US 9332478

US Patent 9,332,478 B2 — Verified Summary

Bibliographic data (confirmed across Google Patents, FreePatentsOnline, and USPTO-adjacent records):

Abstract (verbatim):

A method includes transmitting communications information from a first DMA gateway of a DMA gateway communication network to a second DMA gateway of the DMA gateway communication network. The communications information is associated with a communication network that is accessible to the first DMA gateway. The method includes receiving a communication from the second DMA gateway via the DMA gateway communication network. The communication network and the DMA communication network are different. The communication is associated with a destination device that is indicated by the communications information to be served by the communication network that is accessible to the first DMA gateway. The method includes routing the communication to the destination device.

Independent claims — plain-language overview

⚠️ Caveat: The full, numbered claims text was not retrievable in full during this session (the authoritative Google Patents page truncates before the claims section, and the fetched specification text did not include the claims). The patent's Detailed Description enumerates several "particular embodiment" paragraphs that mirror the independent-claim structure. Based on that specification language — which is consistent with the abstract and with the family's other granted patents — the independent claims appear to cover approximately the following subject matter (treat the claim-by-claim mapping as approximate, not verbatim):

1. Method claim (gateway-to-gateway routing). A method in which a first DMA gateway transmits communications information about a network (a DMA server network or a legacy communications network) that the first DMA gateway serves to a second DMA gateway over a DMA gateway communications network; the first gateway later receives from the second gateway a communication destined for a device on the served network and routes that communication to the destination device. Key limitation: the served communication network and the DMA gateway network are different networks.

2. Computer-readable-medium claim (receiving gateway side). A non-transitory storage medium storing instructions causing a processor at a first DMA gateway to receive communications information about a network accessible by a second DMA gateway, store it in memory (e.g., a home DMA register), receive a communication directed to a device served by the second gateway, and route it by relaying it to the second DMA gateway over the DMA gateway communications network.

3. Computer-readable-medium claim (satellite-gateway outbound call). Instructions causing a DMA server, while in a service area of a first DMA gateway (e.g., an orbiting satellite), to receive first routing instructions; to send a first call (received from a mobile device via an integrated wireless transceiver) to the first DMA gateway over a private IP network; the call being placed to a destination device reachable through a legacy communications network accessible via that first DMA gateway.

4. Computer-readable-medium claim (satellite-gateway inbound call). Instructions causing an orbiting satellite to receive a call on a first interface adapted to communicate with a legacy communications network (the call directed to a mobile device served by a first DMA server), and to route the call to that DMA server via a second interface adapted to communicate with a private IP network, where the DMA server includes a wireless transceiver for the mobile device.

5. DMA gateway apparatus claim. A first DMA gateway having a first interface to a legacy communications network, a second interface to a private IP network, and a third interface to a DMA gateway communications network, plus logic to receive legacy-network information from a second DMA gateway (indicating that legacy network is in the second gateway's range) and to forward communication received from a DMA server to the second DMA gateway for placement to a destination on the legacy network. Embodiment: the first DMA gateway is an orbiting satellite.

6. DMA server apparatus claim. A DMA server coupled to a wireless transceiver, with a private-IP interface and a satellite-communications interface, and logic to receive a call from a mobile device via the transceiver and send call information to a DMA gateway that is an orbiting satellite, for placement to a destination on a legacy network accessible via that gateway.

7. Orbiting-satellite apparatus claim. A satellite with a legacy-network interface and a private-IP interface and logic to receive a call via the legacy-network interface (directed to a mobile device served by a DMA server) and route it to the DMA server via the private-IP interface, the DMA server having the wireless transceiver for the device.

CAFC 2026 docket search result

I searched for a Federal Circuit 2026 docket tied specifically to 9,332,478. I did not find any CAFC proceeding citing this exact patent number.

What I did find is a Federal Circuit decision reported April 10, 2026 in which Lemko Corp. lost its bid to revive claims in a "distributed mobile architecture" patent after the Federal Circuit backed a PTAB invalidity finding favorable to Microsoft. However, the underlying PTAB documents I retrieved tie that appeal (Federal Circuit No. 24-2336, Lemko Corp. v. Microsoft Corp.) to family members U.S. Patent 7,855,988 ('988) and U.S. Patent 8,310,990 ('990) in IPR2023-00531 — not to the '478 patent itself. Because '478 is a later continuation in the same family (via '988), I cannot rule out indirect relevance, but I have no evidence that '478 was the patent at issue in that appeal. Treat any linkage between the April 2026 CAFC decision and 9,332,478 as unconfirmed.

There is also separate district-court litigation history associated with the '478 family (e.g., Texas Northern District Court case 3:22-cv-00363, per Unified Patents), but no CAFC 2026 docket specifically naming 9,332,478 was found.

Bottom line: For the bibliographic data, abstract, and technical overview above, I have high confidence. For the exact independent-claim text and numbering, and for any 2026 CAFC involvement specific to this patent, I have low-to-medium confidence — the claim language should be verified against the full USPTO claims text (PatentCenter / Global Dossier) before relying on it.

