US 7031254

I'll verify against live sources before summarizing.

US 7,031,254 — Verification note

I searched for the exact number 7031254 and for CAFC 2026 activity tied to it. Findings:

• Patent document confirmed at Google Patents (https://patents.google.com/patent/US7031254/en) and Justia (https://patents.justia.com/patent/7031254).
• No CAFC 2026 docket entry, appeal, IPR/PGR, or district-court action for US 7,031,254 was returned by my searches. The searches surfaced only unrelated PTAB/Federal Circuit matters (e.g., SoundClear/Amazon, Motorola, Centripetal). I cannot confirm the absence of such proceedings — only that nothing appeared. Treat "no 2026 CAFC activity found" as a search result, not a certified negative.
• The Google Patents bibliographic block contains an anomalous line reading US 5744902 A next to the application number US10/057,449. Per your literal-interpretation rule I am not folding 5,749,902 into the file history; I flag it as a likely metadata/parse artifact (or a separate document) of unconfirmed significance.

---

Bibliographic summary (from the authoritative full text)

Ownership chain (assignment records): Lucent Technologies → Alcatel-Lucent USA → Provenance Asset Group LLC (2017) → RPX Corporation (recorded 2021-12-28, effective 2021-11-29), with security interests held/released by Nokia USA / Nokia US Holdings, Cortland Capital, and Barings Finance.

---

Abstract (as issued)

A rate control system for a link between a first node (e.g., an RNC) and a second node (e.g., a Node B) in a wireless communications system, where a user is given a rate over the link as a function of link load. The system sets at least one user to a reduced rate on the link as a function of the rates of a plurality of users on the link. Rate is controlled by selecting the size of a transport format block (TFB) used to transport a user's data over the link during a transmission interval; TFB size can be changed by changing the number of transport blocks used to form the TFB as a function of total user data to be transported at that time. Instead of always using the largest TFB (current UMTS practice), TFB size for a user is determined from the current status of all RLC buffers that have data to transmit; if there is more data than the Iub link can carry, smaller TFBs are used for some users so aggregate offered traffic stays below the sustainable rate.

---

Independent claim in plain language

Claim 1 — Method of sending data over an RNC→base station link for wireless transmission to two or more users. Steps:

1. Obtain a description of loading on the link.
2. Select a block size for each user from a discrete set of two or more available block sizes, the set including (a) a maximum block size corresponding to maximum data rate and (b) at least one reduced block size corresponding to a reduced data rate.
3. Map the data to blocks of the respectively selected sizes.
4. Send the blocks to the base station over the link.

Constraints in the "wherein" clause:

• the maximum block size is selected for at least one user;
• a reduced block size is selected for at least one user;
• block sizes are selected with reference to the link-loading description; and
• the number of users receiving reduced block sizes is determined with reference to the link-loading description.

Claims 2–5 are dependent (they add the sustainable-rate/R_s/N_L calculation of the FRC algorithm; reduction within a bearer rate; reduction within a bearer and priority class; and TFB mapping where the number of transport blocks is changed as a function of the TFB sizes of at least one other user). Only claim 1 is independent, so claim 1 defines the single independent scope asserted by the patent.

---

Supporting technical content worth noting

• The specification's Fair Rate Control (FRC) algorithm computes sustainable rate per user R_s = R*T/N, tests whether equivalent peak-rate users N_p > T, picks the largest TFB with R_TFj ≤ R_s, and solves for the minimum N_L satisfying equation (1). Rate control instances are triggered by RLC buffer state changes (RCE1/RCE2).
• Alternatives disclosed: Unfair Rate Control (URC) (a special case with R_TFj = 0, i.e., deferral); Approximate Max-Min Fair Rate Allocation over multiple bearer rates driven by lookup Tables 4–6; and a multi-priority-class extension (Table 7).
• The specification states the bearer rate stays the same for a user for the duration of a connection, characterizing the control as intra-bearer-rate rate throttling (which may bear on claim construction of "reduced rate" and "block size").

Uncertainty flags: (i) current ownership is per Google's unverified assignee field and the recorded assignments — I have not confirmed a live chain of title; (ii) the US 5744902 A string in the metadata is unverified; (iii) my searches found no 2026 CAFC docket activity, which is not proof that none exists.

I'll search for litigation involving US patent 7031254.

