US 7552870

I'll search for this specific patent number and its litigation/docket status, being careful not to conflate it with similar numbers.

The litigation search returned substantial district-court results but no CAFC dockets yet. Let me search specifically for the patent's claims and any Federal Circuit activity.

Summary: U.S. Patent No. 7,552,870 ("Trading network resources")

Search note on scope

I searched for the specific number 7,552,870 / 7552870 (application 11/378,500) only. No results for similar numbers (e.g., 7,552,8xx variants) were relied upon.

Bibliographic data (confirmed)

Abstract (verbatim)

A Network Resource Trading Exchange which allows Access Gateway Operators (who are usually also Network User Device Users) to trade usage of Network Resources (e.g. Internet access) via their Access Gateway for usage of Network Resources via other Access Gateway Operators' Access Gateways. Owners of Internet-connected wireless Access Gateways are enabled to trade Internet access via their own infrastructure for roaming access via other Access Gateways anywhere in the world. Access Gateway Operators can use their underutilized Network Resources to pay for Network Resource usage elsewhere, without having to purchase any subscription or pre-payment plan.

Plain-language overview of the invention

The patent describes a centralized "Network Resource Trading System" that lets operators of network access gateways (e.g., private Wi-Fi routers/hotspots) earn credits by providing network access to other users, and then spend those credits to obtain access through other operators' gateways when roaming — effectively a barter/credit exchange for Internet access that avoids conventional per-session payment, subscriptions, or prepaid plans. Key mechanisms include: registered user/gateway/device records with shared-secret authentication; authorization requests and real-time authorize/deny replies (distinguishing it from prior-art buyer/seller matching exchanges such as ANIP U.S. Pat. No. 6,005,926 and Arbinet U.S. Pat. No. 6,731,729); "first type" (user/spending) accounts vs. "second type" (gateway-operator/earning) accounts; credit reservation at session authorization; configurable use of earned credit "First / Last / Never"; and balance adjustment at session termination.

Independent claims — overview and important caveat

Uncertainty disclosure: The full patent text provided to me (Google Patents rendition) does not include the Claims section — it ends mid-way through the Detailed Description. I therefore cannot quote the independent claims verbatim or authoritatively enumerate all of them from the primary source. The claim-level information below comes from secondary litigation/analysis sources (case-complaint analyses for Reframe Technologies LLC v. Tesla Inc., 7:25-cv-00258, and Reframe Tech LLC v. Lucid Group, Inc., 1:26-cv-00331), which characterize claim 1 as representative of the method claims:

• Claim 1 (representative method claim, per litigation analysis): A method of trading network resources comprising: storing data in one or more first type accounts for a network resource user; storing data in one or more second type accounts for a network resource access gateway operator, wherein credit on a second type account acts as credit on a first type account; receiving from a requester a request to authorize provision of network resources, the request identifying a specific first type account and a specific second type account; sending a reply authorizing or denying provision of the resources, wherein the decision is at least partially dependent on the balance of the identified first type account; and adjusting the balances of the identified first and second type accounts based on the network resources used.

The '870 patent's specification also contemplates system/AAA-style embodiments (database tables for users, access gateways, network devices, account records, and sessions-in-progress; message protocols for Request/Response/Terminate-Session). I could not verify from authoritative sources whether additional independent system claims exist or how many independent claims total — if you need the exact claim set, the USPTO Patent Center / full-text PDF of US 7,552,870 (Claims section) would be the authoritative source.

CAFC / Federal Circuit 2026 docket search

No Federal Circuit (CAFC) appeal involving 7,552,870 was found in my searches of 2026 docket coverage. All litigation located is at the district court level, filed 2025–2026 by Reframe Technologies LLC against numerous defendants (e.g., Marriott International, PayPal, Tesla, Trax Technology Solutions, BC Brands, The Bernard Group, Lifetime Brands, Lucid Group, MOO Inc., Toad & Co., Odacite). Several of these cases show early resolutions — e.g., Marriott (W.D. Tex. 7:25-cv-00251) dismissed with prejudice via joint stipulation (Oct. 2025); Lucid (D. Del. 1:26-cv-00331) voluntarily dismissed (April 2026). I did not locate any notice of appeal or CAFC docket number for this patent as of today's search date; absence of a hit does not definitively prove none exists, so treat this as "not found in the searched sources" rather than a confirmed negative.

I'll research the litigation involving US patent 7552870 across the available docket and litigation databases.

Good progress. Let me now search for the remaining specific cases I haven't yet confirmed the parties/status for.

I've reached the search limit for this session. Based on the docket and litigation sources retrieved (RPX Empower, PacerMonitor, Justia, Ex Parte complaint analyses, IP Verse, PatSnap, Unified Patents litigation data as mirrored on Google Patents), here is the known litigation involving US patent 7,552,870 ("Trading network resources").