Litigation Involving U.S. Patent No. 9,332,478

Based on searches of the Unified Patents litigation database, district court dockets (N.D. Tex.), and related PTAB records, one principal district-court case is associated with U.S. Patent 9,332,478:

Lemko Corporation v. Microsoft Corporation and Affirmed Networks, Inc.

• Plaintiff(s): Lemko Corporation
• Defendant(s): Microsoft Corporation and Affirmed Networks, Inc.
• Jurisdiction / Court: U.S. District Court for the Northern District of Texas, Dallas Division (Judge Sam A. Lindsay; Magistrate Judge Rebecca Rutherford through referral "L-BT")
• Case number: 3:22-cv-00363-L-BT (docket No. 3:2022cv00363 / gov.uscourts.txnd.359096)
• Filing date: Complaint filed February 14, 2022 (the Unified Patents case list shows February 13, 2022; docket entries reflect February 14, 2022)
• Cause of action: Patent infringement (35 U.S.C. § 271)
• Patents at issue: Lemko asserted multiple DMA-related patents; the live pleadings describe eight asserted patents (the '763, '414, '988, '409, '111, '980, '478, and '931 patents). The '478 patent description in the pleading ("transmitting communications information from a first DMAG to a second DMAG in a DMAG communication network...") corresponds to U.S. 9,332,478, and Unified Patents links U.S. 9,332,478 to this case.
• Accused products: Microsoft's and Affirmed Networks' products/services centered on Azure Stack Edge (ASE) and Virtual Evolved Packet Core (vEPC) / Azure Operator Nexus. (Note: Microsoft acquired Affirmed Networks in March 2022.)

Outcome / current status:

• Unified Patents currently lists the case status as "Closed."
• However, the most recent substantive docket activity reflects an ongoing case through mid-2025:
  • On June 4, 2025, the court (Judge Lindsay) entered an order accepting the Magistrate Judge's Findings and Recommendations on Defendants' Motion to Dismiss Plaintiff's Second Amended Complaint (Doc. 94). The motion was granted in part and denied in part: Lemko's contributory patent-infringement claims were dismissed with prejudice, and the motion was otherwise denied (i.e., direct and induced-infringement claims survived). Lemko was not permitted further amendment.
  • Lemko's April 15, 2025 press release confirms the court denied Microsoft's motion to dismiss Lemko's direct and indirect (induced) infringement claims.

Because the Unified Patents status field says "Closed" while the last docket order (June 2025) still resolves motions in a live case, I cannot confirm from the available records whether the case has since been finally terminated (e.g., by settlement, judgment, or stipulation) — the "Closed" flag may reflect a recent termination, but no terminating document or judgment appears in the search results I retrieved.

Related PTAB proceedings (for context)

Microsoft and Affirmed Networks filed inter partes review petitions against Lemko DMA patents, e.g., IPR2023-00529 (U.S. 7,653,414) and IPR2023-00531 (U.S. 7,855,988). I did not find any IPR specifically directed to U.S. 9,332,478 in the available search results, and I do not want to speculate as to one.

Caveats

• I did not locate any other district-court or ITC matter naming U.S. 9,332,478. If additional cases exist, they are not reflected in the sources I searched (Unified Patents, CourtListener, Justia, Patexia, Docket Alarm, and Lemko's own announcements).
• The Unified Patents case-list entry for patent 9,332,478 shows only "Affirmed Networks Inc" as defendant in its summary row, but the actual case caption and docket include both Microsoft Corporation and Affirmed Networks, Inc. as defendants.
• Best sources to confirm the terminal disposition (if any): PACER / the N.D. Tex. docket for 3:22-cv-00363-L-BT and the CAFC docket if an appeal was filed.

Proceedings overview

Count: 0 AIA trial proceedings on US 9332478. The USPTO Open Data Portal (canonical source for this analysis) returns no IPR, PGR, or CBM naming US 9332478, and independent web searches corroborated that — I found no PTAB petition, institution decision, or Final Written Decision directed at this patent number. The only proceeding that surfaced in searches against Lemko's DMA patent family (Microsoft / Affirmed Networks IPRs on the parent and sibling patents) are not against 9,332,478 and are described separately below so they are not mistaken for proceedings on this patent.

Defensive posture: Because no PTAB trial has ever been instituted on 9,332,478, all of its claims remain in force and are PTAB-untested — no claim has been canceled, no claim has been "hardened" by a FWD, and no § 315(e) estoppel attaches to this patent. A defendant facing assertion of 9,332,478 today is not bound by any IPR estoppel and can raise any § 102/§ 103 ground; however, the patent owner (Lemko, now subject to a 2025-03-25 security assignment to Piccadilly Patent Funding LLC) has won every PTAB merits decision on the closely related family patents, which is a meaningful warning about how hard this family is to knock out.

