US 7324487

I'll verify the current litigation/status posture via search, then summarize from the authoritative patent text provided.

Let me try to retrieve the actual claim text, since the fetched document omits the claims section.

US Patent 7,324,487 — Verification & Summary

Searched: USPTO/Google Patents full text, PTAB records, Federal Circuit (CAFC) dockets, ITC notices, district court litigation, UPC. Below I distinguish what is authoritative from what is inferred.

1. Bibliographic Data (confirmed against the patent full text)

Assignment chain (per USPTO assignment records on the patent page):
Hitachi, Ltd. → Hitachi Consumer Electronics Co., Ltd. (2012) → Hitachi Maxell, Ltd. (2014) → Maxell, Ltd. (2018) → Maxell Holdings, Ltd. (2021, merger) → Maxell, Ltd. (2021, change of name).

Literal-identifier caution: The string "7324487" is not unique globally. JP Patent No. 7324487 (特許第7324487号) is an unrelated Japanese patent to Kabushiki Kaisha TRI-TEC for an electric clamp and gravity-fed infusion controller (registered 2023-08-02). I did not conflate the two; everything below concerns US 7,324,487.

2. Abstract (verbatim from the granted patent)

"In a wireless communication system including base transceiver stations connected to a wired network and wireless system terminals, the wireless system terminals are individually registered in a plurality of communicable base transceiver stations and at the same time set in a multilink connection state. Then, a base transceiver station that has a lower communication load and in a more stable communication state at current point in time is selected from among the multilink base transceiver stations connected to a wireless system terminal, thereby performing data communication with the wireless system terminal. With this arrangement, when an individual fixed connection is switched, no load is produced at the time of roaming, and stabilization of communication and balancing of communication loads resulting from switching of the base transceiver stations for use are realized. The throughput of the overall wireless communication system is thereby improved."

3. Plain-Language Overview of the Disclosure

The patent's stated problem: in a conventional WLAN, a terminal picks the access point with the strongest signal, causing (a) frequent "roaming" handoffs that consume computational resources, and (b) load concentration on one AP. The three prior approaches it criticizes are (1) communication-load monitoring with roaming, (2) querying the number of connected terminals, and (3) fixing a single registered connection destination (which prevents roaming altogether).

Core idea — the "multilink": a terminal registers in, and is concurrently connected to, multiple APs ("multilink base transceiver stations") at one instant. Per the specification, "multilink" is expressly defined as "simultaneous connection of a wireless system terminal to a plurality of base transceiver stations at a certain point of time." Because link establishment is already done for all of them, switching among them requires no new link control — reducing roaming overhead.

Three disclosed selection mechanisms for which AP actually carries the data:

• §3-1 Busy-rate checking signal — probe all multilink APs; shortest response time, or smallest reported buffered-data amount, wins.
• §3-2 Buffering at all APs — the same data is buffered at every multilink AP; the fastest/least-busy one delivers it. Duplicate reception is handled by (a) wired discard commands from the winning AP, (b) a data-identifier/time-stamp table at the terminal, or (c) a compound of both.
• §3-3 Least busy status record — the terminal treats the AP that most recently delivered data as least busy (used for uplink).

Implementation details include a multilink table, SSID/encryption-key filtering, time-division channel switching for a single radio, "rounds" of polling with transmission-advance-notice signals (FIGS. 14–15), and optional server-based AP load ranking.

4. Independent Claims (plain language)

⚠️ Uncertainty flag: the claims section was not present in the authoritative full-text extract supplied to me, and the claim language reproduced below comes from a secondary aggregator (RPX Insight patent record) plus the PTAB petition caption. I could not independently verify the complete claim set or the exact independent-claim count (the gaps between claims 1, 6, 10, 11, 18, and 19 suggest dependent claims 2–5, 7–9, 12–17, and 20+ exist, but I have not confirmed their text or the final claim number).

If any of these six is in fact dependent in the granted document, the overall claim-architecture conclusion is unaffected but the per-claim characterization would need correction — treat the table as high-confidence on substance, medium-confidence on independent/dependent status.

5. Litigation and Enforcement Posture (searched; no 2026 CAFC docket found)

• PTAB: IPR2019-00462 was filed against US 7,324,487 (petitioner: Unified Patents; attorney docket 35548-0079IP1; "In re Patent of: Saito"). Google Patents records it as terminated by settlement. Prior-art exhibits cited included Uhlik (US 6,760,599), Yamashita (US 6,108,547), Wong & Lim "Soft Handoffs in CDMA Mobile Systems," Arvidsson (WO2002/49391), and others.
• District court (as asserted by Maxell): E.D. Tex. 5:18-cv-00033 and 5:18-cv-00034 (Huawei matters, per the IPR exhibit list); W.D. Tex. 6:22-cv-00334 (Maxell v. Lenovo, filed 2022-03-30, with US 7,324,487 identified as "无线局域网系统和漫游方法").
• ITC: 337-TA-1312 and 337-TA-1324 listed on the patent's page.
• CAFC 2026: No Federal Circuit docket, appeal, or Rule 36/opinion referring to US 7,324,487 was found. Searches for Maxell and the '487 patent in 2026 returned only unrelated Maxell matters — e.g., video-processing-patent IPRs against Samsung, ITC Inv. No. 337-TA-1432 (which asserts US 8,130,280; 11,490,004; 11,750,915; 11,509,953; 12,108,103; 11,445,241 — not 7,324,487), UPC proceedings against Samsung over EP 2 061 230 B1, and lithium-ion battery appeals. I state this as a negative search result, not as proof of absence.

Practical note: because the listing shows the patent expired 2025-10-05, any 2026 appeal activity on this patent would be unusual; if a live 2026 CAFC docket exists, it would most plausibly be a residual damages/validity appeal in a case filed earlier. I have no authoritative confirmation of one.

6. Explicit Uncertainties

1. Claim set incomplete — sourced from a secondary database, not the authoritative claims text; independent-claim count and dependent claims unverified.
2. No authoritative CAFC 2026 docket record located for this patent.
3. "Expired – Lifetime" status is Google Patents' own disclaimer-laden representation ("legal status is an assumption and is not a legal conclusion"); the adjusted expiration date of 2025-10-05 (presumably reflecting a §154(b) term adjustment) was not independently confirmed against USPTO PatentCenter.
4. The filing date (2003-02-10) postdates the JP priority date (2002-02-12); no foreign family members were verified in this pass.

I'll research litigation involving US patent 7324487. Let me start with targeted searches.

Let me dig deeper into the specific cases associated with patent 7324487.

Let me verify the Eastern District of Texas cases and the IPR listed on the patent's litigation record.

Based on my searches across patent-litigation databases (Unified Patents/Google Patents litigation data, Docket Alarm, RPX, PACER-derived dockets, ITC Federal Register notices, and PTAB records), here is what I found for US Patent 7,324,487 ("Wireless LAN system and method for roaming in a multiple base station"). I searched the exact number and excluded similarly-numbered patents.

Important caveat up front

Google Patents' litigation tab for this patent (which draws on Unified Patents data) links six proceedings to US 7,324,487. I could confirm details for all of them, but I was only able to directly confirm that the '487 patent was among the asserted patents for the ITC/lenovo matters and the PTAB IPR. For the two 2018 E.D. Tex. cases, the specific patent list per complaint was not fully retrievable in my searches — those are flagged below.

Known litigation / proceedings involving US 7,324,487

Detail on each

1 & 2 — Maxell v. Huawei (5:18‑cv‑00033) and Maxell v. ZTE (5:18‑cv‑00034), E.D. Tex.