Confirmed cases (all brought by Reframe Technologies LLC, the current owner/assignee)

Cases listed in litigation databases but with unconfirmed details

Notes and caveats

• No Federal Circuit / CAFC appeal involving 7,552,870 was found in any searched source; all identified litigation is at the district-court level. (Absence of a hit is not a confirmed negative.)
• Litigation pattern: Reframe Technologies LLC (assignee of record as of the Oct. 15, 2025 recorded assignment; New Mexico LLC; represented by Rabicoff Law LLC, and in Delaware by Silverman, McDonald & Friedman / Brian Lutness) filed a first wave of W.D. Tex. cases in June 2025, a second wave (S.D.N.Y./E.D. Tex.) in October 2025, Delaware cases in late Dec. 2025, and further Delaware/S.D.N.Y. cases in March–April 2026. Several cases resolved quickly by dismissal (Marriott — with prejudice; Lifetime Brands — with prejudice; Lucid — voluntary).
• Sources: RPX Empower (litigation.rpxcorp.com), PacerMonitor (cdn.pacermonitor.com), Justia Dockets (dockets.justia.com), Ex Parte complaint analyses (ai-lab.exparte.com), IP Verse (ipverse.greyb.com), PatSnap (patsnap.com), and Unified Patents litigation data as reproduced on the Google Patents page for US7552870B2.
• If you need the unconfirmed defendants for 2:25-cv-00589, 1:26-cv-00332, and 1:26-cv-00334, the authoritative source would be the Unified Patents litigation portal (portal.unifiedpatents.com/litigation/caselist) or PACER directly, which I was unable to query further within this session's search limits.

Based on the canonical structured data block (USPTO Open Data Portal: no AIA trial proceedings on file) plus four independent web-search passes (PTAB proceeding-number formats, petitioner/patent-owner names, "Final Written Decision"/"institution decision" strings, and reexamination control numbers), I can now report with high confidence.

Proceedings overview

There are zero AIA trial proceedings — no IPRs, PGRs, or CBMs — ever filed against US 7,552,870 (breakdown: 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denied), which means the patent is entirely untested before the PTAB and no claim has been canceled, narrowed, or even formally construed by the Board — a defendant facing assertion today gets no FWD to leverage, but also faces no § 315(e)(2) estoppel and no Board-issued adverse constructions from any prior petitioner.

Per-proceeding details

There are no proceedings to itemize. Verified search results supporting this conclusion:

• USPTO ODP structured data (canonical): No AIA trial proceedings returned for this patent as of the most recent ingest.
• Ex Parte complaint analysis, Reframe Tech LLC v. Tesla Inc. (7:25-cv-00258, filed 2025-06-02): expressly states "The complaint does not mention any prior litigation, inter partes review (IPR) proceedings, or licensing history related to the patent-in-suit" (ai-lab.exparte.com/case/dct/txwd/7:25-cv-00258/doc/analysis/1).
• Searches for IPR/PGR/CBM proceeding numbers (IPR20xx-xxxxx / PGR / CBM patterns) against "7,552,870," "7552870," "Reframe Technologies," and inventor "Adrian Jones": no hits.
• Searches for PTAB Final Written Decisions / institution decisions citing the patent: none located. All litigation found is at the district-court level (2025–2026 Reframe Technologies LLC wave: Marriott, PayPal, Radisson, Tesla, Lifetime Brands, Trax, BC Brands, Bernard Group, Lucid, Odacite, MOO, Toad & Co.), with several early dismissals and no Federal Circuit appeal of any PTAB decision — consistent with there being no PTAB decision to appeal.

Strategic summary

Claims status: CANCELED — none. SUSTAINED — none. UNTESTED — all of them. Because no petition has ever been filed, every claim of the '870 patent (the litigation sources describe independent claim 1 as a representative method claim covering first-type/second-type accounts, authorize/deny replies, and balance adjustment; the full claim set was not verifiable from the truncated Google Patents rendition — see the earlier caveat in the patent summary) remains in full force and has never been subjected to PTAB merits review, claim construction, or a § 102/§ 103 patentability determination. This is a double-edged sword for a defendant: you cannot point to any Board cancellation, but you also inherit a completely clean slate — no prior petitioner has locked in an unfavorable construction, and no estoppel of any kind attaches from PTAB practice on this patent.

Estoppel and available grounds. With zero IPRs, § 315(e)(2) estoppel is a non-issue — no ground has been raised or "reasonably could have been raised" by anyone before the Board, so every prior-art ground remains fully available in district court. The operative constraints are statutory timing, not estoppel: (i) § 315(b) bars IPR institution if the petition is filed more than one year after the petitioner/RPI/privy is served with an infringement complaint — a defendant served in the June 2025 wave has already passed that bar (≈June 2026); defendants served later (Dec 2025 wave → ≈Dec 2026; Mar–Apr 2026 wave → ≈Mar–Apr 2027) still have an open window, and any newly served defendant gets a fresh one-year clock; and (ii) forum constraints — because the '870 patent claims priority to a 2006-03-16 filing (pre-AIA), PGR is statutorily unavailable (post-March 16, 2013 effective-filing-date requirement), and the CBM program sunset on 2020-09-16 — so IPR is the only AIA trial vehicle that could ever be brought, and none has been. Also note IPR cannot reach § 101, so an Alice eligibility attack would have to be made in district court regardless.

