US 6002675

I'll search for this specific patent number and any related litigation.

Let me check for any litigation involving this patent and verify the family/related applications.

Search summary

I ran targeted searches for exactly 6002675 / 6,002,675 across patent databases (Google Patents, FreePatentsOnline, Justia Patents) and web-indexed court/PTAB/CAFC material. I found no litigation, PTAB proceeding, or CAFC 2026 docket entry involving this patent. Details and caveats are at the end.

---

Bibliographic data — US 6,002,675

Source: https://patents.google.com/patent/US6002675/en ; https://www.freepatentsonline.com/[6002675](/patent/6002675).html ; https://patents.justia.com/patent/6002675

---

Abstract (verbatim from the patent)

"A repeater which introduces a minimum of delay in the transfer of a data packet from a source LAN to a destination LAN has an arbiter which grants access to a bus for a port to broadcast a received data packet onto the bus for loading the data packet into a transmit buffer for each outgoing port. The arbiter grants access only after all ports signal that their transmit buffers are ready to load the data packet from the bus. A transmit buffer signals that it is ready to be loaded at the passage of the collision window as it is transmitting a packet previously loaded into the transmit buffer. Accordingly, the transmit buffer is simultaneously transmitting the previously loaded packet and loading the new packet after the arbiter responds to the signal."

---

Plain-language overview of the independent claims

Claim 1 — A repeater (apparatus). A repeater has multiple ports, each with its own separate receive buffer (for packets coming in from the attached Ethernet collision domain) and its own separate transmit buffer (for packets going out). An internal bus lets ports talk to each other: a port's receive buffer broadcasts a packet onto the bus so that the other ports' transmit buffers can load it. An arbiter controls who gets the bus. A port that has a freshly received packet (the "requesting port") is granted the bus only when every other port asserts a "transmit buffer ready" signal. The key twist is when that ready signal is asserted: a port that is currently transmitting a stored packet asserts "transmit buffer ready" after the collision window has passed but before its transmit buffer has finished emptying. So the transmit buffer is allowed to be re-filled while it is still draining.

Claim 3 — The same thing as a method. The method counterpart of claim 1, expressed as steps: hold a received packet in a per-port receive buffer; hold a to-be-transmitted packet in a per-port transmit buffer; broadcast a received packet from a receive buffer onto the shared bus for other ports' transmit buffers; have the arbiter grant bus access to a requesting port only upon assertion of "transmit buffer ready" from every other port; and have a transmitting port assert that ready signal after passage of the collision window while it is emptying its transmit buffer and before the buffer is empty.

Claim 5 — An apparatus for communicating between different networks. Broadly drafted: multiple ports corresponding to multiple different networks; a receive buffer per port; a transmit buffer per port; a set of "fullness signals," one per transmit buffer, each indicating the relative fullness of its transmit buffer; a bus coupled to the transmit and receive buffers to move data between buffers; and an arbiter coupled to the bus that controls each port's access to the bus in response to the fullness signals. Note this claim is drafted more generically than claim 1 — it recites "fullness signals" rather than a collision-window-timed "transmit buffer ready" signal, and does not require the Ethernet-collision-domain framing.

Dependent claims observed

• Claim 2 (depends on 1): adds means for the transmitting port's transmit buffer to begin loading a new packet from the requesting port's receive buffer once the arbiter grants access, such that the new packet is loaded before the old packet finishes draining out.
• Claim 4 (depends on 3): method counterpart of claim 2.
• Claim 6 (depends on 5): each fullness signal is set to indicate that the port associated with the corresponding transmit buffer has completed a predetermined stage of transmission of data to its coupled network.

Uncertainty: the claim text I retrieved is complete through claim 6, but I could not confirm the total claim count for certain from the retrieved sources; I did not see claims numbered 7 or higher, but absence in an excerpt is not proof.

---

Context from the specification (useful for claim construction)

• The commercial embodiment named in the patent is the Digital Semiconductor 21340 10/100-Mb/s Ethernet Switched Repeater chip, and the disclosure is built around a "Buffered Repeater" architecture — the BREP chip (4 Ethernet ports each, 4 KB RX FIFO, 2 KB TX FIFO per MAC) — plus a switch engine bridging multiple segment buses.
• The collision window is given as 512 bit times by IEEE 802.3: ~51.2 µs at 10 Mb/s, ~5.12 µs at 100 Mb/s. The ready signal is asserted as soon as transmission time passes that window, because retransmission is no longer required after it.
• FIG. 20 is the timing diagram for the transmit-buffer-ready signal; TX-FIFO-RDY is described as asserted one clock cycle after the end of the collision window.
• The same specification also carries considerable material on address filtering (internal uni-address register, V-bit/F-bit, T1/T2 timers, FIG. 13; external CAM mode, FIG. 14) and credit-based flow control / SFZ control-packet signaling (FIGS. 5, 7B–7G, 15–19), which are the subject matter of several sibling applications.
• Related applications named in the patent, all filed on the same date: Ser. Nos. 08/779,897 ("Automatic Detection of a Similar Device..."); 08/779,880 ("Buffered Repeater with Independent Ethernet Collision Domains"); 08/779,883 ("Ethernet Network with Credit Based Flow Control"); 08/779,884 ("Management of a Computer Network Switched Repeater"); and 08/779,877 ("Adaptive Address Filtering" by Menachem and Perlman).

⚠️ Internal inconsistency I am flagging rather than "correcting": the FreePatentsOnline record lists the related application "Adaptive Address Filtering," Ser. No. 08/779,877, while both Google Patents and Justia list 08/779,877 as the application number of US 6,002,675 itself. These cannot both be right; one source contains an error. Per your rule, I am reporting both literally and not silently resolving it.

---

Litigation / CAFC docket search — no match found

I could not confirm any court or agency proceeding involving US 6,002,675 specifically. What the searches returned instead were different patents sharing the "'675" shorthand, which I am explicitly not treating as this patent:

• SoundClear Technologies v. Amazon.com (E.D. Va., No. 1:24-cv-01283) / SoundClear v. Google (E.D. Va., No. 3:24-cv-00540) and related IPRs — these concern U.S. 11,244,675, a voice-content-control patent (https://cases.justia.com/federal/district-courts/virginia/vaedce/3:2024cv00540/[557525](/patent/557525)/50/0.pdf).
• Aerin Medical v. Neurent Medical (D. Del.) — U.S. 10,610,675.
• Thermage v. Alma — U.S. 7,267,675 (later dropped from that case).
• Luxer v. Quadient (D. Del.) — U.S. 11,625,675.
• Extreme Networks v. Enterasys Networks, 395 F. App'x 709 (Fed. Cir. 2010) — the appeal involved the '727 and '181 patents, not the '675 patent (https://www.courtlistener.com/opinion/[206816](/patent/206816)/extreme-networks-inc-v-enterasys-networks-inc/).
• The 21 June 2005 Enterasys v. Extreme Networks / Foundry Networks suit asserted six patents (reported as covering, among other things, VLANs and multi-protocol routing), but I found no source naming US 6,002,675 among them, so I cannot say whether it was asserted there.
• A hit for the literal string "6002675" appeared in a Brazilian state-court appeal (Apelação Cível nº 6002675-24.2025.4.06.9999/MG, TRF-6). That is a case-number coincidence with a Brazilian procedural numbering scheme and has no relationship to the U.S. patent.

