US 10484511

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

Let me search for the specific CAFC 2026 docket and the reexamination/PTAB outcomes, plus the patent's claims.

Let me search for the specific Federal Circuit docket 26-1836 and the patent's claim text.

US Patent 10,484,511 — Analyst Summary

A note on scope: I searched for this exact number (10484511 / US 10,484,511 B2). All results below match that identifier. Where a source discusses related but different patents in the same family, I flag it explicitly so the records are not conflated. One caution up front: I could not retrieve the verbatim granted claim text for the '511 patent in the sources available to me, so the claim overview below is derived from the patent's own Summary of Invention/Abstract and from administrative records — not from quoted claim language. Treat the claim overview as descriptive, not as a quote.

1. Bibliographic data

Priority chain (from the patent's Cross-Reference section): Continuation of Ser. No. 15/957,950 (filed Apr 20, 2018) → continuation of Ser. No. 14/025,109 (filed Sep 12, 2013, issued as US 10,069,936) → divisional of Ser. No. 12/836,059 (filed Jul 14, 2010, issued as US 8,560,604) → provisional Ser. No. 61/249,624 (filed Oct 8, 2009).

2. Abstract (as published)

"A system designed for increasing network communication speed for users, while lowering network congestion for content owners and ISPs. The system employs network elements including an acceleration server, clients, agents, and peers, where communication requests generated by applications are intercepted by the client on the same machine. The IP address of the server in the communication request is transmitted to the acceleration server, which provides a list of agents to use for this IP address. The communication request is sent to the agents. One or more of the agents respond with a list of peers that have previously seen some or all of the content which is the response to this request (after checking whether this data is still valid). The client then downloads the data from these peers in parts and in parallel, thereby speeding up the Web transfer, releasing congestion from the Web by fetching the information from multiple sources, and relieving traffic from Web servers by offloading the data transfers from them to nearby peers."

3. Plain-language overview of the disclosure (and independent claims)

The patent describes a distributed "acceleration" overlay for Internet traffic built on four roles performed by software on ordinary communication devices:

• Client — intercepts an application's outgoing request (e.g., a browser's HTTP GET) on the same machine.
• Acceleration server — holds a database mapping destination-server IP addresses to "agents"; returns a list of candidate agents (the examples use five) chosen by numerical proximity of IP address to the destination server.
• Agent — assigned to a range of destination servers/IPs; knows which chunks of a given URL have been seen and by which peers, validates freshness, and returns chunk checksums plus a peer list (and, for the first batch, the original HTTP headers).
• Peer — any device that has cached the chunks (keyed by checksum) and can serve them.

Data is split into fixed-size chunks (16 KB in the example), addressed by checksum, and downloaded in parallel from multiple peers. Requests are validated against the origin server when the cache may be stale (FIG. 12). FIGS. 14–15 describe an alternative embodiment applying the same scheme to TCP/IP-level traffic (connect/read/write interception).

The Summary of Invention recites two claim families, which are the clearest indication of the independent claims:

1. A network/system claim — a network for accelerating data communication comprising: (a) at least one client communication device originating a data request; (b) at least one agent communication device assigned to the data server, which receives the request and keeps track of which client devices have received responses to requests from that assigned server; (c) at least one peer communication device storing portions of the data received in response to the request and able to transmit those portions on request; and (d) at least one acceleration server that decides which agent is assigned to which data server and provides that information to the client.

2. A communication-device (apparatus) claim — a device with a memory and a processor configured to: originate a data request to a data server; be assigned to a data server; receive a request from a separate device in the network and track which client devices have received responses from the assigned data server; and store portions of the received data so they can later be served on request.

Uncertainty: I cannot confirm from the material I retrieved the exact independent-claim numbering or count for the '511 patent. Administrative records indicate 30 original claims (referenced as claims 1–30), with the independent claims appearing to fall in the ranges around claims 1, 9, 14, 17, 20 and 25 (based on reexamination and IPR challenge lists), but this is inference, not verified claim text. If verbatim claim language is required, the issued patent's claim set should be pulled directly from USPTO PatentCenter or Google Patents.

4. Post-grant and litigation status (specific to the '511 patent)

• Inter partes review IPR2021-00122 (Petitioner: Teso LT, UAB et al.; Patent Owner: Luminati Networks Ltd.) — filed Oct 29, 2020; institution denied (procedural). Claims listed as challenged: 1, 2, 3, 4, 5, 9, 14, 17, 20, 21, 22, 25, 27, 28, 29, 30.
• Ex parte reexaminations 90/014,624 and 90/014,827 (third-party requested; both concern the '511 patent, Art Unit 3992). Per the final Office action mailed June 21, 2022:
  • Claims 1–30 subject to reexamination;
  • New claims 31–66 canceled in the proceeding;
  • Claims 6–8, 10, 13, 15–16, 18–19, 23 confirmed/patentable;
  • Claims 1–5, 9, 11–12, 14, 17, 20–22, 24–30 rejected.
• District court: The '511 patent was asserted in Bright Data Ltd. v. Code200, UAB, et al., No. 2:19-cv-00396 (E.D. Tex.). The broader family litigation includes Luminati Networks Ltd. v. UAB Tesonet, 2:18-cv-00299; Bright Data/Luminati v. Teso LT, UAB, 2:19-cv-00395; Bright Data v. Tefincom S.A. d/b/a NordVPN, 2:19-cv-00414; and 2:20-cv-00188.
• Federal Circuit, 2026: Google Patents' litigation data lists a Court of Appeals for the Federal Circuit case, docket 26-1836, associated with this patent (link: portal.unifiedpatents.com/litigation/Court of Appeals for the Federal Circuit/case/26-1836). I could not retrieve the docket contents to confirm the parties or issues. Given the timeline (a Nov 25, 2025 PTAB decision affirming the examiner's rejection in the '511 reexamination, per Oxylabs' public case timeline), 26-1836 is plausibly the appeal from that reexamination decision — but I want to be explicit that this linkage is an inference, not a confirmed fact.

Important disambiguation — do not conflate: The high-profile Supreme Court denial of Feb 23, 2026 (petition No. 25-779, Bright Data Ltd. v. Code200, UAB, et al.) and the Aug 1, 2025 Federal Circuit consolidated judgment (Nos. 2023-2144 et al.) concerned US 10,257,319, 10,484,510, 11,044,342 and 11,044,344 — not the '511 patent. Those decisions involved role-based vs. hardware-based construction of "client device" and "second server" and invalidation over the "Crowds" reference. Any statement that the '511 patent was invalidated by those rulings would be incorrect on the record I retrieved.

5. Confidence and gaps

• High confidence: number, title, inventors, assignee chain, filing/issue dates, priority chain, abstract, and the reexamination/IPR administrative facts above (each is corroborated by Google Patents, Unified Patents, Docket Alarm and Oxylabs' published timeline).
• Medium confidence: the Nov 25, 2025 PTAB reexamination outcome and its connection to CAFC docket 26-1836.
• Low confidence / not verified: the exact independent-claim numbering and verbatim claim language. I did not obtain the granted claims text, so the plain-language overview is reconstructed from the patent's own summary and specification, and should be checked against the issued claims before being relied upon for a validity or infringement analysis.

If you need the verbatim claims or the 26-1836 docket contents, the authoritative sources would be USPTO PatentCenter (application 16/278,109 and reexamination controls 90/014,624 / 90/014,827) and the Federal Circuit's PACER docket for 26-1836.

I'll research litigation involving US Patent 10,484,511 using the sources you specified.

Let me dig into the specific cases and confirm which ones actually assert '511.

Litigation involving U.S. Patent No. 10,484,511 ("System providing faster and more efficient data communication")

Important preliminary note on ownership: The patent issued to Web Spark Ltd. and was assigned to Luminati Networks Ltd. (assignment recorded 2019‑12‑05); Luminati then changed its name to Bright Data Ltd. (recorded 2021‑03‑31). So "Web Spark," "Luminati," and "Bright Data" are the same patent‑owner lineage, and the plaintiff name changed mid‑litigation in several of these cases. All cases below were filed in the U.S. District Court for the Eastern District of Texas, Marshall Division, before Judge Rodney Gilstrap (‑JRG).

I found the following matters that specifically involve the '511 patent (not merely sibling patents in the same family). I have been careful not to conflate the many related patents ('319, '510, '614, '968, '342, '344, '622, '034, '953) that appear in the same campaign.