Pattern signals. There is no repeated-petitioner pattern, no PTAB litigation by the patent owner to study, and no defensive-aggregator IPR (e.g., Unified Patents) in the chain — which is itself notable: Unified Patents and similar actors have a track record of challenging heavily-asserted NPE patents, and the absence of any such petition despite ~13 district-court cases across 2025–2026 suggests either (a) the available prior art against this 2006-era method claim is perceived as weak or risky under Fintiv-style discretion, (b) defendants are prioritizing § 101 / early settlement over the cost of an IPR, or (c) cases are resolving (dismissals with prejudice in Marriott and Lifetime Brands; voluntary dismissal in Lucid) before the one-year IPR window is meaningfully used. Those are hypotheses, not findings — the only verified fact is that no petition exists. The complaint analyses' silence on IPR history in every reviewed case corroborates that the patent has no PTAB baggage.

Recommended next steps

1. If you are a defendant: There is no FWD to quote and no canceled claim to lean on — do not represent to a court or adversary that this patent has been "weakened" at the PTAB; it has not. The absence of PTAB activity is affirmatively usable only in the sense that no § 315(e)(2) estoppel exists and no Board claim construction binds you.
2. Check your § 315(b) calendar immediately. If you were served within the last 12 months, an IPR window may still be open (deadline = one year from your service date, not from the first case in the wave). If you were served before ≈2025-09-04 and have not filed, you are likely time-barred — verify against your service date before spending on a petition.
3. If an IPR is still available to you, build the petition knowing the Board's current discretionary-denial posture (the 2025–2026 Stewart-era factors penalize petitioners who knew of the patent and delayed, and require claim-construction positions harmonized with any parallel district-court litigation — a real risk here given active Reframe cases in D. Del., E.D. Tex., W.D. Tex., and S.D.N.Y.).
4. If you are time-barred or choose not to petition, your strongest non-PTAB levers are: (a) § 101 / Alice — unreachable in IPR, fully available in court, and a natural fit for a method claim of this abstraction level; and (b) § 112 / indefiniteness and claim construction, none of which has been tested anywhere — the PatSnap analysis of the Marriott case notes there has been "no court-issued claim construction," so the claims stand at their full filed breadth, making a Markman-stage narrowing motion a high-value target.
5. Monitor, don't assume: Re-verify against the USPTO PTAB E2E portal (developer.uspto.gov/ptab-web) and the ODP API before any filing deadline, since a petition could be filed by another defendant tomorrow — and note that institution of an IPR by anyone within your own § 315(b) window does not reset your clock; you must file your own petition to get your own estoppel shield.
6. Do not cite any PTAB proceeding in your invalidity contentions — there are none to cite, and padding a contentions chart with nonexistent IPR history would be sanctionable. Ground your § 102/§ 103 case in the actual prior art (the '870 specification itself distinguishes ANIP U.S. Pat. No. 6,005,926 and Arbinet U.S. Pat. No. 6,731,729 — start there, plus the RADIUS/DIAMETER/CRANE/CIBER/TAP/IPDR standards literature the specification concedes as background).

Inventors

Original assignee

• Entity named on the issued patent: None — the USPTO/Google Patents record lists the Original Assignee as "Individual" (Adrian Jones). The patent issued to the inventor personally on 2009-06-30.
• Product embodying the claims: N/A — an individual inventor; no product, no operating company in the chain at any point.
• Line of business: The inventor's own background is not disclosed in any source I reached. The invention itself (a network-resource trading exchange for Wi-Fi gateway operators) was never commercialized by any assignee of record.
• Current status: The inventor assigned the patent out on 2025-10-15 to Reframe Technologies LLC and is no longer the owner of record.

Assignment timeline

I could not reach the USPTO Assignment Center's search interface directly within this session (web-index limitations), and no source I reached published the reel/frame number or the correspondent of record for the 2025 transfer. Reel/frame and correspondent fields below are therefore marked unknown — do not treat the blanks as verified. What follows is limited to the assignment event affirmatively shown in the Google Patents legal-events block (which mirrors USPTO assignment data) and the litigation docket record.