Let me search more targeted litigation databases.

Answer: No litigation identified for U.S. Patent No. 7,031,254

I searched for litigation specifically naming U.S. Patent No. 7,031,254 (the Lucent Technologies "Rate control system and method for a link within a wireless communications system" patent, filed 2002‑01‑25, granted 2006‑04‑18). I found no district court, Federal Circuit, or ITC case asserting or otherwise involving this patent. Below I explain what I checked, what I found, and important caveats.

What I checked

Important: do not confuse this with other "'254" patents

Search engines returned several unrelated patents that are routinely abbreviated to "'254." These are not U.S. 7,031,254 and I am excluding them:

• U.S. 9,351,254 — SEVEN Networks, LLC v. Google Inc., E.D. Tex., filed 5/17/2017 (dismissed 2/11/2019). Different patent (network/device resource management).
• U.S. 10,015,254 — SynKloud Technologies (external storage). Different patent.
• U.S. 10,992,254 — Shoals Technologies v. Voltage, LLC, ITC Inv. No. 337‑TA‑1365 and M.D.N.C. 1:25‑cv‑00026. Different patent (photovoltaic connectors).
• U.S. 10,879,254 — YMTC v. Micron (3D NAND). Different patent.
• Kenall Mfg. v. Genlyte Thomas Group (2006) — a lighting patent also referred to as the "'254 patent." Different patent.

None of these involve U.S. 7,031,254.

Ownership history (relevant context, from the patent record itself)

While there is no litigation, the assignment chain is worth noting because it explains why assertion is unlikely:

• 2002‑01‑25 — Assigned to Lucent Technologies Inc. (inventors Abraham, Chuah, Sampath, Saraydar).
• Change of name to Alcatel‑Lucent USA Inc.; security interests with Credit Suisse AG (2013), released 2014.
• 2017‑09‑13 — Assigned to Provenance Asset Group LLC (from Nokia Technologies Oy, Nokia Solutions and Networks BV, Alcatel Lucent SAS), with security interests to Nokia USA Inc. and Cortland Capital Market Services.
• 2021‑12‑28 — Assigned to RPX Corporation (from Provenance Asset Group LLC); subsequent liens to Barings Finance LLC in 2023/2024.
• Legal status per Google Patents: Expired – Lifetime, adjusted expiration 2024‑08‑22.

RPX is a defensive patent aggregator (RPX Clearinghouse), which acquires patents to take them out of assertion rather than to sue on them. That is consistent with the absence of any litigation.

Bottom line

• Plaintiff(s): None identified.
• Defendant(s): None identified.
• Jurisdiction / Case number / Filing date: N/A.
• Outcome / status: No litigation is known. The patent is recorded as expired (adjusted expiration 2024‑08‑22) and is currently held by RPX Corporation.

Caveats and confidence

• I have high confidence that no litigation is publicly indexed under this exact patent number, but moderate confidence overall that no litigation ever existed: I could not directly query PACER, Docket Alarm's full-text docket search, or Westlaw/Lexis, which are the most complete sources for older or obscure cases. The patent is a 2002‑priority Lucent/UMTS "Iub" rate‑control patent that expired in 2024 and never seems to have been asserted.
• No IPR, PGR, or CBM proceeding was found either, but I did not exhaustively search PTAB dockets for this number.

Suggested verification steps (to close the gap)

1. Docket Alarm full-text search for "7,031,254" and "7031254" across all federal district dockets.
2. Unified Patents litigation portal (https://portal.unifiedpatents.com/litigation/caselist) filtered by patent number.
3. RPX Insight (https://insight.rpxcorp.com) — as current owner, RPX's own records would show any assertion or license activity.
4. PACER / CourtListener RECAP search for the patent number in complaint text.
5. USPTO Patent Center "Litigation" tab and 35 U.S.C. § 290 notices filed against the patent.

If you would like, I can draft a short "no known litigation" representation suitable for an IP due‑diligence memo, with the appropriate qualification that it reflects searches of publicly available sources as of today's date.