• Both filed March 2, 2018, before Judge Robert W. Schroeder III. The complaints describe a licensing campaign dating to 2013/2016 and reference earlier Maxell suits (5:16‑cv‑00178 vs. Huawei; 5:16‑cv‑00179 vs. ZTE).
• Huawei matter: terminated April 4, 2019 as "Dismissed – Settled."
• ZTE matter: terminated March 12, 2019 (docket shows CLOSED/CONSOL).
• ⚠️ Confidence note: these two docket numbers appear in Google Patents'/Unified Patents' litigation list for the '487 patent, and the timing aligns with Huawei's IPR2019‑00462 against the '487 patent (below). However, I could not retrieve the complete asserted-patent list from the 2018 complaints, so I cannot state with certainty that the '487 patent itself was asserted in these two complaints rather than in related Maxell portfolio suits between the same parties.

3 — ITC Inv. No. 337‑TA‑1312 ("Certain Mobile Electronic Devices")

• Based on Maxell's March 30, 2022 complaint against Lenovo Group Ltd. (China), Lenovo (United States) Inc. (Morrisville, NC), and Motorola Mobility LLC (Libertyville, IL).
• Asserted patents included U.S. 7,324,487 (the '487 patent), together with 7,199,821; 8,170,394; 8,982,086; 10,129,590; and 10,244,284. The '487 patent was described as directed to "a method of controlling communication between a wireless terminal (such as a smartphone) and a plurality of base stations."
• ALJ Monica Bhattacharyya presided. Maxell withdrew this complaint and the investigation was terminated in 2022, expressly to avoid "overlapping demands or conflicting decisions" with the concurrently running 337‑TA‑1324.

4 — ITC Inv. No. 337‑TA‑1324 ("Certain Mobile Electronic Devices")

• Same complainant and respondents; the investigation was instituted Aug 16, 2022 (Federal Register notice of Aug 22, 2022; USITC notice 87 FR 51445‑46).
• The Commission's notice of investigation lists claims 1, 3, and 4 of the '487 patent among the asserted claims, alongside the '821, '394, '086, '590, and '284 patents.
• Accused products were Motorola-brand mobile devices (e.g., Moto E (2020), Razr 2nd gen, Motorola Edge/Edge+/Edge 5G UW, Moto G family, Motorola One 5G family, etc.).
• Progressive terminations: '590 and '284 patents dropped Mar 6, 2023; '394 patent dropped May 15, 2023; the entire investigation terminated Aug 18, 2023 based on settlement (final ID No. 35, dated Jul 16, 2023).

5 — Maxell v. Lenovo, 6:22‑cv‑00334 (W.D. Tex.)

• Filed the same day as the 1312 complaint (Mar 30, 2022) before Judge Alan D. Albright. Defendants were Lenovo Group Ltd., Lenovo (United States) Inc., and Motorola Mobility LLC — the mirror-image of the ITC respondents.
• This is the district-court companion to the ITC actions (asserting the same six patents, including the '487 patent). The case was stayed pending the ITC determination and was terminated/closed June 29, 2023, consistent with the ITC settlement.
• Related background: this was Maxell's second W.D. Tex. action against Lenovo (the first being 6:21‑cv‑01169, filed Nov 2021, asserting eight other portfolio patents).

6 — IPR2019‑00462 (PTAB)

• Petitioner: Huawei Technologies Co., Ltd.; respondent/patent owner: Jun Saito (the inventor; the patent was owned by Maxell at the time). Challenged US 7,324,487 (application 10/360,713).
• Filed January 4, 2019; terminated as settled on April 17, 2019. Petitioner's counsel: Michael Hawkins; patent-owner's counsel: Robert Pluta. (This is a PTAB administrative trial, not a district-court litigation, but it is recorded in the patent's litigation/PTAB linkage.)

Additional data point I could not resolve

The Google Patents record also carries a "First worldwide family litigation filed" entry pointing to Darts‑ip family fingerprint 27776993. I was unable to retrieve that Darts‑ip page in these searches, so I cannot enumerate any non‑U.S. (e.g., German or Japanese) actions tied to the '487 patent specifically. Maxell has run large parallel campaigns abroad, but the foreign cases I surfaced involved other Maxell patents, not the '487 patent.

What I did not include

The Maxell v. Samsung, Maxell v. Apple, and Maxell v. LG matters that dominate current Maxell news involve different patents (e.g., '757, '493, '161, '394, '086, '590, '284, and various display/video patents) and do not assert US 7,324,487. Per your instruction, I excluded those as not matching this patent number.

Note on patent status

US 7,324,487 is recorded as "Expired – Lifetime," with an adjusted expiration of October 5, 2025 (original assignee Hitachi, Ltd.; current assignee Maxell, Ltd. via a chain of assignments through Hitachi Consumer Electronics, Hitachi Maxell, and Maxell Holdings). No U.S. district-court case currently pending on the '487 patent surfaced in my searches — the last such district case (6:22‑cv‑00334) closed in 2023, and the ITC enforcement action settled the same year.

Sources: Google Patents litigation record for US 7,324,487; USITC notices for Inv. Nos. 337‑TA‑1312 and 337‑TA‑1324 (Federal Register 87 FR 51445‑46; USITC notice 337_1324_notice_08182023sgl.pdf); Docket Alarm dockets for 5:18‑cv‑00033, 5:18‑cv‑00034, and 6:22‑cv‑00334; RPX/insight.rpxcorp.com; Patexia; IPR2019‑00462 records (ipverse.greyb.com).

If you want, I can attempt to pull the full asserted-patent lists for the 2018 E.D. Tex. complaints (5:18‑cv‑00033 / ‑00034) to confirm whether the '487 patent itself was pleaded in those cases, or try again to resolve the Darts‑ip family 27776993 entry for foreign litigation.

I'll reconcile the structured ODP block (which reports no AIA proceedings) against the litigation data embedded in the patent record, which references a PTAB case. Let me search.

Key finding: the patent record's litigation data lists IPR2019-00462, but attributes it to "Unified Patents." The actual petitioner per the Joint Motion is Huawei. Let me dig into the petition grounds and check for any other proceedings.

Let me pin down the challenged claims/grounds and confirm whether any other AIA proceeding exists on this patent.

PTAB / AIA Trial Landscape — US 7,324,487 ("Wireless LAN system and method for roaming in a multiple base station"; Saito; Hitachi → Hitachi Maxell → Maxell, Ltd.)

Important reconciliation note up front. The structured "PTAB proceedings on file" block supplied for this task — sourced from the USPTO Open Data Portal — states that the ODP returns no AIA trial proceedings for this patent. The default therefore is "no PTAB activity on file."

However, web research surfaced a proceeding that the Google Patents litigation record for this patent explicitly flags and that is confirmed by the parties' own filings on the public docket: IPR2019-00462, [Huawei Technologies Co., Ltd.](/litigations/by-plaintiff/Huawei%20Technologies%20Co.%2C%20Ltd.) v. Maxell, Ltd. It was filed 2019-01-04 and terminated 2019-04-17 by settlement before institution, which is the most likely reason it does not appear in an ODP AIA-trial dataset. I flag it as required. I found no other AIA proceeding (IPR, PGR, or CBM) on the '487 patent in any source I could reach.

Proceedings overview

Publicly surfaced count: 1 AIA trial proceeding — 0 active, 0 with claims invalidated, 0 with claims sustained, 1 settled/terminated pre-institution, 0 institution denials (ODP's canonical count: 0). Net defensive posture: the '487 patent has never been tested on the merits at the PTAB — its claims were neither canceled nor confirmed by the Board, so there is no IPR FWD to hide behind and no IPR-based estoppel to exploit. Two facts dominate instead: (1) Huawei's full petition and expert declaration are public, giving any defendant a ready-made invalidity roadmap with zero estoppel attached, and (2) the patent's recorded legal status is "Expired – Lifetime, expires 2025-10-05" — it has been expired for roughly a year as of today (2026-09-28), so exposure is limited to past damages inside the § 286 lookback.

