access to client computers using application protocols the gateway is configured to support.
• § 102 relevance: § 102(a)/(b)/(e) as a gateway-art reference; in the related prosecution it was the basis for § 103 obviousness rejections that Nomadix ultimately overcame by amending to require "packet translation learned during self configuration."

⚠️ Attribution warning (this is the important part): the Sitaraman rejection is documented for the '554 patent = US 7,195,554, a sibling Nomadix patent litigated in the same case — not verified as a citation in the file history or front page of US 7,554,995. Sibling patents in this family share a common specification (they descend from the same 1997/1998/2000 priority chain) and were prosecuted by the same examiner art unit, so the same references very plausibly appear — but "plausibly" is not "verified." Do not represent Sitaraman as a § 102 reference for the '995 until the '995 front page or file wrapper confirms it.

3. Prior art asserted by defendants at litigation (existence confirmed, contents not retrieved)

The Kleinrock/UCLA document set hosted at UCLA contains defense invalidity contentions for the Nomadix patents, including this entry:

"Exhibit C3, Claim Comparison for U.S. Patent No. 7,554,995 filed in Defendants' Supplemental Joint Invalidity Contentions, NOMADIX, Inc. v. ..."
Source: https://www.lk.cs.ucla.edu/data/files/Kleinrock/slides/US%20Patent%208594107.pdf

This confirms that a document exists mapping specific references to claims of the '995 patent (a defense invalidity chart). I could not read the contents. That exhibit is very likely the single richest source of candidate § 102 art for the '995 claims, and it should be the first thing you pull.

The '995 was asserted in Nomadix, Inc. v. Hewlett-Packard Co., No. CV 09‑8441 (C.D. Cal., filed 2009‑11‑17) and in Nomadix, Inc. v. Solutioninc Technologies Ltd., No. 2:10‑cv‑00381 (C.D. Cal.) — https://npe.law.stanford.edu/patent/7554995 — both of which produced invalidity contentions.

4. Why the § 102 analysis for the '995 is unusually date-sensitive

This is the piece that matters more than any single citation, and it is specific to this patent:

The '995 is a continuation of a CIP of a CIP. That produces three candidate effective filing dates, and pre-AIA § 102 applies (applied for 2005‑04‑01, i.e., before the AIA's 2013‑03‑16 first-to-file switch):

Consequences for § 102:

• A cited reference that postdates 1997‑03‑12 but predates 1998‑03‑12 or 2000‑10‑06 is § 102 art only against claims whose subject matter was first disclosed in the 1998 or 2000 CIP, respectively. A reference cannot be § 102(b) art against a claim entitled to the 1997 date unless it predates 1996‑03‑12.
• Because this is pre-AIA, § 102(e) patents (US patents/ publications granted on applications filed before the applicant's date of invention) are also available as prior art and are a common basis for rejections in this family.
• § 102(g) prior invention and public-use/on-sale activity by others (the 1997–2000 window was the commercial birth of public-access gateway products) is a live defense theory in the Nomadix cases.
• The "intercept without being addressed" limitation (Proxy ARP / promiscuous mode) is dated 1997 in the specification, so it is the hardest element to attack with late-1990s art, whereas the DHCP-interception, passive-learning and manual/SNMP-configuration limitations appear later in the chain and have a later critical date.

5. What I am explicitly telling you I could not confirm (do not treat as findings)

• The front-page "References Cited" list (U.S. patent documents, foreign patent documents, other publications) for US 7,554,995 — not retrieved. No count, no numbers.
• The count/gist of independent claims of the '995 — unchanged from the prior section's caveat.
• Whether any IPR/CBM/PGR was filed against the '995 — not established.
• The contents of Exhibit C3 (defense § 102 charts) — existence only.

6. Where to pull the authoritative citation list (do this before anyone relies on a § 102 table)

1. USPTO Patent Public Search / PatentCenter full-text and the issued front page
   https://ppubs.uspto.gov/dirsearch-public/print/downloadPdf/7554995 and https://patentcenter.uspto.gov (search 7554995 — the printed patent's "[56] References Cited" is the examiner-citation list).
2. Google Patents "Patent Citations" + "Cited By" tables
   https://patents.google.com/patent/US7554995/en — scroll below the description to Patent Citations / Cited By / Similar Documents. (These tables are what my retrieval kept truncating.)
3. Justia front page (mirrors the "[56]" list and "Referenced By")
   https://patents.justia.com/patent/7554995
4. The rule-waiver file history for 11/097,925 in PatentCenter → "Documents & Transactions," which shows Form PTO‑892 (Notice of References Cited) with examiner-applied art.
5. Defendants' invalidity contentions, Exhibit C3 (see UCLA link above) for the litigation-side art that was not before the examiner — often the strongest § 102 candidates.