Proceedings overview

Total AIA trial proceedings on file for US 7,031,254: 0 (no IPRs, no PGRs, no CBM reviews — active: 0 / claims invalidated: 0 / claims sustained: 0 / settled: 0 / institution denied: 0). The structured USPTO ODP "PTAB proceedings on file" block returns no AIA trials, and my independent web searches of PTAB decision and petition sources surfaced nothing challenging this patent. Bottom-line defensive posture: the opposite of "hardened." There is no PTAB record here at all — claims 1–5 have never been tested in an AIA trial, and the patent's real weakness is not an adverse PTAB decision but the calendar: Google Patents records the patent as "Expired – Lifetime," with an adjusted expiration of 2024-08-22 (20 years from the 2002-01-25 filing plus PTA). Any assertion now can only reach pre-expiration conduct, subject to the 6-year damages lookback of 35 U.S.C. § 286.

I found no proceeding to profile. Rather than invent one, the section below documents exactly what the absence means and the caveats around it.

No proceedings to report

• Type: n/a
• Filed: n/a
• Status: n/a — the ODP block supplied in this prompt states no AIA trial proceedings are on file as of the most recent ingest.
• Judge panel: n/a
• Petition grounds: n/a
• Institution decision: n/a
• Final Written Decision: n/a — there is no FWD, so not one claim of US 7,031,254 has been canceled, confirmed, or even construed by the Board.
• Settlement / termination: n/a
• Appeal: n/a — no FWD means no CAFC appeal of a PTAB decision on this patent. I found no CourtListener / Federal Circuit docket referencing US 7,031,254.
• Defensive value: Neutral-to-favorable but unexamined. You have no IPR record to point at, so you cannot say "claims 1–5 are canceled." You also carry no petitioner estoppel — the field of § 102/§ 103 art is entirely open to you. Your strongest non-PTAB defenses (expiration, § 286 damages window, § 102(b) art from the 1990s UMTS/ATM Iub flow-control literature) have never been crystallized by a Board panel.

Strategic summary

Claim status on US 7,031,254. The patent has 5 claims (1 independent method claim and claims 2–5 as dependents; claim 2 recites the §‑style sustainable-rate / RTFj / NL computation described in the specification, claim 3 the bearer-rate-within-class limitation, claim 4 the priority-class limitation, and claim 5 the transport-format-block mapping limitation). Because there is no PTAB proceeding of any kind, the correct characterization is: methods claims 1–5 are all UNTESTED at the PTAB — none canceled, none sustained. Contrast this with patents that have been through an IPR and narrowed by cancellation or disclaimer; there is no such narrowing here. The patent has, however, reached the end of its term (adjusted expiration recorded as 2024-08-22), and the maintenance-fee record shows the 4th, 8th, and 12th-year fees were paid, so the patent was in force through expiration.

Estoppel landscape. § 315(e)(2) estoppel is a non-issue: estoppel only attaches to a petitioner (and its RPOs/privies) after an FWD, and there is no petitioner. No prior-art ground is foreclosed to a current defendant. That said, if you are considering filing your own IPR, three AIA realities now bite: (1) the patent is expired, so the Board will construe claims under Phillips and there is no possibility of a claim amendment — you are litigating for a judgment that has little practical commercial value; (2) if you have already been served with a complaint asserting the patent, the § 315(b) one-year bar controls your filing window; and (3) the current Director-level institution regime (all IPR/PGR institution decisions centralized with the Director since the October 2025 Squires memorandum) has made institution materially harder and less predictable, including via "settled expectations" considerations that disfavor challenges to older patents. None of those help a patent owner here, because the patent is expired.

Pattern signals. There is no serial-petitioner pattern, no patent-owner PTAB appeal pattern, and no Unified Patents-style defensive IPR in the chain. The ownership history is the notable signal: Lucent Technologies → Alcatel-Lucent USA → Provenance Asset Group LLC (2017-09-12 assignment, i.e., the Nokia/Alcatel PAE/spin-out vehicle) → RPX Corporation (assignment recorded 2021-12-28, effective 2021-11-29), with matching security-interest and release filings involving Nokia USA/US Holdings and Cortland Capital, and a later RPX–Barings collateral agreement. In other words, the patent was absorbed into a defensive aggregator (RPX) rather than retained by an active litigating PAE — consistent with the near-total absence of district-court or PTAB assertion activity. The technical lineage is also relevant to any validity fight: the same R&D group published on this exact subject matter ("Impact of rate control on the capacity of an Iub link: single service case," Saraydar et al. 2003), and the specification itself concedes that the TB-size selection schemes are the conventional UMTS Release 99 approach (3GPP TS 25.322/25.302-style TFS mapping, AAL2/ITU-T I.363.2 framing). That is fertile § 102/§ 103 territory — but a defending party simply doesn't need an IPR to reach art this old and this standardized.