No proceeding on US 9332478 — but related family activity you should know about (flagged)

The ODP block is authoritative: there is no PTAB proceeding on US 9332478. Web search did, however, surface a coordinated 2023 PTAB campaign by Microsoft Corporation and Affirmed Networks, Inc. against Lemko family patents asserted in the same litigation where 9,332,478 was in suit (Lemko Corp. v. Microsoft Corp. et al., 3:22-cv-00363-L-BT, N.D. Tex., filed 2022-02-13, per Unified Patents: Closed). Because 9,332,478 is a continuation of the '988 patent challenged in IPR2023-00531 (chain: 7,855,988 → 12/942,642 → 13/648,704 = 9,332,478), these outcomes are highly probative. None of the following is a proceeding on 9,332,478:

Highlights from the record I could verify:

• Judge panel (IPR2023-00529/00531): Frances L. Ippolito, Brian D. Range, and Scott Raevsky; oral hearing held 2024-07-11 (transcript on Docket Alarm). Petitioners were represented by Winston & Strawn (Robert N. Kang, Michael Rueckheim); Lemko by Kramer Levin (James Hannah, Aaron Frankel).
• Grounds on the parent '988 (IPR2023-00531): claims 1, 3–9, and 20 alleged anticipated by U.S. 2007/0021120 ("Flore") and obvious over Flore combined with 3GPP TS 23.002 v7.1.0 (per the Proctor declaration and exhibit lists).
• Outcome pattern: Bloomberg Law (2024-09-04) summarized that the Board found challenged Lemko claims not unpatentable as obvious; Unified Patents docket data shows FWDs in both instituted cases and denial of institution in both others. Per GreyB's case page, IPR2023-00531's status is "Final Written Decision – Appealed" — but I could not confirm a Federal Circuit docket number or disposition in the sources available, so do not treat an appeal as verified beyond that status flag.
• Why no IPR on 9,332,478 despite it being asserted in 3:22-cv-00363: I could not confirm the petitioners' rationale from public sources. Plausible explanations include claim-specific strategy or the one-year § 315(b) bar timing (the family IPRs were filed 2023-02-14/15, right at the deadline from the 2022-02 service), but this is inference, not verified fact.

Strategic summary

Claims of 9,332,478 — CANCELED / SUSTAINED / UNTESTED: Every claim of 9,332,478 is UNTESTED at the PTAB. No AIA petition has been filed against this patent number, no institution decision exists, and no FWD has touched its claims. They are fully in force. The nearest PTAB activity is on the parent '988 patent (whose claim 1 is substantially similar in wording to the '478 abstract/claim language), where the challenged claims survived — the Board held them not unpatentable over Flore + 3GPP TS 23.002. That is a favorable signal for Lemko on overlapping claim scope, but it is not a merits determination on the '478 claims and should not be cited as one.

Estoppel landscape: § 315(e) estoppel is proceeding- and patent-specific. Because no IPR was ever brought on 9,332,478, no petitioner is estopped as to the claims of 9,332,478 — not even Microsoft/Affirmed Networks, who are estopped only as to the claims of the patents they actually challenged ('988, '414) in the district court. A new defendant facing 9,332,478 today has the full universe of § 102/§ 103 prior art available: the Flore / 3GPP TS 23.002 / Waylett / Mauer-type references that failed against the family are not off-limits against the '478 claims, though the Board's family rulings (including its treatment of "distributed mobile architecture" claim construction, argued at the 2024-07-11 hearing) signal that Lemko's constructions will be tough to overcome with the same art. A materially stronger, materially different art combination would be needed.

Pattern signals: Microsoft + Affirmed Networks (the defendants in the now-Closed N.D. Tex. case where 9,332,478 was asserted) ran a four-petition IPR campaign in 2023 against Lemko's DMA family — and lost across the board: two denials of institution and two FWDs sustaining patentability. This is a defensive-aggregator-adjacent pattern only in the sense that Unified Patents' portal tracks the campaign; Unified is not a petitioner here. Lemko has shown it will litigate family validity aggressively through the PTAB and (per GreyB's "Appealed" flag) possibly beyond. The district court case also shows Lemko's pleadings were trimmed (contributory infringement claims dismissed with prejudice, 2025-06-04, Doc. 94), and the case is closed per Unified Patents — check the docket for settlement terms if you are evaluating a license.

Recommended next steps

• If you are a defendant receiving a demand on 9,332,478: The PTAB record on this exact patent is a blank slate. Do not argue "this patent was already tested and survived" — it wasn't. Instead, run a fresh § 102/§ 103 search on the DMA-gateway routing claims, and note that no § 315(e) estoppel constrains you. Be aware, however, that the family FWDs (IPR2023-00529, FWD 2024-08-30; IPR2023-00531, FWD 2024-09-10, both viewable via USPTO PTAB E2E / PTAB decision search and Docket Alarm) show the Board crediting Lemko's "distributed mobile architecture" framework — reuse of the same failed references is unlikely to get you to institution.
• If you are Microsoft, Affirmed Networks, or a privy: Confirm the preclusive effect of the family IPRs and the Closed district court case (3:22-cv-00363) on your specific situation before filing anything on the '478; the "Appealed" flag on IPR2023-00531 means the parent-patent validity fight may still be live at the Federal Circuit — locate that docket before relying on the FWD as final.
• Verify current status before acting: The ODP ingest shows no PTAB case, but the district court litigation involving this patent was active into 2025 (Doc. 94, 2025-06-04) and the ownership/security chain changed on 2025-03-25 (Piccadilly Patent Funding LLC security assignment). Pull the USPTO Assignment database and the N.D. Tex. docket (CourtListener docket 63000037) to confirm who currently controls enforcement and whether any covenant/settlement binds the patent.
• No pending PTAB deadlines apply to 9,332,478 because no proceeding exists; if you file a new IPR, expect the standard timeline — institution decision ~6 months from filing, and a FWD within 12 months of institution under the § 316(a)(11) statutory deadline.