1. District court cases asserting the '511 patent

A. Bright Data Ltd. (f/k/a Luminati Networks Ltd.) v. Code200, UAB, Oxysales, UAB, and Metacluster LT, UAB

• Case No.: 2:19‑cv‑00396‑JRG (E.D. Tex., Marshall Div.) ("the Code200 Action")
• Filed: December 6, 2019 (the docket order states Dec. 6, 2019; one PTAB petition brief states "12/31/2019" — a discrepancy I cannot resolve from the sources reviewed)
• Asserted patents: U.S. 10,484,511 ('511), and U.S. 10,637,968 ('968) added June 9, 2020
• Accused products: Oxylabs Data Center Proxy Service and Real‑Time Crawler
• Status/outcome:
  • Claims against Metacluster LT, UAB dismissed with prejudice (joint stipulation dated June 30, 2021; dismissal entered July 2, 2021) after Metacluster represented it did not perform the accused acts.
  • Case stayed (mediation stay entered ~November 2021; a further sua sponte stay on February 28, 2023 pending the USPTO proceedings).
  • The '511 and '968 claims survived Defendants' § 101 Alice motion to dismiss (court denied the Rule 12(b)(6) motion).
• Sources: https://www.docketalarm.com/cases/PTAB/IPR2021‑01492/NetNut_Ltd._v._Bright_Data_Ltd/docs/01‑06‑2023‑Patent_Owner/Exhibit‑2023‑EX_2023___EDTX_2_19_cv_00396_98.pdf ; https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1549278](/patent/1549278)/download-documents?artifactId=FLhvL3s4im6OcDdm3Um5DXtqMfbgc6iDbFUltyX9WCqRwzrAVIjcFO0 ; https://oxylabs.io/legal-timeline

B. Luminati Networks Ltd. v. NetNut Ltd.

• Case No.: 2:20‑cv‑00188‑JRG‑RSP (E.D. Tex., Marshall Div.)
• Filed: June 11, 2020
• Asserted patents: U.S. 10,484,511 ('511) and U.S. 10,637,968 ('968) — confirmed by the AO‑120 patent filing notice listing both
• Status/outcome: Case stayed (July 23, 2021 order) pending ex parte reexaminations. On summary judgment the court (Report and Recommendation, Dkt. 211) addressed NetNut's non‑infringement of the '511 patent after Dec. 21, 2020, when NetNut moved its proxy servers outside the U.S.; the court found a justiciable controversy and granted summary judgment of non‑infringement for the post‑relocation period. A later document on the '511 file history reflects dismissal of "all claims and causes of action asserted by Luminati … against NetNut" without prejudice, though I could not confirm from the sources reviewed whether this closed 2:20‑cv‑00188 or relates to the related NetNut action 2:21‑cv‑225 (which asserts '319/'510, not '511).
• Sources: https://storage.courtlistener.com/recap/gov.uscourts.txed.[198166](/patent/198166)/gov.uscourts.txed.198166.4.0.pdf ; https://storage.courtlistener.com/recap/gov.uscourts.txed.198166/gov.uscourts.txed.198166.211.0.pdf

C. Luminati Networks Ltd. v. Tefincom S.A. d/b/a NordVPN

• Case No.: 2:19‑cv‑00414‑JRG (E.D. Tex., Marshall Div.)
• Filed: ~December 31, 2019 (the complaint PDF is dated 2019‑12‑31)
• Asserted patents: includes U.S. 10,484,511 ('511) — the complaint contains "COUNT IV (Infringement of the '511 Patent)" — plus '510 and '968
• Status/outcome: Not confirmed in the materials reviewed. (Note: the IPR2021‑00122 institution decision identifies this as one of the three related district‑court litigations involving the '511 patent, alongside the Code200 and NetNut actions.)
• Sources: https://luminati.io/static/patent/2019‑12‑31_NordVPN_Complaint.pdf ; https://www.docketalarm.com/cases/PTAB/IPR2021‑00122/Teso_LT_UAB_v._Luminati_Networks_LTD/docs/04‑20‑2021‑Board/Institution_Decision‑12‑Decision_Denying_Institution_of_Inter_Partes_Review.pdf

Related but NOT asserting '511 (listed to avoid confusion):

• Bright Data Ltd. v. Teso LT, UAB, et al., No. 2:19‑cv‑00395 (E.D. Tex.) — asserts '319, '510, and '614. Jury verdict Nov. 5, 2021 (patents valid, willfully infringed); '319 and '510 later invalidated by PTAB and affirmed by the Federal Circuit (Aug. 1, 2025), with Supreme Court review denied (~Feb./Mar. 2026). This case does not involve '511.
• Luminati Networks Ltd. v. BI Science (2009) Ltd., No. 2:19‑cv‑00397 (E.D. Tex.) — dismissed. I could not confirm that '511 was asserted here; it is listed at the family level by Google Patents but no source I reviewed ties '511 to this case.
• Metacluster LT, UAB v. Bright Data Ltd., No. 2:22‑cv‑00011 (E.D. Tex.) — Oxylabs' offensive suit against Bright Data asserting Oxylabs' own patents ('948 etc.); not a '511 case.

2. PTAB post‑grant proceeding

Teso LT, UAB; Code200, UAB; Metacluster LT, UAB; and Oxysales, UAB v. Luminati Networks Ltd.

• Proceeding: IPR2021‑00122, U.S. Patent No. 10,484,511
• Filed: October 29, 2020 (filing date accorded Nov. 17, 2020)
• Claims challenged: 1–5, 9, 14, 17, 20–22, 25, 27–30
• Outcome: Institution DENIED — Decision Denying Institution of Inter Partes Review, April 20, 2021 (denial based in part on the advanced stage of the parallel Code200 litigation). Panel: Sheila F. McShane, Thomas L. Giannetti, Russell E. Cass.
• Sources: https://portal.unifiedpatents.com/ptab/case/IPR2021‑00122 ; https://www.docketalarm.com/cases/PTAB/IPR2021‑00122/Teso_LT_UAB_v._Luminati_Networks_LTD/docs/04‑20‑2021‑Board/Institution_Decision‑12‑Decision_Denying_Institution_of_Inter_Partes_Review.pdf

3. USPTO ex parte reexamination (not "litigation," but arising from the dispute)

Ex Parte Reexamination Control Nos. 90/014,827 and 90/014,624 (U.S. Patent No. 10,484,511)

• Filed: December 4, 2020 (requester associated with the Oxylabs defendants; correspondence via Liston Abramson LLP and Charhon Callahan Robson & Garza)
• Outcome/status: Final Office Action mailed June 21, 2022 rejecting all claims asserted against Oxylabs; a PTAB decision on November 25, 2025 affirmed the rejection.
• Sources: https://www.docketalarm.com/cases/PTAB/IPR2021‑01492/NetNut_Ltd._v._Bright_Data_Ltd/docs/03‑20‑2023‑[Petitioner](/litigations/by-plaintiff/Petitioner)/Exhibit‑1128-Ex_1128___Final_Office_Action_mailed_June_21,_2022,_Reexamination_Nos_90014,827__90014,624.pdf ; https://oxylabs.io/legal-timeline

4. Federal Circuit appeal

In re / appeal concerning U.S. Patent No. 10,484,511

• Google Patents lists a U.S. Court of Appeals for the Federal Circuit case, Case No. 26‑1836, associated with this patent (docket link from the Google Patents page). This is consistent with an appeal from the November 25, 2025 PTAB reexamination decision on the '511 patent.
• Note/caveat: I could not independently confirm the parties, filing date, or outcome of CAFC No. 26‑1836 from the sources reviewed; and I want to flag that the consolidated Federal Circuit appeals Nos. 2023‑2144 through 2023‑2443 (which produced the Aug. 1, 2025 affirmance and the ~Feb./Mar. 2026 Supreme Court denial) involve the '342, '319, '510, and '344 patents — not '511.

Assessment

• Yes, there is known litigation involving U.S. 10,484,511: the primary district‑court matters are 2:19‑cv‑00396 (Code200/Oxysales/Metacluster), 2:20‑cv‑00188 (NetNut), and 2:19‑cv‑00414 (Tefincom/NordVPN), plus the IPR2021‑00122 PTAB petition, the 90/014,827 & 90/014,624 reexaminations, and the CAFC No. 26‑1836 appeal.
• Current posture: The '511 patent's asserted claims were cancelled/rejected in ex parte reexamination, a rejection affirmed by the PTAB on November 25, 2025, and the related district‑court cases have been stayed or resolved in view of the USPTO proceedings. The '511 patent's anticipated expiration is July 14, 2030 (per Google Patents).
• Uncertainty flags (don't treat as settled): (i) the exact filing date of the 2:19‑cv‑00396 complaint (Dec. 6 vs. Dec. 31, 2019); (ii) the current docket status of the Tefincom/NordVPN action; (iii) whether the NetNut '511 claims were dismissed without prejudice in 2:20‑cv‑00188 or in the separate 2:21‑cv‑225 action; and (iv) whether '511 (as opposed to other family patents) was asserted in the dismissed BI Science action (2:19‑cv‑00397).