IPR2019-00462 — Huawei Technologies Co., Ltd. v. Maxell, Ltd.

• Type: Inter Partes Review
• Filed: 2019-01-04 (petition; patent owner response never reached — settled pre-institution)
• Status: Terminated – Settled (termination decision 2019-04-17, per the Board's Termination Decision Document — Termination Settlement Before Institution of Trial, Paper No. 9). Plain English: the Board never instituted, never reached the merits, and never issued a Final Written Decision.
• Judge panel: Not publicly identified in the sources I could access. Because the case settled before institution, no merits panel was ever assigned in any published decision. (The Board issued an order at Paper No. 6 on 2019-03-01, referenced in the Joint Motion.) Petitioner's counsel of record: Michael Hawkins; Patent Owner's counsel: Robert Pluta (per GreyB/ipverse case record — treat as unverified secondary sourcing).
• Petition grounds: Challenged claims 1–5, all under pre-AIA 35 U.S.C. § 103(a) (the '487 patent has a 2002-02-12 priority date; Huawei's expert applied that date). Nine grounds:
  • Ground 1 — claims 1–3 obvious over Uhlik (US 6,760,599) in view of Yamashita (US 6,108,547)
  • Ground 2 — claim 2 over Uhlik + Yamashita + Wong (D. Wong & T.J. Lim, Soft Handoffs in CDMA Mobile Systems, IEEE Personal Communications, Dec. 1997)
  • Ground 3 — claim 4 over Uhlik + Yamashita + Arvidsson (WO 2002/049391)
  • Ground 4 — claim 5 over Uhlik + Yamashita + Lee (US 6,842,619)
  • Ground 5 — claims 1–2 over Wong + Douzono (US 5,574,983)
  • Ground 6 — claim 3 over Wong + Douzono + Neumiller (WO 2000/27141)
  • Ground 7 — claim 3 over Wong + Douzono + Neumiller + Feder (EP 1 133 208)
  • Ground 8 — claim 4 over Wong + Douzono + Neumiller + Feder + Arvidsson
  • Ground 9 — claim 5 over Wong + Douzono + Neumiller + Feder + Li (US 2002/0163879)
  • Supporting expert: Declaration of Mr. Mark R. Lanning (EX1003). Additional exhibits gathered but not all pressed into the nine grounds include EP 1 032 237 (Shakhgildian), US 5,940,762 (Lee2), US 7,089,009 (Fauconnier), US 6,151,502 (Padovani), WO 2002/001893 (Toskala), US 7,085,260 (Karaul), and 3GPP TS 25.133. Petitioner also relied on Maxell's own preliminary infringement contentions from Maxell, Ltd. v. Huawei Device USA, Inc., No. 5:18-cv-00033 (E.D. Tex.).
• Institution decision: None. The proceeding was terminated before the Board decided whether to institute. There is no institution decision, no Board reasoning on the merits, and no claim construction ruling.
• Final Written Decision: None issued. No claim of the '487 patent was canceled or held patentable in this proceeding. Do not describe any claim of this patent as "invalidated by the PTAB" — that would be inaccurate.
• Settlement / termination: The parties filed a Joint Motion to Terminate on 2019-04-10 under 35 U.S.C. § 317(a) and 37 C.F.R. § 42.74, and the Board terminated on 2019-04-17. The motion was one of nine substantially identical joint motions filed the same day covering nine Huawei–Maxell IPRs (IPR2019-00192, -00462, -00464, -00575, -00576, -00622, -00640, -00656, -00661), reflecting a global settlement of the Huawei–Maxell dispute. A confidential settlement agreement was filed as EX1021 and the parties jointly requested business-confidential treatment under § 317(b); the terms are not public.
• Appeal: None. With no FWD and termination by settlement, there was nothing appealable. No Federal Circuit docket exists for this proceeding.
• Defensive value: For a defendant today, this proceeding is a free gift rather than a shield. Because it ended pre-institution, no § 315(e)(2) estoppel attaches to Huawei or anyone else — but equally, there is no invalidity judgment to lean on. The petition, the Lanning declaration, and the full prior-art exhibit set (Uhlik, Yamashita, Wong, Douzono, Neumiller, Feder, Arvidsson, Lee, Li) are all publicly available at zero cost, and a fresh petitioner can run those same grounds, or better ones, without any procedural bar.

Strategic summary

Claim status — CANCELED vs. SUSTAINED vs. UNTESTED. Every claim of the '487 patent is UNTESTED at the PTAB. Claims 1–5 were challenged in IPR2019-00462 but never adjudicated; the Board never instituted, so no claim was canceled and no claim was confirmed. There are no surviving-by-IPR claims to enumerate, because there was no IPR judgment. Any statement that the PTAB "upheld" or "struck down" claims of this patent is unsupported by the record. The one thing that has changed the patent's commercial posture is not the PTAB at all — it is expiration: Google Patents records the legal status as Expired – Lifetime, with an adjusted expiration of 2025-10-05. Practically, that eliminates injunctive exposure and confines any assertion to back damages.

Estoppel landscape. There is no estoppel. Under 35 U.S.C. § 315(e)(2), estoppel arises only "after a final written decision." IPR2019-00462 terminated before institution, so Huawei — and by extension anyone in privity — is not barred from raising before a district court or the PTAB any ground it raised, or reasonably could have raised, including the Uhlik/Yamashita and Wong/Douzono combinations. Conversely, Maxell obtained no protection from the settlement: the art Huawei assembled remains fully available. For a defendant now facing assertion, the entire prior-art space is open, and the public Huawei petition is a useful, already-vetted starting point rather than a boundary.

Pattern signals. Huawei was an aggressive, serial petitioner against Maxell, not a one-off: the 2019 settlement swept up nine IPRs in a single joint motion, and Huawei had earlier filed on related Maxell patents (e.g., IPR2018-00246 on US 7,671,901). The petitioner here was Huawei Technologies Co., Ltd. directly — not a defensive aggregator such as Unified Patents. (Note a sourcing trap: the Google Patents page's litigidation block lists IPR2019-00462 adjacent to a "Unified Patents PTAB Data" citation, but that is a data-license attribution, not the petitioner; the parties' own Joint Motion confirms the petitioner is Huawei.) Maxell, for its part, has not pursued any PTAB appeal on this patent (there was nothing to appeal) and instead drove a broad, serial assertion campaign on the '487 patent across forums: E.D. Tex. against Huawei (5:18-cv-00033) and ZTE (5:18-cv-00034), W.D. Tex. (6:22-cv-00334), and two ITC investigations against Lenovo/Motorola — 337-TA-1312 and 337-TA-1324. 337-TA-1324 was terminated in its entirety based on settlement (Commission notice, 2023-08-18, declining review of the ALJ's ID No. 35). The consistent pattern is settlement-before-adjudication, which explains why this patent has a rich assertion history but a completely blank PTAB merits record.

