US 7426392

I'll verify this patent number against live sources and check for any 2026 litigation activity.

US Patent 7,426,392 — Verified Summary

I ran two live searches: one for the patent itself (Google Patents/Justia/FreePatentsOnline corroborate the number and content), and one for USPTO/CAFC docket activity in 2026. Results below, with the authoritative full text in your message treated as controlling.

Bibliographic data

Abstract (as issued)

"A search receiver within a wireless communicator uses an adaptive path detection threshold. By using a continually updated path detection threshold, a communicator may be adapted to the energy of a base station. A search receiver continually updates the path detection threshold to reduce a false alarm rate and enhance the accuracy of the search detector."

Plain-language overview of the claims

Claim 1 (the only independent claim) — a wireless communicator with three cooperative parts:

1. Two communication receivers plus a handoff context. A first communication receiver talks to a first base station before a handoff; a second communication receiver talks to a second base station after the handoff. The search and the traffic receivers are thus tied together by a handoff event.
2. A search receiver made of three elements:
   • a searcher that searches for base stations (including the second base station) by summing detected multipath components on a per-base-station basis for components that exceed a variable path detection threshold, and refraining from summing components that do not exceed it;
   • a power profile memory storing measured energy levels associated with the detected multipath components for each base station;
   • a path threshold calculator that dynamically varies the variable path detection threshold on a per-base-station basis over a plurality of searches for each base station, based on the measured energy levels in the power profile memory.
3. The handoff hook. Based on those stored measured energy levels, a handoff decision is made causing the second communication receiver to communicate with the second base station and the first communication receiver to refrain from communicating with the first base station.

In plain terms: the threshold that decides whether a multipath "path" is counted is not fixed — it is recomputed per base station from that base station's own measured energy, repeatedly over successive searches, and the result drives a handoff.

Dependent claims 2–7 (2 depends on 1; 3, 4, 5, 6, 7 depend on 2):

• Claim 2 — narrows the multipath components to pilot signal portions; has the searcher identify the second base station as a handoff candidate; has the calculator update the threshold to exclude multipath components that do not significantly contribute to base-station selection; has the searcher repeat searching with the updated thresholds; requires that the calculator either refrain from updating when no paths are detected or reset to a default value when no paths are detected; and states the threshold is increased for greater measured energy and decreased for lower measured energy.
• Claim 3 — measured energy is a total received energy estimate (sum of detected path energies), and the threshold is varied using measured energy levels stored for all base stations, not just the one being thresholded.
• Claim 4 — measured energy is the strongest detected multipath component, and the threshold is set by multiplying that strongest-path energy by a predetermined factor.
• Claim 5 — the search receiver is a multi-dwell search receiver that reduces search time by skipping follow-on dwells for components below threshold.
• Claim 6 — the second receiver communicates with the second base station after selection for handoff and just prior to handoff, concurrently with the first receiver; both receivers are coupled to a common receive antenna.
• Claim 7 — the calculator varies the threshold by multiplying a total measured energy value for each base station by a predetermined factor.

Note on scope narrowing. The issued claim 1 is materially narrower than claim 1 of the published application US 2003/0220120 A1, which recited simply "a search receiver to search for a base station by detecting multipath components … that exceed a path detection threshold, said search receiver including a path threshold calculator to adapt said path detection threshold over time based on a measured energy associated with the base station." The issued claim added the dual-receiver/handoff architecture and the per-base-station summing/refraining limitation. This pattern is consistent with amendment during prosecution, but I do not have the file wrapper, so I cannot say with confidence what was amended, when, or why.

Specification support for the adaptation schemes (from the detailed description): a threshold set a fixed ratio (X dB) below the base station's total detected energy; a fixed ratio below the strongest detected path; a ratio made dependent on total energy; and a floor of the form max(X dB below strongest path, worst case), where "worst case" is the minimum possible threshold value, to avoid dropping below the noise floor. The spec also notes the threshold need not be updated every search (e.g., every third search), and updates may draw on detected path information from all currently searched base stations.

Litigation / 2026 docket check

I found no USPTO or CAFC 2026 docket entries involving US 7,426,392. My CAFC scheduled-cases, opinion, and Federal Circuit practice-update searches returned only unrelated 2026 matters (e.g., Extremity Medical v. Nextremity, No. 25-1160; FedEx v. Qualcomm; Spero v. Volkswagen; Wyeth v. AstraZeneca). None referenced the '392 patent. The only forward citations of record in the patent text are to non-analogous fields:

• US 2008/0162985 A1 (published 2008-07-03, George Henry Ahrens) and its sibling US 7,756,048 B2 (2010-07-13, IBM) — customizable surveillance of network interfaces;
• US 2010/0029232 A1 (2010-02-04, Siemens VDO Automotive) and US 8,588,716 B2 (2013-11-19, Continental Automotive) — searching for signals among interference signals in a multi-channel radio receiver.

This absence of activity is consistent with the patent having lapsed in 2016, a decade before the current date. I cannot rule out unindexed district court or PTAB filings, but nothing surfaced in my searches.

Notable prior art cited on the face of the patent

The 13 examiner citations include reference points a validity analyst would want to check: US 6,229,842 B1 (Ericsson, "Adaptive path selection threshold setting for DS-CDMA receivers," 2001-05-08), US 6,510,143 B1 (Alcatel, "Adaptive path searcher in a CDMA receiver," 2003-01-21), US 6,580,749 B1 (NEC, controllable search range), US 6,465,606 B1 (Qualcomm, search acquisition in multi-carrier), US 6,745,034 B2 (Nokia, "adaptively selecting a handoff threshold"), and US 2002/0114378 A1 (Philips, "Signal discriminator for a spread spectrum system"). Two non-patent citations are Fukumoto et al. papers in IEICE Trans. on Communications (E81-B(7), July 1998) and IEICE Trans. Fundamentals (E83-A(11), 2000) on matched-filter rake combining and pilot-symbol-assisted coherent rake path search.

The claimed distinction over this art is the per-base-station, over-time-variable threshold derived from the base station's own measured energy, combined with the summing/non-summing effect on the base station's total energy estimate and the handoff consequence — whereas the Ericsson and Alcatel references are generally characterized in the patent's background as using a fixed threshold tied to total input power (worst case) or to a network-specified base station report threshold.

Uncertainty flags

• Title spelling: "theresholds" is reproduced literally from the record; there is a 2008-12-02 certificate of correction on file, but I do not know authoritatively what it corrected (it may or may not relate to the title or claims). Do not assume the claims as reproduced above are post-correction without checking the correction document.
• Assignee chain: Google Patents' "current assignee (Micron Technology Inc)" reflects recorded assignment data; I have not independently confirmed present ownership or any subsequent transfers after 2013.
• Expiration dates: the record shows both an "adjusted expiration" entry of 2023-06-06 and a maintenance-fee lapse effective 2016-09-16. These are not inconsistent (the statutory/adjusted term and the actual lapse are different concepts), but the practical expiration is the 2016 nonpayment of maintenance fees.
• Claim-construction/amendment history: not verified; the file wrapper was not retrieved.
• No 2026 litigation found — stated as a negative search result, not as proof of absence.