I could not directly query the USPTO Assignment Center database through my available tools, and web searches did not surface reel/frame numbers or correspondent names for this patent. Below I reconstruct the record from the authoritative Google Patents legal-events feed embedded in the full patent text, the Unified Patents litigation portal, and FreePatentsOnline. Where a specific data field (reel/frame, correspondent, execution date) was not retrievable, I say so rather than fabricate it.

Inventors

• ShaoWei Pan — sole named inventor. Employer at time of filing: Lemko Corporation (founder and principal; listed as Kildeer, IL, US on family records). All patent family prosecution in the Lemko DMA portfolio (US 7,855,988, 8,310,990, 8,349,667, etc.) is by the same inventor.
• Unusual-pattern check: not present — there is no multi-inventor roster departing an assignee post-filing. The portfolio is a single-inventor/founder operation, which is a different (and weaker) troll tell than a departing-inventor pattern.

Original assignee

• Lemko Corporation (Itasca, IL, US) is the named assignee on the issued patent (confirmed by FreePatentsOnline and Google Patents legal events).
• Line of business: design and sale of "distributed mobile architecture" (DMA) / femtocell-type cellular infrastructure for rural and remote telephony; network equipment provider.
• Product-shipping status: unclear from available records. Unified Patents classifies Lemko as an "Operating Company" (High-Tech) in its 2022 litigation, which supports operating-company status, but I found no independent evidence of current commercial product shipments.
• Current status: active entity; patent remains in force (adjusted expiration 2029-01-10) and Lemko remains the owner of record subject to a 2025 security interest (below).

Assignment timeline

The USPTO Assignment Center record for this patent was not fully retrievable in this session (reel/frame numbers and correspondent-of-record names could not be verified). The two recorded events below are those surfaced by the Google Patents legal-events feed (which is drawn from USPTO assignment records). Both are therefore real recorded events, but the reel/frame citations must be verified against the Assignment Center before relying on them.

• Executed (date not retrievable; recorded 2012-10-10) — Reel/frame not retrievable

  • Conveyance: Assignment of Assignor Interest (see document for details)
  • Assignor: ShaoWei Pan (inventor)
  • Assignee: Lemko Corporation
  • Correspondent: not retrievable
  • Context: Standard inventor-to-company assignment, recorded at the time the continuation application was filed; no fire-sale or reorg character.

• Executed (date not retrievable; recorded 2025-03-25) — Reel/frame not retrievable

  • Conveyance: Patent Security Agreement
  • Assignor: Lemko Corporation
  • Assignee: Piccadilly Patent Funding LLC, as Security Holder
  • Correspondent: not retrievable
  • Context: A security agreement, not a conveyance of title — Lemko remains owner, with a secured party holding a collateral interest. Consistent with patent-backed debt financing/litigation funding, not a transfer-to-asserter.

No other recorded assignments (no mergers, changes of name, releases, or subsequent transfers) were surfaced for this patent. Note: the patent's first enforcement (below) was brought by Lemko itself while it was the unencumbered owner, so no standing-related assignment preceded the suit.

Timeline diagram

timeline
    title Ownership of US 9332478
    2008 : Priority application filed by Lemko
    2012 : Inventor assignment recorded
    2016 : Patent US9332478 issued
    2022 : Lemko sues Affirmed Networks
    2025 : Security agreement to Piccadilly

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only post-issuance event is a 2025-03-25 Patent Security Agreement to "Piccadilly Patent Funding LLC, as Security Holder." A security agreement does not transfer title; Lemko remains the assignee of record. No "IP Holdings / Licensing / Ventures" LLC has taken ownership. (The entity name suggests a funding vehicle, but the conveyance type alone is dispositive: it is a security interest, not an assignment.)

2. Known asserter in the chain — not present / unclear. The asserting party in the only known litigation (3:22-cv-00363, N.D. Tex., filed 2022-02-13, closed) is Lemko Corp itself, which Unified Patents codes as an "Operating Company," not an NPE. Piccadilly Patent Funding LLC is not on the classic Acacia/Marathon/IV/Wi-LAN/Mosaid asserter lists I can verify, and no RPX/Unified asserter-directory match for it was found in this session.

3. Repeat correspondent across the chain — unclear / no data. Correspondent-of-record names were not retrievable from the available sources. Without the Assignment Center correspondent data, this signal cannot be assessed; it should be checked directly at assignmentcenter.uspto.gov (search by patent 9332478).

4. Cascading transfers — not present. Only two recorded events over 13 years: one inventor assignment (2012) and one security agreement (2025). No chained LLC transfers.