• Executed: unknown (likely 2025) / recorded 2025-10-15 — Reel/frame not available in searched sources (verify at https://assignmentcenter.uspto.gov/ or https://assignment.uspto.gov/patent/index.html by patent number 7552870 / application 11/378,500)
  • Conveyance: Assignment of Assignor's Interest (per Google Patents legal event: "ASSIGNMENT OF ASSIGNOR'S INTEREST"; Assignors: JONES, ADRIAN PAUL)
  • Assignor: Adrian Paul Jones (inventor)
  • Assignee: Reframe Technologies LLC
  • Correspondent: Unknown — not disclosed in any source reached. Note: Isaac Rabicoff / Rabicoff Law LLC appears as litigation counsel for Reframe across the 2025–2026 cases, but that is not evidence that Rabicoff was the recording correspondent; do not conflate the two without the Assignment Center record.
  • Context: Transfer-to-asserter — after 16 years of dormant personal ownership, the sole-inventor patent moved to an LLC that had already begun (or immediately began) a multi-defendant assertion campaign (see NPE signals below).

No other recorded assignments were found. The Google Patents event history shows no assignment at filing, no assignment between issuance (2009) and 2025, and no licenses, security agreements, mergers, or changes of name. If the Assignment Center search confirms this, the chain is exactly two links: inventor (2006–2025) → Reframe Technologies LLC (2025–present). Given that my search was limited to indexed sources, a direct Assignment Center query is the authoritative check for additional recorded documents (e.g., a pre-suit execution date that predates the 2025-10-15 recordation).

Timeline diagram

timeline
    title Ownership of US 7552870
    2006 : Filed by Adrian Jones
    2009 : Patent issued to inventor
    2025 : Assigned to Reframe Technologies LLC
         : First infringement suits filed
         : Two cases dismissed with prejudice

NPE / troll-pattern signals

1. Shell-entity transfer — present (weak-to-moderate evidence). The patent moved from a non-operating individual (no products, no business of record) to a single-purpose LLC (Reframe Technologies LLC) that sells no product and exists to assert the patent. The "LLC with no products + immediate litigation" evidence is concrete; the "registered-agent address / single-member structure" details are unclear — no source I reached published Reframe's formation documents, member list, or registered agent. PatSnap's litigation database independently characterizes Reframe as a "Patent assertion entity — holder of US7552870B2." Supporting dates: assignment recorded 2025-10-15; first infringement complaints June 1–2, 2025 (W.D. Tex. 7:25-cv-00251/00255/00256/00258).

2. Known asserter in the chain — present. Reframe Technologies LLC matches the public NPE/PAE definition as reflected in litigation databases: PatSnap labels it a patent assertion entity, and the docket record shows 13+ cases across W.D. Tex., E.D. Tex., D. Del., and S.D.N.Y. (2025–2026) against commercially unrelated defendants — Marriott, Radisson, PayPal, Tesla, Lucid, Lifetime Brands, Trax, BC Brands, Bernard Group, MOO, Toad & Co., Odacite. Defendants spanning hospitality, EV, payments, retail, and cosmetics is the signature of a broad PAE campaign, not competitor enforcement. (Reframe does not appear on the specific Acacia/Marathon/IV-type rosters I checked, but the multi-district campaign plus third-party PAE characterization is sufficient.)

3. Repeat correspondent across the chain — unclear / not present. Only one recorded assignment exists in the chain, so there is no recurrence to establish. The recording correspondent is unknown; Rabicoff Law LLC is litigation counsel only. No finding either way without the Assignment Center correspondent field.

4. Cascading transfers — not present. The chain is a single hop (inventor → Reframe). No chained LLCs, no <24-month cascade.

5. Pre-litigation transfer — present with a timing nuance. The recordation date (2025-10-15) actually postdates the first complaints (June 2025), which means one of two things: (a) the assignment was executed shortly before the June 2025 filings and recorded ~4 months later, or (b) Reframe sued before recordation and perfected standing later. Either way, the transfer was arranged to enable assertion — the patent sat unused for 16 years and moved to the asserting entity precisely as the campaign began. The execution date (unavailable) is the fact that would pin this down; verify on the Assignment Center record.

6. Bankruptcy fire-sale — not present. No bankruptcy anywhere in the chain; the assignor is an individual, not a failed operating company.

7. Privateering — not present. No operating company transferred the patent to an NPE to attack competitors; the assignor is the original inventor. This is classic inventor-to-NPE monetization, not privateering.

8. Defensive aggregator (anti-NPE) — not present. The chain terminates at Reframe Technologies LLC, an asserting entity. No RPX, AST, LOT, Unified Patents, or OIN involvement.

Verdict

NPE — high confidence.

Two strong signals drive the call: (1) known asserter — Reframe Technologies LLC is characterized as a patent assertion entity by third-party litigation databases and has filed 13+ cases across four districts against commercially unrelated defendants within ~12 months of the 2025-10-15 assignment from the inventor; and (2) transfer-to-asserter / shell-entity transfer — a patent dormant for 16 years under personal inventor ownership moved to a no-product LLC timed to an assertion campaign, with no operating company, bankruptcy, or defensive aggregator anywhere in the chain. The principal data gaps are the reel/frame number and correspondent of record (USPTO Assignment Center only) and Reframe's formation details; neither gap undermines the conclusion, which rests on the recorded assignment event plus the verified litigation record.