I'll search for litigation involving US patent 7426392.

Let me run more targeted searches on litigation databases and patent dockets.

Summary Answer

I found no litigation involving U.S. Patent No. 7,426,392 ("Search receiver using adaptive detection thresholds"). Based on the searches I ran, there is no district court action, ITC investigation, PTAB proceeding, or CAFC appeal that I can identify as involving this specific patent.

I want to be explicit: this is "not found" rather than a verified negative. I could not get direct search access to the primary litigation databases you named (Unified Patents' litigation caselist, PACER, Docket Navigator, RPX Insight), so my conclusion rests on open-web searching, which is a weaker basis. Treat the absence as reported/observed rather than confirmed.

What I checked and what came back

The strongest specific hits were all false positives from other jurisdictions/registries, which I flag so they aren't mistaken for the U.S. patent:

• JP 7426392 B2 — a Japanese patent owned by Nike Innovate C.V. for a "Loop structure with pocket space for storage" (a garment patent). Completely unrelated to U.S. 7,426,392.
• JP 7426392 A (1992) cited in EP 0563841 — a Japanese application for a magnetic recording/reproducing unit. Also unrelated.
• A Brazilian municipal register listing an individual with registration number "7426392." Unrelated.

Corroborating context from the patent's own record

The authoritative patent text you supplied itself points away from litigation:

• Prosecution/ownership history: Filed 2002-05-23 by inventors Tsofnat Hagin-Metzer and David Ben-Eli, originally assigned to D.S.P.C. Technologies Ltd., then to Intel Corporation (2006), then to Micron Technology, Inc. (recorded 2013-07-04, effective 2011-11-22).
• No litigation entries appear in the Google Patents record you provided. The only "family"/related activity is the pre-grant publication US 2003/0220120 A1 (2003-11-27) and a certificate of correction (2008-12-02).
• Status: expired. Maintenance fees lapsed; the patent expired 2016-09-16 for nonpayment of maintenance fees under 37 CFR 1.362, with a Google-noted adjusted expiration of 2023-06-06. An expired, fee-lapsed patent that was never asserted is consistent with the absence of litigation.
• Cited-by references are to non-patent literature and unrelated third-party patents (IBM, Continental Automotive), not to litigation documents.

Case table

Because no litigation was located, there are no cases to populate:

Important caveats and how to verify

1. My tooling was limited. I could not query PACER, Docket Navigator, Unified Patents' caselist, or RPX Insight directly. If you need a defensible negative, those are the sources to run (plus the PTAB's Patent Trial and Appeal Board Endpoint / PatentCenter "Litigation" and the ITC's EDIS for Section 337 complaints).
2. Assignees to watch when checking: D.S.P.C. Technologies Ltd., Intel Corporation, and Micron Technology, Inc. Suits, if any, would most plausibly have been brought by or against those entities during 2008–2016 (before the fee lapse), or in a later NPE campaign — but I found no evidence of either.
3. Don't confuse with patent numbers that differ by a digit. In my searches I repeatedly encountered U.S. Patent No. 7,426,380 (Samsung) and other "7,426,3xx" numbers in unrelated matters; none involve 7,426,392.

If you'd like, I can next attempt targeted searches framed around the likely parties (e.g., "Micron 7,426,392"), around PACER-style docket strings, or around NPE/patent-assertion campaigns — but on the current record, my answer is that no known litigation involves U.S. Patent No. 7,426,392.

Proceedings overview

Zero AIA trial proceedings are on file for US 7,426,392. The structured PTAB/OPS dataset in this prompt returns no IPR, PGR, or CBM for this patent, and targeted web searches (Google Patents, Justia, PTAB E2E/ptacts.uspto.gov, Docket Alarm PTAB dockets, PTAB Bar Association practitioner lists) surfaced none either. The count is 0 active, 0 invalidated, 0 sustained, 0 settled, 0 institution-denied. Bottom line for a defendant: there is no IPR roadmap and no canceled claim to hide behind — but there is also no live patent to be sued on, because 7,426,392 lapsed for failure to pay maintenance fees on 2016-09-16 and has been statutorily expired since then.

No proceeding to report — verification detail

• Type: N/A (no Inter Partes Review, Post-Grant Review, or Covered Business Method Review identified)
• Filed: N/A
• Status: N/A — the absence of any AIA trial entry is the finding.
• Sources checked: the canonical "PTAB proceedings on file" block supplied with this task (USPTO ODP ingest, current); Google Patents family/legal-events record for US 7,426,392 B2 (which lists cited by and citing documents but no US…B2/A1 PTAB publications and no reexamination certificates); Justia's patent page for US 7,426,392; USPTO PTAB E2E/P-TACTS queries; Docket Alarm PTAB coverage.
• False-positive alert for whoever runs the next search: searching "7426392" or "'392 patent" on PTAB dockets heavily returns IPR2022-00529, [Apple Inc.](/litigations/by-plaintiff/Apple%20Inc.) v. Scramoge Technology Ltd., involving US 10,193,392 — a different patent (wireless power transfer, claims 1–8, § 103 grounds). That is not this patent. Similarly, the DE 60016306 T2 and JP 7426392 B2 hits are unrelated documents that merely cite or happen to share the number. Do not import those proceedings into a 7,426,392 analysis.
• Judge panel / petition grounds / institution decision / FWD / settlement / appeal: inapplicable — nothing to report, and I will not fabricate a panel or a disposition.
• Defensive value: Neutral-to-favorable in an unusual way. The absence of IPRs is not a signal that the patent is hardened — it is a signal that by the time AIA trials became widely used (2012–2014), the patent was in Intel-then-Micron's hands, was not being asserted against anyone, and its 8-year maintenance fee went unpaid in 2016.

Strategic summary

Claim status: all seven claims are UNTESTED — none canceled, none confirmed. Claims 1–7 (1 independent; 2–7 dependent) have never been the subject of an AIA trial, a reexamination certificate, or (as far as public records show) an asserted infringement case. There is no FWD to quote, no claim-by-claim verdict, and no surviving-narrowed claim set. Anyone who tells you "claims 1–5 were canceled" or "the patent survived two IPRs" is thinking of a different patent (most likely US 10,193,392 in the Scramoge campaign).

The real gating fact is expiry, not validity. From the structured record: filed 2002-05-23, granted 2008-09-16, assigned D.S.P.C. Technologies Ltd. → Intel Corporation (recorded 2006-11-08, effective 2006-09-26) → Micron Technology, Inc. (recorded 2013-07-04, effective 2011-11-22). The legal events show a maintenance-fee reminder on 2016-04-29 and lapse for failure to pay maintenance fees on 2016-09-16, with status "Expired – Fee Related" and an adjusted expiration of 2023-06-06. A patent that lapsed in 2016 cannot be infringed by conduct today, and a complaint filed in 2026 could not reach back six years (35 U.S.C. § 286) to any window when the patent was in force. So the practical defense is not an IPR — it is the patent's own face.