Recommended next steps

• If you are a defendant today: lead with expiration, not invalidity. The patent's recorded status is Expired – Lifetime with adjusted expiration 2025-10-05. Assert only-past-damages framing (six-year lookback under § 286 from the date of suit), and check whether the asserted conduct predates your own notice/limitations windows. An injunction demand is not credible on an expired patent.
• Front-load the public Huawei record. Pull the petition and the Lanning declaration from the docket and re-use them: Petition for Inter Partes Review, filed 2019-01-04 and the Joint Motion to Terminate, filed 2019-04-10. Grounds 1–9 (above) are a turnkey § 103 attack on claims 1–5 with no estoppel baggage. USPTO PTAB E2E search: https://ptacts.uspto.gov/ptabweb/ ; full case docket: https://www.docketalarm.com/cases/PTAB/IPR2019-00462/Inter_Partes_Review_of_U.S._Pat._7324487/ .
• There are no pending trial-stage milestones to track. No AIA proceeding on the '487 patent is active, so there is no institution deadline, no oral hearing, and no statutory one-year FWD date running. Any IPR you file now would be a fresh proceeding on an expired patent — weigh that cost against a straight § 282 invalidity/unenforceability defense and the expired-patent damages cap.
• Verify the record before relying on this memo. The ODP structured feed returns zero AIA trials for this patent while the public docket shows IPR2019-00462; before filing anything, re-confirm both the ODP/PTAB E2E status and whether any post-2023 IPR or PGR was filed by the Lenovo/Motorola or Amperex/ATL defendants. I could not confirm such a filing and did not find one, but my searches were not exhaustive and I will not assert its absence categorically.
• Do not overstate the PTAB outcome to a court. Any brief that characterizes IPR2019-00462 as resolving claim validity — in either direction — misstates the record. The correct characterization is: challenged claims 1–5; terminated by settlement before institution; no FWD; no estoppel; no adjudicated claim status.

Verification caveats (stated explicitly, per instructions): (a) The structured ODP block says no AIA proceedings — that is the canonical feed, and IPR2019-00462 is flagged only because independent sources (parties' filings, Google Patents litigation record, Docket Alarm) confirm it. (b) The judge panel is not determinable from public decisions because no merits decision issued. (c) Counsel names come from a secondary aggregator (GreyB/ipverse) and should be confirmed against the E2E docket. (d) I did not locate any Federal Circuit appeal, and none would be possible absent a FWD.

I'll start by gathering corroborating data on the litigation, the IPR, and any assignment records I can surface.

Let me try to surface the specific reel/frame recording data for this patent.

Assignment & Ownership Analysis — US 7,324,487 B2

Patent: US 7,324,487 B2 ("Wireless LAN system and method for roaming in a multiple base station")
Application: 10/360,713 · filed 2003‑02‑10 · priority 2002‑02‑12 · granted 2008‑01‑29 · status: Expired – Lifetime, adjusted expiration 2025‑10‑05
Verification hub: USPTO Assignment Center and Assignment Search — search 7324487. Underlying file: Google Patents US7324487B2.

Source limitation up front. The authoritative record I could retrieve (Google Patents legal‑events feed, mirrored from USPTO assignment data) gives the conveyance, assignor, assignee and recording date for each link but does not expose the reel/frame numbers for this patent — those were not recoverable through the tools available to me. I therefore do not print reel/frame values I could not verify. Where I cite a reel/frame below, it is explicitly labeled as belonging to a different Maxell patent that was recorded the same day by the same assignor, used only as circumstantial context. To obtain the actual reel/frames, query the Assignment Center by patent number 7324487 and by application 10/360,713.

Inventors

Unusual‑pattern check: I could not verify departure timing for the inventor. There is no evidence in the retrieved sources that the inventor left Hitachi within 12 months of filing, and a single‑inventor, employer‑assigned patent of this vintage does not exhibit the "all inventors bail out" precursor to a portfolio fire sale. Treat this item as not present / unverified, not as a signal.

Original assignee

Hitachi, Ltd. (Tokyo, Japan) — named on the issued patent; the original application was filed by Hitachi, Ltd. and the inventor's rights were assigned to Hitachi, Ltd. by 2003‑05‑02.

• Line of business: diversified global industrial/electronics conglomerate (rail, power, industrial, IT, and — at the time — consumer electronics and mobile). The '487 wireless‑LAN roaming disclosure sits in Hitachi's then‑current networking/mobile R&D.
• Did it ship a product embodying the claims? Hitachi was an operating manufacturer with wireless LAN and mobile product lines; the patent was prosecuted and maintained as part of the corporate portfolio, but I found no product literature tying the specific claims to a shipped Hitachi SKU. Treat "shipped an embodying product" as unverified.
• Current status: Operating. Hitachi, Ltd. remains a large publicly traded conglomerate. It is no longer the owner — the patent was moved out of Hitachi in the 2012–2014 restructurings below. Hitachi is referenced in later litigation discovery as "prior owner of the asserted patents" (Maxell v. Apple, E.D. Tex. 5:19‑cv‑00036, Rule 26 disclosure naming Hitachi, Ltd. and its employee Satoshi Matsuo for "licensing of Maxell patents").

Assignment timeline

Dates below are the recording date reported in the USPTO‑sourced legal‑events feed. Execution dates were not exposed in the retrievable record except where noted. Reel/frame is not available in my sources — flagged on every line rather than guessed.

• 2003‑02‑12 (priority) / application filed 2003‑02‑10 by Hitachi, Ltd.
• 2003‑05‑02 (recorded) — Reel not retrievable
  • Conveyance: Assignment of Assignors' Interest (inventor → employer)
  • Assignor: Jun Saito
  • Assignee: Hitachi, Ltd.
  • Correspondent: Not exposed in retrievable record.
  • Context: Ordinary employee‑inventor assignment to the original employer at filing.
• 2008‑01‑29 — Patent granted to Hitachi, Ltd. (no assignment).
• 2012‑07‑13 (recorded) — Reel not retrievable
  • Conveyance: Assignment of Assignor's Interest
  • Assignor: Hitachi, Ltd.
  • Assignee: Hitachi Consumer Electronics Co., Ltd.
  • Correspondent: Not exposed in retrievable record.
  • Context: Intra‑Hitachi corporate reorganization (consumer‑electronics business carved into a Hitachi subsidiary).
• 2014‑09‑08 (recorded) — Reel not retrievable
  • Conveyance: Assignment of Assignor's Interest
  • Assignor: Hitachi Consumer Electronics Co., Ltd.
  • Assignee: Hitachi Maxell, Ltd.
  • Correspondent: Not exposed in retrievable record.
  • Context: Intra‑group restructuring — the consumer‑electronics portfolio is consolidated into Hitachi Maxell (Hitachi's recording‑media/battery affiliate). (A witness claim in a later Samsung IPR transcript references Maxell paying ~$35M to acquire "the entirety of patents that Hitachi's mobile division had before April 2004" — I could not verify this is the same transaction; treat as corroborating color only.)
• 2018‑01‑25 (recorded) — Reel not retrievable; execution date on the parallel Maxell recording = 2017‑10‑01
  • Conveyance: Assignment of Assignor's Interest
  • Assignor: Hitachi Maxell, Ltd.
  • Assignee: Maxell, Ltd.
  • Correspondent: Not retrievable for the '487 record. A sibling recording made the same day, by the same assignor, for a different Maxell patent (app 12/028,885) shows correspondent of record Nadeen Nsouli, 71 South Wacker Drive, Chicago, IL 60606 (Reel 045142/0208, 141 pages). Flag: this is a bulk corporate rebranding recording; the same Chicago correspondent recurs across Maxell's 2017–2018 recordings, but I could not confirm she appears on the '487 reel, so this is context, not a finding.
  • Context: Corporate rebranding — Hitachi Maxell renamed itself Maxell, Ltd. (Hitachi divested its stake); recorded as an assignment of interest.
• 2021‑11‑29 (recorded) — Reel not retrievable
  • Conveyance: Merger
  • Assignor: Maxell, Ltd.
  • Assignee: Maxell Holdings, Ltd.
  • Correspondent: Not exposed in retrievable record.
  • Context: Internal holding‑company reorganization (merger).
• 2021‑12‑03 (recorded, 4 days later) — Reel not retrievable
  • Conveyance: Change of Name
  • Assignor: Maxell Holdings, Ltd.
  • Assignee: Maxell, Ltd.
  • Correspondent: Not exposed in retrievable record.
  • Context: Name restoration — the operating name "Maxell, Ltd." is re‑adopted. Net effect of the 11‑29 / 12‑03 pair: the same economic entity, Maxell, Ltd. (Kyoto, Japan), is the current owner.