Verification links: USPTO Patent Assignment Search — https://assignmentcenter.uspto.gov/ (search patent no. 7552870 or application 11/378,500); Google Patents legal-events mirror — https://patents.google.com/patent/[US7552870](/patent/US7552870)/en.

Prior-Art Analysis — US 7,552,870 B2 ("Trading network resources")

Important verification caveats (read first)

Before giving the analysis, I need to be transparent about what I could and could not verify, because it affects the reliability of every "which claim" conclusion below:

1. The full text of US 7,552,870 supplied for this analysis (Google Patents rendition) is truncated — it ends mid-Detailed-Description and does not contain the Claims section or the front-page "References Cited" list. I therefore cannot quote the actual claims verbatim or confirm the examiner's formal citation list from the primary source. (This was flagged in the earlier patent summary and remains unresolved.)
2. My search quota was exhausted after the searches run for this task. I could not reach the USPTO full-text database, the USPTO Assignment/Patent Center, the Google Patents "References Cited" section for US7552870B2, or the EP/CN/JP family search reports. All "dates" for the cited references below that are not directly quoted from the '870 specification are therefore approximate and flagged as unverified.
3. What I can confirm from the authoritative specification text (provided verbatim in the patent text): the Background of the Invention / Summary of the Prior Art section expressly names three U.S. patents and several industry standards as the prior art the invention distinguishes itself from. These are the only "citations" I can affirmatively tie to 7,552,870 without speculation about the examiner's front-page list.
4. The claim set used for the § 102 analysis is the representative claim 1 characterization from the plaintiff's own litigation analyses (Reframe Tech LLC v. Tesla, 7:25-cv-00258; Reframe Tech LLC v. Lucid, 1:26-cv-00331), which describe claim 1 roughly as: (a) storing data in "first type" accounts for a network resource user; (b) storing data in "second type" accounts for a gateway operator, where credit on a second-type account acts as credit on a first-type account; (c) receiving a request to authorize provision of network resources identifying a specific first-type and second-type account; (d) sending a reply authorizing or denying provision, the decision at least partially dependent on the balance of the identified first-type account; and (e) adjusting balances of the identified first/second-type accounts based on resources used. This is a complaint characterization, not the claim text — treat it as indicative, not authoritative.

Category A — Patent references cited in/against US 7,552,870

The three U.S. patents below are the only patent references I can verify as being tied to 7,552,870 — each is quoted verbatim in the '870 specification as prior art. Because the examiner's front-page "References Cited" block was not available to me, I cannot rule out additional examiner-cited references; verify at the links at the end.

A1. U.S. Patent No. 6,005,926 ("ANIP")

A2. U.S. Patent No. 6,731,729 ("Arbinet")

A3. U.S. Patent No. 6,526,390 (Wang & Gorrepati)

Category B — Non-patent literature discussed in the '870 specification

The specification concedes as background the AAA/accounting standards for wireless roaming:

• RADIUS — RFC 2865/2866
• DIAMETER — RFC 3588
• CRANE — RFC 3423
• CIBER, TAP, and IPDR (ipdr.org)

The '870 spec's own distinction: these "specify how accounting data is moved around a network but do not address the actual settlement of charges" and "do not incorporate the concept of trading Network Resources." § 102 anticipation potential: negligible alone for the trading-credit elements; they are § 103 building blocks (and, notably, the spec's own admission that AAA-style authorize/deny message flows were known is a double-edged sword for claim validity on the authorization elements — a defendant will argue the only novelty left is the credit-trading overlay).

Category C — References that cite 7,552,870 (not prior art — do not confuse)

My search results surfaced later documents that cite 7,552,870 / its publication US2007/0215684A1 (e.g., Barkan's wireless-Internet family US2010/0296441A1 and its Russian counterpart RU2483487C2; Wefi's US8,000,276B2 and family; Sun Microsystems' US7,490,225B2 in a RU translation). These are later-filed and therefore cannot be § 102 prior art against 7,552,870 (effective filing date March 16, 2006). They are useful only to show the technology's downstream evolution (credential-sharing, hotspot federations such as Wefi/TruConnect), which in turn can supply secondary-considerations or claim-scope context — not anticipation.

Bottom line on § 102 anticipation

The most honest one-line answer: no reference I can verify was cited in/against 7,552,870 appears to anticipate the full representative claim as characterized; U.S. 6,731,729 is the reference a defendant would press hardest on the credit-trading elements, and 6,526,390 + 6,731,729 in combination is the natural § 103 story — but none of this has ever been tested, because (per the prior sections) the patent has zero IPRs, no court-issued claim construction, and no FWD, and the actual claim text was not available to me in the primary source.