5. Pre-litigation transfer — not present. The only relevant date pattern runs the wrong direction: the suit (filed 2022-02-13) was brought by the original owner Lemko, and the security agreement postdates the suit by ~3 years (recorded 2025-03-25). No transfer occurred within six months before first suit.

6. Bankruptcy fire-sale — not present. No bankruptcy proceedings involving Lemko were found.

7. Privateering — not present. No operating company transferred this patent to an NPE to assert against competitors. Lemko asserted in its own name against Affirmed Networks Inc. (and separately pursued IPR appeals against Microsoft on family members '988/'990 in CAFC No. 24-2336 — note: that appeal concerns family patents, not '478 itself, per earlier research).

8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN; it remains at Lemko subject to a secured lender's interest.

Verdict

Operating-company assertion

The record shows the original operating assignee (Lemko Corporation) asserting this patent directly against a market participant (Affirmed Networks Inc., 3:22-cv-00363, N.D. Tex., filed 2022-02-13), with Unified Patents classifying the plaintiff as an "Operating Company." The sole post-issuance recorded event is a 2025-03-25 Patent Security Agreement to Piccadilly Patent Funding LLC — a debt-financing event that leaves title with Lemko and is not a transfer to an NPE, a shell, or a defensive aggregator. No shell-entity transfer, cascading chain, pre-suit assignment, bankruptcy, or privateering signal is present. Caveats: I could not verify product shipments, reel/frame numbers, or correspondent identities from the available sources, and the 2025 security interest means enforcement decisions may now be influenced by a secured lender — verify the full record at the USPTO Assignment Center (https://assignmentcenter.uspto.gov — search patent 9332478) before relying on this assessment.

I have reached the search-step limit, so I will now answer based on the information retrieved. Below is my best-effort prior-art analysis for US 9,332,478 B2, with explicit statements about what could and could not be verified in this session.

1. Scope and reliability caveats (read first)

• Source of the citation list. The complete, authoritative USPTO citation record (front-page "References Cited," the PTO-1449 forms, and the examiner's search history in the file wrapper) was not directly retrievable in this session. The citation list below comes from the FreePatentsOnline full-text mirror of US 9,332,478 (freepatentsonline.com/9332478.html), which reproduces the patent's "US Patent References" section.
• The retrieved list is truncated. The FreePatentsOnline snippet I obtained shows the reference list beginning mid-way at US 2015/0081812 A1 and cutting off at US 7,856,233 B2. References appearing earlier in the alphabetically/chronologically sorted list (older art, potentially the most important for § 102) and the tail end of the list were not captured. I could not confirm the "Other References" (non-patent literature) section, if any.
• Claims text. As flagged in the earlier patent summary, the full numbered claims of the '478 patent were not retrievable in this session (the Google Patents fetch truncates before the claims; FreePatentsOnline snippets did not surface them). Claim mapping below is therefore tied to the approximate independent-claim structure reconstructed from the specification's "particular embodiment" paragraphs (method claim, receiving-gateway CRM claim, satellite outbound/inbound CRM claims, gateway apparatus claim, DMA-server apparatus claim, orbiting-satellite apparatus claim). Treat claim-number mapping as indicative, not verbatim.
• Statutory framework. The '478 application was filed October 10, 2012 — before the AIA effective date of March 16, 2013 — so pre-AIA § 102 governs. The patent claims priority to July 14, 2008 (via 12/942,642 → 7,855,988). A reference can only anticipate under pre-AIA § 102(a)/(b)/(e) if it predates the invention date (presumed to be the effective filing date of the claim, July 14, 2008 for claims adequately supported by the parent disclosure).

2. Confirmed citations appearing on the face of US 9,332,478

A. Lemko / ShaoWei Pan family references (same or overlapping disclosure as '478)

These share the DMA specification and, critically, most claim the same July 2008 priority. Under pre-AIA § 102, a same-family, same-disclosure publication is not anticipatory against claims that are entitled to the July 14, 2008 effective filing date, because it is neither prior in time nor a "prior art" reference for subject matter fully supported in the parent. They were nonetheless cited (typically as part of the IDS/continuation prosecution chain) and would only become § 102 candidates if the '478 claims were held not entitled to the 2008 priority date for some limitation.

B. Third-party references (actual independent prior art candidates)

These are the references an examiner would treat as genuine § 102/§ 103 art — but all confirmed ones post-date July 14, 2008, so with the 2008 priority date intact they are § 103 (obviousness) candidates or § 102 art only for any claim limitation not entitled to the 2008 date.

3. Most relevant prior art overall (ranked)

1. US 2012/0044908 A1 (Spinelli/Cisco; priority 2007-10-24) — if the 2007 priority date is confirmed, this is the only confirmed cited reference that predates the '478 effective filing date and squarely covers gateway interworking/mobile-node routing. It is the strongest § 102(e) anticipation candidate against the gateway apparatus claim (≈ claim 5: legacy-network interface + private-IP interface + DMA-gateway-network interface + forwarding logic) and the method claim (≈ claim 1) to the extent the DMA-gateway network maps to the inter-gateway interface. Recommendation: pull the full Spinelli publication and map element-by-element.
2. US 8,018,907 B2 (Kubler; grant 2011) — best third-party art for distributed, non-hierarchical wireless routing to telephony destinations; § 103 strong, § 102 only if priority is lost.
3. US 2011/0060853 A1 (Pan; the '478's own parent publication) — identical disclosure; dispositive only in a priority-loss scenario.
4. US 2011/0320576 A1 (Lauer) — best art for the satellite/air-to-ground IP relay embodiments (claims 3, 4, 6, 7), on obviousness grounds.