Current assignee: Maxell, Ltd., 1 Koizumi, Oyamazaki, Oyamazaki‑cho, Otokuni‑gun, Kyoto 618‑8525, Japan (address as pleaded in ITC Inv. No. 337‑TA‑1338). No assignment has been recorded since 2021‑12‑03.

Assertion events (for sequencing, not assignments)

Correction worth recording: Google Patents tags IPR2019‑00462 with the "Unified Patents PTAB Data" licensing label. That is the data provider, not the petitioner. The PTAB joint motion to terminate identifies the petitioner as Huawei Technologies Co., Ltd. — do not log Unified Patents as an adversary on this patent.

Timeline diagram

timeline
    title Ownership and assertion of US 7324487
    2002 : Priority date 12 February
    2003 : Application filed by Hitachi Ltd
    2003 : Inventor assignment to Hitachi Ltd
    2008 : Patent granted to Hitachi Ltd
    2012 : Reassigned to Hitachi Consumer Electronics
    2014 : Reassigned to Hitachi Maxell Ltd
    2018 : Reassigned to Maxell Ltd
    2019 : Huawei IPR settled and terminated
    2021 : Merged into Maxell Holdings Ltd
    2021 : Name changed back to Maxell Ltd
    2022 : Suits filed against Lenovo and Motorola
    2025 : Patent expired

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. Every assignee in the chain is a real, product‑bearing Japanese corporation: Hitachi, Ltd. → Hitachi Consumer Electronics Co., Ltd. (2012‑07‑13) → Hitachi Maxell, Ltd. (2014‑09‑08) → Maxell, Ltd. (2018‑01‑25) → Maxell Holdings, Ltd. (2021‑11‑29, merger) → Maxell, Ltd. (2021‑12‑03, change of name). No "IP / Holdings / Ventures" licensing shell, no registered‑agent service address, no single‑member LLC. The 2021 pair are a merger and a name change within one corporate group — reorg, not a sell‑off.

2. Known asserter in the chain — Present (moderate). The current assignee, Maxell, Ltd., does not appear on the classic NPE lists you named (Acacia, Marathon, IV, Wi‑LAN, Pendrell, RPM, etc.), but it is surfaced by RPX as a high‑frequency patent plaintiff with a sustained licensing‑and‑litigation campaign — e.g. the RPX Insight litigation feature on the West Texas Maxell–Lenovo dispute, and the run of complaints at ITC (337‑TA‑1312, 337‑TA‑1324, 337‑TA‑1338) and in district court (E.D. Tex. actions vs ZTE, Huawei, Apple, Blackberry; W.D. Tex. 6:22‑cv‑00334 vs Lenovo/Motorola; C.D. Cal. 2:21‑cv‑06758 vs VIZIO). Maxell itself describes its posture as a portfolio "actively enforces … through licensing and, potentially, litigation."

3. Repeat correspondent across the chain — Unclear (no finding). Only one correspondent is corroborated anywhere in the Maxell family's 2017–2018 recordings: Nadeen Nsouli, 71 South Wacker Drive, Chicago, IL 60606 (Reel 045142/0208, recorded 2018‑01‑25) — and that is a different patent. The correspondents on the '487's own six recordings were not retrievable. A single appearance is not the recurrence pattern this signal requires, so mark it unclear, not present. (If the Assignment Center pull shows the same Chicago correspondent across the 2012, 2014 and 2018 '487 recordings, this would flip to present — worth resolving.)

4. Cascading transfers — Not present (weak at most). There is one rapid pair — 2021‑11‑29 (merger into Maxell Holdings, Ltd.) and 2021‑12‑03 (change of name back to Maxell, Ltd.), 4 days apart — but these are a merger and a name change inside a single corporate family, not consecutive assignments through chained LLCs with common principals or a shared correspondent address. Flag it as a timing curiosity only.

5. Pre-litigation transfer — Not present. The last recorded assignment (2021‑12‑03) lands ~4 months before the 2022‑03‑30 Lenovo/Motorola complaints (ITC 337‑TA‑1312/1324; W.D. Tex. 6:22‑cv‑00334). But that recording is a change of name restoring "Maxell, Ltd." — no new party acquired the patent, no venue‑shopping transfer, and Maxell had already asserted related‑family patents against Huawei, ZTE, Apple and Blackberry from 2016 onward (the '487 was itself the subject of the 2019 Huawei IPR). No clean‑standing transfer engineered for the 2022 suits.

6. Bankruptcy fire‑sale — Not present. No Chapter 7/11 proceeding by Hitachi or any Maxell entity appears in the record; the transfers were corporate restructurings, not distressed sales.

7. Privateering — Not present. There is no transfer of the patent to a third‑party NPE asserting on Maxell's behalf; Maxell asserts in its own name as patent owner and complainant.

8. Defensive aggregator — Not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents or OIN. (Unified Patents' only touchpoint is as a data source label on the Google Patents page, not as an owner or petitioner — the IPR petitioner was Huawei.)

Verdict

NPE — moderate confidence.

The chain itself is clean — a genuine operating‑company lineage (Hitachi, Ltd. → Hitachi Consumer Electronics, 2012‑07‑13 → Hitachi Maxell, Ltd., 2014‑09‑08 → Maxell, Ltd., 2018‑01‑25 → Maxell Holdings, Ltd., 2021‑11‑29 merger → Maxell, Ltd., 2021‑12‑03 name change) with no shell‑LLC hop, no fire sale, and no third‑party enforcer, which rules out the high‑confidence category. What tips the call to moderate NPE is a single strong signal: the current assignee Maxell, Ltd. is tracked by RPX as a high‑frequency patent plaintiff running a licensing‑driven campaign against non‑competitor smartphone/TV makers (Lenovo/Motorola via ITC 337‑TA‑1312/1324 and W.D. Tex. 6:22‑cv‑00334; Huawei via IPR2019‑00462; plus the VIZIO and earlier ZTE/Huawei/Apple/Blackberry actions), while the '487 claims (a wireless‑LAN roaming method in client devices) describe products Maxell no longer manufactures. Note the hybrid framing: this is an operating‑company‑backed enforcer rather than a classic shell‑LLC troll, so a reviewer who insists Maxell is an operating company could reasonably record it as Operating‑company assertion instead — but under your definition, which requires the current assignee to ship products embodying the claims, "moderate NPE" is the better fit. Finally, the patent expired 2025‑10‑05, so its assertion value is now historical.

Verify independently: search patent 7324487 at the USPTO Assignment Center / Assignment Search to capture the reel/frame and correspondent fields I could not retrieve, and confirm whether the Chicago correspondent recurs across the 2012/2014/2018 links (which would move signal #3 from unclear to present).

I'll search for the authoritative citation record for US 7,324,487, since the fetched full text omitted the "References Cited" section.

I've located the front page reference list in the granted-PDF. Let me pull the complete "References Cited" block, including foreign documents and other publications.