Estoppel landscape: nothing is estopped, because no one ever petitioned. 35 U.S.C. § 315(e)(2) estoppel attaches only to a petitioner (and its privies/real parties in interest) that obtained a final written decision. There is no such party here. Every prior-art ground — including art that would have been raised in a hypothetical IPR — remains freely available in district court under §§ 102/103/112, subject to the ordinary clear-and-convincing standard. There is likewise no § 315(b) one-year clock running against any defendant, and no General Plastic or § 325(d) baggage on this patent. No defensive aggregator (Unified Patents, RPX, etc.) appears anywhere in the chain; the patent has simply been dormant in Micron's portfolio.

Pattern signals: none. No repeat petitioner, no serial IPR campaign, no PTAB appeal history, no litigation-driven validity fight. The cited-by set (two third-party references, both unrelated — network interface surveillance and multi-channel radio receiver searching) confirms the patent was never a hot target. The most notable pattern is the opposite of a troll narrative: a large operating company (Micron, ex-Intel/DSPC) let the patent die for nonpayment of fees rather than monetize it.

Recommended next steps

• If you are a defendant or demand-letter recipient: do not spend a PTAB budget here. Plead the patent's status directly — quote the legal-events entry showing lapse for failure to pay maintenance fees on 2016-09-16, which produced "PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362." Confirm current chain of title at the USPTO Assignment database (the record shows Micron Technology, Inc. as current assignee, with the 2013-07-04 assignment effective 2011-11-22) — a demand letter asserting claims 1–7 in 2026 from any entity warrants a title and standing inquiry.
• Damages reality check: because the patent lapsed in 2016 and today is 2026-09-27, the § 286 six-year recovery window contains no period of enforceability. Even on the most plaintiff-favorable view, there is no accruable past damages and no prospective relief.
• If a client insists on a validity record: an IPR (or ex parte reexamination) against an expired patent is procedurally possible but provides essentially no value — the patent owner cannot amend, and the resulting paper does not help a defense that is already complete on expiration. I would advise against it and would say so in writing.
• If you nonetheless want to confirm the zero-proceeding finding independently: query the PTAB E2E / P-TACTS "Patent Number" field with 7426392 (not "392"), and search Docket Alarm's PTAB index for the same. Watch specifically for any newly filed petition that the ODP ingest has not yet picked up — that is the only event that would change this analysis, and given the 2016 lapse it is very unlikely.

Reliability caveat: my searches found no PTAB activity, but absence of search results is weaker evidence than a positive docket query. I did not have a live API response from PTAB E2E in this session; the finding rests on the supplied structured block plus public sources. If the stakes are high, run the PTAB E2E patent-number query yourself before relying on the zero-count in a filing.

I'll verify the assignment record and check for litigation/assertion history before writing.

Let me verify the Intel→Micron transaction, check for any litigation, and look for the assigning correspondent.

Ownership & NPE Analysis — US 7,426,392 B2

"Search receiver using adaptive detection thresholds" · Appl. 10/154,948 · Filed 2002-05-23 · Granted 2008-09-16

Source note / limitation on this report: The USPTO Assignment Center (https://assignmentcenter.uspto.gov/) and the legacy Assignment Search (https://assignment.uspto.gov/patent/index.html) are interactive query tools that could not be scripted in this pass. The reel/frame strings below are taken from the assignment records as reproduced verbatim in the legal-events feed for this patent (https://patents.google.com/patent/US7426392/en), which carries the USPTO REEL/FRAME values. The correspondent-of-record field was not retrievable for any of the three recordings — I flag that explicitly in Signal 3 rather than infer it. No assignment data has been fabricated.

Inventors

Departure pattern — none. This is the inverse of the fire-sale tell. Both inventors continued filing in the same technical lineage for years after this application: Ben-Eli appears as a named inventor on Marvell World Trade-assigned patents as late as 2016 (e.g., US 9,344,303) and Hagin-Metzer on Marvell World Trade-assigned US 8,724,597 (granted 2014-05-13). In other words, the inventors' work followed the business unit (Intel cellular baseband → Marvell, 2006), while the patent did not. That divergence is the single most analytically interesting fact in this chain and is discussed under Signal 2.

Original assignee

D.S.P.C. Technologies Ltd. (Israel) — the entity named on the 2002 assignment, REEL/FRAME 012944/0314. The face of the issued patent instead names Intel Corporation (Santa Clara, CA) because of the 2006 intra-group assignment below (https://patents.justia.com/patent/[7426392](/patent/7426392)).

• Business: fabless developer of digital signal processors and CDMA baseband chipset products for cellular handsets; NASDAQ-listed successor to DSP Communications (ticker DSPC). Founded 1987 by Davidi Gilo.
• Shipped product embodying the claims: Yes — the invention is a search receiver (baseband searcher / multi-dwell path-detection logic) for a CDMA/W-CDMA handset modem, which is precisely the DSPC baseband product line. Intel bought the company expressly to enter cellular baseband (Craig Barrett's "bridgehead" rationale).
• Status: dissolved / absorbed. Intel acquired DSPC in 1999 for ~$1.6B cash. Intel then sold the cellular baseband and applications-processor business (the former DSPC operation, ~1,400 employees, ~500 in Israel) to Marvell Technology Group for ~$600M, announced 2006-06-27, closing ~November 2006.
  • Sources: https://www.haaretz.com/2006-06-27/ty-article/seven-years-after-buying-dspc-intels-selling-it-for-half-the-price/0000017f-e099-d38f-a57f-e6dbe9d50000 ; https://finance.walla.co.il/item/[930859](/patent/930859)
• Current owner of the patent: Micron Technology, Inc. (an operating semiconductor manufacturer, Boise, ID) — see timeline.

Assignment timeline

Three recorded conveyances. All three are ordinary corporate assignments; none is an LLC/security/license recording.

• 2002-05-23 (executed) / recorded 2002-05-23 — Reel 012944/0314

  • Conveyance: Assignment of assignors' interest
  • Assignor: Hagin-Metzer, Tsofnat; Ben-Eli, David
  • Assignee: D.S.P.C. Technologies Ltd. (Israel)
  • Correspondent: not retrievable from the indexed record (recorded at filing; the reel/abstract text shows only the three named assignors).
  • Context: Ordinary inventor-to-employer assignment, executed and recorded on the filing date.

• 2006-09-26 (executed) / recorded 2006-11-08 — Reel 018499/0629

  • Conveyance: Assignment
  • Assignor: DSPC Technologies Ltd.
  • Assignee: Intel Corporation (California)
  • Correspondent: not retrievable from the indexed record.
  • Context: Intra-group/internal reorganization — DSPC's patents were perfected into Intel's name seven years after the 1999 acquisition. Note the timing anomaly: this recording lands in the exact window of Intel's sale of the DSPC business to Marvell (announced 2006-06-27; closed ~Nov 2006), yet the patent was recorded to Intel, not to Marvell. The patent was carved out of the divested business.