Caveats on the litigation search:

• Your available tooling here was general web search, not direct queries of USPTO Patent Public Search, USPTO Patent Trial and Appeal Board (PTAB) APIs, PACER, or the CAFC docket. I therefore cannot state with authoritative certainty that no proceeding exists — only that I found no evidence of one.
• The patent expired 2017-01-06, which makes a 2026 CAFC appeal on this patent inherently unlikely; the 2026 litigation material surfaced in these searches all relates to other, younger "'675" patents.
• Note a date discrepancy in the request itself: the prompt states the current date as April 26, 2026, while the patent record supplied to me was fetched 2026-09-30. I have not adjusted any date information in reliance on either.

If you need a definitive litigation/PTAB negative, the authoritative next steps would be a PACER docket search on "6,002,675" and a PTAB proceeding search in USPTO Patent Public Search / the PTAB Endpoint API, both of which are outside the tools available to me in this session.

I'll research litigation involving US patent 6002675. Let me search several sources.

Let me search more targeted sources.

Let me search Unified Patents and CourtListener directly.

Summary of Findings

I could not confirm any litigation specifically asserting U.S. Patent No. 6,002,675. My searches of general web and litigation-oriented sources did not return any case record, docket, complaint, or judgment in which US 6,002,675 was identified as an asserted patent. I want to be explicit about this rather than imply a match that I did not actually verify.

What I searched

• Targeted queries for the patent number and its formatted variants ("6002675", "6,002,675") combined with litigation/defendant terms.
• Party-based queries (assignees/owners of the patent).
• General results surfaced from patent-litigation aggregators and CourtListener opinion pages.

Important caveat on my search

I did not get a clean, direct hit from the specific litigation databases you named (Unified Patents litigation caselist, PACER, CAFC docket search). The searches returned either (a) unrelated cases, (b) the patent's prosecution/assignment record, or (c) party-based litigation involving the patent's corporate family but not naming the '675 patent. Any one of those hits could be relevant, but none actually names US 6,002,675, so I will not present them as confirmed litigation for this patent.

Relevant context (not a litigation finding)

The assignment/prosecution record on the patent's front page shows a chain of ownership that is relevant to who could have asserted it:

The patent also reached its anticipated expiration on 2017-01-06 (per the record's legal-status entry, "Expired – Lifetime"), and its current listed assignee is Extreme Networks Inc.

I did observe that these same parties were involved in separate patent litigation in the general time frame — for example, Enterasys Networks, Inc. v. Foundry Networks, LLC et al, No. 1:05-cv-11298 (D. Mass., filed June 21, 2005), and the related appeal Extreme Networks, Inc. v. Enterasys Networks, Inc., 395 F. App'x 709 (Fed. Cir. 2010). However, the material I retrieved ties those to other patents (e.g., a '727 patent discussed in the Federal Circuit opinion), not to US 6,002,675. I therefore cannot state that the '675 patent was at issue in them, and I will not do so.

Bottom line

• No litigation involving US 6,002,675 is known to me on the basis of the searches performed.
• I cannot assert with high confidence that none exists — only that my searches did not surface any case naming this patent, and I was unable to complete verification against PACER/CAFC/Unified Patents directly.

Recommended verification steps (to be done with direct docket access)

1. Unified Patents litigation caselist — search the exact patent number "6002675."
2. PACER / CourtListener docket search — search the patent number as a text string across district-court complaints involving the likely plaintiffs (Enterasys Networks, Cabletron Systems, Extreme Networks).
3. USPTO Patent Center "Legal" / assignment records — confirm any recorded licenses or litigation-related assignments.
4. The specific 2005–2010 Extreme/Enterasys/Foundry cases above — pull the complaints to check the asserted-patent lists directly.

If you can share any of those docket documents, or confirm which case you believe this patent appeared in, I can review and report the plaintiff(s), defendant(s), jurisdiction, case number, filing date, and outcome with confidence. If you would like, I can also run additional targeted searches against specific case captions.

Proceedings overview

Total AIA trial proceedings on file for US 6,002,675: 0. The USPTO Open Data Portal structured block returns no IPR, PGR, or CBM for this patent, and no proceeding surfaced in web search that I could tie to this patent number — so the breakdown is active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0. The bottom line for a defendant is unusual and worth stating precisely: there is no PTAB-based invalidity record to borrow, but there is also no live patent to defend against — the patent's term expired on 2017-01-06 and its legal status is "Expired - Lifetime," so the operative defensive posture is not "hardened patent" or "dead claims," but "expired patent with an untested PTAB history."

A critical disambiguation before anything else: search results for "the '675 patent" and IPR2025-00673 are a different patent — US 11,244,675 (SoundClear Technologies), "Word Replacement in Output Generation for Detected Intent by Voice Classification," asserted against Amazon and Google. That is not this patent. Similarly, the ex parte reexamination of the Netlist '608 patent (Application No. 90/015,449) and the Wilus/HP '035 and '879 patent litigation are unrelated. None of those proceedings involve US 6,002,675.

No proceedings to enumerate

There is no proceeding number to report, and I will not invent one. For completeness, here is what would normally populate this section and what I found instead:

• Type: N/A — no Inter Partes Review, Post-Grant Review, or Covered Business Method review on record.
• Filed: N/A.
• Status: N/A (USPTO ODP: no AIA trial proceedings; Google Patents legal-events timeline likewise shows only prosecution, assignment, and expiration events).
• Judge panel: N/A.
• Petition grounds: N/A.
• Institution decision: N/A.
• Final Written Decision: N/A — no FWD exists, so no claim was canceled or sustained at the PTAB.
• Settlement / termination: N/A.
• Appeal: No PTAB-originating Federal Circuit appeal identified. (The patent-owner-side assignment history — Cabletron → Digital Equipment → Cabletron → Enterasys → Extreme Networks, with the 2015-08-26/2015-09-03 Extreme Networks assignments — reflects corporate transactions, not litigation or appeal activity.)
• Defensive value: A defendant cannot cite an FWD to knock out any claim, because none exists. Any invalidity case must be built from scratch in district court or at the ITC, subject to the patent's expired status (see below).