Scope and method note (read first)

I searched for the literal identifier 7324487 and returned only that patent. Two caveats up front:

1. The "References Cited" block was absent from the authoritative full text I was given (the Google Patents fetch stops before the claims and front-page citations). I therefore had to retrieve the citation list from the granted-PDF front page, and my retrieval is partial (OCR snippets) — I did not obtain the complete list, and I obtained none of the foreign patent documents or the "Other Publications" entries. I flag every gap rather than fill it.
2. Literal-identifier collisions. A search for the string "7324487" also returns US 4,929,653 (stabilized pigmented polypropylene fiber) — this is a Google Patents family-page artifact in which "US7324487A" appears only as an internal token alongside the unrelated application serial 07/073,244; it is not the target patent. Separately, JP Patent No. 7324487 is an unrelated Japanese patent (electric clamp / infusion controller). Neither is conflated below. The target is US 7,324,487 B2, application 10/360,713, "Wireless LAN system and method for roaming in a multiple base station" (Saito; Hitachi).

1. Confirmation of the identified patent

Source: https://patents.google.com/patent/[US7324487B2](/patent/US7324487B2)/en

2. U.S. patent documents cited on the face of US 7,324,487 (as retrieved, partial)

Retrieved verbatim from the grant front page (examiner/applicant citation column headed "References Cited — U.S. PATENT DOCUMENTS"). Dates and examiner-applied class tags are reproduced as printed; asterisks denote documents the examiner considered and applied against a subclass.

Critical timing observation: entries 3–12 all issued after the '487 filing date (2003-02-10), several by 3–4 years. They can therefore only be prior art under §102(e) (their own earlier U.S. filing/provisional dates) — not under §102(a)/(b) as printed. Any §102 analysis of these must be run against their filing dates, which I did not retrieve. Treat entries 3–12 as examiner-cited art of unverified §102 date, not as confirmed prior art.

Anomaly flag (do not auto-correct): entry 6 is printed as "1/2005" for a 6,985,xxx-series patent. A grant number in that range would ordinarily issue in January 2006; this is likely OCR truncation of "1/2006" or a printed typo. I have not corrected it.

Explicitly NOT retrieved: the Foreign Patent Documents column and the Other Publications column of the '487 front page. I therefore cannot enumerate the JP/EP/WO publications the examiner cited. (A likely JP family member, JP 3904462 B2, appeared adjacent to the '487 in search results, but I could not confirm it as the priority document — flagged as unverified.)

3. Prior art actually asserted against the '487 in IPR2019-00462 (petitioner-cited — different list)

This is not the patent's own citation list. It is the evidence set of the IPR petition (Huawei Technologies v. Saito, filed 2019-01-04, terminated–settled 2019-04-17). These are, in practice, the references the challenger judged most relevant, so they belong in any "most relevant prior art" answer:

Sources: https://www.docketalarm.com/cases/[PTAB](/ptab)/IPR2019-00462/Inter_Partes_Review_of_U.S._Pat._7324487/docs/01-04-2019-Petitioner/Petition-2-Petition_for_Inter_Partes_Review.pdf ; https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2019-00462. (Note: the petition names Huawei as petitioner; Google Patents lists the petitioner as Unified Patents — the two records conflict and I have not resolved which caption governs.)

Note the anti-similar-number caution applies here too: US 6,842,619 ("Lee") and US 5,940,762 ("Lee2") are distinct references, as are US 5,574,983 (Douzono) and US 6,151,502 (Padovani). I did not substitute one for another.

4. §102 anticipation mapping (to the claims of the '487)

Threshold caveat, stated plainly: §102 anticipation requires a single reference to disclose every element of the claim as arranged. All twelve front-page U.S. documents were considered and overcome during prosecution — the claims issued over them. On the record I retrieved, I cannot defend an anticipation rejection for any claim against any of them. The honest characterization is that the front-page citations were cited as background/technical context, not as anticipatory art, and the realistic challenge to the '487 is §103 obviousness (which is exactly how the IPR petitioner framed it).

Claim architecture (from the prior section of this analysis, carried forward — claim text from a secondary aggregator, unverified): independent claims 1 (concurrent multi-AP connection + communicate with one), 6 (every multi-AP buffers; least-loaded transmits), 10 (managing server detects load, selects bufferer/transmitter), 11 (system; concurrent multi-AP; one performs the transmit/receive), 18 (system with managing server selecting relay AP by load), 19 (terminal-side: link table + buffered-data discard signal to the others), with 20 also appearing.

On claims 19/20 (terminal-side discard of buffered data): none of the front-page U.S. references I retrieved is classed to data-duplication avoidance. The buffered-data-discard feature (terminal sends discard signal to the other connected APs) is the element the examiner apparently found no showing for — so §102 against claim 19/20 is the weakest attack on the retrieved record.

5. What I could not verify (explicit uncertainties)

1. Incomplete citation list. The U.S. foreign patent documents and other publications columns of the '487 front page were not retrieved. My §102 mapping covers only the partial U.S. list plus the IPR evidence set.
2. No verified §102(e) filing dates for the eleven post-filing-issuance U.S. citations — required before any could be called prior art at all.
3. Descriptions of 6,729,929 / 7,006,472 / 7,023,803 / 7,110,388 / 7,171,206 and the exact titles of 5,854,786 and 6,591,099 are unverified; I declined to guess.
4. "6,985,475 … 1/2005" is reproduced literally and flagged as anomalous rather than corrected.
5. IPR petitioner identity conflict (Unified Patents vs. Huawei) unresolved.
6. No U.S. PatentCenter "References Cited" screen was directly retrieved — the front-page PDF is my closest authoritative proxy for the citation of record.

If the exact foreign-document and non-patent-literature citations are needed, the definitive source is the USPTO PatentCenter "References Cited" / PTAB Exhibit EX1002 (File History) for application 10/360,713 — which I was unable to open in this pass.

I'll ground the prior-art references first. The page's "Prior Art" content for this patent is thin (only keywords and a Definitions block), but the IPR petition for this patent (IPR2019-00462) supplies the actual prior-art exhibit set. Let me verify those references.

Key prior-art hits confirmed. Let me pin down the remaining references and the actual IPR2019-00462 grounds.

Obviousness Analysis — US 7,324,487 B2 under 35 U.S.C. § 103

Patent: US 7,324,487 B2, "Wireless LAN system and method for roaming in a multiple base station" (Saito; appl. 10/360,713; filed 2003‑02‑10; JP priority 2002‑02‑12; issued 2008‑01‑29; now Maxell, Ltd.)

0. Source basis and caveats (read before the analysis)

Contradiction flagged: the earlier "Litigation summary" section in this file reached the same IPR conclusion but named [[Huawei Technologies Co.](/litigations/by-plaintiff/Huawei%20Technologies%20Co.), Ltd.](/litigations/by-plaintiff/Huawei%20Technologies%20Co.%2C%20Ltd.) as petitioner; the Google Patents page's PTAB block prints "Petitioner: 'Unified Patents PTAB Data'". The joint motion to terminate is captioned "HUAWEI TECHNOLOGIES CO., LTD., Petitioner v. MAXELL, LTD., Patent Owner" (Docket Alarm, IPR2019‑00462, Paper 7), and ipverse confirms Petitioner = Huawei, Respondent = Saito, Jun, Status = Terminated‑Settled (Apr 17, 2019). "Unified Patents" on the Google page is the data-source label, not the petitioner. Treat Huawei as petitioner.

Two further caveats I will not paper over:

1. The full claim text was not in the authoritative extract. The claim architecture below is reconstructed from the IPR's ground structure plus the ITC's asserted-claim list; see §2.
2. I could not independently pull the full text of US 6,760,599 (Uhlik) in this session. Its characterization below is derived from (a) the petition's ground/exhibit structure, (b) how petitioners have characterized an "Uhlik" exhibit in parallel PHS‑related PTAB filings (configuration‑request / configuration‑message frame exchange, a frequency‑hopping control bit governing terminal reception, multi‑antenna SDMA base station), and (c) the ArrayComm broadcast‑channel family record. Confidence is medium on Uhlik's specifics, high on the combination structure.