Recommended next steps

• Do not assume the patent is dead just because it has no PTAB record. The absence of an IPR is a neutral signal for a patent of this vintage, not proof of strength. Well-asserted patents attract IPRs; a patent sitting with RPX and never asserted is far more likely to have been unasserted than unassailable.
• Lead with expiration, not invalidity. Confirm the 2024-08-22 adjusted expiration against the USPTO PatentCenter / Patent Term Adjustment record for application 10/057,449. If confirmed, frame your response as a § 286 damages-window problem: on 2026-09-30, the recoverable window runs only from 2020-09-30 to 2024-08-22, and any demand letter asserting ongoing infringement of claims 1–5 as of today is legally incoherent.
• If you do want a PTAB ruling (e.g., a customer indemnity demand forces a decision), recognize that an IPR against an expired patent yields a Phillips construction and no claim amendment, and that the § 315(b) clock starts on service of the complaint. Because no estoppel exists and no ground is barred, any § 102/§ 103 art from the pre-2002 UMTS Iub/AAL2/transport-format literature is fully available to you.
• Verify the negative before you rely on it. Re-check PTAB E2E / the PTAB "Patent Trial" search for US 7,031,254 (https://ptacts.uspto.gov/) and confirm no ex parte reexamination certificate is on file (the Google Patents record for US7031254B2 lists no reexamination entry, but that page is not authoritative on post-issuance filings). ODP ingest can lag newly-filed petitions, so a proceeding filed in the last few months could be missing from the structured block.
• Preserve your own record. Because the record is empty, discovery responses or IPR papers that represent a claim-by-claim PTAB outcome for US 7,031,254 would be a misstatement. Nothing in this patent's file supports any such representation.

Confidence and caveats. High confidence that (a) the ODP canonical list is empty as stated, and (b) the assignment chain and "Expired – Lifetime" / 2024-08-22 adjusted-expiration data shown on the Google Patents fetch are accurate as of 2026-09-30. Lower confidence (moderate) that literally no AIA trial has ever been filed and terminated in a way ODP would have purged — I could not locate any PTAB or Federal Circuit document naming US 7,031,254, but I was not able to run a direct PTAB E2E party/patent query, and I did not locate any district-court docket identifying an asserting plaintiff. I have deliberately not supplied any proceeding number, panel, filing date, or claim-level disposition, because none exists in the sources I could reach; fabricating one would be worse than the gap.

I'll start by pulling whatever public records I can to corroborate the assignment chain and look for litigation/correspondent data.

Good — the chain is revealing. Let me now check whether this specific patent was ever asserted, and dig into the RPX leg and correspondents.

Ownership Reconstruction — US 7,031,254 B2

"Rate control system and method for a link within a wireless communications system"
App. No. 10/057,449 · Filed 2002-01-25 · Issued 2006-04-18 · Expired 2024-08-22 (adjusted)
Family: US 2003/0156580 A1 · EP 1331768 A1 (withdrawn)

Verification entry point: https://assignment.uspto.gov/patent/index.html (search 7031254) and https://assignmentcenter.uspto.gov/

Scope caveat up front: the source set I could actually reach exposes reel/frame, conveyance, assignor, assignee and execution/recording dates (via Google Patents legal events, which mirror the Assignment Center's recorded data). It does not expose the correspondent of record field for this patent, and I could not retrieve that field from the Assignment Center/PEDS for any of the 14 records below. I have therefore marked correspondent as not retrieved on every entry rather than guess — see signal 3.

---

Inventors

Basis: all four appear together as assignors on reel 012546/0915, assigning to Lucent Technologies Inc., New Jersey, signed 2002-01-21 through 2002-01-23 — i.e., contemporaneous with (indeed days before) the 2002-01-25 filing. An assignment executed by the inventors to a corporate assignee pre-filing is the standard indicium of employee-inventor status; I found no contrary evidence. Corroboration for at least one inventor: an Ashwin Sampath with a Somerset, New Jersey address appears as an inventor on a contemporaneous Lucent Technologies EP filing (EP 01304724.6, published 2002), consistent with the Bell Labs Wireless Research organization in New Jersey.