Confidence note: the ODP structured block is the canonical source and is explicit. My independent web searches (Google Patents, FreePatentsOnline, PTAB-related queries) likewise produced no proceeding for this patent number. I could not run a direct free-text query against PTAB E2E within this session, so I cannot rule out a very old, never-indexed petition with absolute certainty — but there is no affirmative evidence of one, and the default per the task instructions ("no PTAB activity on file") applies.

Strategic summary

Claim status: every claim is UNTESTED at the PTAB, and the entire patent is expired. No claim of US 6,002,675 has been canceled, confirmed, or even challenged in an AIA trial. The independent claims are claim 1 (a repeater apparatus) and claim 3 (a method of operating a repeater), with claim 2 depending from claim 1 and directed to the "means for said transmitting port transmit buffer to begin loading a new packet… before said transmitting port transmit buffer is emptied." Because no FWD exists, there is no "surviving claims" list to report — the claims survive only in the sense that nothing at the PTAB killed them, and they are no longer enforceable because the patent expired on 2017-01-06 (legal status: Expired - Lifetime). Note that I am reporting the claim set only to the extent the public record shows it; I did not independently verify the total claim count, and none of it has been adjudicated.

Estoppel landscape: §§ 315(e)(2) and 325(e)(2) are irrelevant here. Estoppel attaches only after a petitioner obtains an FWD, and no IPR/PGR ever reached that stage for this patent. There is therefore no petitioner estopped, no privity chain to trace, and no prior-art ground that a hypothetical petitioner "reasonably could have raised" and is now barred from raising. Practically speaking, that means a defendant preparing an invalidity case has the entire universe of prior art available — but that freedom is mostly academic given expiration. The more relevant constraint is temporal: IPRs became available on 2012-09-16, creating a ~4.3-year window (2012-09-16 to 2017-01-06) during which this patent could have been challenged; no one filed, which strongly suggests the patent was not being commercially asserted during that window.

Pattern signals: none of the usual markers. There is no repeat petitioner, no IPR-vs.-ITC or IPR-vs.-district-court race, and no defensive aggregator (Unified Patents or similar) in the record — a notable absence, since Unified frequently targets networking patents. The ownership trail (Cabletron Systems → Digital Equipment → Cabletron → Enterasys → Extreme Networks, with security interests from Wells Fargo, Obsidian, Silicon Valley Bank, and Bank of Montreal) is a chain of corporate acquisitions and financings, not litigation behavior. The absence of any IPR, combined with expiration and the change of assignee to a large operating company (Extreme Networks), points away from the patent having been a monetization vehicle.

Recommended next steps

1. Do not assume PTAB cover exists. The value of an IPR-based defense is that you can borrow a prior FWD to estop or preempt the patent owner. Here there is no FWD to borrow, so any challenge must be litigated on its own record. If you need a prior-art read, it must be commissioned from scratch.

2. Lead with the expiration defense. Confirm the term via USPTO PatentCenter / PTAB E2E for US 6,002,675 and the Patent Term Adjustment record; the Google Patents record shows anticipated expiration 2017-01-06 with status "Expired - Lifetime." An expired patent cannot support ongoing infringement relief, which generally ends the dispute if the demand letter is premised on current sales.

3. If a demand letter cites this patent, pressure-test the patent number first. Demand letters in this space have been known to conflate patents with similar last-three digits. Verify against the claims actually quoted in the letter: claim 1 of this patent recites a repeater with "a separate receive buffer for each port," "a separate transmit buffer for each port," a bus, an arbiter granting access "upon assertion of a 'transmit buffer ready' signal from each port… other than said requesting port," and means for asserting that signal "after passage of a collision window." If the letter cites different language, you are dealing with a different patent — most plausibly a same-numbered '675 in a different family (e.g., the voice-classification US 11,244,675 involved in IPR2025-00673, which is not this patent and is an active proceeding worth checking if your product is a voice assistant).

4. Check the foreign counterparts, which are the only proceedings-adjacent activity on record. The family includes WO 98/33283 A2, EP 1 012 988 A2, AU 714667 C, and CA 2 277 097 C. None of these are PTAB matters, but if you face ex-US assertions, the EPO and CA oppositions/reexaminations are the places to look, and the Google Patents "Also published as" links give the entry points.

5. If you genuinely need a "no PTAB activity" certification for a litigation position, cite the ODP record (no AIA trial proceedings) and pull the PTAB E2E case-search result for the patent number as of today's date, since the ODP ingest is not real-time.

I'll research the assignment record for US 6,002,675, focusing on USPTO Assignment Center records, reel/frame details, and correspondents.

Let me try more targeted searches for the specific assignment records.

Let me dig for reel/frame specifics and the assignment correspondents.

Ownership Reconstruction — US 6,002,675

Data-provenance note up front: I was able to reconstruct the chain from the patent's own legal-events record (mirrored on the Google Patents page for US6002675A, which reflects USPTO assignment data). I was not able to pull the individual Assignment Center entries for this patent during this session — my searches against assignment.uspto.gov / assignmentcenter.uspto.gov returned only general documentation, not the per-record reel/frame table. Two reel/frame references are recoverable from the legal-events text itself (see below), but I could not independently confirm the reel/frame, execution dates, or correspondents for the remainder. Anywhere below where I lack a reel/frame, I say so rather than guessing.

---

Inventors

Employer basis (medium-high confidence): The inventors are the assignors on the very first recorded transfer, dated on the filing date (1997-01-06), to DIGITAL EQUIPMENT CORPORATION — the classic inventor-to-employer assignment executed at filing. The specification independently confirms the DEC origin: it names "the Digital Equipment Corporation product, Digital Semiconductor 21340 10/100‑Mb/s Ethernet Switched Repeater chip" as an example chip incorporating the invention, and the related-application docket prefixes ("PD96‑0183," "PD25618," "PD25619," "PD25620," "PD25621") are Digital's internal docket format. The inventor names are consistent with DEC's Israel-based semiconductor/networking design organization, though I could not verify a specific DEC site from the sources retrieved (low confidence on that detail).

Unusual-pattern check: No evidence of inventors departing the original assignee within 12 months of filing. The opposite is documented: the 1998-03-25 DEC→Cabletron transfer is an entity-level portfolio transfer, not an inventor departure. Mark: not present.