• 2011-11-22 (stated effective date) / recorded 2013-07-04 — Reel 030747/0001

  • Conveyance: Assignment
  • Assignor: INTEL CORPORATION
  • Assignee: MICRON TECHNOLOGY, INC. (Idaho)
  • Correspondent: not retrievable from the indexed record.
  • Context: Unclear. The record is a straight assignment of assignor's interest, but the stated effective date precedes the USPTO recording by ~19 months (2011-11-22 → 2013-07-04). I searched for an Intel→Micron corporate transaction dated circa November 2011 and could not identify one; the documented Intel/Micron activity in that window is the IM Flash NAND joint venture (Micron–Intellectual Ventures license, 2011-05-31; IM Flash restructuring, 2012-02-28 / 2012-04-06). I cannot state what consideration or agreement produced this conveyance — treat the nature of the deal as an open item.

Non-assignment legal events for completeness (not part of the chain): certificate of correction 2008-12-02; maintenance fee payment Year 4 on 2011-09-22; maintenance fee reminder 2016-04-29; lapse for failure to pay maintenance fees 2016-09-16 (STCH/FP entries 2016-10-14, 2016-10-17, 2016-11-08). The Google Patents status line separately shows an "adjusted expiration" of 2023-06-06; that computed date is inconsistent with the 2016 lapse events and should be verified against ODP/PatentCenter before being relied on.

Timeline diagram

timeline
    title Ownership of US 7426392
    2002 : Inventors assign to DSPC Technologies Ltd
    2006 : DSPC Technologies assigned to Intel Corp
    2008 : Patent issues naming Intel as assignee
    2008 : Certificate of correction recorded
    2011 : Intel executes assignment to Micron
    2013 : Micron assignment recorded at USPTO
    2016 : Lapse for unpaid maintenance fee

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. All three links run between operating companies: inventors → D.S.P.C. Technologies Ltd. (012944/0314) → Intel Corporation (018499/0629) → Micron Technology, Inc. (030747/0001). No "IP / Patents / Holdings / Licensing / Ventures" suffix anywhere in the chain, no registered-agent address, no single-purpose LLC. Micron is a publicly traded, product-shipping chipmaker.

2. Known asserter in the chain — NOT PRESENT (with one adjacency worth noting). No Acacia / Marathon / IV / Wi-LAN / Conversant / Vringo / Pendrell / MPHJ / Lumen View entity appears as an assignor or assignee on any reel/frame above. The only contact point with a public NPE is adjacency, not assignment: Micron and Intellectual Ventures announced an IP license agreement on 2011-05-31 — a portfolio license plus mutual patent-purchase rights (https://www.intellectualventures.com — Micron/IV license press release). That announcement sits ~6 months before the 2011-11-22 effective date of REEL/FRAME 030747/0001, but it is directed the other way (Micron licensing in and acquiring from IV for defensive purposes) and there is no evidence in any source consulted that US 7,426,392 was transferred to IV or to any IV-sponsored entity. I decline to draw that inference. Likewise, I was unable to verify in this pass any link between this patent and Round Rock Research (a Micron-origin NPE vehicle) — in any event Round Rock's relationship to Micron runs the opposite direction, and Micron is the assignee here. Call: not present; no supporting reel/frame.

3. Repeat correspondent across the chain — UNCLEAR / NO DATA. This is the signal I most wanted and could not deliver: the correspondent field was not retrievable from the indexed record for 012944/0314, 018499/0629, or 030747/0001. For the record, and explicitly distinguishing it from an assignment correspondent, the patent's prosecution attorney of record is listed as Schwegman, Lundberg & Woessner, P.A. (https://patents.justia.com/patent/7426392) — that is a large general-practice IP firm that services both operating companies and assertion-side clients, so its presence as prosecution counsel carries no NPE weight on its own. Two of the three recordings are by a large operating company's in-house/outside IP counsel on a single acquisition, and there is no recurrence pattern to test. Recommend a direct Assignment Center pull of the three reel/frames to fill this column before any NPE call is made on related Intel/DSPC-lineage patents.

4. Cascading transfers — NOT PRESENT. Three assignments spread across 2002, 2006, and 2011 — ~4.3 years and ~5.2 years apart. No chained LLCs, no <24-month cascade, no shared correspondence address across links, no common principals evidenced.

5. Pre-litigation transfer — NOT PRESENT. I located no infringement litigation naming US 7,426,392 in the sources searched. The last ownership change (030747/0001) was executed 2011-11-22 / recorded 2013-07-04, and the maintenance-fee lapse follows in 2016 — the record is the profile of a patent that was held and then dropped, not one that was moved to fund a suit.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 proceeding involving DSPC, Intel, or Micron is associated with this patent. DSPC was acquired (not liquidated); Intel divested the baseband business as a going concern; the Intel→Micron conveyance is a private corporate transfer, not a bankruptcy sale.

7. Privateering — NOT PRESENT on this record. The classic privateering tell is an operating company arming an NPE to sue its competitors. Here the patent moved from one operating semiconductor company (Intel) to another (Micron) and, as far as sources show, was never asserted against anyone. The 2006 Marvell carve-out is a divestiture structuring fact, not privateering: Intel kept the patent while selling the people and the product line.

8. Defensive aggregator — NOT PRESENT. The chain terminates at Micron Technology, Inc., not at RPX, AST, LOT, Unified Patents, or OIN. Inverse-signal check: no defensive-aggregator terminal link exists, so the patent was not "neutralized" by acquisition into a defensive pool.

Verdict

Insufficient data — read narrowly, and not as an NPE signal.

Justification. The ownership chain is fully documented and is a pure operating-company lineage: inventors → D.S.P.C. Technologies Ltd. (Reel 012944/0314, 2002) → Intel Corporation (Reel 018499/0629, executed 2006-09-26) → Micron Technology, Inc. (Reel 030747/0001, effective 2011-11-22). Every link is a genuine corporate assignment between chipmakers; zero of the eight NPE signals are present, and the transfer cadence (4–5 years between events) is antithetical to an assertion-vehicle chain. I select "Insufficient data" rather than "Operating-company assertion" only because the taxonomy's operating-company verdict requires an active assertion against competitors, and no suit on US 7,426,392 was located — the honest description is operating-company chain with no assertion activity observed, which the menu does not offer. Two caveats drive the residual uncertainty: (a) the correspondent-of-record field could not be retrieved for any of the three reel/frames, so the repeat-correspondent test is unrun; and (b) the nature of the Intel → Micron conveyance effective 2011-11-22 is unidentified, and its ~19-month recording lag (recorded 2013-07-04) is unexplained in the available record.

Verify directly at: https://assignmentcenter.uspto.gov/ (search "7426392") and https://assignment.uspto.gov/patent/index.html — pull the correspondent field for Reel 012944/0314, Reel 018499/0629, and Reel 030747/0001.

I'll research US 7426392 and its cited prior art before building the analysis.

Prior-Art Analysis — US 7,426,392 B2

1. Patent record (verified against the authoritative full text provided)

Since the filing date is 2002-05-23, pre-AIA 35 U.S.C. §§ 102/103 apply.

Note on claim scope. The granted claim 1 is materially narrower than published claim 1 of US 2003/0220120 A1. Granted claim 1 requires (a) first and second communication receivers bracketing a handoff, (b) a searcher that sums detected multipath components on a per-base-station basis that exceed a variable threshold and refrains from summing those that do not, (c) a power profile memory storing measured energy levels per base station, (d) a path threshold calculator that varies the threshold on a per-base-station basis over a plurality of searches based on those stored measured energy levels, and (e) a handoff decision driven by those stored energy levels. Any anticipation analysis must be run against that combination, not the broader published claim.