4. What I could not verify

• The full citation list (list truncated in retrieval both at its head — where older, potentially § 102(a)/(e)-relevant references would sort — and its tail) and the "Other References"/NPL section.
• The verbatim claims of '478 (I am relying on the approximate claim structure from the earlier summary).
• Exact filing/priority dates for every reference (e.g., Spinelli's 2007 priority date comes from Unified Patents metadata, not the PTO record, and should be confirmed).
• Whether the examiner placed a § 102 rejection over any specific reference during prosecution (file-wrapper history not retrieved).

Recommended next step: pull the PTO-1449 / front-page "References Cited" from USPTO Patent Center for 13/648,704, and obtain the full text of US 2012/0044908 A1, before relying on any § 102 anticipation conclusion.

I'll need to locate the prior-art record for this patent (cited references and prosecution/PTAB art) and the actual claim language before building an obviousness analysis. Let me search.

I have useful context on the family's PTAB/CAFC history. Now I need (a) the actual claims of the '478 patent, and (b) the prior-art references cited against it (the "Prior Art" / Citations section of the patent's record). Let me search for those.

I've reached the search limit for this session, so I'll now synthesize the obviousness analysis from what was retrieved, with explicit confidence flags where the record is incomplete.

Obviousness Analysis — U.S. Patent 9,332,478 B2 (pre-AIA 35 U.S.C. § 103)

0. Scope, governing law, and evidence-integrity caveats

Governing law. The '478 application was filed October 10, 2012 — before the March 16, 2013 AIA effective date — and claims priority through the '988 patent to July 14, 2008. Pre-AIA 35 U.S.C. § 103 therefore governs. The analysis framework is Graham v. John Deere (383 U.S. 1 (1966)) as refined by KSR Int'l Co. v. Teleflex (550 U.S. 398 (2007)): (i) scope/content of prior art; (ii) differences between the claims and the prior art; (iii) level of ordinary skill; (iv) objective indicia — with KSR making clear that "the combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results," and that no rigid teaching-suggestion-motivation test applies.

Claim text caveat (important). I was unable to retrieve the numbered claims of the '478 from the truncated record. The abstract, the specification's "particular embodiment" paragraphs, and the family's IPR papers strongly indicate the independent claims track this structure (per the earlier summary, treat the claim-by-claim mapping as approximate):

• Claim-type A (method/CRM, gateway-to-gateway): a first DMA gateway transmits/receives "communications information" describing a network (a DMA-server network or a legacy network) accessible to a DMA gateway over a DMA gateway communications network, the information indicating devices served; a communication to a destination device is then routed by relaying gateway-to-gateway over that DMA-gateway network; the served network and the DMA-gateway network are different.
• Claim-type B (satellite gateway): calls routed through an orbiting-satellite DMA gateway having a legacy-network interface and a private-IP-network interface, to/from a DMA server that has an integrated wireless transceiver (and, per the spec, MSC/BSC modules) for the mobile device.

Prior-art-record caveat (important). The authoritative Google Patents "Citations/Prior Art" tab for the '478 was not present in the fetched full text (it truncates before claims/citations), and my searches did not recover the examiner's citation list for the '478 itself. What I did verify is: (a) the PTAB/IPR record for the grandparent '988 (IPR2023-00531, primary reference Flore, US 2007/0021120 A1, with 3GPP TS 23.002 v7.1.0 as background art); (b) the sibling IPR2023-00529 (patent 7,653,414) using the combination Waylett + Mauer; and (c) a large "Patent Art" list associated with the '478 in the Unified Patents litigation database (which may reflect examiner citations and/or invalidity contentions from Lemko v. Microsoft, N.D. Tex. 3:22-cv-00363). References from list (c) are identified below as context but are not individually verified as examiner-cited for the '478. Specific PTAB finding: the Board held '988 claims anticipated by Flore, and on April 10, 2026 the Federal Circuit affirmed in Lemko Corp. v. Microsoft Corp., No. 2024-2336 (Rule 36), squarely rejecting Lemko's construction arguments on "communications information" and "distributed mobile architecture." That result does not bind the '478, but the language of the '478's gateway-routing claims is materially the same as '988 claim 1, so the same art is the primary threat here.

1. Person of ordinary skill in the art (POSITA)

Both experts in IPR2023-00531 agreed on the profile (Proctor Decl. ¶ 56; Cooklev Decl. ¶ 21, both dated ~2008/2023): a bachelor's degree in electrical/computer engineering (or equivalent) plus ~2 years of telecom-network industry experience — encompassing mobile network architecture, gateways, servers, standards-compliant radio access, and interworking with legacy networks. The Board in IPR2023-00531 also accepted that "the prior art may reflect an appropriate level of skill in the art." A POSITA in this field would be intimately familiar with: 3GPP/GPRS/UMTS/EPS network elements (SGSN, GGSN, MSC, HLR/VLR), inter-network roaming and location updating, IP backhaul, and satellite backhaul/trunking for remote coverage.