Data anomaly worth flagging (not an ownership issue): The specification's "related applications" section lists "PD25621, 'Adaptive Address Filtering,' by Avraham Menachem and Shuki Perlman, Ser. No. 08/779,877" — i.e., it recites this application's own serial number under a different title and inventor set than the issued patent ("Method and apparatus for controlling transmission of data over a network," Ben-Michael, Perlman, Menachem). This is either a drafting/OCR artifact or a title change during prosecution. It does not change the assignment chain but is worth noting if you are matching adjacent-family patents by serial.

---

Original assignee

Two different notions of "original" matter here, and they diverge:

• Assignee on the issued patent (front page): Cabletron Systems, Inc. This is why Google Patents lists "Original Assignee: Cabletron Systems Inc."
• First recorded assignee (on the filing date): Digital Equipment Corporation — the inventors assigned to DEC at filing, and DEC then conveyed to Cabletron.

Cabletron Systems, Inc. — Rochester/Portsmouth, New Hampshire networking-equipment maker, founded 1983 by Craig Benson and Robert Levine; at peak ~6,600 employees; shipped routers, switches (SmartSwitch, SmartSTACK, Spectrum management software). It did ship products in the relevant technology (buffered-repeater / switched-repeater hardware — the BREP chip this patent describes was a real Digital Semiconductor part marketed by Cabletron after it bought DEC's networking business). In 2000 Cabletron reorganized under an Amended and Restated Transformation Agreement (June 3, 2000) into four businesses — Enterasys Networks, Riverstone Networks, Aprisma Management Technologies, and Global Network Technology Services — and subsequently renamed itself Enterasys Networks (Cabletron is referred to as "formerly known as Cabletron" in later Enterasys filings). So Cabletron did not "dissolve" so much as become Enterasys.

Successor chain status: Enterasys went through an SEC accounting investigation and restatement (settled 2003), was taken private by The Gores Group / Tennenbaum Capital (Nov 2005, ~$386M), and was acquired by Extreme Networks (announced 2013-09-12, closed late 2013). Extreme Networks remains an operating networking-equipment vendor (switches, wireless, network security).

---

Assignment timeline

Records exist for this patent. Chronological list of recorded conveyances (dates as listed in the patent's legal-events record; reel/frame shown only where recoverable):

• 1997-01-06 (executed/recorded on filing date) — Reel/frame not retrieved

  • Conveyance: Assignment of Assignors Interest
  • Assignor: Simoni Ben-Michael; Avraham Menachem; Shuki Perlman
  • Assignee: Digital Equipment Corporation
  • Correspondent: not retrieved
  • Context: inventor-to-employer assignment at filing (DEC-origin patent).

• 1998-03-25 — Reel/frame not retrieved

  • Conveyance: Assignment of Assignors Interest
  • Assignor: Digital Equipment Corporation
  • Assignee: Cabletron Systems, Inc.
  • Correspondent: not retrieved
  • Context: acquisition — Cabletron's purchase of DEC's networking products business; entity-level portfolio transfer.

• 2000-10-19 — Reel/frame not retrieved

  • Conveyance: Assignment of Assignors Interest
  • Assignor: Cabletron Systems, Inc.
  • Assignee: Enterasys Networks, Inc.
  • Correspondent: not retrieved
  • Context: internal reorg — Cabletron's 2000 transformation/asset-contribution into the Enterasys subsidiary.

• 2006-05-23 — Reel 017656 / Frame 0552 (reel/frame recovered from the later release document's own citation)

  • Conveyance: Security Agreement
  • Assignor: Enterasys Networks, Inc.
  • Assignee: Wells Fargo Foothill, Inc.; Obsidian, LLC
  • Correspondent: not retrieved
  • Context: securitization — collateral grant tied to the Gores/Tennenbaum going-private LBO financing, not a transfer of title to the patent.

• 2010-11-10 — Reel 025339 / Frame 0875 (reel/frame recovered from the later termination document's own citation)

  • Conveyance: Grant of Security Interest in U.S. Patents
  • Assignor: Enterasys Networks Inc.
  • Assignee: Wells Fargo Trust Corporation Limited, as Security Agent
  • Correspondent: not retrieved
  • Context: securitization — replacement collateral-agent security interest.

• 2010-11-30 — Reel/frame not retrieved

  • Conveyance: Release and Reassignment of Patents and Patent Applications at Reel/Frame No. 17656/0552
  • Assignor: Enterprise Communications Funding GmbH (as successor in interest to Obsidian, LLC); Wells Fargo Capital Finance, Inc. (formerly Wells Fargo Foothill, Inc.)
  • Assignee: Enterasys Networks, Inc.
  • Correspondent: not retrieved
  • Context: securitization release — clearing the 2006 collateral lien; title stays with Enterasys.

• 2013-11-01 — Reel/frame not retrieved

  • Conveyance: Termination and Release of Security Interest in Patents at Reel/Frame No. 25339/0875
  • Assignor: Wells Fargo Trust Corporation Limited
  • Assignee: Enterasys Networks Inc.
  • Correspondent: not retrieved
  • Context: securitization release — clearing the 2010 collateral lien; title stays with Enterasys.

• 2015-08-26 — Reel/frame not retrieved

  • Conveyance: Assignment of Assignors Interest
  • Assignor: Enterasys Networks, Inc.
  • Assignee: Extreme Networks, Inc.
  • Correspondent: not retrieved
  • Context: acquisition — the Extreme/Enterasys merger patent transfer.

• 2015-09-03 — Reel/frame not retrieved

  • Conveyance: Assignment of Assignors Interest
  • Assignor: Enterasys Networks, Inc.
  • Assignee: Extreme Networks, Inc.
  • Correspondent: not retrieved
  • Context: acquisition — second/confirmatory recording of the same Enterasys→Extreme transfer (the duplicate pair is common for merger omnibus + confirmatory filings).

• 2017-07-14 — Reel/frame not retrieved

  • Conveyance: Second Amended and Restated Patent and Trademark Security Agreement
  • Assignor: Extreme Networks, Inc.
  • Assignee: Silicon Valley Bank
  • Correspondent: not retrieved
  • Context: securitization — operating-company credit facility collateral.

• 2018-05-01 — Reel/frame not retrieved

  • Conveyance: Security Interest
  • Assignor: Extreme Networks, Inc.
  • Assignee: Bank of Montreal
  • Correspondent: not retrieved
  • Context: securitization — replacement credit-facility collateral.