If you need definitive docket‑level confirmation (e.g., PACER docket entries, final judgments, or the CAFC 26‑1836 briefing), those are the places to pull the authoritative records, since some of the above statuses come from secondary aggregators (CourtListener, Docket Alarm, Unified Patents, Oxylabs' own litigation timeline, Google Patents).

I'll start by verifying the structured-data baseline against what the web shows, since the patent page itself already flagged one PTAB item.

Two things jump out: the patent page itself lists IPR2021-00122 (not instituted), and a court docket references a PTAB decision denying IPR of the '511 patent. Let me dig into that proceeding plus the parallel ex parte reexamination.

Good — I've confirmed the IPR and the petitioner. Let me now pull the denial reasoning, the claim count, and the reexamination outcome.

Proceedings overview

Total AIA trial proceedings on file for US10484511: 1 — and it never made it past the institution threshold. Breakdown: 0 active, 0 claims invalidated in an AIA trial, 0 claims sustained in an AIA trial, 0 settled, 1 institution denied (IPR2021-00122, denied 2021-04-20). The bottom line for a defendant: the '511 patent has never been tested on the merits at the PTAB — it survived one IPR by a discretionary § 314(a) denial, not by winning on the art. That cuts both ways: there is no FWD canceling any claim (so no "claim 1 is dead" argument), and there is no § 315(e)(2) estoppel binding the challenger group either. Critically, the more consequential validity event for this patent is happening outside the AIA regime — an ex parte reexamination (Control Nos. 90/014,624 & 90/014,827) in which all asserted claims were finally rejected on 2022-06-21, with a PTAB appeal decision reported as affirming that rejection on 2025-11-25.

⚠️ Data-integrity flag. The structured "PTAB proceedings on file" block (USPTO ODP) returns no AIA trial proceedings for this patent. That is incomplete: Google Patents' own page for US10484511 lists "PTAB case IPR2021-00122 filed (Not Instituted - Procedural)," and Docket Alarm, Patexia, and GreyB/IPVerse all show the proceeding. I am surfacing it as the canonical proceeding of record notwithstanding the ODP gap. Also note: the "Petitioner:" field in Google Patents' litigation block is followed by the Unified Patents data-license attribution — that is a rendering artifact, not evidence that Unified Patents was the petitioner. The petitioner of record is the Teso/Code200/Metacluster/Oxysales group.

IPR2021-00122 — Teso LT, UAB; Code200, UAB; Metacluster LT, UAB; Oxysales, UAB v. Luminati Networks Ltd. (now Bright Data Ltd.)

• Type: Inter Partes Review (35 U.S.C. §§ 311–319)
• Filed: 2020-10-29 (filing date accorded by Notice mailed 2020-11-17, Paper 6)
• Status: Institution Denied — Google Patents: "Not Instituted - Procedural"; Patexia: "Terminated-Denied"; GreyB IPVerse: "Institution Denied." Terminated without a trial on the merits.
• Judge panel: Thomas L. Giannetti, Sheila F. McShane, and Russell E. Cass, Administrative Patent Judges; opinion authored by APJ McShane (full text excerpt)
• Petition grounds: Challenged claims 1–5, 9, 14, 17, 20–22, 25, and 27–30 of the '511 patent (16 claims), supported by the Declaration of Michael Freedman, Ph.D. (Ex. 1010). Statutory basis was § 102/§ 103 on patents and printed publications (the § 311(b) universe). Petition exhibit list includes U.S. Patent No. 5,974,566 (Ault) and U.S. Patent No. 6,185,625 (Tso), among others; I could not retrieve the complete ground-by-ground § 102/§ 103 table for this petition, so I am not characterizing the individual references beyond those confirmed exhibits.
• Institution decision: Denied 2021-04-20 (Paper 12, Decision Denying Institution of Inter Partes Review, 35 U.S.C. § 314). The panel's bottom line, quoted verbatim: "For the reasons that follow, we exercise our discretion under 35 U.S.C. § 314(a) to deny institution of inter partes review." The decision recites the parallel E.D. Tex. matters as related matters — Luminati Networks Ltd. v. Code200, UAB, No. 2:19-cv-00396-JRG; Luminati Networks Ltd. v. NetNut, Ltd., No. 2:20-cv-00188-JRG; and Luminati Networks Ltd. v. Tefincom S.A. d/b/a NordVPN, No. 2:19-cv-00414-JRG — and notes the then-denied sibling petitions IPR2020-01266 (re '319) and IPR2020-01358. The Patent Owner Preliminary Response (Paper 10, with the Rhyne Declaration) was built around the district-court posture (claim construction order, docket control order, and an order denying a motion to stay) plus a stipulation, which is the signature of a Fintiv-type § 314(a) discretionary denial. I have verified the outcome and the § 314(a) basis, but not the factor-by-factor weighing — that reasoning is in Paper 12, available via PTAB E2E. A second plausible but unverified factor is the co-pending ex parte reexamination covering all asserted claims.
• Final Written Decision: None. No trial was instituted, so no claim was adjudicated unpatentable or patentable. Any statement that claims of '511 were "held valid" or "held invalid" in this IPR is incorrect.
• Settlement / termination: No settlement. The proceeding ended by denial; Petitioner filed a Request for Refund of Post-Institution Fee on 2021-07-19 (Paper 13), and the Board issued a Notice of Refund on 2021-07-21 (Paper 14).
• Appeal: None. A denial of institution is non-appealable under 35 U.S.C. § 314(d) (Cuozzo), so there is no Federal Circuit docket for this proceeding and no CAFC opinion. (The CAFC activity in this family — consolidated Appeal Nos. 2023-2144, -2145, -2146, -2147, -2414, -2415, -2442, -2443 — concerns other Bright Data patents, e.g., 10,257,319 and 10,484,510, not the '511.)
• Defensive value: This proceeding gives a defendant no claim-cancellation leverage — nothing was adjudicated, so no offensive estoppel or "dead claim" argument exists. But it is affirmatively useful in the other direction: because institution was denied, no § 315(e)(2) estoppel attached to Teso/Code200/Metacluster/Oxysales, and the denial was discretionary rather than merits-based, meaning the Board has expressed no view on the patentability of any '511 claim. It also establishes that a § 314(a) discretionary-denial argument has once worked against this patent when it was already in parallel E.D. Tex. litigation.

(Not an AIA trial — flagged for completeness) Ex Parte Reexamination Control Nos. 90/014,624 & 90/014,827 — U.S. Pat. No. 10,484,511

This is not a PTAB trial, so it is outside the structured "PTAB proceedings" list, but it is the single most important validity development on this patent and no defendant should evaluate '511 without it.

• Type: Ex parte reexamination (35 U.S.C. §§ 301–307), filed by a third-party requester
• Filed: 2020-12-04 (requester's filing date as stated by NetNut Ltd. in its E.D. Tex. stay papers: "NetNut accordingly pursued EPR to challenge the validity of the '511 and '968 patents on December 4, 2020"). Examiner: Eron J. Sorrell, Art Unit 3992 (CRU). Requester correspondence: Liston Abramson LLP.
• Status: Request granted on all asserted claims "plus more" (per NetNut's stay motion: "Both EPRs have been granted, and all asserted claims in this case (plus more) are being reexamined"). A Final Office Action issued 2022-06-21 rejecting the claims, and per Oxylabs' published litigation timeline, a PTAB decision on 2025-11-25 affirmed the rejection. ⚠️ The 2025-11-25 affirmance is sourced only to a party's published litigation timeline (oxylabs.io/legal-timeline) — I could not independently verify it against the PTAB Appeal Center or a reexamination certificate. Which claims were rejected, and whether a reexamination certificate has issued canceling them, are unverified.
• Claim-level outcome: Unverified. The 2022-06-21 FOA is on the record (filed as Exhibit 1128 in IPR2021-01492); reportedly it rejected all claims asserted against Oxylabs and more.
• Estoppel: Ex parte reexamination does not trigger § 315(e)(2) estoppel — only an IPR/PGR final written decision does.
• Defensive value: If the 2025-11-25 affirmance is accurate and a reexamination certificate cancels (or narrows) the asserted claims, that is a far stronger defense than anything in IPR2021-00122 — but only after it is confirmed in the Office record and any CAFC appeal window has run. Confirm before relying on it.

Strategic summary

Claim status — CANCELED vs. SUSTAINED vs. UNTESTED. Every claim of US10484511 is UNTESTED in an AIA trial: claims 1–5, 9, 14, 17, 20–22, 25, and 27–30 were challenged in IPR2021-00122 but never instituted on, and the remaining claims were never challenged at all. There is no FWD and therefore no claim held unpatentable and none held patentable by the Board in an AIA proceeding. The narrowing, if any, is coming through the parallel ex parte reexamination (90/014,624 & 90/014,827), where the examiner finally rejected the claims on 2022-06-21 and, as reported, the Board affirmed on appeal on 2025-11-25. Until that certificate is confirmed, the safe planning assumption for a defendant is that the original granted claims are still in force — the patent's anticipated expiration is 2030-07-14, so roughly four years of life remain.