2. Verified prior-art foundation

Prior-art status of the Lemko family itself: the Unified Patents "Patent Art" list also includes Lemko's own earlier DMA publications (US 7,840,230 B2 priority 2005; US 2007/0032245-family apps from 2006, e.g., US 2007/0232267 A1; US 2012/0252444 A1; US 2010/0008306 A1). Under pre-AIA § 103(c), commonly-owned prior art can be disqualified in some circumstances, so these are weaker levers and I do not build the primary grounds on them.

3. Proposed obviousness combinations

Ground 1 — Flore alone or Flore + 3GPP TS 23.002 → claim-type A (gateway-to-gateway routing of communications information)

Element mapping (claim-type A):

• First/second DMA gateway participating in a DMA gateway communications network → Flore's interworking gateway nodes (AGW) that communicate to bridge distinct core networks (3G ↔ 4G). 3GPP TS 23.002 supplies the standard "gateway" node concept (GGSN, SGSN, MSC-GW) interconnecting different networks.
• "Communications information" indicating devices accessible via a DMA server or legacy network → Flore's PDP context / location-update messaging identifying the UE and the network through which it is reachable (this is precisely the disclosure the Board credited as anticipating '988 claim 1, over Lemko's narrower proposed construction that was rejected).
• Storing at a gateway / home DMA register → HLR/VLR registers in the standard 3GPP architecture (TS 23.002) that a gateway accesses to locate subscribers.
• Receiving a communication directed to a device served by the other gateway's network and routing/relaying it gateway-to-gateway → Flore's call/session continuity across the interworked networks.
• The served network and the gateway network are different → inherent in Flore (a gateway's served radio network differs from the inter-gateway transport/core network).

Motivation / rationale. The '478's own Background identifies the problem: DMA networks deployed in rural/remote areas cannot reach one another or legacy/PSTN users. Flore was published in 2007 precisely to solve inter-network service continuity across different radio/core technologies through gateway interworking, and TS 23.002 taught the standard mechanism of register-based location and gateway routing. A POSITA seeking to let two independently deployed DMA or legacy networks exchange calls would look directly to (i) Flore's interworking-gateway architecture and (ii) the 3GPP register/gateway toolset. Combining them to route a call toward a device whose location is advertised through a gateway is the predictable application of known inter-network routing — no inventive leap. The Board already found this disclosure sufficient to anticipate the near-identical '988 claim 1; a fortiori, an obviousness ground under § 103 (which tolerates distributed disclosure across references) would be at least as strong against the '478's parallel claims. The practical knockout risk to claim-type A is high, subject to any claim-text differences between '478 and '988 that this analysis could not verify.

Ground 2 — Flore + satellite-communications gateway art → claim-type B (orbiting-satellite DMA gateway with legacy and private-IP interfaces)

Element mapping:

• Orbiting satellite as the DMA gateway → conventional satellite communications platforms (satellite backhaul/trunking for cellular, in-orbit switching/processing) well documented by 2008.
• First interface to a legacy communications network; second interface to a private IP network; routing a call between them to reach a mobile device served by a DMA server with an integrated wireless transceiver → Flore's gateway logic (legacy/3G side ↔ IP/packet side) mounted on the satellite platform, serving a remote DMA server over the satellite's IP link.

Motivation / rationale. The patent itself motivates this exact combination: DMA servers are deployed where "a private IP network is not available," and the spec proposes DMA gateways "implemented as orbiting satellites" precisely to bridge such remote DMA servers and legacy networks (spec ¶ "when a private IP network is not available... DMA gateways may be implemented as orbiting satellites"). Satellite backhaul of cellular traffic and satellite-to-ground IP connectivity were mature, predictable technologies by 2008. KSR's "known element substitution" and "design choice" rationales apply: putting an already-known inter-network gateway (Flore) on an already-known satellite transport platform to obtain coverage where terrestrial IP links are absent is the kind of combination a POSITA would make with a reasonable expectation of success — and the specification itself concedes the satellite form factor is just an implementation choice ("DMA servers also may be embodied in any other suitable form"). The satellite-dependent claims (B) are the '478's most distinctive additions, but the addition is a placement of known gateway functionality on a known platform for a known purpose (remote/rural connectivity), not a new technical effect.

Ground 3 — Flore (or Waylett) + distributed-processing references (Motorola 2004/0014466, Samsung CN 1427637, Ericsson 6,795,444) → DMA-server node claims

Element mapping:

• DMA server with integrated MSC and BSC modules, wireless transceiver/BTS interface, IP interface → Motorola's distributed real-time radio processing and Samsung's distributed mobile-network processing units show distributing MSC/BSC-type control onto multiple processing nodes was known; Ericsson '444 shows voice-over-packet with gateway interworking to legacy networks.
• Routing calls received at the server's transceiver to a destination via a DMA gateway over a private IP network → Flore's gateway-routing over packet networks applied to the distributed node.