Once you paste the [56] list and any Form PTO‑892 entries to me, I will produce the requested table — for each reference: full citation, publication/filing date, brief description, and the specific claim(s) it potentially anticipates under 35 U.S.C. § 102(a), (b), or (e) — with the effective-date tier (1997 / 1998 / 2000) applied to each claim so the anticipation theory is date-correct.

I'll start by locating the actual prior-art references cited on the '995 patent's face and in its family, since the authoritative text I was given did not include a prior-art listing.

Good leads. Now let me pull the actual cited references and the claim text for the '995 patent.

Threshold note before the analysis

Two things you need to know about my inputs, because they bound everything below:

1. The "Prior Art" section of the page I was given is not a reference list. What the supplied US 7,554,995 record contains under/around that label is: "Prior art keywords — address; corresponds; target; network; network packet," the prior-art date (1997‑03‑12), the Background Art narrative, the "Definitions" block, and the "Landscapes" classifications. There is no "References Cited" table (no US/foreign patent list, no "Other Publications" list) and, as the prior section already flagged, the text stops before the Claims. I attempted to retrieve both the face-of-patent citations and the verbatim claim language and could not — my searches either returned duplicated specification text or ran out. So: I do not have claim language, and I do not have a verified list of what the examiner cited against this patent.

2. What I can ground the analysis in, and did:

• The specification's own admissions (statements about Mobile IP, about routers/firewalls/proxies in FIGS. 7A–7F) — admissions are prior art regardless of citation.
• The face-of-patent citation lists of sibling Nomadix family members (US 9,548,935; US 7,822,873; US 7,752,334; US 9,491,136), which list the same reference pool this family was prosecuted against.
• The litigation record, which is where the real §103 attacks on the '995 patent were actually made.

I flag each source's reliability inline. Where I'm asserting a reference's content from my own knowledge rather than a fetched document, I say so.

I. Governing law and the level of ordinary skill

• The '995 patent issued from an application filed 2005‑04‑01 with earliest priority 1997‑03‑12, so it is a pre‑AIA patent: 35 U.S.C. § 103(a) (pre‑AIA) governs, with pre‑AIA § 102(g)/(e)/(f) available for swearing behind or § 102(e) art. (Any later family member whose claims lack pre‑2013 support would move to AIA § 103; not the case here as far as I can tell.)
• Controlling framework: Graham v. John Deere; and under KSR Int'l v. Teleflex, predictable combinations, "known technique" substitution, and design incentives matter as much as explicit teaching.
• POSITA — I can cite an actual, on-the-record definition from this very patent family: in Nomadix, Inc. v. Second Rule LLC, No. CV 07‑01946 DDP (VBKx), amended claim construction order dated Oct. 15, 2008, the court adopted: "a post‑high school education in computer science, or related area, but not necessarily a four‑year degree, as well as at least two years working in developing computer applications, or equivalent education and work experience." (order text) That is a low skill level for a networking patent — which cuts strongly in favor of obviousness, because it means the knowledge base charged to the artisan is broad, practical, and includes ordinary network administration practice (NAT, proxy ARP, firewall rules, DHCP).

II. The critical-date problem (do this analysis before the merits)

The cross-reference section (which I do have verbatim) is:

continuation of Ser. No. 09/684,937 (filed Oct. 6, 2000, now US 7,088,727); which is a CIP of Ser. No. 09/041,534 (filed Mar. 12, 1998, now US 6,130,892); which is a CIP of Ser. No. 08/816,174 (filed Mar. 12, 1997, abandoned).

The two CIP links mean the '995 claims get the 1997 date only for subject matter actually disclosed in the 08/816,174 parent. Every limitation that first appears in the 1998 or 2000 or 2005 specifications is entitled only to that later date. This is decisive for §103 because a great deal of this field published in 1996–1999. In particular, the following "unknown network / unknown user device" features read as candidate new matter:

• DHCP interception when no DHCP server exists on the foreign network;
• passive/promiscuous-mode network learning, including electing an unused IP and switching if another device takes it;
• port-table remapping to support multiple hosts behind one router address;
• application-layer translation of DNS server names / FTP-embedded addresses / proxy server addresses.