Unusual patterns: none visible in the record. There is no evidence in the assignment record of inventors departing Lucent within 12 months of filing, and no re-assignment by any inventor back to the company, no confirmatory assignment, and no nunc pro tunc correction — all of which are the markers one would look for. The inventors' later career movements are not part of the recorded chain and I will not assert them.

---

Original assignee

Lucent Technologies Inc. (Murray Hill, New Jersey) — named on the issued patent, and still shown as the original assignee in the Assignment Center record via reel 012546/0915.

• Line of business: telecommunications equipment, and at the relevant time a top-tier UMTS/3G radio access network vendor. Lucent shipped UMTS RAN product — Node B base stations and Radio Network Controllers (the Flexent™ family) — and the patent is expressly directed to exactly that product: rate control over the Iub interface between an RNC and a Node B (spec., Tables 1–7; claims directed to "sending data over a link from a node of a radio access network to a base station").
• Product embodying the claims? Yes — the claimed subject matter is UMTS Iub transport-format-block selection, a function implemented in Lucent's RNC/Node B products. Note all five claims are method claims, so direct practice sits with the network equipment vendor/operator, not with end users.
• Current status: not operating under that name. Lucent merged with Alcatel in 2006 to form Alcatel-Lucent; the US entity was renamed Alcatel-Lucent USA Inc. (change of name effective 2008-11-01, recorded 2019-02-07, reel 049887/0613). Alcatel-Lucent was acquired by Nokia in 2016; Alcatel-Lucent USA Inc. survives as a Nokia subsidiary/affiliate entity. No bankruptcy proceeding is involved at any point in this chain.

---

Assignment timeline

Fourteen recorded events. Ordered by recording date; execution/effective dates noted.

1. 2002-01-21 to 2002-01-23 (executed) / recorded 2002-01-25 — Reel 012546/0915

• Conveyance: Assignment
• Assignor: Santoshi Abraham; Mooi Choo Chuah; Ashwin Sampath; Cem Uygur Saraydar (individually)
• Assignee: Lucent Technologies Inc., New Jersey
• Correspondent: not retrieved (see signal 3)
• Context: original inventor-to-employer assignment, executed pre-filing.

2. 2013-01-30 (executed) / recorded 2013-03-07 — Reel 030510/0627

• Conveyance: Security Interest (security agreement)
• Assignor: Alcatel-Lucent USA Inc.
• Assignee: Credit Suisse AG, New York
• Correspondent: not retrieved
• Context: corporate-level securitization — Alcatel-Lucent pledged its patent estate (including this patent) as collateral for a credit facility. Not a transfer of ownership.

3. 2014-08-19 (executed) / recorded 2014-10-09 — Reel 033949/0531

• Conveyance: Release by Secured Party
• Assignor: Credit Suisse AG
• Assignee: Alcatel-Lucent USA Inc.
• Correspondent: not retrieved
• Context: release of the lien on refinancing/repayment. Ownership never left Alcatel-Lucent USA Inc.

4. 2017-09-12 (executed) / recorded 2017-09-13 — Reel 043877/0001

• Conveyance: Assignment
• Assignors: Nokia Technologies Oy; Nokia Solutions and Networks B.V.; Alcatel Lucent SAS
• Assignee: Provenance Asset Group LLC, Connecticut
• Correspondent: not retrieved
• Context: the pivotal transaction — a bulk, multi-seller divestiture of roughly 4,000 US grants and ~12,000 assets worldwide under a Patent Purchase Agreement, to a newly formed patent-monetization vehicle. Note the chain-of-title anomaly: the assignors of record are the three foreign Nokia/ALU entities, while the US patent's recorded owner of title was still Lucent Technologies Inc./Alcatel-Lucent USA Inc.; the US name-change record clearing that gap was not recorded until 2019 (entry 7).

5. 2017-09-13 (executed) / recorded 2017-09-13 — Reel 043879/0001

• Conveyance: Security Interest
• Assignors: Provenance Asset Group Holdings, LLC; Provenance Asset Group LLC
• Assignee: Nokia USA Inc., California
• Correspondent: not retrieved
• Context: seller-retained lien — Nokia financing its own divestiture and retaining a secured claim on the sold assets. Recorded one day after the sale closed.