2. Threshold legal caveat before the mappings

The 13 references below are the references cited on the face of the patent (i.e., considered by the examiner). Two things must be stated plainly:

1. On the record as I have it, none of the 13 U.S. patent citations appears to disclose every element of granted claim 1 as arranged in the claim. They are best characterized as § 103 obviousness art (individually or in combination), not as clean § 102 anticipations. I flag below where an anticipation argument is at least arguable, and where it is not.
2. The examiner's citation of a reference is not evidence that it anticipates; examiners routinely cite art under § 102 that is later defended as § 103 art.

For each reference I give: full citation, dates, description, and the potential § 102 mapping (with an explicit "no" where none exists).

3. The 13 cited U.S. patent references

3.1 — US 6,229,842 B1 — the most on-point citation

• Full citation: US 6,229,842 B1, "Adaptive path selection threshold setting for DS-CDMA receivers," Telefonaktiebolaget LM Ericsson (inventors Schulist, Klang, He). Priority 1998-07-16; PCT filing 1999-07-08; granted 2001-05-08.
• Disclosure: A path selection unit identifies valid peaks in a delay power profile (DPP). The system estimates the noise level in the DPP to form an accurate SNR estimate, which is mapped through a threshold table or mapping function to set the path selection threshold. The threshold is formed by multiplying the noise level by a threshold factor, where the factor varies linearly/non-linearly with SNR (e.g., constant factor below SNR ≈ 2.5, then y = 0.1822x + 1.444). SNR is estimated iteratively by removing N strongest peaks and evaluating the residual.
• Claim mapping: The reference discloses an adaptive path-selection threshold and a "power profile"-type construct, but the adaptation variable is estimated noise/SNR, not the measured energy of the base station's own detected multipath components stored in memory. That distinction defeats anticipation of claim 1 and claim 3.
  • Arguable § 102: claim 4 (threshold derived from the strongest path) is the only claim where a § 102 argument is even colorable, because '842 removes the strongest peaks to estimate noise — but it uses the strongest path to characterize noise, not to set the threshold, so even here I would expect it to fail as § 102 and succeed as § 103.
  • Best characterization: § 103 — primary reference against claims 2, 3, 4, 7.

3.2 — US 6,510,143 B1 — second most on-point citation

• Full citation: US 6,510,143 B1, "Adaptive path searcher in a CDMA receiver," Alcatel. Filed 1999-07-01; granted 2003-01-21 (DE 69906262 T2 counterpart).
• Disclosure: Path selection in a CDMA Rake receiver by comparing transmission channel energy with a threshold that is dynamically determined as a function of the mean of noise and interference samples derived from the transmission channel profile. Noise/interference samples are obtained by suppressing at least the most important (~10%) estimated paths. The noise distribution is modeled as a centered chi-squared variable; the threshold is set so the probability of a noise sample exceeding it is very low. The specification expressly discusses and criticizes Fukumoto's "maximum of two thresholds" approach.
• Claim mapping: Adapts a path selection threshold over time — conceptually close — but the adaptation basis is the noise/interference mean from the profile, not a measured energy level of the base station (claim 1(d), claim 3). It also lacks the two-communication-receiver-before/after-handoff architecture.
  • Best characterization: § 103, potentially combinable with 3.1. No clean § 102 anticipation of any granted claim.

3.3 — US 6,466,606 B1 — relevant to claim 5 (multi-dwell)

• Full citation: US 6,466,606 B1, "Method and apparatus for performing search acquisition in a multi-carrier communication system," Qualcomm Incorporated. Priority 1999-12-22; granted 2002-10-15.
• Disclosure: Multi-carrier pilot acquisition using a large search window with a first detection threshold and, when energy exceeds it, a smaller "zoom" window; uses a first dwell threshold T₁ and a validation (second) threshold, with noncoherent accumulation over M integrations.
• Claim mapping: The closest art on claim 5 (multi-dwell search receiver). But the thresholds are fixed protocol thresholds applied to hypotheses/windows, not per-base-station thresholds dynamically varied from stored measured base energy. No anticipation of claims 1–4, 6, 7.
  • Best characterization: § 103 against claim 5; possibly § 102 against a hypothetical bare "multi-dwell searcher" claim, but claim 5 is not that bare.

3.4 — US 6,580,749 B1

• Full citation: US 6,580,749 B1, "CDMA receiver having a controllable search range and method for controlling the same," NEC Corporation. Priority 1999-05-10; granted 2003-06-17.
• Disclosure: Controls the range over which a CDMA receiver searches, rather than the detection threshold; search-window/range management.
• Claim mapping: Relevant only to the general "searcher" context. Does not disclose a variable per-base-station path detection threshold based on measured energy. No § 102 anticipation. § 103 background art at best.

3.5 — US 6,275,186 B1

• Full citation: US 6,275,186 B1, "Device and method for locating a mobile station in a mobile communication system," Samsung Electronics Co., Ltd. Priority 1998-12-10; granted 2001-08-14.
• Disclosure: Mobile-station position location using pilot/PN measurements.
• Claim mapping: Addresses location, not adaptive path detection thresholds. No § 102 anticipation. Background art / possibly § 103 for the "measured energy" measurement techniques.

3.6 — US 6,801,511 B2

• Full citation: US 6,801,511 B2, "Apparatus and method for performing handoff in a mobile communication system," [Samsung Electronics Co.](/litigations/by-defendant/Samsung%20Electronics%20Co.), Ltd. Priority 1999-12-30; granted 2004-10-05.
• Claim mapping: Touches the handoff context of claim 1, but discloses no variable, energy-adaptive, per-base-station path detection threshold. No § 102 anticipation.

3.7 — US 6,961,398 B2

• Full citation: US 6,961,398 B2, "Base station synchronization for wireless communication systems," InterDigital Technology Corp. Priority 2000-04-07; granted 2005-11-01. (Post-filing grant; potential § 102(e) art only on its earlier effective filing date.)
• Claim mapping: Base-station synchronization, not path-threshold adaptation. No § 102 anticipation of claims 1–7.

3.8 — US 6,894,996 B2

• Full citation: US 6,894,996 B2, "Apparatus and method for searching a base station in an asynchronous mobile communications system," Samsung Electronics Co., Ltd. Priority 2000-09-09; granted 2005-05-17.
• Claim mapping: Base-station searching in an asynchronous (W-CDMA-type) system — relevant technical field, but I have not retrieved text showing a per-base-station adaptive threshold keyed to stored detected-path energy. No § 102 anticipation based on available material; § 103 context only. (Flagged: full-text not retrieved; treat as provisional.)