Recommended verification steps (to close the gaps)

1. Pull the actual claims and front-page "References Cited" of US 7,552,870 from the USPTO full-text database (PatFT/Patent Center) or the Google Patents "Claims/References" tabs — the supplied text omits both. The exact claim language is a prerequisite to any defensible element-by-element § 102 mapping.
2. Check the PCT/EP family search reports — the WO2007/106535A2 international search report and the EP2002392A4 supplementary search report likely contain the EPO/ISA examiner's X/Y citations, which often surface art more relevant than the three U.S. patents the applicant chose to discuss. (Priority chain: US 11/378,500 → PCT/US2007/006470 → EP 07772876 → JP 2009-500466 → CN 200780014457.6.)
3. Confirm citation dates for 6,005,926 / 6,526,390 / 6,731,729 on USPTO or Google Patents — all dates above not directly quoted from the '870 specification are unverified approximations.
4. If you are a defendant, note that the absence of PTAB/claim-construction history means § 102/§ 103 arguments here are untested and the claims stand at full filed breadth — the plaintiff's own complaint charts (e.g., Tesla, Ex. 2) and the PatSnap Marriott analysis are the only "maps" of claim 1 in existence, and they are advocacy documents, not constructions.

§ 103 Obviousness Analysis — U.S. Patent 7,552,870 ("Trading network resources")

0. Critical scope caveats before the analysis

Three honesty constraints shape everything below; I flag them rather than paper over them:

1. No verbatim US claim set is available in the materials. The Google Patents rendition of US7552870B2 provided for this analysis is truncated and does not include the Claims section. The only claim-level description I can ground on is the independent claim 1 characterization quoted in the Ex Parte complaint analysis for Reframe Tech LLC v. PayPal Inc. (7:25-cv-00255) and mirrored in the earlier sections of this file. I therefore analyze obviousness of the claim 1 features as characterized, not of a verified verbatim claim. The family publication WO2007/106535A2 (published claims 5–15, e.g., "applying network usage credits to a user based upon usage by at least one third party of a first network resource operated by said user; substantially immediately processing a request … for usage of a second network resource operated by a fourth party and allowing access … if said user has at least a predetermined amount of said network usage credits") provides a parallel, verified claim formulation that corroborates the same feature set.
2. No separate "Prior Art" search-results block was supplied with this task. The only prior-art inventory available is the one the specification itself admits in "SUMMARY OF THE PRIOR ART" (U.S. 6,526,390; RADIUS/DIAMETER/CRANE/CIBER/TAP/IPDR; ANIP U.S. 6,005,926; Arbinet U.S. 6,731,729) plus the commercial roaming-aggregator background (Boingo, GoRemote, T-Mobile, Wayport, WeRoam, Orange, BTOpenzone, iPass). My web searches for the patent's examiner-cited references returned no usable citation list, so the analysis below uses the specification-admitted prior art — which is legally significant, because a patentee's own admissions in the Background section can be used against it.
3. Reference details beyond what the '870 specification discloses are from general knowledge and are not independently re-verified this session. Where a reference's teaching matters, I state both what the '870 spec admits and what I otherwise know, so the reader can weight each leg of the mapping.

1. Legal framework

Obviousness under pre-AIA 35 U.S.C. § 103 (the '870 patent claims priority to a March 16, 2006 filing) is governed by Graham v. John Deere (383 U.S. 1 (1966)): (i) scope and content of the prior art; (ii) differences between the prior art and the claimed invention; (iii) level of ordinary skill; and (iv) secondary considerations. Under KSR Int'l Co. v. Teleflex Inc. (550 U.S. 398 (2007)), a combination of known elements "according to known methods" that "yields predictable results," a "design need or market pressure," and "obvious to try" reasoning all support a finding of obviousness — and a patentee's admission that each component existed in the prior art, with the only asserted advance being their recombination, is a textbook § 103 posture.

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

Based on the specification's own implementation details (a TCP/IP server with relational database tables; RADIUS-like request/response/notification protocol messages; shared-secret authentication; account/reservation ledger logic), the POSITA is a person with a B.S. (or equivalent experience) in computer science/engineering plus 2–4 years designing network AAA (authentication, authorization, accounting) and billing/settlement systems — i.e., someone familiar with RADIUS/DIAMETER, prepaid/postpaid rating and balance management, and clearinghouse settlement between access providers. That person would be equally at home in the "wireless hotspot billing" art and the "telecommunications bandwidth trading" art, which the '870 spec itself treats as the two relevant fields.

3. Prior-art inventory (as admitted/identified by the '870 specification)

4. Representative independent claim 1 (as characterized) — feature map

The mapping shows that no single feature is new; the only arguable novelty is the selection and wiring together of known components — the paradigm KSR obviousness posture.