6. 2017-09-13 (executed) / recorded 2017-09-13 — Reel 043967/0001

• Conveyance: Security Interest
• Assignors: Provenance Asset Group Holdings, LLC; Provenance Asset Group, LLC
• Assignee: Cortland Capital Market Services, LLC, Illinois
• Correspondent: not retrieved
• Context: second, parallel lien from a third-party collateral agent — i.e., acquisition financing layered on the same day as the acquisition itself.

7. Effective 2008-11-01 / recorded 2019-02-07 — Reel 049887/0613

• Conveyance: Change of Name
• Assignor: Lucent Technologies Inc.
• Assignee: Alcatel-Lucent USA Inc.
• Correspondent: not retrieved
• Context: change of name only — a retroactive chain-of-title cleanup filed ~18 months after the 2017 divestiture. Corrects the US-side gap flagged at entry 4.

8. Effective 2018-12-20 / recorded 2019-02-14 — Reel 048370/0682

• Conveyance: Assignment and Assumption Agreement
• Assignor: Nokia USA Inc.
• Assignee: Nokia US Holdings Inc., New Jersey
• Correspondent: not retrieved
• Context: internal Nokia reorganization of the lienholder entity (not a transfer of the patent itself).

9. Effective 2021-11-01 / recorded 2021-11-30 — Reel 058983/0104

• Conveyance: Release by Secured Party
• Assignor: Cortland Capital Markets Services LLC
• Assignees: Provenance Asset Group LLC; Provenance Asset Group Holdings LLC
• Correspondent: not retrieved
• Context: lien discharge in connection with the wind-down/sale of the Provenance holding.

10. Effective 2021-11-29 / recorded 2021-11-30 — Reel 058363/0723

• Conveyance: Release of Security Interest
• Assignor: Nokia US Holdings Inc.
• Assignees: Provenance Asset Group LLC; Provenance Asset Group Holdings LLC
• Correspondent: not retrieved
• Context: Nokia's retained lien discharged the same week the asset was sold on. Together with entry 9, this clears both encumbrances as a precondition to a clean sale.

11. 2021-11-29 (executed) / recorded 2021-12-28 — Reel 059352/0001

• Conveyance: Assignment
• Assignor: Provenance Asset Group LLC
• Assignee: RPX Corporation, California
• Correspondent: not retrieved
• Context: transfer to a defensive patent aggregator. Executed the same day the liens were released (entry 10). This is the current ownership event.

12. Effective 2022-01-07 / recorded 2023-04-22 — Reel 063429/0001

• Conveyance: Patent Security Agreement
• Assignor: RPX Corporation
• Assignee: Barings Finance LLC, as Collateral Agent, North Carolina
• Correspondent: not retrieved
• Context: encumbrance — RPX pledged its patent collateral for corporate financing. Ownership unchanged.

13. Effective 2024-08-02 / recorded 2024-08-05 — Reel 068328/0278

• Conveyance: Release of Lien on Patents
• Assignor: Barings Finance LLC
• Assignee: RPX Corporation
• Correspondent: not retrieved
• Context: release of the 2022 charge.

14. Effective 2024-08-02 / recorded 2024-08-06 — Reel 068328/0674

• Conveyance: Patent Security Agreement
• Assignors: RPX Corporation; RPX Clearinghouse LLC
• Assignee: Barings Finance LLC, as Collateral Agent
• Correspondent: not retrieved
• Context: immediate re-pledge of the same collateral, now also covering RPX Clearinghouse LLC. Purely a financing event; no change in beneficial ownership.

Net chain: Lucent Technologies Inc. → (Alcatel-Lucent / Nokia corporate succession) → Provenance Asset Group LLC (2017) → RPX Corporation (2021, current). Every other entry is a lien or lien release.