3.9 — US 6,745,034 B2 — relevant to claim 1's handoff-decision clause

• Full citation: US 6,745,034 B2, "Apparatus, and associated method, for adaptively selecting a handoff threshold in a radio communication system," Nokia Corporation. Priority 2000-12-22; granted 2004-06-01.
• Disclosure: Adaptively selects the handoff (hysteresis) threshold using mobile-station position and velocity plus mean and standard deviation of measured RF propagation data; pilot signal-strength measurements of serving and target base stations; handoff requested when the difference exceeds the (adaptive) threshold.
• Claim mapping: Directly relevant to the final "wherein" clause of claim 1 (handoff decision based on measured energy) — but the "threshold" adapted here is a handoff decision threshold, not a path detection threshold used by a searcher to sum detected multipath components. No § 102 anticipation of claim 1 (different threshold, different role). Useful § 103 art on the handoff-decision element.

3.10 — US 2002/0114378 A1

• Full citation: US 2002/0114378 A1, "Signal discriminator for a spread spectrum system," Koninklijke Philips Electronics N.V. Priority 2001-02-21; published 2002-08-22 (published after the 2002-05-23 filing date; potential § 102(e) art on its 2001-02-21 filing date).
• Claim mapping: I have not retrieved the substance of this reference beyond its title/assignee. On its face ("signal discriminator") it appears directed to discriminating signal from noise in a spread-spectrum receiver, which may overlap the thresholding concept, but I cannot responsibly assert a claim-by-claim § 102 mapping. Treat as unverified — flag for full-text review.

3.11 — US 6,915,076 B1 — analogous-art problem

• Full citation: US 6,915,076 B1, "System and method for adaptively selecting a signal threshold of an optical link," Ciena Corporation. Priority 2001-05-14; granted 2005-07-05.
• Claim mapping: Adaptively selects a signal threshold, but in an optical link using optical-link performance data (BER/OSNR-type criteria) — a different art. Its citation is best explained as support for the general proposition that "adaptively selecting a threshold" was known. No § 102 anticipation of any claim; § 103 use would require a successful analogous-art argument, which is doubtful for a CDMA searcher context.

3.12 — US 6,565,209 B1

• Full citation: US 6,565,209 B1, "Method for frequency division duplex communications," AT&T Wireless Services, Inc. Priority 1997-02-06; granted 2003-05-06.
• Claim mapping: FDD communications method; no disclosure of per-base-station adaptive path detection thresholds. No § 102 anticipation.

3.13 — US 6,625,132 B1

• Full citation: US 6,625,132 B1, "Idle intersystem roaming determination and system reselection in a CDMA wireless communication system," Nortel Networks Ltd. Priority 1997-11-13; granted 2003-09-23.
• Claim mapping: Idle-mode roaming/reselection — touches the "search for base stations and decide" context (cf. the patent's discussion of idle-mode autonomous handoff), but no adaptive detection threshold. No § 102 anticipation.

4. Non-patent literature (both "cited by examiner")

4.1 — Fukumoto et al. (1998)

• Full citation: S. Fukumoto, M. Sawahashi, F. Adachi, "Matched filter-based Rake combiner for wideband DS-CDMA mobile radio," IEICE Transactions on Communications, vol. E81-B, no. 7 (July 1998), pp. 1384–1390.
• Relevance (verified indirectly via US 6,510,143, which describes it): Proposes a threshold depending on the lowest noise level, and — critically — derives a second threshold from a less-important sample, then uses the maximum of the two thresholds to decide whether a path must be selected.
• Claim mapping: This is the closest thing in the record to the patent's own "max(X dB below strongest path, worst case)" embodiment described in the specification. It is § 102-adjacent against the disclosed embodiment, not the granted claims — claim 1 recites a per-base-station threshold varied from stored measured energy levels over a plurality of searches, which Fukumoto's per-profile max-of-two-thresholds scheme does not show. Best used as § 103 art ("max of a signal-derived threshold and a floor threshold" was known).

4.2 — Fukumoto et al. (2000)

• Full citation: S. Fukumoto et al., "Path search performance and its parameter optimization of pilot symbol-assisted coherent Rake receiver for W-CDMA mobile radio," IEICE Transactions on Fundamentals, vol. E83-A, no. 11 (2000), pp. 2110–2119.
• Relevance: Path-search performance and parameter optimization — i.e., trading off detection threshold against false alarm / non-detection and search time. This is the conceptual antecedent for the patent's stated motivation (reducing false-alarm rate and search time via a better threshold).
• Claim mapping: No § 102 anticipation of the granted claims; strong § 103 motivation/background art, and potentially a § 102(b) "printed publication" reference for the underlying problem statement (not the claimed solution).

5. Bottom line

Most relevant prior art to US 7,426,392 B2:

1. US 6,229,842 B1 (Ericsson) — adaptive path-selection threshold from estimated SNR/noise in a delay power profile. Primary § 103 reference.
2. US 6,510,143 B1 (Alcatel) — adaptive path-selection threshold from noise/interference mean derived from the channel profile. Secondary § 103 reference; and it self-identifies the Fukumoto two-threshold approach as prior art.
3. Fukumoto et al. (1998 and 2000) — the two-threshold / parameter-optimization NPL, the acknowledged conceptual ancestor of the "max(X dB below strongest path, worst case)" embodiment.
4. US 6,466,606 B1 (Qualcomm) — multi-dwell, multi-threshold search (relevant to claim 5).
5. US 6,745,034 B2 (Nokia) — handoff decision threshold adaptation (relevant to claim 1's final "wherein" clause).

§ 102 assessment: On the record available to me, no single cited reference anticipates any of granted claims 1–7 in full. The differentiator is that every cited threshold-adaptation reference adapts on noise/SNR or interference estimates, whereas granted claims 1, 2, 3, 4 and 7 require adaptation on measured energy levels of the base station's own detected multipath components (total detected energy, or strongest detected path × predetermined factor), stored per base station in a power profile memory, and varied per base station across a plurality of searches, with the handoff decision driven by those stored levels. That is a genuine, documented gap in the cited art.

§ 103 assessment: Claims 2, 3, 4, 5 and 7 are the most exposed, and the strongest combination is US 6,229,842 + US 6,510,143 (adaptive path-selection threshold based on profile statistics), optionally + Fukumoto (1998) (threshold from strongest/less-important samples with a floor) + US 6,466,606 (multi-dwell thresholds) + US 6,745,034 (adaptive handoff threshold).

6. Things I could not verify — stated explicitly rather than assumed

• US 2002/0114378 A1 (Philips) and US 6,894,996 B2 (Samsung): I did not retrieve substantive text; the descriptions above are provisional and should be confirmed by full-text review before any § 102 assertion.
• I did not independently verify inventors, exact filing dates, or full claims of US 6,656,209, US 6,625,132, US 6,275,186, US 6,580,749, US 6,801,511, US 6,961,398, or US 6,915,076; the dates used are those in the patent's own front-page citation list.
• Litigation note (lower confidence): a third-party source reports that US 7,426,392 was among the patents asserted by XR Communications against Ruckus Wireless / ARRIS (alongside US 7,062,296, US 7,877,113, US 6,611,231, US 7,729,728). I did not confirm this from a court docket, so treat it as unverified context rather than fact.
• Per the operating rule, I interpreted the identifiers literally: the title spelling "theresholds" and the family identifier US 2003/0220120 A1 are reproduced as they appear.

I'll analyze the prior art referenced on the page. Let me first verify the disclosures of the key references.