5. Combination analyses

Combination A (strongest two-reference case): Arbinet '729 + Wang '390

Claim 1 elements supplied by Arbinet. Arbinet teaches an exchange in which a party that provides telecommunications/network resources earns a credit balance (an account receivable) and can apply that earned credit toward usage of network resources provided by others — i.e., fungibility between the "earning/operator" side and the "spending/user" side. Applied to a single multi-homed operator, this is the claim's "credit on a second type account acts as credit on a first type account" — the operator's gateway-earned credits (second-type, or "earned balance" in the spec's table 100/400 nomenclature) fund the operator's own roaming usage (first-type). The '870 spec itself concedes Arbinet "provides for credit balances … earned through provision of Network Resources to be traded for usage of Network Resources."

Claim 1 elements supplied by Wang. Wang teaches what Arbinet does not: a per-session, real-time authorization/denial decision for a specific roaming user on a specific foreign access gateway, where the decision is at least partially dependent on the sufficiency of the user's pre-existing credit/account, followed by settlement/charge adjustment after usage. This is precisely the "no matching requirement … authorization/denial component, which requires sending a reply to the requester, explicitly authorizing or denying provision of Network Resources by the seller" that the '870 spec identifies as its own point of departure from Arbinet/ANIP. Wang's roaming scenario (GSM, extendable per the spec to "other wireless technologies") supplies: the requester (gateway) asking a central system, an identified user account with a balance, an authorizing/denying reply, and post-session balance adjustment.

Why the differences over each reference are exactly what the other reference fills.

• Over Arbinet alone: Arbinet is an exchange that matches counterparties from a plurality; it does not disclose the immediate authorize/deny-and-settle transaction flow for an already-identified user/gateway pair. Wang discloses that flow.
• Over Wang alone: Wang is a pay-per-use roaming billing system — the user pays with a credit card, not with credits earned by operating a gateway. Arbinet discloses the earned-credit trading economy. A POSITA combining the two would simply ask: "why make the roaming user pay cash when, like Arbinet's sellers, he can pay with credits he earned by hosting other roamers?" Implementing Wang's authorization decision against an Arbinet-style earned-credit account balance rather than a credit card yields claim 1: first-type user accounts (balance checked), second-type operator accounts (credits earned), fungibility of operator credits into user spending power, request identifying both, authorize/deny reply gated on the user-account balance, and post-session adjustment of both accounts.

Motivation / reason to combine (articulated, not conclusory).

1. Same field, same players. Both references sit in the "charging/settlement for network resource usage" art that the '870 Background itself merges. A POSITA designing a roaming-hotspot settlement system in 2006 would have before him (i) the trading-exchange art (Arbinet/ANIP) that solved how to value and settle reciprocal provisioning without cash, and (ii) the AAA/billing art (Wang; RADIUS) that solved how to authorize a specific roamer in real time. The '870 patent's own statement of the problem — "turning … wireless Access Gateways into hotspots by providing an alternate means of authorization and settlement" — describes joining those two solutions.
2. Design need / market pressure. The Background complains that each BSP's coverage is "extremely sparse" and that per-user payment is a barrier. The obvious engineering response to "we need authorization + settlement without conventional payment" is to reuse Wang's authorization plumbing and swap Arbinet's credit-trading settlement into it. That is a KSR "known problem in the field … with a finite number of identified, predictable solutions."
3. Predictable result. Neither reference has to be modified in any surprising way: Arbinet's account/debit-credit ledger logic maps one-to-one onto Wang's session flow. The combination produces no new capability beyond the sum of its parts; it merely substitutes the medium of settlement (earned credit instead of cash/credit card) while keeping the authorization mechanics identical.

Combination B: Arbinet '729 (or ANIP '926) + RADIUS/DIAMETER AAA framework

Claim 1 and dependent-claim elements supplied by the AAA leg. The RADIUS/DIAMETER standards (admitted prior art in the spec) disclose the complete message architecture the '870 patent later re-describes in its own protocol: an Access-Request from a network access server (gateway) identifying the user/device; an Access-Accept/Access-Reject reply authorizing or denying the specific session; Accounting-Start/Interim/Stop messages, including a session-termination notification carrying usage (the analog of the spec's Terminate-Session Notification with Session Length); and subscriber account/balance data consulted before acceptance — all "readily available" server-side data enabling sub-second processing, which the spec itself says commercial DB/transaction software (Oracle, SQLServer) handles in "a fraction of a second."

Dependent claims. The family/WO claim set adds: database tables storing the account/user/gateway/device/session data (WO claims 6–7, 9–10, 12–13), and use of "a request for authorization for immediate provision …, a response to authorization …, and a terminate-session notification" data messages (WO claims 8, 11, 14; US dependent claims per the same family). These are standard relational-database design (the spec names off-the-shelf Oracle/DB2/SQLServer) and standard AAA message taxonomy (RADIUS Access-Request/Access-Accept/Accounting-Stop). There is nothing inventive in storing trading accounts in tables or wrapping the Arbinet credit model in RADIUS-style messages — a POSITA would do exactly that as a matter of routine implementation.