Estoppel landscape. § 315(e)(2) estoppel is absent. Estoppel attaches only upon a final written decision, and IPR2021-00122 died at institution. That means: (a) Teso LT, Code200, Metacluster, Oxysales and their privies are not barred from raising in district court any § 102/§ 103 ground on '511, including grounds they raised or could have raised in the petition; and (b) any new defendant — including one not in privity with the 2021 petitioner group — retains the full § 311(b) prior-art universe for a fresh IPR. The practical constraints on a new petition are discretionary, not statutory: the Board's General Plastic/Fintiv practice now cuts against a serial attack from the same E.D. Tex. defendant group, and § 325(d) is now the real obstacle — the '511 reexamination record has put a substantial body of prior art before the Examiner and the Board, so a new petitioner must satisfy Advanced Bionics by showing the Office materially erred (or rely on art the reexam never considered). NetNut's § 112 and non-prior-art defenses (indefiniteness, enablement, written description, on-sale bar) remain entirely outside AIA/reexam estoppel. One litigation-side caveat: in the NetNut action, NetNut offered a Veraseal-type stipulation not to assert at trial the prior-art grounds raised in its EPRs if a stay were granted — a stipulation tied to the stay motion, not a statutory estoppel.

Pattern signals. This is a serial-challenger family fight, not a defensive-aggregator story. The same E.D. Tex. defendant cohort (Teso LT / Code200 / Metacluster / Oxysales) filed IPRs against a portfolio of Bright Data/Luminati patents in this family — '319, '510, '344, '317, and others — and won several of those on the merits (e.g., IPR2022-00135 and IPR2022-00138, where all challenged claims were held unpatentable and which the Federal Circuit affirmed in the consolidated 2023-2144 et al. appeal, with cert. denied). '511 is the outlier where the challengers were turned away at the institution stage. Separately, NetNut went the reexamination route on '511 (and '968) deliberately, telling the district court it chose EPR over a follow-on IPR because of the Board's policy disfavoring serial petitions. There is no Unified Patents IPR on '511 — the Unified Patents name on Google Patents is a data-license attribution. No PGR or CBM proceeding exists on '511 (and CBM is unavailable for this subject matter post-SAS/Oil States era/fee sunset anyway).

Recommended next steps

1. Do not rely on IPR2021-00122 for anything but estoppel and posture. Pull Paper 12 (Decision Denying Institution, 2021-04-20) from PTAB E2E to confirm the § 314(a) reasoning — it is the strongest available template for a renewed discretionary-denial argument if '511 is asserted against a defendant in pending E.D. Tex. litigation. Proceeding link: https://portal.unifiedpatents.com/ptab/case/IPR2021-00122 · docket mirror: https://www.docketalarm.com/cases/PTAB/IPR2021-00122/ · PTAB E2E: https://ptacts.uspto.gov/ptabweb/
2. Verify the reexamination status first — it may moot everything. Confirm on the USPTO reexamination record (Patent Center / PTAB Appeal Center) for Control Nos. 90/014,624 and 90/014,827: (i) the exact claims rejected in the 2022-06-21 FOA; (ii) whether the reported 2025-11-25 PTAB appeal decision affirmed all rejections, and the appeal number; (iii) whether a reexamination certificate has issued, and (iv) whether any CAFC appeal was noticed. Bring the FOA (already public as Ex. 1128 in IPR2021-01492) and the appeal decision to any infringement analysis.
3. If you are a defendant being asserted now: there is no FWD to cite and no canceled claim to attack the demand letter with — but there is also no estoppel against you. Budget a fresh IPR only if the reexam has not already canceled the asserted claims; if it has, lead with the reexamination certificate (and 35 U.S.C. § 307(b) effect) rather than an AIA filing. Check whether you are in privity with Teso/Code200/Metacluster/Oxysales, because that is what governs your § 315(e) exposure.
4. Watch for a unicorn. A patent asserted across at least five E.D. Tex. actions that has attracted exactly one (denied) IPR and one reexamination is unusual; if the reexam does not cancel the claims and assertion continues, a well-funded defendant with clean hands and new art not before the Examiner is a strong candidate for a successful IPR. Feed that timing into the § 315(b) one-year bar from service.
5. Litigation context links: Bright Data v. Code200 (2:19-cv-00396, where the IPR denial was noticed on the docket): https://www.courtlistener.com/docket/16555815/bright-data-ltd-v-code200-uab/. NetNut's stay briefing on the EPRs (2:20-cv-00188, Dkt. 77) is the best narrative source for why the reexam route was chosen.

Confidence notes: (i) The existence, filing date (2020-10-29), petitioner group, panel, challenged claim set, denial date (2021-04-20), and § 314(a) basis for IPR2021-00122 are well corroborated. (ii) The precise Fintiv factor weighing in the denial is not verified. (iii) The 2025-11-25 reexam appeal affirmance and the "all asserted claims rejected" characterization rest on a party-published timeline and remain unverified; treat as a lead, not a fact. (iv) I found no PGR, CBM, or additional IPR on '511; if one exists it is not in ODP, Google Patents, or the docket aggregators I searched, which is itself worth a direct PTAB E2E name search.

I'll research the assignment chain, litigation, and correspondence of record for US 10484511.

Now let me try to pin down the recorded assignment details (reel/frame, correspondent) and the inventors' background.

Assignment & Ownership Analysis — US 10,484,511 B2

Patent: "System providing faster and more efficient data communication"
App. No. 16/278,109 · Filed 2019-02-17 · Issued 2019-11-19 · Priority 2009-10-08 (prov. 61/249,624)
Family: continuation of 15/957,950 → 14/025,109 (US 10,069,936) → 12/836,059 (US 8,560,604)

Source caveat (read first). I could retrieve the conveyance events for this patent from the Google Patents legal-events register (which mirrors the same assignment data the Assignment Center surfaces), but I could not retrieve the reel/frame numbers or the correspondent of record for the individual recordings from the sources available to me in this session. I have not invented reel/frame values or correspondent names. Every reel/frame field below is marked [not retrieved] and should be filled in from the USPTO Assignment Center before this memo is relied upon. If the Assignment Center turns out to have no records, the Google Patents events below would then be the only trace and the analysis collapses to "Insufficient data."

Inventors

Unusual patterns: none of the classic red flags. The inventors did not bail out of the original assignee: both are named as assignors into Web Spark Ltd. on the filing date, and the same two-name inventor pair runs across the entire continuation family (12/836,059 → 14/025,109 → 15/957,950 → 16/278,109), so there is no evidence of a post-filing inventor exodus preceding a fire-sale. What does exist is a corporate-family reshuffle (Web Spark → Luminati → Bright Data), which is an entity-consolidation pattern, not an inventor-abandonment pattern.

Original assignee

Web Spark Ltd. — named as the original assignee of the issued patent. Google Patents lists Original Assignee = Web Spark Ltd. and Current Assignee = Bright Data Ltd.

• Primary line of business: not established on the face of the record. Web Spark Ltd. appears in the chain only as a title-holding/vehicular entity for the Shribman–Vilenski patent family; the operating business (proxy infrastructure / web-data collection) is carried by Luminati Networks Ltd. and later Bright Data Ltd. I found no evidence of Web Spark selling a product, and no evidence of its dissolution.
• Shipped a product embodying the claims? Unclear. The commercial embodiment of these claims is plainly the Luminati/Bright Data proxy platform (the asserted claims recite per-request selection of a client-side IP address from a stored group and relaying content through it — exactly a residential-proxy rotation service), but that product is attributed to Luminati/Bright Data, not to Web Spark.
• Current status: superseded as owner by the 2019-12-05 transfer to Luminati Networks Ltd.; no bankruptcy, acquisition, or dissolution record surfaced for Web Spark itself.

Assignment timeline

Three recorded events. Reel/frame values not retrieved — verify at the Assignment Center (links at the end).

• 2019-02-17 (event/recording date) / recorded 2019-02-17 — Reel [not retrieved] / [not retrieved]

  • Conveyance: Assignment
  • Assignor: Derry Shribman; Ofer Vilenski
  • Assignee: Web Spark Ltd.
  • Correspondent: [not retrieved] — unable to confirm whether this is the same recording attorney as the later links; flag for follow-up.
  • Context: standard inventor-to-company assignment executed for a newly filed continuation, giving Web Spark record title on the day of filing.