Obviousness Analysis — U.S. Patent No. 7,426,392 B2

Title (as published): "Search receiver using adaptive detection theresholds"
App. No.: 10/154,948 | Filed/Priority: 2002-05-23 | Granted: 2008-09-16
Inventors: Tsofnat Hagin-Metzer; David Ben-Eli
Original assignee: Intel Corp. (assignment chain: D.S.P.C. Technologies Ltd. → Intel → Micron Technology, Inc.)
Status (per page): Expired – Fee Related; adjusted expiration 2023-06-06; lapsed 2016-09-16 for failure to pay maintenance fees.

Source of record: https://patents.google.com/patent/[US7426392](/patent/US7426392)/en

1. Bottom Line Up Front

Under 35 U.S.C. § 103 and the Graham v. John Deere / KSR Int'l v. Teleflex framework, claims 1–7 are, in my assessment, more likely than not invalid as obvious. The three "adaptive path selection threshold" citations already of record — US 6,229,842 (Ericsson), US 6,510,143 (Alcatel), and the Fukumoto non-patent literature — teach the core inventive concept (a searcher threshold that varies based on measured energy/noise in the delay profile) almost verbatim in language. The remaining elements (two communication receivers for pre-/post-handoff operation; a power-profile memory; a threshold calculator; multi-dwell staging) are each independently and expressly disclosed in other references of record (Samsung US 6,801,511; Nokia US 6,745,034; Qualcomm US 6,466,606).

The one element that gives me some pause is the feedback topology of claim 1/3: the threshold is varied based on the energy of the paths already detected for that base station (i.e., output of the searcher fed back to the threshold calculator), rather than on a noise/SNR estimate. That is a narrower, more defensible reading, but the specification itself (and the Chinese-language survey of "多径门限设定算法" surfaced in the art) treats this as a routine design alternative to the known "threshold = strongest path / α" formulation, which renders it an obvious design choice under KSR.

2. Legal Framework Applied

1. Scope and content of the prior art — all references below were filed or published before the 2002-05-23 priority date, so all qualify under § 102(b) or § 102(e) (with US 6,745,034 and US 6,801,511 qualifying as § 102(e) art by their 2000/1999 filings).
2. Differences between the prior art and the claims — analyzed element-by-element in § 4–5 below.
3. Level of ordinary skill — a POSITA here is a communications/DSP engineer with 2–3 years of spread-spectrum (DS-CDMA/W-CDMA) receiver design experience, familiar with Rake searchers, delay power profiles (DPP), multi-dwell acquisition, and soft-handoff state machines.
4. Motivation to combine — KSR controls: where design need or market pressure exists, and where a known technique has been used to improve one device in a foreseeable manner, combination is obvious. Here the motivation is express in the art itself (each reference states the same objective: reduce false alarms without increasing non-detection).

3. The Asserted Novelty Point (as the patent frames it)

The specification's own statement of the problem is important because it narrows, rather than broadens, the gap over the art:

"In the past, the path detection threshold was either a fixed threshold relative to the total input power received by the communicator, based upon a worst-case scenario, or it was calculated based on a base station report threshold specified by the corresponding network."

The patent then asserts that "a path detection threshold is adapted during operation of a communicator based on a measured energy level associated with a corresponding base station." That is the whole invention. There is no new hardware, no new signal-processing mathematics, and no new metric — only the selection of a feedback variable for a threshold that the art already taught to adapt.

4. Claim 1 — Element-by-Element Mapping

Conclusion on claim 1: Every element is disclosed, and the only genuine arguable distinction is that the adaptive-threshold input in the cited art is a noise/SNR quantity rather than the detected-path energy sum. That is a difference of degree, not of kind — and it is a design alternative the patent's own background section acknowledges as the known "report threshold" formulation.

5. Combinations That Render the Claims Obvious

Ground 1 — Ericsson '842 in view of Samsung '511 (and optionally Nokia '034) → claim 1

Primary reference: US 6,229,842 B1 (Schulist et al., Telefonaktiebolaget LM Ericsson), filed 1998-07-16, granted 2001-05-08 — "Adaptive path selection threshold setting for DS-CDMA receivers."

• Teaches a DS-CDMA searcher whose path selection threshold is adaptive, mapped from an estimated SNR, expressly to "optimize the probability of both non-detections and false alarms" — the identical stated objective as the '392 patent.
• Teaches the feedback/iterative structure: noise estimate → threshold factor → path selection threshold → peak removal → refined noise estimate. This is structurally the same closed loop the '392 claim 1 recites, with the loop variable swapped.
• Fig. 7 is literally captioned "illustrating setting a variable path selection threshold."

Secondary reference: US 6,801,511 B2 (Samsung) — supplies the two-receiver, pre-/post-handoff architecture and the "searcher reports energy → controller makes handoff decision" logic that claim 1 recites.

Motivation to combine: Both references are in the same field (DS-CDMA mobile stations), address the same problem (searching for neighbor base stations in a multipath channel), and are combinable with no change in principle of operation. Ericsson's adaptive path-selection unit is a direct, plug-in substitute for the fixed detector in Samsung's searcher. KSR — "a patent composed of several elements is not proved obvious merely by demonstrating that each of its elements was, independently, known in the prior art"; the inquiry is whether the combination is "beyond the skill of an ordinary artisan," which it plainly is here. Nokia '034 supplies an alternative or additional motivation by expressly teaching that a handoff threshold should be dynamically adjusted, and that a fixed threshold causes "selection of initiation of a handoff … when a handoff is not necessary."

Ground 2 — Ericsson '842 in view of Alcatel '143 → claim 1 (and claim 3)

US 6,510,143 B1 (Alcatel), filed 1999-07-01, granted 2003-01-21 — "Adaptive path searcher in a CDMA receiver."

Alcatel is arguably the stronger secondary reference because it teaches the profile-based (rather than noise-only) threshold derivation:

• "the threshold is dynamically determined as a function of the mean of noise and interference samples derived from the transmission channel profile" — i.e., from the same accumulation of correlated energy the searcher produces.
• "the path searcher performs a summation of the instantaneous profiles. Such a summation allows detecting the true paths of the channel by averaging the spurious noise."
• "This thresholding of the detected paths is obtained by comparison of their energy with the energy of the total CIR or by comparison with the most important path."

That last sentence maps directly onto '392 claims 3 and 4 ("total received energy estimate … determined by summing energies associated with the detected multipath components" and "energy level associated with a strongest multipath component … multiplied by a predetermined factor"). Alcatel teaches both alternatives expressly.

Motivation: Alcatel and Ericsson are both directed to the identical problem the '392 specification identifies, and Alcatel expressly criticizes the fixed threshold the '392 patent criticizes: "If the fixed threshold is too low, there is a non-negligible probability of erroneously selecting a large number of only noise samples … if the fixed threshold is too high, some of the channel taps with the lower power are not selected."

On claim 3's "all base stations" limitation: Alcatel computes a single noise/interference estimate from the overall profile, and Ericsson estimates SNR across the search; using pooled statistics from all currently searched bases to set each base's threshold is the natural extension and is expressly contemplated by the reference's "robust for all values of SIRs" teaching.