Motivation. The AAA standards are the lingua franca of hotspot/gateway authorization in 2006 (the spec concedes this). Once a POSITA adopts Arbinet's credit-trading settlement, the path of least resistance to deploy it over real access gateways is to express the credit-check and settlement as AAA request/response/accounting messages — no design insight required, only implementation.

Combination C: Wang '390 + Arbinet '729 + ANIP '926 (three-reference belt-and-suspenders)

ANIP is a weaker but independent source of the trading/settlement leg and would be cited where a challenger wants redundancy on "earned credits from providing services are applied to consuming services." The combination rationale is the same as Combination A; ANIP is interchangeable with Arbinet for the trading leg and would be relied on for its earlier date and independent disclosure of buyer/seller account settlement. Note the '870 spec treats ANIP and Arbinet as one undifferentiated prior-art cluster, which is itself an admission that the trading concept was well-settled by 2006.

Combination D (admitted-background route): BSP roaming clearinghouse + Arbinet credit trading

The spec admits the commercial aggregators (Boingo, iPass, T-Mobile, Wayport, WeRoam, Orange, BTOpenzone) already operated centralized BSP platforms that authenticate roaming users, authorize them on third-party hotspots, and settle between the user, the BSP, and the gateway operator. The only missing piece versus claim 1 is that settlement used "direct or indirect payment" rather than gateway-operator-earned credits. Combining that admitted architecture with Arbinet's admitted earned-credit trading (substituting an earned-credit ledger for the cash settlement leg) produces the claimed system. This is a useful fallback theory and also supports the "combination of known elements for predictable results" narrative, because it shows both halves of the combination were commercially practiced in the same industry by 2006.

6. Why the patentee's own "distinction" is not a patentable difference

The '870 spec's only articulated distinction over Arbinet/ANIP is that the prior art "involves matching suitable buyers and sellers from among a plurality … rather than operating as an authorization/denial process for a specific buyer/seller pair," and that the present invention "instead [has] an authorization/denial component, which requires sending a reply to the requester." On this record that is precisely the feature Wang '390 (and, structurally, RADIUS) already supplied in the roaming-billing context:

• Wang: credit-card authorization of a specific roamer on a specific foreign network, with an allow/deny decision and post-usage settlement — the claimed authorize/deny component, unmodified.
• Arbinet: reciprocal credit trading between providers and users — the claimed two-account fungibility, unmodified.

A POSITA combining them is not exercising invention; he is selecting the authorization engine from one known system and the settlement ledger from another known system in the same field and plugging them together, with predictable results. Under KSR, that is the archetype of obviousness. The claim's "at least partially dependent on the balance of the identified first type account" is also broad enough to cover Wang's credit-sufficiency check and AAA balance checks — the balance dependency is not a new kind of decision, only a new source of the balance being checked.

7. Secondary considerations (Graham factor 4)

On the available record there are no secondary indicia of non-obviousness:

• No commercial success: the patent sat dormant under personal inventor ownership for ~16 years (2009–2025) with no commercialization identified; no product, licensee, or standard adoption appears in the record.
• No long-felt need / praise / copying evidence: none identified; the litigation wave (2025–2026 Reframe Technologies suits) is an assertion campaign, not evidence of industry recognition of inventiveness.
• No unexpected results: the system is a straightforward clearinghouse; the spec claims speed only via off-the-shelf databases.

The absence of secondary indicia (combined with the fact that the only post-issuance activity is NPE litigation) leaves the Graham balance tilted firmly toward obviousness. (Caveat: this is a conclusion from an absence of evidence in the searched record, not a verified finding that no such evidence exists.)

8. Bottom line

The strongest § 103 case is Arbinet U.S. 6,731,729 in view of Wang U.S. 6,526,390 (optionally plus the RADIUS/DIAMETER AAA framework and/or ANIP U.S. 6,005,926), rendering the representative independent method claim — and, a fortiori, the routine dependent limitations (database tables; request/response/terminate-session data messages; credit reservation; balance expiry; earnings consolidation, all of which are conventional ledger/AAA mechanics described in the spec) — obvious to a POSITA. The motivation is concrete and contemporaneous: the industry had (i) per-session roaming authorization and settlement (Wang, RADIUS) and (ii) cashless reciprocal credit trading of network resources (Arbinet, ANIP); the '870 patent simply swaps the payment leg of the former for the credit leg of the latter in a centralized clearinghouse, a combination of known elements yielding predictable results under KSR.

Standing caveats for any use of this analysis: (1) confirm the actual claim language from the USPTO full-text PDF/Patent Center — the element mapping above tracks claim 1 as characterized in the 7:25-cv-00255 complaint analysis, not a verified verbatim claim; (2) the prior-art details for '729/'926/'390 beyond the '870 spec's own admissions are drawn from general knowledge and should be verified against the patents before filing anything; and (3) I found no record of any PTAB or court obviousness determination on this patent (zero IPRs; district-court cases still early), so this is a de novo theory, not a restatement of any adjudicated holding.