---

Timeline diagram

timeline
    title Ownership of US 7031254
    2002 : Filed by Lucent Technologies
         : Inventors assign rights to Lucent
    2008 : Lucent renamed Alcatel-Lucent USA
    2013 : Credit Suisse security interest recorded
    2014 : Credit Suisse lien released
    2017 : Assigned to Provenance Asset Group
         : Nokia USA security interest recorded
         : Cortland Capital lien recorded
    2019 : Lucent name change recorded
         : Nokia USA rights move to Nokia US Holdings
    2021 : Cortland and Nokia liens released
         : Assigned to RPX Corporation
    2022 : Barings security agreement recorded
    2024 : Barings lien re-recorded for RPX

---

NPE / troll-pattern signals

1. Shell-entity transfer — present.
Reel 043877/0001 (executed 2017-09-12, recorded 2017-09-13) moves the patent from three operating Nokia/Alcatel-Lucent entities to Provenance Asset Group LLC, a Connecticut vehicle formed in September 2017 — contemporaneously with the deal — by Dan McCurdy (CEO), Tim Lynch (President) and Laura Quatela, funded with $40M from GCM Grosvenor, with offices in Essex, CT and Rochester, NY. It manufactures nothing and its business is licensing/selling patent rights. Honest qualification: this is a purpose-built monetization vehicle, not a bare registered-agent shell — it had named officers, published a business model, and publicly stated that assertion was not part of the model. The "shell" label should therefore be applied to the function (non-practising aggregation vehicle) rather than to the absence of operations.

2. Known asserter in the chain — present (partial / moderate).
Provenance Asset Group LLC does not appear on the classic high-frequency plaintiff lists itself; RPX, the current assignee, is the opposite of an NPE. The finding here rests on Provenance's outbound transferees, which include confirmed asserting entities: Sisvel International S.A. (April 2019), Wireless Innovations LLC (described as an NPE with ties to VIA Licensing, a Dolby Labs subsidiary, March 2019), and Proven Networks, LLC — formed as RAK Investments III, LLC in October 2019, which received patents from Provenance in February 2020 and then litigated them in In re: Proven Networks, LLC, Patent Litigation, MDL No. 2959 (JPML transfer order, Sept. 2020), where the panel found common factual issues "relating to the patents' transfer history and associated valuation, damages, and standing issues." This patent was not among the Proven Networks asserted assets and no suit asserting US 7,031,254 was located (see signal 5), so the signal is about the conveyor, not this asset.

3. Repeat correspondent across the chain — unclear / not determinable.
I could not retrieve the correspondent-of-record field for any of the 14 entries; Google Patents' legal-events mirror omits it and the Assignment Center field was not reachable in my search set. I will not invent attorney names. What should be checked on the Assignment Center record: whether a single outside firm (frequently a Connecticut, Delaware or New York IP boutique) appears as correspondent on 043877/0001, 043879/0001, 043967/0001, 058983/0104, 058363/0723 and 059352/0001. If one firm/attorney recurs across those six, this signal upgrades to present on its own. For context only — and expressly not a record for this patent — Lucent of this era routed its filings through an in-house IP division correspondent (e.g., Martin I. Finston, Lucent Technologies docket administrator) rather than an outside firm, which is why the 2002 entry would likely show a corporate correspondent rather than a named private practitioner.

4. Cascading transfers — present.
Three records land within a ~24-hour window on 2017-09-13 (043877/0001, then 043879/0001 and 043967/0001 the same day), and four records land within ~48 hours in November–December 2021 (058983/0104 and 058363/0723 on 2021-11-30; 059352/0001 executed 2021-11-29, recorded 2021-12-28). Two simultaneous liens on the day of purchase (Nokia + Cortland), and two simultaneous releases on the day of sale, is the signature of acquisition financing and a pre-cleared disposal, not of ordinary corporate housekeeping.

5. Pre-litigation transfer — not present.
No infringement action naming US 7,031,254 was located. The 2021-11-29 Provenance→RPX assignment is not tied to any filing I could find; to the contrary, the transferee is a non-asserting aggregator. The patent also expired 2024-08-22, which independently forecloses most assertion value. Absence of litigation is a finding here, but I note it is inferred from the sources I could reach rather than from a full docket sweep.

6. Bankruptcy fire-sale — not present.
Neither Lucent/Alcatel-Lucent nor Nokia sought Chapter 7 or Chapter 11 protection. The 2017 divestiture was a portfolio-streamlining sale following Nokia's 2016 acquisition of Alcatel-Lucent, not a distressed sale.