• 2019-12-05 / recorded 2019-12-05 — Reel [not retrieved] / [not retrieved]

  • Conveyance: Assignment
  • Assignor: Web Spark Ltd.
  • Assignee: Luminati Networks Ltd.
  • Correspondent: [not retrieved]
  • Context: internal corporate consolidation — title moved from the holding vehicle into the operating proxy company, recorded one day before Luminati filed its first two suits naming the '511 patent.

• 2021-03-31 / recorded 2021-03-31 — Reel [not retrieved] / [not retrieved]

  • Conveyance: Change of Name
  • Assignor: Luminati Networks Ltd.
  • Assignee: Bright Data Ltd.
  • Correspondent: [not retrieved]
  • Context: change of name only — Luminati rebranded to Bright Data; no change in beneficial ownership. (Independently corroborated: court filings in E.D. Tex. are styled "Bright Data Ltd. (formerly known as Luminati Networks Ltd.)", e.g. Dkt. 272 and 315 in Luminati Networks Ltd. v. BI Science, and Safe-T Group's 2022 release uses the same "f/k/a" formulation.)

No other transfers — no security agreements, no licenses recorded against title, no foreclosure, no releases.

Timeline diagram

timeline
    title Ownership of US 10484511
    2009 : Priority provisional filed
    2010 : Parent application filed
    2019 : Continuation filed
         : Inventors assign to Web Spark
         : Patent issues 19 Nov
         : Assigned to Luminati Networks
         : First infringement suits filed 6 Dec
    2021 : Luminati renamed Bright Data

NPE / troll-pattern signals

1. Shell-entity transfer — not present. Both transferees are operating entities, and the economically meaningful transfer (2019-12-05) moved the patent into the operating company (Luminati), the opposite of the operating-company → naked-licensing-LLC direction. No "IP / Holdings / Licensing" suffix appears, no registered-agent service address surfaced, and no single-purpose LLC exists in the chain.

2. Known asserter in the chain — not present. No assignee matches the reference NPE lists (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Spangenberg entities). The current owner, Bright Data Ltd., is an operating web-data/proxy platform. It is a high-frequency plaintiff — its cases are tracked on the Unified Patents litigation portal and it holds "over 760 granted patent claims" per its own marketing — but it sues its own market competitors, which is the operating-company-assertion profile, not the NPE profile.

3. Repeat correspondent across the chain — unclear. Correspondent of record was [not retrieved] for all three recordings, so the recurrence test cannot be run. I want to be explicit that this is a data gap, not a negative finding: if the three recordings show the same recording attorney, that would be ordinary for a single corporate family and would still not, on its own, indicate an NPE.

4. Cascading transfers — not present. Two transfers over ~25 months, both within one corporate family, no unrelated chained LLCs, no common-address clustering.

5. Pre-litigation transfer — present (timing), but benign in context. The assignment to Luminati Networks Ltd. recorded 2019-12-05 precedes by one day the first complaints asserting the '511 patent, filed 2019-12-06: Luminati Networks Ltd. v. Code200, UAB et al., No. 2:19-cv-00396 and Luminati Networks Ltd. v. BI Science (2009) Ltd., No. 2:19-cv-00397 (both E.D. Tex., Judge Gilstrap). The AO-120 forms in those cases list the '511 patent with Luminati Networks Ltd. as holder. This is a pre-assertion title cleanup to establish standing — the same timing you'd see in an NPE set-up, but here the transferee is the operating plaintiff itself, not a paper licensor. I could not distinguish execution date from recording date on the 2019-12-05 entry, which matters: if the assignment was executed earlier in 2019, the "within 6 months of suit" gloss weakens further.

6. Bankruptcy fire-sale — not present. No Chapter 7/11 record for Web Spark or Luminati. (An Israeli receivership/receiver filing does appear in the litigation record — an August 2021 Application for Instructions by Receiver Adv. Ziv Ironi in the Execution Office, Tel Aviv — but that is a judgment-enforcement proceeding tied to the BI Science dispute against the owner, not a corporate insolvency sale of the patent.)

7. Privateering — not present. Bright Data asserts on its own behalf and for its own platform; no evidence of a third-party operating company directing assertions.

8. Defensive aggregator — not present, and the inverse holds. The chain terminates at an active litigant asserting the patent, not at RPX, AST, LOT, Unified, or OIN. The patent is very much alive: IPR2021-00122 (Teso LT, UAB; Code200, UAB; Metacluster LT, UAB; Oxysales, UAB v. Luminati Networks Ltd.), petition filed 2020-10-29, institution denied 2021-04-20; a jury verdict of validity and infringement in 2:19-cv-395; a stipulated permanent injunction granted 2024-01-02 in Luminati Networks Ltd. v. BI Science Inc. (2:18-cv-00483, Dkt. 348) covering the '511 patent; and a confidential 2022 settlement with NetNut Ltd. (Safe-T Group).

Verdict

Operating-company assertion

The chain is short, documented, and runs entirely inside one operating corporate family: inventors → Web Spark Ltd. (2019-02-17) → Luminati Networks Ltd. (2019-12-05) → Bright Data Ltd. via Change of Name (2021-03-31), per the Google Patents legal-events register. Bright Data is a real proxy/web-data platform whose commercial service reads on the asserted claims, and it has asserted the '511 patent against actual competitors in that market (Code200/Oxysales/Metacluster, 2:19-cv-00396; BI Science, 2:19-cv-00397; Teso/Tesonet, 2:19-cv-395) — that is the textbook operating-company profile, and none of the eight NPE signals is present. The only signal worth flagging is the one-day gap between the 2019-12-05 transfer to Luminati and the 2019-12-06 first suits, which is a timing artifact of pre-suit title housekeeping rather than evidence of an assertion vehicle.

Two things to verify before this memo is final:

1. Reel/frame numbers and the correspondent of record for all three recordings — I did not retrieve them and have left them blank rather than guess.
2. Execution dates vs. recording dates, particularly on the 2019-12-05 Luminati assignment, which determines how tight the pre-suit window really is.

Verify at:

• USPTO Assignment Center — https://assignmentcenter.uspto.gov/
• Mirror search — https://assignment.uspto.gov/patent/index.html (search patent 10484511)
• Google Patents legal events — https://patents.google.com/patent/US10484511/en

I'll search for the specific cited references on this patent, since the "Patent Citations" sections were not included in the patent text provided to me.

The claims are now confirmed (thanks to the Espacenet result), but I still need the cited prior-art references from the patent's "(56) References Cited" section. Let me pull those directly.

I have the reexamination references now, but the task asks for the references cited on the face of the patent ("(56) References Cited"). Let me try to retrieve that specific list.

Prior-Art Analysis — US 10,484,511 B2

⚠️ Two things to read before the substance

1. Hard limitation on the sources. I searched for this exact number (10484511 / US 10,484,511 B2) and did not retrieve the verbatim "(56) References Cited" list that is printed on the front face of the issued patent — Google Patents' structured citation block was stripped from the full text I was given, and Espacenet/Justia returned the biblio, abstract and claims but not the front-page U.S.-patent-document citations. I am therefore not going to fabricate a front-page citation list. What I did retrieve, and what this memo is built on, are three distinct and separately-sourced prior-art sets that bear on the '511 patent:

• (A) the ex parte reexamination SNQ references (third-party requester, Control Nos. 90/014,624 & 90/014,827) — fully cited and dated, and the Office applied them to the claims;
• (B) the IPR2021‑00122 petition exhibit references (partially retrieved);
• (C) references listed against a sibling family member (US 10,637,968) — flagged as not confirmed for '511.

If you need the literal front-page "(56)" list, pull the granted PDF (front page) or the file wrapper for App. 16/278,109. The claim set, however, I do now have verbatim (Espacenet), and it changes one thing the earlier sections said.

2. Contradiction flag (must resolve). The previously-generated Patent summary section stated: "the independent claims appear to fall in the ranges around claims 1, 9, 14, 17, 20 and 25." That is incorrect. Both the Espacenet claim text and the reexamination Office action are explicit: claim 1 is the only independent claim; claims 2–30 are all dependent.

• Espacenet claim listing shows claims 2–17 each reciting "The method according to claim 1/2/3/6/11/14…" (Espacenet claims, US10484511 B2).
• The examiner's FOA states verbatim: "The '511 Patent issued on November 19, 2019 having claims 1- 30 with claim 1 being the only independent claim." (FOA, Ex. 1128, IPR2021‑01492)

This matters for your §102 mapping: only one claim (claim 1) can be anticipated as an independent claim. All other claims are dependent limitations, and prior art "anticipates" them only if it also discloses the parent claim plus the added limitation.