Motivation / rationale. The cost driver stated in the '478's Background — a conventional MSC/BSC/HLR system "can cost over $2.0 million" and needs ~10,000 subscribers — was the well-known motivation to collapse MSC/BSC functions onto low-cost, IP-connected software nodes ("network-in-a-box" and distributed BSC/MSC architectures were active pre-2008). A POSITA combining a distributed-processing reference with Flore's inter-network gateway would implement exactly the claimed node: a software MSC/BSC on an IP-connected server whose calls egress through a gateway to other DMA or legacy networks.

Ground 4 — Sprint US 7,130,670 / register-based roaming + Flore → register-sharing/roaming-specific limitations

If any dependent claim requires explicit sharing of HLR/VLR-type register data between DMA gateways or a "central register database," Sprint's roaming-agreement database between private and public wireless networks, read together with Flore and TS 23.002, supplies the mechanism: database-mediated determination of where a subscriber is reachable across network boundaries, followed by routing. Motivation: standard roaming practice in wireless (HLR/VLR interrogation across operator networks) — a POSITA would automatically apply register lookups to any multi-network gateway routing scheme.

4. Why a POSITA would combine — consolidated rationale (Graham/KSR)

1. Same field, complementary disclosures. Flore, TS 23.002, and the register/gateway references all address inter-network mobility and call delivery in cellular systems — the same technical field and problem as the '478.
2. Known, finite design choices, not new functions. Each claimed building block (gateway bridging networks, register-based device location, packetized voice routing to a legacy network, satellite transport, distributed call processing) existed before 2008. The '478 rearranges them for a rural-deployment business goal the specification itself acknowledges. Under KSR, combining them "according to known methods to yield predictable results" is obvious.
3. Express motivation in the specification. The patent's own text identifies the motivation to deploy satellite DMA gateways when private IP networks are unavailable, and to bridge DMA and legacy networks — motivations a POSITA reading the prior art would share.
4. The PTAB/CAFC reality check. In IPR2023-00531, the Board (affirmed by the Federal Circuit on April 10, 2026) found Flore sufficient to anticipate the grandparent's identically-scoped gateway-routing claim 1 under the Board's construction of "communications information." Since obviousness under § 103 is a lower bar than § 102 anticipation (multiple references may be combined; Flore need not be a single-source perfect match), any ground Flore-based that succeeded as anticipation on '988 is prima facie at least as threatening as an obviousness ground against the substantively identical '478 claims.

Countervailing considerations (for balance):

• If litigation claim construction of "communications information" for the '478 requires the information to identify the specific communications network through which the target device is reachable (Lemko's position, rejected for the '988), and if the '478's claims include satellite-specific limitations absent from Flore (orbiting-satellite gateway with legacy and private-IP interfaces; DMA server with integrated transceiver), then a challenger must rely on Grounds 2–4 rather than Flore alone — which are still strong but require layering Flore with satellite and distributed-processing art.
• Objective indicia (long-felt need for low-cost rural telephony, licensing/industry interest) could weigh against obviousness, but Lemko has not yet shown such evidence for the '478 in the retrieved record, and the family's IPR loss undercuts a strong secondary-indicia position.
• No IPR specifically on the '478 was located in this session; estoppel therefore does not currently constrain a petitioner (and the district-court litigation — Lemko v. Microsoft, N.D. Tex. 3:22-cv-00363 — remains the live forum, with a June 2025 order dismissing only contributory-infringement claims).

5. Bottom line

The most credible § 103 case against US 9,332,478 is: Flore (US 2007/0021120 A1) as the primary reference, alone or combined with 3GPP TS 23.002 (standard gateway/register architecture) for claim-type A; combined with satellite-communications gateway art for the orbiting-satellite claims (type B); and combined with distributed call-processing references (e.g., Motorola US 2004/0014466, Samsung CN 1427637, Ericsson US 6,795,444) plus register/roaming references (e.g., Sprint US 7,130,670) for the DMA-server-node and register-sharing limitations. The motivation to combine is supplied by the identical problem the '478 specification itself identifies (interconnecting low-cost distributed cellular nodes with each other and with legacy/PSTN networks, including via satellite when terrestrial IP is unavailable) and by the standard roaming/interworking toolkit of 3GPP. The family's history — the Board's Flore-anticipates finding on the near-identical '988 claim 1, affirmed by the Federal Circuit on April 10, 2026 — makes Flore-based grounds the highest-probability obviousness theory for the '478.

Confidence flags: (1) Numbered '478 claims were not verified verbatim; the element mapping above relies on the abstract and specification paragraphs and on the IPR description of '988 claim 1. (2) The examiner's actual citation list for the '478 was not recovered; several references above come from the Unified Patents "Patent Art" list associated with the '478 in litigation and should be re-verified against the patent's Certificate/IDS before use in a formal proceeding. (3) Content details for Waylett and Mauer were not retrieved; they are referenced only as the combination used in sibling IPR2023-00529. A formal opinion should confirm each reference's disclosure against the exact claim language via USPTO PatentCenter/Global Dossier and the PTAB IPR2023-00531 final written decision.