• 2018-05-01 — Reel/frame not retrieved

  • Conveyance: Release by Secured Party
  • Assignor: Silicon Valley Bank
  • Assignee: Extreme Networks, Inc.
  • Correspondent: not retrieved
  • Context: securitization release — SVB lien terminated on the BMO refinancing.

On correspondents: none of the recorded conveyances exposed a correspondent in the material I could retrieve. The only attorney-of-record data point I surfaced was on FreePatentsOnline, which lists "THLANG, ERIC S" with firm "Haley Guiliano LLP (New York, NY, US)" in the attorney/agent field. I flag this cautiously: FPO's attorney field typically reflects the current correspondent of record rather than the 1997–2000 prosecuting firm, and I could not confirm which (if any) of the recorded assignments this attorney filed. Treat as unverified, and do not read it as a recurring-NPE-correspondent hit on the strength of one data point.

---

Timeline diagram

timeline
    title Ownership of US 6002675
    1997 : Filed 1997-01-06
         : Inventors assign to Digital Equipment Corp
    1998 : Digital assigns to Cabletron Systems
    1999 : Patent issued 1999-12-14
    2000 : Cabletron transfers to Enterasys Networks
    2006 : Security agreement to Wells Fargo Foothill
    2010 : Wells Fargo Trust security interest recorded
         : Release and reassignment to Enterasys
    2013 : Security interest terminated and released
    2015 : Enterasys assigns to Extreme Networks
    2017 : Patent expires 2017-01-06
         : Security agreement with Silicon Valley Bank
    2018 : Bank of Montreal security interest
         : Release by Silicon Valley Bank

---

NPE / troll-pattern signals

1. Shell-entity transfer — not present. Every title-holding assignee is an operating company: Digital Equipment Corp → Cabletron Systems → Enterasys Networks → Extreme Networks. There is no "IP / Holdings / Ventures / Licensing" entity anywhere in the chain, and no single-purpose LLC took title. The non-operating names that appear — Wells Fargo Foothill (reel 017656/0552), Obsidian, LLC, Wells Fargo Trust (reel 025339/0875), Enterprise Communications Funding GmbH, Silicon Valley Bank, Bank of Montreal — are lenders/collateral agents on security agreements, not title holders.

2. Known asserter in the chain — not present. None of Digital, Cabletron, Enterasys, or Extreme appears on the Acacia / Marathon / IV / IPNav / Wi-LAN / Conversant / Vringo / Pendrell / Innovatio / MPHJ / Lumen View / Round Rock / Spangenberg or Unified Patents / RPX high-frequency-plaintiff lists as a patent-assertion entity. Each is a network-equipment manufacturer.

3. Repeat correspondent across the chain — unclear. I could not retrieve correspondents for any of the recorded conveyances, so recurrence cannot be tested. The single FPO attorney data point (Eric S. Thlang / Haley Guiliano LLP) is not enough to support the signal, and Haley Guiliano is a general-practice IP firm — one appearance is expressly not a finding under your criteria.

4. Cascading transfers — not present. The title transfers are spaced years apart (1998, 2000, 2015) and are all operating-company events; there are no chained LLCs and no <24-month cascade. The only closely-spaced recordings are the 2015-08-26 and 2015-09-03 Enterasys→Extreme pair, which is a duplicate/confirmatory recording of a single merger transfer — not a cascade.

5. Pre-litigation transfer — not present (and unverifiable anyway). No litigation asserting US 6,002,675 was confirmed in the earlier litigation research, so there is no "6 months before first suit" transfer to identify. The last title transfer (2015) post-dates the patent's commercial relevance and precedes expiration by ~18 months.

6. Bankruptcy fire-sale — not present. Enterasys experienced severe financial distress (SEC investigation, 2002–2003 restatement, criminal convictions of executives 2004–2006) and was taken private by Gores/Tennenbaum in 2005, but this was a leveraged buyout, not a Chapter 7/11 proceeding, and the patents were not sold out of an estate. The 2006 security agreement (reel 017656/0552) is LBO collateral, not a fire-sale conveyance.

7. Privateering — not present. No operating company transferred title to an NPE to assert against competitors. Extreme Networks retained direct ownership through the patent's expiration.

8. Defensive aggregator — not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at Extreme Networks, an operating vendor.

---

Verdict

Operating-company assertion.

Justification: Every title-holding link in the chain is an operating network-equipment company — inventors → Digital Equipment Corp (1997-01-06), Digital → Cabletron Systems (1998-03-25), Cabletron → Enterasys Networks (2000-10-19), Enterasys → Extreme Networks (2015-08-26 and 2015-09-03) — and the only non-operating names in the record are secured lenders on collateral filings (reel 017656/0552 and reel 025339/0875), which transfer no ownership. There are zero NPE signals present: no shell LLC, no known asserter, no cascading LLC chain, no bankruptcy sale, no privateering, and no defensive-aggregator terminus. I caveat the "assertion" prong: the previously generated litigation research found no confirmed infringement suit naming US 6,002,675, and the patent reached its anticipated expiration on 2017-01-06 ("Expired – Lifetime"), so "operating-company assertion" describes the ownership character of the chain, not proven litigation activity — on the evidence available this is a clean operating-company chain, not an NPE vehicle.

Verification links:

• USPTO Assignment Center: https://assignmentcenter.uspto.gov/ (search patent number 6002675)
• Legacy Patent Assignment Search: https://assignment.uspto.gov/patent/index.html (search 6002675)
• Patent legal events (source used here): https://patents.google.com/patent/US6002675/en

Open items I could not close: the reel/frame and correspondent for the 1997, 1998, 2000, 2010 (release), 2013, 2015, 2017, and 2018 recordings; and independent confirmation of the current correspondent of record (FPO shows Haley Guiliano LLP, unverified). If you can supply the Assignment Center result table for 6002675 (or its PDFs), I can fill those fields and definitively test signal #3 (repeat correspondent).

I'll research the prior art cited in US 6,002,675. Let me run targeted searches.

Let me try to retrieve the actual "References Cited" list from patent databases.

Let me search for the patent's own citation list via Justia and other sources.

Prior-Art Analysis — US 6,002,675

1. Identification of the patent number (as required, literally)

I searched the literal string 6002675 without normalizing or auto-correcting it. The only U.S. patent matching that exact number is:

Source: https://patents.google.com/patent/US6002675/en

Similar-number results I am deliberately excluding (per your "do not return results for similar numbers" instruction — but you should know they exist so you don't mistake them for hits):

• JP 6002675 A (= JPS51136628A), "Process for preparing benzylalchol derivatives," JP application 6002675, filed 1975-05-19 — a Japanese chemical patent, unrelated. (https://patents.google.com/patent/JPS51136628A/en)
• DE 6002675 D1 / T2 — a German utility/patent publication number appearing in an unrelated membrane-filtration family.
• The "'675" patents flagged in the prior litigation section (US 11,244,675; 10,610,675; 7,267,675; 11,625,675) — all different numbers.