1. The operative claim — claim 1 (verbatim, Espacenet)

1. A method for fetching, by a first client device, a first content identified by a first content identifier and stored in a web server, for use with a first server that stores a group of IP addresses, the method by the first server comprising: receiving, from the first client device, the first content identifier; selecting, in response to the receiving of the first content identifier from the first client device, an IP address from the group; sending, in response to the selecting, the first content identifier to the web server using the selected IP address; receiving, in response to the sending, the first content from the web server; and sending the received first content to the first client device, wherein the first content comprises a web-page, an audio, or a video content, and wherein the first content identifier comprises a Uniform Resource Locator (URL).

The elemental breakdown used below: (a) first server stores a group of IP addresses; (b) receives a URL/content identifier from a client device; (c) selects an IP from the group; (d) sends the identifier to the web server using the selected IP; (e) receives the content; (f) returns it to the client.

Dependent-claim clusters (verbatim from Espacenet):

• 2–5: each group IP is associated with a respective client device → the request "steps out" through that client device (proxy-through-user design).
• 6–8: server communicates with / stores the client-device IPs; receives a "keep alive" message.
• 9: group IPs are for currently-online client devices.
• 10: establishing a TCP connection with each of those client devices.
• 11–13: server sends keep-alive and waits for responses; removes non-responsive IPs; periodic sending.
• 14–17: selection according to a criterion — numerical closeness to the web-server IP, geographic location, or response time.
• 18+ (partial): software application with a version number pushed to a client device (auto-update).

2. Most relevant prior art

(A) The ex parte reexamination references — the references the Office actually applied

These are from the third-party Request for Ex Parte Reexamination (Control Nos. 90/014,624 & 90/014,827, granted filing date 2020‑12‑04; order granting mailed 2021‑01‑21), as reproduced in the file record (CourtListener Dkt. 73‑2, 2:20‑cv‑00188). The examiner rejected claim 1 and its dependents under pre-AIA 35 U.S.C. § 102(b) and/or § 103(a) (FOA mailed 2022‑06‑21).

Also appearing in the reexam rejection (partially retrieved): the FOA excerpt shows a rejection grounded on "…[U.S. Pub.] No. 2006/0212584 … in view of Kageyama (U.S. Pub. No. …)". I have the shape of that ground (a U.S. publication 2006/0212584 combined with a Kageyama U.S. publication) but not the full citation string — the number for the Kageyama reference was truncated and I am not going to guess it. Flag for verification against the FOA.

(B) IPR2021‑00122 petition references (partial)

From the Teso LT / Code200 / Metacluster / Oxysales petition (filed 2020‑10‑29; institution denied 2021‑04‑20), the petition exhibit list includes, among others, U.S. Patent No. 5,974,566 (Ault) and U.S. Patent No. 6,185,625 (Tso). I could not retrieve the ground-by-ground § 102/§ 103 table, so I am not assigning specific claims to these two references. What I can say with confidence: the petition challenged claims 1–5, 9, 14, 17, 20–22, 25 and 27–30 (16 claims) and was never instituted, so no prior-art reference here was ever adjudicated. (IPR2021‑00122 case page)

(C) References listed against a sibling — NOT confirmed for '511

A Justia page for the sibling patent US 10,637,968 (same family/spec) lists "Patent Art (189)" including: US 2007/0142036 A1 (Nokia), US 2012/0124173 A1 (IBM, Content Delivery Using Multiple Sources Over Heterogeneous Interfaces), US 2001/0054020 A1 (Sidestep), US 2010/0235438 A1 (Juniper, Variable Rate Media Delivery System), US 2014/0189802 A1 (Navteq/Here), US 2010/0085977 A1 (Cisco), US 2002/0133621 A1 (PeerApp/Zephyrtel, Method and System for Accelerating Receipt of Data in a Client to Client Network), US 2004/0107242 A1 (Microsoft, Peer-to-peer Content Broadcast Transfer Mechanism), US 8,719,430 B2 (Piece Future), US 2011/0264809 A1 (Philips, Method for Exchanging Data), US 2009/0217122 A1 (Saturn Licensing), US 2008/0125123 A1 (Motorola). These are citations on US 10,637,968, not on the '511 patent. I am listing them only as candidate art for a fresh challenge, expressly flagged as unverified for '511.

3. § 102 mapping summary (claim 1 and its clusters)

Bottom line on anticipation: the only reference retrieved that plausibly anticipates claim 1 alone is Crowds (Reiter & Rubin, 1998) — it maps to every element of claim 1 (group of IPs belonging to other users; selection of one IP; submission of the request to the web server via that IP; return of content). The remaining references (Luotonen, RFC 2187, Squid, Wessels & Claffy, RFC 791, Gupta '374) are best characterized as § 103 combination art, which is exactly how the examiner used them (the FOA rejected claims 1–5, 9, 11–12, 14, 17, 20–22, 24–30 under § 102(b) and/or § 103(a), while confirming 6–8, 10, 13, 15–16, 18–19 and 23).

4. What I could not verify — do not treat as settled

1. The face-of-patent "(56) References Cited" list. Not retrieved. The list above is not the printed front-page citation list; it is the reexam + IPR + sibling-derived set. Pull the granted PDF front page / App. 16/278,109 file wrapper.
2. US 6,701,374 (Gupta) — number is confirmed as an SNQ reference; title and content are not confirmed in my retrieved excerpts.
3. The Kageyama reference number in the FOA rejection — number truncated in the retrieved excerpt.
4. The complete ground-by-ground table for IPR2021‑00122 (Ault '566, Tso '625 assigned to specific claims) — not retrieved; the petition was never instituted.
5. Applicability of pre-AIA vs. AIA § 102/103. The reexam applied § 102(b)/§ 103(a) to pre-2008 printed publications, consistent with pre-AIA law (all claims carry a pre‑2013 effective filing date via the 2009 priority chain). Confirm against the FOA before relying on any AIA-specific § 102(a)(1)/(a)(2) analysis.
6. The previously-generated Patent summary's statement about multiple independent claims (1, 9, 14, 17, 20, 25) is contradicted by both Espacenet and the reexamination FOA and should be corrected to "claim 1 is the only independent claim."

I'll research the prior art on this patent's face and in the IPR2021-00122 petition, then analyze the § 103 combinations.

The Espacenet claims page gave me the actual granted claim text, which differs from what the earlier summary inferred. Let me pull the remaining claims and the prior art of record.

Critical finding: the reexam request relied on Luotonen, "Web Proxy Servers" (Prentice Hall, 1998). Let me pull the full prior-art list and the remaining claims.

§ 103 Obviousness Analysis — U.S. Patent No. 10,484,511 B2

0. Threshold correction: the claim record in the earlier sections is wrong, and it changes the entire § 103 analysis

Before any obviousness analysis, I have to flag a direct contradiction between the earlier-generated "Patent summary" section and the actual granted claims, because it would send a § 103 analysis off in the wrong direction:

The "Patent summary" section reconstructed the claims from the specification's Summary of Invention and Abstract — but those passages are not the granted claims. The granted claim 1 (quoted verbatim from the Espacenet claims page, https://si.espacenet.com/publicationDetails/claims?CC=US&NR=[10484511B2](/patent/10484511B2)&KC=B2&FT=D) reads:

"1. A method for fetching, by a first client device, a first content identified by a first content identifier and stored in a web server, for use with a first server that stores a group of IP addresses, the method by the first server comprising: receiving, from the first client device, the first content identifier; selecting, in response to the receiving of the first content identifier from the first client device, an IP address from the group; sending, in response to the selecting, the first content identifier to the web server using the selected IP address; receiving, in response to the sending, the first content from the web server; and sending the received first content to the first client device, wherein the first content comprises a web-page, an audio, or a video content, and wherein the first content identifier comprises a Uniform Resource Locator (URL)."

This is confirmed independently by the Google Patents "Prior art keywords" field for US10484511 — server, client, address, client device, internet — which is the vocabulary of an address-selection proxy claim, not a chunk/checksum claim. It also matches the observation buried in the earlier "Assignment history" section ("the asserted claims recite per-request selection of a client-side IP address from a stored group and relaying content through it"). Two of the earlier sections disagree with each other; the Espacenet claim text controls.

Consequence: the specification's chunk/checksum/agent/peer scheme (and therefore BitTorrent art, the Cohen reference, ICP cache-miss art for chunks, etc.) is not what claims 1–18 recite. A § 103 attack must be aimed at egress-IP selection on a per-request basis at an intermediary server. Claims 19–30 were not retrieved verbatim (see § 9).

1. Governing framework, critical date, and level of ordinary skill

Statute. The reexamination file is explicit: the Office recorded "AIA (FITF) Status: No" (PTOL-471G, Control Nos. 90/014,624 & 90/014,827; https://storage.courtlistener.com/recap/gov.uscourts.txed.[198166](/patent/198166)/gov.uscourts.txed.198166.73.2.pdf). The reexam applied pre-AIA 35 U.S.C. §§ 102(b) and 103(a). The analysis below therefore uses pre-AIA § 103(a) and the Graham v. John Deere framework as refined by KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007).