Consequence: a challenger should not concede the 1997 date. If any of those limitations carries a 1998‑03‑12 or 2000‑10‑06 effective date, references published in 1996–1998 move from "too late" to "§102(b) art." A single Jun. 14, 1997 item is illustrative: the Check Point FireWall‑1 White Paper v.3 (June 1997) — on the family's reference list (sibling patent list) — is too late for a 1997‑03‑12 claim but squarely prior art for anything dated 1998 or later. I flag that as a live, date-sensitive lever, not as a conclusion.

III. The prior-art universe, by reliability tier

Tier A — admitted prior art inside the '995 specification (unimpeachable):

• RFC 2002, Mobile IP (Perkins, Oct. 1996) — the spec expressly states: "The internet RFC 2002 Mobile IP protocol specifies the mapping between permanent and temporary IP addresses. The unique aspect of the nomadic router is that the Mobile IP protocols are not necessarily running in, or supported by, the host CPU but rather are internal to the nomadic router." That sentence is the applicant conceding the entire permanent-address/temporary-address mapping, and identifying the only asserted point of novelty as locus (in the router, not the host).
• RFC 2002's home agent / foreign agent / care-of-address / encapsulation model, also admitted.
• The spec's own FIGS. 7A–7F admissions that routers translate at the network layer, firewalls filter at network/transport layers, and proxies/gateways "manipulate packets at the application level."

Tier B — references of record on sibling family patents (verified as listed; contents largely from my own knowledge):

Tier C — prior art actually asserted in litigation (the real §103 battleground):
From the Nomadix docket, Defendants' Supplemental Joint Invalidity Contentions in Nomadix v. Hewlett‑Packard, No. 09‑CV‑8441 DDP (VBKx), and Nomadix v. Solutioninc, No. 2:10‑CV‑00381 DDP (VBKx), included Exhibit C3, "Claim Comparison for U.S. Pat. No. 7,554,995" (25 pages), and Exhibit B, "Supplemental Prior Art List" (source; source). Named art in that record includes:

• RFC 1919, "Classical versus Transparent IP Proxies" (Clausell, Mar. 1, 1996);
• Chapman, "Network (In)Security Through IP Packet Filtering," USENIX Security, Sept. 1992;
• Chapman & Zwicky, "Building Internet Firewalls," O'Reilly, Sept. 1995;
• Check Point FireWall‑1 White Paper v.3, June 1997;
• Casey, "Realizing Mobile Computing Personae," Ph.D. Thesis, Oct. 1995;
• the "SPINACH" prior art, the Vos publication, the Ko Ikudome prior-art work, and John Birrell, Jr.

I could not retrieve the contents of the SPINACH, Ikudome, Vos, or Birrell references — these are the specific §103 combinations the defendants actually built, and my analysis below is therefore a reconstruction of the most likely combinations, not a report of theirs.

IV. Limitation-by-limitation mapping (against the disclosure, since I lack the claims)

Because the claim text is unavailable, I map to the six operative elements the prior section identified from the specification. Everything in this table is inference about claim scope, expressly labeled as such.

V. The five combinations I would run, with motivations

Combination 1 — Mobile IP (RFC 2002) + NAT (RFC 1631) + Transparent Proxy (RFC 1919).
Motivation, expressly cognizable under KSR: RFC 2002's own deployment model required either host-side Mobile IP stack support or a cooperating foreign agent on the visited network — i.e., it required the foreign network's participation. Both RFC 1919 and the firewall literature were directed at exactly the opposite problem: serving a client that cannot be modified. A POSITA facing the admitted problem ("the host is configured for its home network and will not find its gateway") had two known tools — transparent interception and address translation — and the express design incentive (hotels, airports, multi-tenant buildings, all named in the '995 specification) to put them in a box in the host's path rather than in the host. The only asserted novelty over Mobile IP is where the mapping lives, which is a locus-of-implementation choice.

Combination 2 — US 5,371,852 (Attanasio) + DHCP (RFC 1541/2131) + RFC 1256 Router Discovery.
Motivation: Attanasio supplies gateway-performed, bidirectional address substitution transparently, on behalf of hosts that are not individually addressed on the outside network — structurally the same as "the router pretends to be the host." DHCP and Router Discovery supply the automatic acquisition of the other unknown (network configuration). Putting an auto-configuring translator on the network so that roaming hosts need no manual reconfiguration is the natural, predicted use of both. Result is a fully predictable aggregation of known functions.