---

2. ⚠️ Critical limitation on this task: I could not retrieve the (56) "References Cited" list

This is the honest answer to the core of your request, and I want to state it plainly rather than pad the response with plausible-looking citations.

Your request was to "look at each patent citation for 6002675" and map each to § 102. To do that I need the patent's own front-page (56) References Cited list. I could not obtain it. Specifically:

• The authoritative full text supplied to me for this analysis contains no (56) section. It runs from the bibliographic block/classifications straight into "Definitions," the Abstract, and the Description. Google Patents renders the citation lists in a tab/sidebar that was not captured in the fetched text.
• My searches were constrained to general web search — I have no tool access to USPTO Patent Public Search, USPTO Patent Center, the PTAB API, or Espacenet's citation tables, which are the authoritative sources for a (56) list.
• The searches I ran returned the inverse relationship repeatedly: pages for other patents that cite US 6,002,675 in their "Cited By" (forward-citation) tables — e.g. US 6,335,939 ("Apparatus and method for selectively supplying data packets between media domains in a network repeater"), US 7,143,221 B2, US 7,305,008 ("Parallel bus LAN," 18 patent citations), and US 7,376,127 (Broadlight, high-speed internal bus architecture). Forward citations tell me who built on this patent; they say nothing about what this patent's examiner cited against it.

I therefore cannot give you a per-reference § 102 anticipation mapping keyed to actual citations, because I do not have the actual citation list. Fabricating a table of patent numbers, dates, and "potentially anticipates claim X" entries would be the single most damaging thing I could do here, so I am not doing it.

---

3. What IS grounded and citable as prior art / admitted background

The specification itself identifies prior art and incorporates standards by reference. These are verifiable from the patent text you supplied.

Cross-reference / contradiction carried forward: as flagged in the previously generated sections, FreePatentsOnline lists "Adaptive Address Filtering," Ser. No. 08/779,877, while Google Patents and Justia list 08/779,877 as the application number of US 6,002,675 itself. The full text you supplied here lists 08/779,884 as the Management application and enumerates 08/779,897; 08/779,880; 08/779,883; 08/779,884; and 08/779,877. That raises, rather than resolves, the discrepancy — one of the secondary sources is wrong. I am reporting the identifiers literally and not resolving it.

---

4. Framework for the § 102 analysis once the (56) list is obtained

Because the application was filed 1997-01-06, pre-AIA 35 U.S.C. § 102(a)–(g) governs (AIA §§ 102/103 first-inventor-to-file provisions do not apply). Concretely, for the independent claims:

• Claim 1 (repeater with per-port RX/TX buffers, internal broadcast bus, arbiter gated on "transmit buffer ready" from every other port, ready asserted after the collision window and before the TX buffer empties) — the anticipatory reference must disclose all of: (i) per-port separate receive and transmit buffers; (ii) a shared internal bus with receive-buffer broadcast to other ports' transmit buffers; (iii) an arbiter gating on per-port readiness; and (iv) the collision-window-timed assertion of readiness enabling simultaneous drain-and-fill. Elements (iii)+(iv) are the narrow point. Expect most cited art to fail on (iv).
• Claim 3 — method counterpart of claim 1; same element set, same analysis.
• Claim 5 — materially broader. It recites only "fullness signals … indicating the relative fullness" of each transmit buffer, and an arbiter controlling bus access "in response to the fullness signals." It omits the collision-window timing and the Ethernet-collision-domain framing entirely. Claim 5 is by far the claim most exposed to § 102, because generic per-port TX-FIFO-fullness arbitration in buffered repeaters/switches was known art. If any cited reference discloses buffer-fullness-gated bus arbitration on a shared broadcast bus without the collision-window timing, claim 5 is the one it would anticipate — and claims 1/3 would survive.
• Claim 2 / Claim 4 — add loading a new packet into a still-draining TX buffer before the old packet finishes. Anticipation requires the reference to disclose the overlap explicitly (cut-through re-fill), not merely double-buffering.
• Claim 6 — recites the fullness signal indicating completion of "a predetermined stage of transmission." This re-imports a timing milestone into claim 5's branch, which is where the collision-window rationale would come back in as the "predetermined stage."

Claim-count caveat (carried forward): the claim text available to me runs through claim 6; I could not confirm the total claim count. A complete § 102 mapping needs the full claim set from the USPTO record.

---

5. What I recommend, and what would settle this definitively

To produce the per-reference citation table you asked for, retrieve the (56) References Cited block from one of these (none available to me in this session):

1. USPTO Patent Public Search (https://ppubs.uspto.gov) — full-text record for 6,002,675, "References Cited" (U.S. Patent Documents / Foreign Patent Documents / Other Publications) and Examiner Cited References.
2. USPTO Patent Center — the issued patent PDF itself; the (56) block is on the face of the printed patent (page 1, left column). https://patentcenter.uspto.gov
3. Espacenet "Also published as / Cited documents" and the E PO Register / A1 search report for EP 1 012 988 A2 (EP application 98905942), the European family member — its Article 54(3)/search report will independently identify the art the EPO considered.
4. WO 98/33283 A2 — the PCT publication's International Search Report will list the art the ISA considered, which for this family is a strong proxy for the examiner's (56) list.
5. Google Patents "Patent Citations" panel — https://patents.google.com/patent/[US6002675A](/patent/US6002675A)/en (the "Patent Citations" table, distinct from the "Cited By" table my searches kept returning).

Once that list is in hand, the analysis is mechanical: for each cited reference check (a) publication date vs. 1997-01-06 for § 102(a)/(b), or (b) earlier filing + later publication/issue for § 102(e); then compare disclosure element-by-element against claims 1–6 as set out in § 4 above, with claim 5 as the primary anticipation candidate.

Bottom line: US 6,002,675 is confirmed and correctly identified. Its own cited-reference list could not be retrieved with the tools available to me, so I will not present fabricated citations. The grounded prior art I can cite is the IEEE 802.3 (1996) and IEEE 802.3u (1995) standards incorporated into and admitted by the specification — background art sufficient for § 103 and claim construction, but on my reading not anticipatory of the independent claims, whose novelty lies in the collision-window-timed "transmit buffer ready" arbitration.