Critical date. Priority runs to provisional 61/249,624, filed Oct 8, 2009. The § 102(b) critical date is therefore Oct 8, 2008. Every reference relied on by the Office or by the IPR petitioner predates it (Luotonen 1998; Crowds Nov 1998; RFC 2187 Sept 1997; RFC 791 1981; Wessels IEEE Apr 1998; "The Squid Book" 2004; Gupta US 6,701,374, issued 2004; Ault US 5,974,566; Tso US 6,185,625; Cohen US 6,389,462). A patent owner seeking a pre-2009 invention date carries the burden of swearing behind; absent a § 1.131 declaration, treating Oct 8, 2008 as the critical date is the conservative planning assumption.

Level of ordinary skill (POSITA). I construct this as: a bachelor's degree in computer science, computer engineering, or electrical engineering (or equivalent), plus 2–3 years of experience designing, configuring, or operating networked intermediary systems (HTTP proxies, caches, load balancers, or content-distribution systems), with working knowledge of HTTP/1.1 (RFC 2616), TCP/IP (RFC 791), and inter-cache protocols (ICP). Flag: this is my construction, not the parties'. The sibling IPRs in this family used expert declarations (Freedman for petitioner in IPR2021-00122; Rhyne and Dunham for the patent owner) to establish a level of skill, and the exact formulation should be pulled from those papers before it is relied on.

2. Prior art of record (provenance and confidence)

Note on the assignment of references to the two challenges: IPR2021-00122 (Teso/Code200/Metacluster/Oxysales v. Luminati, filed 2020-10-29, institution denied 2021-04-20, https://portal.unifiedpatents.com/ptab/case/IPR2021-00122) challenged claims 1–5, 9, 14, 17, 20–22, 25, 27–30. The ex parte reexaminations (90/014,624 & 90/014,827, filed 2020-12-04) requested review of claims 1–5, 9, 11, 12, 14, 17, 20–22, 25–30. The two proceedings share a core but are not identical — the reexam added claims 11–12 and the examiner also reached claim 24.

3. Claim-feature map for the grounds below

Claim 1 breaks into: (a) first server stores a group of IP addresses; (b) receives a URL from the client device; (c) selects an IP address from the group in response to the request; (d) sends the URL to the web server using the selected IP address; (e) receives the content; (f) sends it back to the client. Dependent claims add: 2 (each IP is a respective client device's address), 3 (forward the URL to that client device), 4 (that device sends to the web server and receives content), 5 (first server receives content from that device), 6–8 (communicate with / store addresses of / receive 'keep alive' from the devices), 9 (group members currently online), 10 (TCP connection to each), 11–13 (send 'keep alive', wait for response, purge non-responders, periodic), 14–17 (selection per a stored criterion: numerical closeness to the web-server IP / geography / response time), 18 (send a version number of a software application to a client device).

4. Ground 1 — Crowds in view of Luotonen (claims 1–5, 9)

This is the strongest § 103 ground, and it has a powerful track record against this family.

The combination. Luotonen supplies the single-actor architecture the preamble requires ("the method by the first server comprising"): an HTTP proxy server that receives a request identifying content, forwards it to the origin web server, receives the response, and returns it to the requesting client. Luotonen's proxy arrays additionally supply the idea of a controller selecting among a set of downstream proxies per request — that is the "(c) selecting an IP address from the group" step.

Crowds supplies the element Luotonen lacks: using another participating device's IP address as the egress identity for the request. In Crowds, a request is submitted to the end web server from a jondo, so the web server sees the jondo's address — which is precisely limitation (d) ("sending the first content identifier to the web server using the selected IP address"). Crowds also supplies limitation (a): the blender/central node stores the set of member addresses, and claim 9's "currently online" (Crowds maintains the live membership list). Claims 3–5 map directly onto Crowds' jondo behavior: the request is passed to the node addressed by the selected IP, that node sends the identifier to the web server, receives the content, and returns it.

Motivation to combine (KSR).

1. Same field, same problem. Both references address interposing a non-originating network node between a user and a web server. Unification of an anonymity overlay (Crowds) with a proxy-array control plane (Luotonen) is a combination of elements from the same art, addressing the same need.
2. Known, desirable design choice. Crowds pushes routing decisions to the (thick) client jondos and must distribute the member list to every jondo. Centralizing the selection in the intermediary node is a recognized, beneficial simplification: it enables per-request policy (rotation of egress identity), removes the need to replicate the membership list and routing logic at every client, and allows the intermediary to enforce the selection criterion. KSR expressly credits "design incentives" and "market demand" as motivation.
3. Predictable result, low technical risk. Both are software functions over standard HTTP/TCP; the combination is a rearrangement of known elements performing known functions with a predictable outcome, with no missing or unpredictable component. That is the core KSR rational.
4. The patent's own specification concedes the building blocks. The '511 Background admits that a proxy "requests the data from the Web servers … on their behalf" and that in P2P systems "files are stored on computers of consumers … Each consumer can serve up data to other consumers." Those admissions frame the invention as a combination of the two known architectures, which supports motivation rather than defeating it.

Note on why this matters beyond the '511. The earlier "Litigation summary" section records that the Federal Circuit, in the Aug 1, 2025 consolidated judgment (Nos. 2023-2144 et al.) concerning US 10,257,319 / 10,484,510 / 11,044,342 / 11,044,344, resolved "role-based vs. hardware-based construction of 'client device' and 'second server'" and invalidated those sibling claims over "Crowds." The '511 claim set uses the same functional/role-based vocabulary ("first client device," "client device addressed by the selected IP address," "first server") — so the same reference, and possibly the same construction, is squarely available here. Caveat: the '511 patent was not part of that judgment, and the '511 claim 1 recites a different sequence (single-server egress-IP selection). The analogy is strong but is not a holding about the '511.

Predicted patent-owner rebuttals and their strength.

• "Crowds' blender doesn't receive the URL and select; the jondo does." — Real, but it is a location-of-function argument. Under KSR, moving a known function from one known node to another in the same architecture, with no change in the function performed, is a classic obviousness rationale. And Luotonen's proxy-array controller does receive the request and select.
• "Crowds' jondos are not the clients of the first server in the claimed sense (claim 2: each IP is 'associated with a respective client device')." — Crowds' jondos are end-user computers identified by their own IP addresses; that is exactly claim 2. This is the same role-based-vs-hardware-based construction fight the siblings lost.
• "The specification disparages proxies (cost, storage, dynamic content)." — Weak. The disparagement is aimed at globally deployed caching proxies needing Internet-scale storage, not at the claimed per-request egress selection. In re Fulton/KSR: criticism of the general approach, or of a different embodiment, is not a teaching away from the claimed combination.

5. Ground 2 — Ground 1 further in view of RFC 2187 (ICP) in view of Wessels and the Squid Book (claims 6–13)

What it adds. ICP defines a message set by which cooperating caches (i) discover and query one another, (ii) test availability, and (iii) disqualify non-responders — i.e., a keep-alive/liveness discipline over a maintained group of peer addresses. RFC 2187's ICP_PING/response mechanism and Squid's handling of unresponsive peers map onto claims 8, 9, 11, 12 and 13; the Squid Book and Wessels supply the operational configuration (periodic probes, timeouts, purge). Claim 10 (TCP connection established to each group member) is met by ordinary TCP connection establishment (RFC 791 + the Squid hierarchy, which holds persistent connections to peers/parents).

Motivation. Anyone maintaining a pool of selectable nodes — whether Crowds members or Luotonen proxy-array members — must avoid selecting a dead node; otherwise the very benefit of the pool (failover, distribution, anonymity) is defeated. Failure detection by periodic heartbeat is one of the most predictable solutions available, and ICP supplies exactly that mechanism for a pool of cooperating caches. This is the "finite number of identified, predictable solutions" scenario KSR says supports obviousness.

This is not hypothetical. The examiner's own Group I in the reexamination was "RFC 2187 in view of Wessels and further in view of RFC 791 and Squid," applied to claims 1–5, 9, 11, 12, 14 … (Order/office action, Control 90/014,624; https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1549278](/patent/1549278)/download-documents). The 2022-06-21 Final Office Action then rejected claims 1–5, 9, 11–12, 14, 17, 20–22, 24–30 under § 102(b) and/or § 103(a), while confirming 6–8, 10, 13, 15–16, 18–19, 23 (https://www.docketalarm.com/cases/PTAB/IPR2021-01492/NetNut_Ltd._v._Bright_Data_Ltd/docs/03-20-2023-Petitioner/Exhibit-1128-Ex_1128___Final_Office_Action_mailed_June_21,_2022,_Reexamination_Nos_90014,827__90014,624.pdf).