1. Legal framework, priority date, and POSITA

• Governing standard: Graham v. John Deere factors as modified by KSR Int'l v. Teleflex (2007): scope/content of the prior art, differences from the claims, level of ordinary skill, and secondary considerations. A combination of known elements is obvious where it "does no more than yield predictable results," where there is a "finite number of identified, predictable solutions," or where the motivation comes from "the nature of the problem to be solved."
• Critical date: Feb 12, 2002 (JP priority). Any reference published/filed before that date qualifies.
• POSITA: a person with a B.S. in EE/CS and ~2–3 years in wireless data networking / WLAN / cellular MAC‑layer design, familiar with 802.11 AP–station association, roaming/handoff, and with DS‑CDMA soft handoff. (This level, applied by petitioner's declarant Mark R. Lanning, is consistent with the art.)

2. The claim architecture (reconstruction — verify against the printed claims)

From the IPR2019‑00462 petition table of contents:

• Ground 1 — claims 1–3 obvious over Uhlik in view of Yamashita
• Ground 2 — claim 2 obvious over Uhlik + Yamashita + Wong
• Ground 3 — claim 4 obvious over Uhlik + Yamashita + Arvidsson
• Ground 4 — claim 5 obvious over Uhlik + Yamashita + Lee

Exhibit list confirms: EX1001 = the '487 patent; EX1004 = US 6,760,599 ("Uhlik"); EX1005 = US 6,108,547 ("Yamashita").

Two independent corroborations of claim numbering:

• The ITC Inv. No. 337‑TA‑1324 notice of investigation asserted claims 1, 3, and 4 of the '487 patent (alongside US 7,199,821; 8,170,394; 8,982,086; 10,129,590; 10,244,284) — see the CCPIT‑Sichuan report on the ITC filing.
• The earlier summary's secondary-source note that claims 1, 6, 10, 11, 18, and 19 read as independents.

Reconciled working model: claim 1 is the sole independent method claim in the challenged group; claims 2–5 are dependent; claims 6–19+ (if any) were not challenged and therefore carry no §103 record. The prior section's characterization of claims 6/10/11/18/19 as independent is therefore compatible with the IPR structure and does not conflict — but note that if any of 6–19 is in fact dependent on claim 1, every Ground 1–4 conclusion below propagates to it automatically.

Feature mapping inferred from the ground split (this drives the analysis):

3. The prior‑art set actually deployed

Note the Arvidsson overlap is striking: the '487's own §3‑1 discloses exactly two busy‑rate metrics — shortest response time and smallest buffered data amount — and Arvidsson expressly teaches "shortest queue length or shortest response time" as the load metric. That is not a coincidental match; it is the same teaching in a cellular core‑network wrapper.

4. Ground‑by‑ground obviousness

Ground 1 — Claims 1–3: Uhlik + Yamashita

The gap the patent claims to fill. The '487's stated point of novelty is the "multilink," which the specification expressly defines as "simultaneous connection of a wireless system terminal to a plurality of base transceiver stations at a certain point of time," so that "when a multilink process is finished … there is no need to perform the link control processing again" when switching among them.

Why the combination is obvious.

• Yamashita supplies every element of the "multilink." Its mobile station is connected only to a first BTS in normal mode, and in soft‑handoff mode "said mobile station is connected also to a second base transceiver station via a radio channel." Both BTSs are tied to a concentrator by wired lines, and forward traffic is delivered through both (copied from BTS1 to BTS2 over the line). That is simultaneity + wired‑network APs + data actually flowing through one of the two. Yamashita's "concentrator" is the functional analogue of the '487's wired network 300 / server 310.
• Uhlik supplies the per‑terminal frame protocol and link bookkeeping — broadcast burst → request burst → directed configuration message describing available channels, i.e., the terminal and base station know about each other and are already synchronized before data flows. That is precisely the "link establishment already completed" advantage the '487 claims.
• Motivation to combine (explicit and inherent).

1. Same problem, same field‑of‑endeavor logic. KSR: the '487's expressed problem — "enormous computational resources" consumed by re‑association, and unstable connections when a link is weak — is exactly the problem CDMA soft handoff was invented to solve. A POSITA looking to improve WLAN roaming had a strong reason to import make‑before‑break multi‑link operation.
2. Physiological/compatibility (KSR "predictable use"). Both references use base stations with a wired backhaul to a central node; physically, the '487's FIG. 3 architecture (server 310 + wired network 300 + APs A–D) is the WLAN re‑skin of Yamashita's concentrator + BTS1/BTS2 topology. There is no structural re‑engineering required.
3. Reasonable expectation of success. Yamashita demonstrates that the multiple legs can be kept aligned (BTS1 controlling, traffic forwarded over the line); Wong & Lim documents that soft handoff reliably reduces outage. Nothing in the combination requires an unpredictable result.

• Result: claim 1 (concurrent connection to ≥2 wired APs + communicating data with one) reads on Yamashita's soft‑handoff mode; the "which one carries the payload" election is the ordinary routing/traffic‑forwarding decoupling Yamashita and Uhlik both implement.

Where this ground is weakest (see §7): Yamashita is a DS‑CDMA system where the mobile station demodulates multiple legs simultaneously on a single frequency; the '487 is a WLAN in which a single radio communication unit is time‑division channel‑switched across APs ("rounds," FIGS. 14–15). The references do not obviously teach that specific mechanism.

Ground 2 — Claim 2: Uhlik + Yamashita + Wong

Claim 2 (per the reconstructed architecture) imports the link reference value — the '487's "−70 dBm … / S‑N ratio" threshold used to decide which APs qualify as multilink members, and to add/drop them as the value is crossed.

• Wong & Lim supplies the threshold‑based decision. Soft‑handoff literature uses add/drop thresholds (pilot‑strength based) for keeping a leg; this is the same genus as the '487's link reference value.
• Motivation: the '487's own §1‑1 admits the threshold concept in conventional terms and only reframes it; Wong supplies the reason (link‑quality reliability vs. resource cost — the '487's stated §1‑3 trade‑off between stability and load). A POSITA implementing Yamashita's soft handoff in a packet system would necessarily define a "keep/abort this leg" criterion; Wong & Lim is the canonical source of that criterion. Combining is a design choice among a finite set of known signal‑quality metrics, and the '487 concedes that both RSSI and S/N are interchangeable.
• Result: obvious.

Ground 3 — Claim 4: Uhlik + Yamashita + Arvidsson

This is the strongest ground, because claim 4 is the load‑balancing core and Arvidsson is a near‑verbatim anticipation of the '487's §3‑1 teaching.

• Arvidsson's quoted teaching: "estimate the load in each MSC/VLR and then connect a mobile station to the least loaded MSC/VLR," where load = "processor load … queue lengths in selected buffers … or … response times," and "the least loaded … would be the one with the lowest processor load, shortest queue length or shortest response time."
• The '487's §3‑1 teaching: "the response times of the multilink base transceiver stations to the busy rate checking signals are compared. The multilink base transceiver station with the shortest response time is determined to be in the least busy state"; alternatively "buffered data amount information … the station that has transmitted the response signal having the smallest buffered data amount is determined to be in the least busy state."
• These are the same three metrics (processor load / buffer occupancy / response latency) applied to the same kind of decision (pin a mobile to the least‑loaded server). The only difference is the scope of the pool — Arvidsson's pool is a set of MSC/VLR core‑network nodes; the '487's pool is "the set of APs to which this terminal is already multilinked."
• Motivation to combine:

1. Yamashita creates the pool; Arvidsson ranks it. Once a terminal is already multilinked to two or more APs (Ground 1), the immediately apparent next question is "which leg do I actually use?" — and KSR holds that where a technique is known and the improvement is a predictable use of it, the combination is obvious. Arvidsson is that technique, ready‑made.
2. The '487's own AAPA concedes the goal. The Background admits method (1), "communication load monitoring" with roaming to a lower‑load AP. The patent's only complaint is that roaming itself burns resources — a defect the multilink (Yamashita) removes. So AAPA + Yamashita alone gives you claim 4's substance; Arvidsson supplies the explicit "least‑loaded" formulation with citation.
3. Known design incentive / market pressure. Load imbalance across APs is a classic WLAN capacity problem; equalizing it is an articulated, predictable objective — KSR's "predictable results" prong is comfortably met.