I'll first attempt to recover the actual cited-reference record that the Prior Art section flagged as missing, then build the § 103 analysis on whatever is grounded.

Let me try harder to recover the (56) cited-reference list itself, and pin down dates for the candidate references that surfaced.

Obviousness Analysis — US 6,002,675 under pre-AIA 35 U.S.C. § 103

Current date: April 26, 2026. Consistent with the rules, I interpret all identifiers literally and do not auto-correct. Where this section contradicts or extends the previously generated Prior Art section, I flag it explicitly.

---

0. Structural limitation carried forward (must be stated first)

The Prior Art section concluded — correctly, and after a further attempt in this session — that the (56) "References Cited" list for US 6,002,675 could not be retrieved. I re-ran targeted searches (Justia "References Cited," Google Patents "Patent Citations" panel, FPO) and again obtained only forward-citation tables (who cites '675), never the examiner's cited art. The authoritative full text supplied to me contains no (56) block.

Consequence for this § 103 section: I cannot build combinations keyed to the actual cited references, because I still do not have them. What follows is therefore (a) a fully grounded § 103 analysis built only on art that is verifiable on the record before me, plus (b) clearly-labelled candidate references with their art status resolved as far as the sources permit, and (c) a completed framework for the moment the (56) list / PCT ISR / EPO search report is obtained. I will not fabricate a prior-art combination table.

Governing law: Application filed 1997-01-06; the AIA first-inventor-to-file provisions do not apply. Pre-AIA § 103(a) governs, with § 103(c) operating as a carve-out for commonly owned art.

---

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

The specification itself supplies most of the definition. A POSITA here would be a network-ASIC designer with, typically, a B.S. in EE/CS plus ~3–5 years in Ethernet LAN repeater/switch silicon, and would be familiar with:

• the IEEE 802.3 / ISO/IEC 8802-3 CSMA/CD MAC and repeater architectures, including the collision window = 512 bit times (~51.2 µs @ 10 Mb/s; ~5.12 µs @ 100 Mb/s) — expressly incorporated by reference in the '675 specification;
• IEEE 802.3u (100BASE-TX/T4 repeater behaviour; NWay auto-negotiation, Ch. 27–28);
• FIFO buffering in MAC datapaths, and the store-and-forward vs. cut-through tradeoff;
• the then-current industry push toward buffered repeaters with independent collision domains (the Gigabit Ethernet Alliance's "buffered repeater" debate material surfaced in this session's searches is illustrative of that discourse, albeit undated as retrieved).

---

2. Reference ledger — what is actually available, and its § 102/§ 103 status

Observation from the searches worth recording: the PCT/EP family member of this application is titled "BUFFERED REPEATER WITH EARLY FILLING OF TRANSMIT BUFFER" (EP 1 012 988 A2, orig. WO9833283A2 — http://data.epo.org/gpi/EP1012988A2). That title is a better description of the actual point of novelty than the US title, and it matters for claim construction: the inventive core is early filling, i.e. the timing of when the TX buffer accepts the next packet relative to its own drain.

---

3. Claim-by-claim § 103 analysis

Claim text as available runs through claim 6; the total claim count remains unverified (carried forward). Claim 1 (apparatus) and claim 3 (method) are the independent claims described in the earlier sections; claim 5 is the broader apparatus claim.

3.1 Claim 5 — the strongest § 103 target (highest invalidity risk)

Claim 5 recites, generically: multiple ports to multiple networks; a receive buffer per port; a transmit buffer per port; "fullness signals" one per transmit buffer indicating relative fullness; a bus coupled to the buffers; and an arbiter coupled to the bus that controls each port's access in response to the fullness signals. It contains no collision-window limitation and no Ethernet-collision-domain limitation.

Proposed combination:

R1 (IEEE 802.3, 1996) in view of R3 (US 4,500,990) — and, for the arbitration element, in view of any § 102(b) reference disclosing per-port TX-FIFO-fullness-gated local-bus arbitration in a multiport repeater.

Reasoning:

• R1 supplies the multiport repeater environment, independent collision domains per port, and the CSMA/CD buffering context.
• R3 supplies the longstanding recognition that a CSMA/CD packet buffer can overflow and must be monitored and signalled, i.e. that a fullness signal is a conventional CSMA/CD building block.
• Once a POSITA puts per-port transmit FIFOs in a repeater (the very premise of a "buffered repeater" — an admitted architectural goal in the spec's own statement of goals: "avoid any deadlock between its ports; reduce packet loss to very marginal cases only; Ability to receive and transmit at the wire speed"), gating the shared broadcast bus on the transmit ports' readiness is the predictable, mechanical design step. A shared bus cannot be granted to a source until the destinations can accept, or the packet is lost; that constraint is inherent in the architecture, not an inventive insight.

Why claim 5 is more exposed than claims 1/3: the earlier Prior Art section already flagged this, and nothing in this session changes it. Claim 5 reads on generic buffer-fullness-gated bus arbitration, a design pattern that was conventional in buffered repeaters and shared-memory switches by the mid-1990s. If the examiner-cited art (still unrecovered) contains any multiport repeater/switch with per-port TX-FIFO-fullness arbitration, claim 5 is the claim that falls first, and claims 1/3 would survive it because of the collision-window timing element.

3.2 Claim 6 — falls with claim 5

Claim 6 narrows the fullness signal to indicate that the port has "completed a predetermined stage of transmission." Under R1, the 512-bit collision window is precisely such a predetermined stage — it is a fixed, standardized transmit-time milestone after which retransmission is no longer required. R1 in view of R3 renders claim 6 obvious for the same reasons as claim 5, with the added impetus that R1 defines the milestone to be used.

3.3 Claims 1 and 3 — the narrow claims; defensible but not invulnerable

Claims 1/3 require all of:
(i) per-port separate receive and transmit buffers;
(ii) a bus over which a receive buffer broadcasts for loading into other ports' transmit buffers;
(iii) an arbiter granting access upon assertion of "transmit buffer ready" from each port other than the requesting port; and
(iv) assertion of that signal "after passage of a collision window … and at a transmit time before the transmit buffer is emptied" — the simultaneous drain-and-fill.

Proposed combination:

Primary reference: a § 102(b) buffered-repeater reference disclosing (i)–(iii) — per-port RX/TX FIFOs on a shared broadcast bus with arbitration.
In view of R1 (IEEE 802.3, 1996) for element (iv)'s timing milestone.