Interesting reverse-signal: the FOA confirmed claims 6–8 (communicate with the devices / store their addresses / receive keep-alive) and 10 (TCP connections), yet rejected 11–13 (send keep-alive / purge / periodic). That pattern suggests the examiner found a reference teaching a server-initiated probe-and-purge discipline (CP/ICP_PING-style) but did not find art teaching client-initiated keep-alive registration with the intermediary. That is a meaningful gap a challenger must fill, and it is precisely where the confirmed claims cluster.

6. Ground 3 — Grounds 1/2 further in view of CDN/peer-selection art (claims 14–17)

Claims 14–17 require the selection to be governed by a stored criterion: numerical closeness to the web server's IP (15), geography of the selected client device (16), or response time (17).

• Claim 15 (numerical closeness to the web server IP). Luotonen's proxy/cache-array routing (hashing/address-based routing among array members) plus RFC 2187/Wessels peer selection supply selecting among a group by address/network relationship. The additional motivation is topology: choosing an egress node numerically/network-near the origin server shortens the forwarding path and improves the client's experienced latency — a benefit the '511 specification itself asserts.
• Claim 16 (geographic location). Content-distribution art of the era expressly selects serving nodes by client/network geography — e.g., Akamai's Leighton & Lewin, U.S. 6,108,703 ("Global hosting system," 1998/2000) and U.S. 6,185,598 (Farber & Grier, "Optimized network resource location"). Both predate the Oct 8, 2008 critical date by a decade. (Caveat: these are not in the reexam or IPR record I retrieved; I am identifying them as art a challenger would need to add, and their disclosure should be verified against their specifications before being asserted.)
• Claim 17 (response time). Squid/ICP peer selection is explicitly RTT/response-time driven (Wessels & Claffy 1998), and Tso (US 6,185,625, in the IPR record as Ex. 1019) teaches per-request selection among service providers based on measured throughput/bandwidth. This limitation is very likely obvious on the record already.

Motivation for the criterion generally: the patent's own specification states the agent list is built by "finding the communication devices … whose IP address is numerically close to the IP of the destination Web server" and that "the list of peers for each chunk is sorted by geographical proximity to the requesting client." Those are the applicant's own statements that such criteria are the natural, known way to choose among candidate nodes — usable as an admission.

7. Claims that resist § 103 on the current record — and what art would be needed

The FOA confirmed claims 6–8, 10, 13, 15–16, 18–19, 23 as "patentable and/or confirmed." I will not claim a § 103 case exists on the present record for these; that would be irresponsible. The honest read:

• 6–8, 10, 13 were not reached by the art of record. A challenger would need art showing intermediary-server-side registration and maintenance of a pool of participating end-user devices, including client-initiated keep-alive/registration (claims 6–7, 8) and periodically solicited liveness with purge (13). Distributed-computing membership/heartbeat art (e.g., group-membership protocols, cluster heartbeat mechanisms, and peer-registration in volunteer/distributed-computing systems) is the natural hunting ground. I have not identified and verified such a reference in this session; do not assume one exists.
• 15–16 (closeness/geography) are straightforwardly obvious if the CDN/array references above are added to the record; their confirmation likely reflects the references actually before the Examiner, not the state of the art.
• 18–19 (sending a software version number from the server to a client device) is the weakest link in the patent and the most likely target of a new IPR. Automatic software update / version-check mechanisms were ubiquitous well before 2008 (and the '511 specification itself describes the initializer "provid[ing] the version number of the acceleration application software to the acceleration server," with the server returning an indication "whether there is a newer version"). A challenger should search pre-2008 software-distribution/auto-update art. I did not verify any such reference here — flagging the opportunity, not asserting the ground.
• 23 depends from the 20-family; unretrieved (see § 9).

8. Consolidated motivation-to-combine (KSR) analysis

Across all grounds, the same KSR rationales carry the day:

1. Combination of known elements, each performing its known function, yielding a predictable result. Proxy relay (Luotonen/RFC 2616), anonymity routing through others' devices (Crowds), inter-node liveness (RFC 2187/Squid), and destination selection criteria (Squid RTT; Tso throughput; CDN geography) are all known, and their union produces no result beyond the sum of those functions.
2. Finite, identified, predictable solutions. At the 2008 critical date there were a small number of recognized ways to make an origin server observe a different source IP for a request (open/anonymous proxy; proxy chain; onion routing; P2P relaying) and a small number of recognized ways to pick among a pool (round-robin, hashing/address-array routing, RTT, geography). Selecting the known solution from a finite set supports obviousness.
3. Design incentives. Centralized pool management, rotation, failover, and policy enforcement were all recognized advantages; nothing about the claimed arrangement required discovery, only engineering choice.
4. The applicant's own admissions. The '511 Background concedes proxies forward requests on clients' behalf and that P2P systems store/share data on consumers' computers; the Detailed Description concedes that the node list is built by IP numerical closeness and sorted by geographic proximity. These blunt the "unexpected result" and "no motivation" defenses.
5. No criticality of the claimed sequence. Nothing in claim 1 suggests that the "in response to the receiving … selecting" ordering or the "using the selected IP address" mechanism produces an unexpected technical effect; it is ordinary routing.

Secondary considerations. I have seen no evidence of record — no license-nexus evidence, no unexpected-results data, no industry-praise evidence — in the material I retrieved. The patent's commercial success (Luminati/Bright Data's proxy platform) and the family's assertion campaign are not evidence of non-obviousness without a nexus to the claimed subject matter, and here the commercial product plausibly reads on the admitted prior art (a proxy platform that rotates egress IPs).

9. Contradictions, gaps, and confidence

Contradictions with earlier-generated sections (both must be resolved before reliance):

1. Claim scope. As set out in § 0, the "Patent summary" section's claim overview (client/agent/peer; chunks; checksums; "keeps track of which client devices have received responses") is contradicted by the granted claim 1 text. The "Assignment history" section's characterization ("per-request selection of a client-side IP address from a stored group and relaying content through it") is the accurate one. Any prior analysis built on the agent/peer claim model should be re-run.
2. Independent-claim numbering. The "Patent summary" section speculated independent claims sit "around claims 1, 9, 14, 17, 20 and 25." From the retrieved text, claims 9, 14 and 17 are dependent (9 depends from 2; 14 from 1; 17 from 14). Only claim 1 is confirmed independent. Claims 20 and 25 are plausibly independent (the FOA rejects 20–22 and 25–30 as groups and confirms 23), but that is inference from the rejection grouping, not verified claim text.

Unretrieved items I did not guess at: verbatim claims 19–30; the complete text of the 2022-06-21 FOA and the specific claim-chart mappings; the reexam request's SNQ Groups II–VII (my retrieval was truncated after Group I); the disclosures of Gupta US 6,701,374, Ault US 5,974,566 and Cohen US 6,389,462; the CAFC docket 26-1836 contents; and the allegedly reported 2025-11-25 PTAB affirmance (sourced only to Oxylabs' published timeline — treat as a lead, not a fact).

Confidence:

• High: the granted claim 1 text and its limitation structure; the identity of the prior art of record (Luotonen, Crowds, RFC 2187, RFC 791, Wessels, Squid Book, Gupta, Ault, Tso, Cohen); the examiner's Group I ground; the FOA's rejected/confirmed claim split; that pre-AIA § 103(a) governs and the critical date is Oct 8, 2008.
• Medium: that Crowds discloses the egress-IP-for-a-web-request mechanism as needed for limitation (d); the precise teaching of Tso '625.
• Low / not established: that any reference of record teaches claims 8, 11–13's client-initiated keep-alive or claims 18–19's version-number messaging; the existence of a verified motivation statement in the prior art itself (as opposed to a KSR-supplied rationale).

Practical bottom line. The Crowds-based § 103 attack is the strongest and is the one that has already worked in this family — the same reference invalidated the sibling '319/'510/'342/'344 claims at the Federal Circuit in the Aug 1, 2025 consolidated appeal, and the '511's own claims are drafted in the same role-based language. On the existing administrative record, claims 1–5, 9, 11–12, 14, 17, 20–22 and 24–30 stand rejected — first by the examiner's RFC 2187/Wessels/RFC 791/Squid grounds and any parallel Crowds/Luotonen grounds in Groups II–VII, then (per an unverified report) affirmed on appeal. But before any of this is relied on: (i) pull the granted claims 19–30 and the complete FOA/Luotonen/Crowds claim charts from Patent Center (controls 90/014,624 / 90/014,827) and PTAB E2E; (ii) confirm whether a reexamination certificate has issued; and (iii) confirm the status of CAFC 26-1836 — because if the certificate has canceled the asserted claims, the § 103 question is largely academic for enforcement purposes, while remaining live for the confirmed claims (6–8, 10, 13, 15–16, 18–19, 23) and for the rest of the continuation family.