Combination 3 — Firewall/proxy art (Chapman 1992; Chapman & Zwicky 1995; Check Point FireWall‑1 v.3, Jun. 1997) + NAT.
Motivation: Firewalls were already two-interface devices that received traffic not addressed to them, inspected it, and rewrote/filtered it; NAT rewrote addresses and checksums. The '995 specification itself distinguishes firewalls only by saying they "do not manipulate the content of the packet" — a distinction dissolved the moment you add NAT to a firewall, which by 1997 was routine (the white paper above concerns precisely a NAT-capable firewall). Date caveat: the June 1997 Check Point paper works only if the claims read on post‑1997 subject matter (see §II). Chapman 1992/1995 is clean.

Combination 4 — SPINACH / Ikudome / Vos / Birrell art (as actually asserted).
These are the combinations the defendants reduced to 25-page claim charts against the '995 patent (Exhibit C3). I cannot reconstruct them without the documents. Anyone doing a real §103 opinion must obtain Exhibit C3 and Exhibit B from the 09‑CV‑8441 / 2:10‑CV‑00381 dockets before finalizing anything.

Combination 5 — "Obvious to try," narrowly.
Pre-1997, it was known to (a) intercept, (b) translate addresses, and (c) auto-discover configuration. The only genuine question is whether selecting all three at every protocol layer simultaneously was a finite, predictable set of options. Under KSR, a finite number of identified, predictable solutions supports §103 — but In re O'Farrell cuts the other way if the prior art gave no direction as to which of the options to pick. I regard this as the strongest non-obviousness redoubt for the patentee (see below).

VI. Where the patentee has real arguments (put these in the opinion)

1. The "no client software" limitation. If the claims require that the host remain entirely unmodified — no Mobile IP stack, no client agent, no reconfiguration — then RFC 2002's own teaching (host-side Mobile IP, or foreign-agent cooperation) does not suggest eliminating the host-side component, and one could argue it teaches away from pure transparency. That is the strongest defense and it is reinforced by the fact that the HP Markman order construed "absent additional agents implemented by the user's computer" to mean "absent additional special client software implemented by the computer for managing the communication between the computer and the gateway device" (2011 Markman order).
2. Multi-layer translation. RFC 1919 addresses application-layer transparent proxying; NAT addresses network-layer rewriting; no single reference requires both, plus transport-layer port remapping, plus link-layer interception in one device. A patentee will argue the combination is not "additive" because layer-crossing translation introduces ordering interdependencies (the spec itself notes the transport pseudo-header must wait on the application layer).
3. Unknown-both-sides. The specific "learn host config and network config, elect an unused IP, and switch IPs if another device takes it" flow is arguably a specific, non-routine algorithm. But note the date exposure: this is the feature most likely to be 1998/2000 new matter, which makes Casey (Oct. 1995) and the later 1996–1997 art newly available.
4. Secondary considerations to look for (not established by me): commercial success of the Nomadix gateway products (the Nomadix product/patent listing evidences a real product line), industry adoption, and third-party licensing. I did not find a nexus-analyzed record on these points and will not assert one.

VII. Bottom line

• On the record I can verify, the '995 disclosure is highly vulnerable to a §103 attack built from (i) RFC 2002 Mobile IP, which the applicant admitted as prior art and distinguished only on location of implementation, plus (ii) NAT (RFC 1631) and (iii) transparent-proxy/firewall art (RFC 1919; Chapman 1992/1995). The motivation to combine is not merely "same field" — it is the express deployment limitation of Mobile IP itself, plus the named market need (hotels, airports, MDUs) recited in the specification.
• The single most important procedural lever is priority, not references. The dual-CIP chain (08/816,174 → 09/041,534 → 09/684,937) means several of the distinguishing limitations are probably not entitled to the 1997 date, which promotes 1996–1998 publications into §102(b) art. I would not accept the 1997 date without a limitation-by-limitation priority analysis against the 08/816,174 disclosure.
• The actual §103 case the defendants made is in Exhibit C3 (25 pages) and Exhibit B from 09‑CV‑8441 / 2:10‑CV‑00381 — documents I identified but could not retrieve. My Combinations 1–3 are reconstructions of the most probable theories, not a report of theirs.
• Caveat repeated: I never obtained the '995 claims. Every limitation mapping above is against the specification's disclosed elements. An opinion of obviousness cannot be finalized without the claim text — the claim scope determines whether element (1)'s "without being addressed to the device" and element (4)'s application-layer translation are actually recited, and those two elements carry most of the patentee's non-obviousness argument.
• U.S. government rights are recorded (DARPA Contract No. DAAH01‑97‑C‑R179), and the patent's recorded status is Expired – Fee Related, adjusted expiration Dec. 3, 2018 — consistent with the prior section; a present-day §103 challenge would matter chiefly for the still-live family members, not for the '995 claims themselves.