Motivation to combine (this is the crux):

1. The constraint dictates the answer. The spec explains the rationale itself: the TX FIFO contents must be preserved through the collision window because a collision before the window's end requires retransmission; after the window, retransmission is no longer required, so the buffered bytes may be overwritten. Once a POSITA accepts that premise — and it is given by R1, not invented — the earliest safe moment to declare "ready to load" is exactly the passage of the collision window. The '675 spec itself frames the design choice this way (and the same language appears in the sibling family's EP 0 853 406 A2 text: "In cut-though operation the port can load a packet if the loaded packet is being transmitted and has passed the collision window (64 Bytes were already transmitted without incurring a collision)" — see https://patents.google.com/patent/EP0853406A2/en). That is a design choice with a predictable result, which is the KSR touchstone.
2. Cut-through was known art. The specification uses "cut-through operation" as an established term of art and distinguishes it from "store & forward." A POSITA seeking to shave repeater latency had an express, known motivation to begin forwarding before reception/drain completes — the early-fill idea is the buffered-repeater analogue of cut-through, and the market pressure (100 Mb/s repeaters, wire-speed relay) supplied the incentive.
3. Small, discrete implementation step. The change from "assert ready when the FIFO is empty" to "assert ready one clock cycle after the collision window" is a one-milestone change in a state machine, with no new hardware — the kind of improvement KSR classifies as obvious where the art provides the motivation.

Where this combination could fail (be candid):

• Teaching away / alternative resolution. The closest contemporaneous art I could date (R4, US 6,295,281, filed 1997-05-16) resolves the same TX-FIFO capacity problem in the opposite direction: rather than permitting concurrent fill-and-drain, it suspends the local bus when a TX FIFO falls below a full-frame's capacity ("Rather than one of the transmit ports risking being unable to store a complete frame once transmission has begun, the local bus is suspended before the next frame is put on the local bus"). A patent owner can plausibly argue that the art's prevailing teaching was to prevent refill-while-draining, and that the '675 design runs counter to it. That is an argument the § 103 case must rebut, not ignore.
• R4 cannot be used as art (it post-dates the priority date by four months). Its value is evidentiary only — see § 5.

3.4 Claims 2 and 4 — largely consequential to claims 1/3

Claim 2 recites means for the transmitting port's transmit buffer to begin receiving the new packet before the stored packet finishes draining; claim 4 is its method counterpart.

Analysis: if the arbiter in claim 1 grants access on a "ready to load" assertion made before the buffer is empty (which claim 1 already recites), then the loading-before-emptying of claim 2 is the inherent consequence of claim 1's timing, not a separate inventive step. Claim 2 adds essentially nothing that claim 1 does not already imply. Practical effect: claims 2/4 rise or fall with claims 1/3, and they are the least likely to survive independently if the primary reference is good on element (iv). They are correspondingly the weakest targets if the primary reference teaches only double-buffering (two alternating buffers) rather than true concurrent fill-while-drain — that distinction must be checked against the primary reference's figure.

---

4. Motivation-to-combine summary (KSR factors)

---

5. The R4 question — near-simultaneous independent development (evidentiary, not art)

US 6,295,281 (3Com), filed 1997-05-16 — four months after the '675 priority date — independently discloses a buffered repeater whose transmit ports monitor TX FIFO capacity and assert a shared signal that gates local-bus transmission, with round-robin arbitration and cascading across devices. It is not § 102/§ 103 art against the '675 patent.

But it is legitimate objective evidence of the level of ordinary skill in the art and of what was "known or obvious" at the time (KSR expressly permits recourse to such considerations where the reference itself cannot be used). Its teaching is that the buffered-repeater/bus-arbitration architecture of '675 claims 1/3/5 — per-port RX and TX FIFOs, shared bus, arbitration gated on transmit-side capacity — was being independently developed by a competitor on essentially the same timescale. That undercuts any argument that the architectural elements (i)–(iii) of claim 1 were inventive, and confines the entire § 103 dispute to element (iv) — the collision-window timing.

Caveat: I retrieved R4 only via the Justia examiner page listing and the Google Patents record; I have not confirmed its full disclosure against figures. It should be pulled in full before being relied on even evidentially.

---

6. Confidence assessment

Overall: on the art currently on the record, I can articulate a grounded § 103 case against claims 5 and 6, and a combine-with-802.3 case against claims 1/3 that is arguable but not clean. I cannot responsibly represent that claims 1–4 are invalid without the primary reference — and the primary reference is exactly what is missing.

---

7. What would complete this analysis (unchanged from the Prior Art section, now ranked by § 103 utility)

1. (56) References Cited from the printed patent (USPTO Patent Public Search / Patent Center / ESPACENET "Cited documents") — the single highest-value missing item.
2. International Search Report for PCT/US1998/000128 (= WO 98/33283 A2) — the ISA's citation list is the closest proxy for the examiner's art.
3. EPO search report for EP 1 012 988 A2 (EP app. 98905942) — an independent second examiner's view of the same disclosure, and the only prior-art assessment in this family that I can see being published.
4. Filing dates for US 5,999,538 (R5) — resolves whether it is § 102(e) art; if yes it becomes a strong primary-reference candidate.
5. Full text/figures of US 6,295,281 (R4) — evidentiary use only, but important to the "level of skill" argument.
6. The examiner's Reasons for Allowance / any § 103 rejection in the file history — tells us which element the examiner regarded as the point of novelty (I would expect it to be the collision-window timing of claim 1, which would confirm that claims 5/6 were an afterthought and are correspondingly weaker).

---

8. Contradictions and anomalies carried forward (not resolved)

• Serial-number anomaly: the family record lists 08/779,877 both as the application number of US 6,002,675 and, on FreePatentsOnline, as the sibling "Adaptive Address Filtering" application. Both cannot be correct. Reported literally; not reconciled. This matters for § 103 only in that a mis-cited sibling would change the § 103(c) ownership analysis — but on the text supplied, the siblings are same-day filings and therefore not art regardless.
• Title divergence within the family: US title "Method and apparatus for controlling transmission of data over a network" vs. EP/WO title "Buffered repeater with early filling of transmit buffer." This is a genuine, verifiable divergence (EP 1012988 A2 via EPO Global Patent Index) and it materially assists claim construction in favour of the "early filling" reading I applied above.
• Date discrepancy: the task prompt states 2026-04-26; the authoritative record was fetched 2026-09-30. No date-dependent conclusion here relies on either.

Bottom line: the § 103 case is strongest against claim 5 (and claim 6), moderate against claim 1/3, and consequential against claim 2/4 — but it cannot be finalized on the record deployed in the Prior Art section, because the examiner's cited art was never retrieved. Any statement that claims 1–4 of US 6,002,675 are obvious would currently be unsupported, and I am not making it.