Ground 3 — Add Qualcomm '606 and the Fukumoto papers for claims 5 (multi-dwell)

Claim 5 requires the search receiver be a multi-dwell receiver that "[reduces] search time by refraining from searching using multipath components … that do not exceed the variable path detection threshold."

US 6,466,606 B1 (Qualcomm) describes exactly this: "This is the first dwell in a multiple-dwell process. If the accumulated energy is below T₁, the mobile station considers that there is no signal present on that time hypothesis and moves to the next hypothesis." https://patents.google.com/patent/US6466606

Fukumoto, S., et al., "Matched filter-based Rake combiner for wideband DS-CDMA mobile radio," IEICE Trans. on Communications, E81-B(7), July 1998, 1384–1390, and "Path search performance and its parameter optimization of pilot symbol-assisted coherent Rake receiver for W-CDMA mobile radio," IEICE Trans. Fundamentals, Vol. E83-A, No. 11, 2000, 2110–2119 (both of record). The second paper is, by its title alone, directed to optimizing the search parameters of a W-CDMA path searcher — precisely the threshold-selection problem.

Motivation: Combining an adaptive threshold with a multi-dwell searcher is not merely obvious, it is compelled: the entire point of a multi-dwell architecture (per Qualcomm) is to reduce mean acquisition time, and the '392 specification itself concedes the mechanism — "the initial dwell threshold usually allows a higher false alarm probability than the later dwell thresholds … In a system where the path threshold is adaptive, the initial dwell/dwells false alarm probability will typically be reduced."

Ground 4 — Claims 1–2 for the "no paths detected" and "direction of adaptation" limitations

Claim 2 requires that the threshold be "increased for greater measured energy and decreased for lower measured energy," and that the calculator either refrain from updating when no paths are detected or reset to a default.

• The direction-of-adaptation limitation is the definitional behavior of any threshold proportional to measured energy. Ericsson's Fig. 8 ("dependence of the variable threshold factor upon SNR") and its claim language ("selecting the threshold factor according to a first function if the estimated SNR is less than a first value and otherwise" per the DE 699 12 281 T2 counterpart) disclose a monotonic mapping.
• The "no paths → hold or reset" limitation is an obvious error-handling choice with only two predictable alternatives — the classic KSR "obvious to try" situation where the prior art gives a finite number of identified, predictable solutions.

Ground 5 — Claim 6 (concurrent pre-/post-handoff operation, common antenna)

Claim 6 requires the second receiver to communicate "just prior to handoff concurrently with the first communication receiver communicating with the first base station," with both "coupled to a common receive antenna."

• US 6,801,511 (Samsung) describes exactly this concurrent state: "the BS1 and the BS2 provide the mobile station with the service through their allocated channels."
• A single receive antenna feeding multiple Rake receivers is the standard CDMA handset architecture and is disclosed in the Alcatel/EP 1 069 697 family ("at least two path searchers (3a, 3b)"), and is inherent in Samsung's receiver block diagram.
• US 6,965,398 (InterDigital), also of record, is directed to base-station synchronization with multi-path reception — corroborating the common-antenna/multi-path-front-end context.

6. The Genuine Weaknesses in the Obviousness Case (candid assessment)

An examiner or a court could reasonably find a non-obvious distinction in the following, and the patentee would press these:

1. The feedback variable is the base station's own detected-path energy, not an SNR/noise estimate. Ericsson and Alcatel both adapt the threshold based on noise/interference estimates (SNR-mapped), whereas claim 1 requires variation "based on the measured energy levels stored in the power profile memory." Alcatel derives noise from the profile, but by suppressing the strongest paths before estimating noise — the opposite of '392's use of the retained paths. This is a real, if narrow, structural difference.

   • Rebuttal: the specification itself lists "X dB below the total energy of the base station" and "X dB below the strongest detected path" as the two embodiments — both are the well-known "threshold = α × (path energy)" rule that the art (Alcatel's "comparison with the most important path"; the Chinese-language survey's θ = S_max/α) already taught for fixed thresholds. Reducing a known fixed ratio to a running, per-base update is an obvious application of known adaptive techniques.

2. Claim 1 bundles the adaptive-threshold searcher with the dual-receiver handoff architecture. A patentee may argue no reference teaches the combination. But each reference independently states the same motivation (accurate, low-false-alarm detection to drive handoff correctly), so KSR's "predictable use of prior art elements according to their established functions" applies.

3. Nokia '034's dynamic handoff threshold is not a path detection threshold. This is correct and is a fair argument to raise against Ground 1's use of Nokia. Nokia should be used only as secondary evidence of the general motivation to adapt thresholds, not as an anticipation of the path-detection limitation. Ericsson and Alcatel must carry that element.

If I were defending the patent, I would argue claim 1 over the art on element (f), and I would expect a district court to find a genuine issue of material fact absent strong testimony mapping Ericsson's SNR-mapping unit onto "measured energy levels stored in the power profile memory."

7. Secondary Considerations

The record on the page reflects no evidence of objective indicia:

• No evidence of commercial success attributable to the claimed adaptive threshold.
• No unexpected results in the specification — the stated benefits (reduced false alarm rate, reduced search time) are precisely the benefits Ericsson and Alcatel already promise for their adaptive thresholds.
• No long-felt but unsolved need: the Ericsson reference shows the problem was recognized and solved four years before the '392 filing.
• No teaching away: nothing in the art criticizes varying a path threshold based on measured path energy.

Additionally, the failure to pay maintenance fees (lapsed 2016-09-16) means no commercial party has asserted the patent in the last decade — consistent with a weak or narrow claim set.

8. Practical Impact

• For a validity challenge: Ground 1 (Ericsson '842 + Samsung '511) and Ground 2 (Ericsson '842 + Alcatel '143) are the strongest, and both rely entirely on references the examiner already cited — i.e., they were before the examiner and were overcome only by the dependent-claim structure, not by newly discovered art.
• For a § 103 rejection: I would expect claim 1 to be rejectable over Ericsson in view of Samsung, with Alcatel relied on for the "total/strongest path energy" embodiments of claims 3, 4, and 7, Qualcomm '606 for claim 5, and Samsung for claim 6.
• Residual scope: Even if claims 1 and 3 survive, the patent's real teaching — adapt each base station's path-detection threshold using that base station's own retained-path energy, with a floor at a "worst case" minimum — is a narrow implementation detail rather than a platform-level contribution, and is unlikely to block a competent CDMA searcher design that derives thresholds from noise/SNR as Ericsson and Alcatel teach.

Caveat on confidence: I have verified the contents of Ericsson '842, Alcatel '143, Nokia '034, Samsung '511, and Qualcomm '606 against their published texts (URLs above). I was unable to retrieve the full texts of the two Fukumoto papers in this session, and the NEC reference US 6,580,749 was not retrieved — my treatment of those relies on their titles, abstracts, and the citations as they appear on the '392 record, and should be verified against the full documents before filing any paper. I have also not reviewed the file wrapper / prosecution history, which could contain applicant arguments or a certificate of correction (one is noted on the record, 2008-12-02) that materially affect claim construction.