• Secondary note: Arvidsson's priority (WO published 2002‑06‑20; Ericsson) — timing must be checked against the Feb 12, 2002 critical date. If Arvidsson's priority date precedes Feb 12, 2002 it is §102(e)/§102(a)‑eligible; if not, it drops out and Ground 3 must lean on AAPA method (1) instead. This is a real vulnerability in Ground 3 as pleaded and I flag it explicitly.

Ground 4 — Claim 5: Uhlik + Yamashita + Lee

Claim 5 (reconstructed) adds buffering at the APs plus duplicate suppression — the '487's §4‑2, i.e., buffer the same downlink data at every multilink AP and let the fastest/least‑busy one deliver, then (a) issue wire‑side discard commands, (b) keep a data‑identifier table at the terminal, or (c) both.

• The petition used an exhibit identified as "Lee" for this ground. I did not retrieve its citation and will not invent one. Its role, from the ground structure, is to supply the buffer/duplicate‑suppression element.
• Motivation, independent of which "Lee" is used: once the terminal is simultaneously reachable through multiple APs (Yamashita) and one leg is chosen by load (Arvidsson), some mechanism must prevent the terminal from receiving the same packet twice. Discarding duplicate/out‑of‑order copies at the receiver, and flushing redundant buffers once delivery is confirmed, is routine network‑protocol hygiene (it is the ordinary function of sequence numbers, TCP/RLP re‑sequencing, and ARQ discard timers). The '487 itself concedes this is a necessary consequence, not an invention: "a mechanism for avoiding duplicate data reception of the second and subsequent data transmissions becomes necessary." An admission that an element is required by the combination is close to dispositive on obviousness of that element.
• Result: obvious, subject to confirming the Lee exhibit's teaching.

5. Additional/supplementary §103 combinations worth pleading

These are not in IPR2019‑00462 but are available on the same record:

1. AAPA (Background methods 1–3) + Yamashita. Method (1) (load monitoring → roam to lower‑load AP) and method (2) (query connected‑terminal count → join the smallest) are admitted prior art. Yamashita supplies the concurrent multi‑BTS connection. The '487's own "challenge" — "reduce the frequency of roaming … and evenly distribute communication load" — is the union of the two. Combining them is the definition of an obvious design from known elements.
2. Wong & Lim alone as a §102(b) printed publication on the premise. A 1997 IEEE tutorial establishing that maintaining multiple simultaneous radio links to distinct base stations reduces outage and error rate is powerful evidence that concurrent multi‑AP connection was a known technique with known benefits — i.e., the "multilink" is not an inventive concept but an import.
3. 802.11/WLAN roaming art + Yamashita (not retrieved here; the page's Patent Citations table was missing from my extract). A full §103 opinion should enumerate the examiner‑cited references from the '487's own front page and from the "Cited By" list; I could not do that in this pass.

6. Consolidated motivation‑to‑combine (the KSR showing)

7. Where the obviousness case is genuinely weak (the non‑obviousness counter‑argument)

A rigorous analyst must state this, and it is the argument Maxell would run:

1. The single‑radio, time‑division channel‑switching mechanism is not in the references. The '487's §2‑1 states the constraint expressly: "a single wireless communication unit equipped with the wireless system terminal can perform communication over a single channel alone at an arbitrary moment." Its solution is sequential time‑division hopping across APs ("rounds") with transmission‑advance‑notice signals and per‑round timers (FIGS. 14–15). Yamashita is DS‑CDMA, where a single wideband receiver simultaneously demodulates both legs — there is no channel‑switching problem to solve. Wong & Lim is also CDMA. So the art cited in Grounds 1–2 teaches away from, or at least does not supply, the very mechanism the '487 emphasizes. A POSITA combining Yamashita with a WLAN would not automatically arrive at round‑robin channel hopping; arguably the CDMA teaching makes the WLAN port less obvious, because the reference frame supplies a receiver architecture the WLAN terminal does not have.

• Rebuttal: the claims as reconstructed (method: "concurrently connected to at least two APs") are not limited to the time‑division implementation — the specification offers the multi‑radio alternative, and the round mechanism appears only in FIG. 14‑15 detail. If the granted claims do not recite channel switching/rounds, this counter‑argument collapses at the claim‑construction step. This is the single most important thing to resolve by reading the actual printed claims.

1. The "already‑established link, no re‑association" advantage in a packet WLAN is a real difference. 802.11 association/re‑association has state (keys, AID, bridge tables) that CDMA soft handoff does not. Whether importing soft handoff necessarily yields the claimed benefit is arguable.
2. Ground 3's Arvidsson date problem (see above) and Ground 4's unverified "Lee" weaken two of the four pleaded grounds.
3. No merits ruling exists. IPR2019‑00462 was terminated by joint motion under 35 U.S.C. §317(a) on Apr 17, 2019 before any final written decision; the petition never received a Board obviousness holding. Nothing in the record is a legal determination of obviousness, and the settlement gives no admission of invalidity.

8. Secondary considerations

• No objective evidence of non‑obviousness is on this record. There is no unexpected‑results, long‑felt‑need, or failure‑of‑others evidence in the materials reviewed.
• Commercial success / licensing: Maxell's campaign (E.D. Tex. 5:18‑cv‑00033 (Huawei) and 5:18‑cv‑00034 (ZTE), both terminated 2019 on settlement; W.D. Tex. 6:22‑cv‑00334 (Lenovo/Motorola, closed June 2023); ITC 337‑TA‑1312 (withdrawn) and 337‑TA‑1324 (terminated on settlement Aug 18, 2023, asserting claims 1, 3, 4) demonstrates that Maxell extracted licensing value from the '487. Under WBIP/Volvo, a nexus must tie that success to the claimed advance rather than to the portfolio or to other asserted patents ('821, '394, '086, '590, '284) — which is not established here. Note also that the '487 is recorded "Expired – Lifetime" (adjusted expiration 2025‑10‑05), which further undercuts any forward‑looking commercial‑success narrative.

9. Bottom line

Overall: the '487's asserted novelty is the "multilink" — defined in the specification as the simultaneous connection of one terminal to a plurality of base stations. That concept is squarely disclosed by Yamashita (US 6,108,547) in the very same architectural setting (multiple base stations on a wired line to a central node, one controlling, traffic forwarded over the line), a decade before the critical date, and its benefits are independently documented by Wong & Lim (1997). Layering Arvidsson for the least‑loaded‑node election and ordinary duplicate‑suppression for the buffering case produces a combination in which every element is known, every result is predictable, and the motivation comes from the patent's own stated problem. On this record, claims 1–4 (and the reconstructed claim 5) would more likely than not have been held obvious under §103, with the residual uncertainty concentrated in (a) the exact printed claim language — particularly whether the single‑radio time‑division channel‑switching mechanism is recited — and (b) the priority dates of Arvidsson and the unidentified "Lee" exhibit.

Three things I could not verify and would not assert: (i) the printed claim text (all claim mapping above is a reconstruction); (ii) the full text of US 6,760,599 (Uhlik) and of the "Lee" exhibit; (iii) the complete examiner‑cited art, because the Patent Citations/Cited By tables were absent from the fetched page.