7. Privateering — present.
The 2017 transaction saw Nokia sell ~4,000 US grants to Provenance while retaining a security interest through Nokia USA Inc. (reel 043879/0001) and — per contemporaneous reporting — taking a share of Provenance's revenues from the transferred assets; that lien was not released until reel 058363/0723 (effective 2021-11-29), the same week the assets were on-sold. Nokia has an established pattern of transferring assets to licensing/assertion intermediaries (earlier deals with Acacia and Conversant). A seller that both finances the buyer and shares in downstream licensing revenue has retained economic exposure to assertion outcomes — the substantive definition of privateering, regardless of the stated defensive intent.

8. Defensive aggregator (anti-NPE) — present.
The chain terminates at RPX Corporation by assignment executed 2021-11-29 and recorded 2021-12-28 at reel 059352/0001. RPX is a defensive patent-risk-management firm (defensive buying, acquisition syndication, insurance) that does not assert patents; the only records after that date are security interests and releases (063429/0001, 068328/0278, 068328/0674), none of which change beneficial ownership. This is an inverse signal: the asset has effectively been neutralized.

---

Verdict

Defensive / non-asserting.

The chain terminates at a defensive aggregator: RPX Corporation took assignment of this patent by instrument executed 2021-11-29 and recorded 2021-12-28 at reel 059352/0001, after Provenance's and Nokia's liens were simultaneously discharged (reels 058983/0104 and 058363/0723), and RPX has since done nothing with it but pledge it as collateral (reels 063429/0001, 068328/0278, 068328/0674) until the patent expired 2024-08-22. No infringement action naming US 7,031,254 was located.

That said, the verdict is not a clean one, and the record should not be read as "this patent was never part of an assertion machine." The 2017 leg carries three genuine NPE-adjacent markers: a transfer from operating entities to a purpose-built monetization vehicle (043877/0001), same-day seller and third-party liens evidencing acquisition financing (043879/0001, 043967/0001), and a seller-retained security interest plus revenue participation consistent with privateering. Had the chain stopped at Provenance — as it did for the assets that went to Sisvel, Wireless Innovations LLC and Proven Networks LLC and were litigated in MDL-2959 — the correct call would be NPE — moderate confidence. The RPX acquisition is what breaks the pattern here, and it is the terminal event, so the defensive characterization controls.

Verification: https://assignment.uspto.gov/patent/index.html → search 7031254 (also indexed at https://assignmentcenter.uspto.gov/). The two open items to close on that record are (a) the correspondent of record on reels 043877/0001, 043879/0001, 043967/0001, 058983/0104, 058363/0723 and 059352/0001, which could upgrade or eliminate signal 3, and (b) the Patent Purchase Agreement between Nokia Technologies Oy, Nokia Solutions and Networks BV, Alcatel Lucent SAS and Provenance Asset Group LLC, a copy of which is on file in the USPTO record at https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1549234](/patent/1549234) — it defines the "Security Interests" (expressly citing the Cortland 2017-09-13 and Nokia US Holdings 2019-02-14 recordings) and the encumbrances carried by the assigned patents.

I'll verify the cited references against live sources before analyzing each one.

Let me confirm the remaining bibliographic details for two references.

Prior-art analysis — US 7,031,254 B2

Scope and method. As instructed, I looked only for the identifier 7031254 and its own citation list; I did not pull in similarly-numbered patents. I relied on (a) the authoritative full text supplied above, whose "Patent Citations (5)" / "Citations (5)" block is the examiner/applicant citation list of record as mirrored by Google Patents, and (b) live retrieval of each cited document. Note: I reached these documents through Google Patents, uspto.report, patentimages, Justia and EPO/WIPO mirrors, not through a direct PatentCenter/PTAB API query. The prior-art date is the pre-AIA statute, since the application was filed 2002-01-25.

The five cited references (literal, as listed for US 7,031,254)

The asterisk convention ("cited by examiner" vs. "cited by third party") is imperfect in the fetched rendering; treat the examiner/third-party split as probable, not verified.

---

1. US 4,771,391 A — "Adaptive packet length traffic control in a local area network" (IBM)

• Citation: US 4,771,391 A; filed 1986-07-21; issued 1988-09-13. Text: https://patentimages.storage.googleapis.com/45/9f/0e/1ace9eb4f30721/US4771391.pdf
• Description: Token-ring LAN traffic control in which a ring adapter measures normalized offered traffic and the node processor computes a delay metric, then changes the packet size (and hence the effective rate) in response to measured traffic load. The search hit confirms formulaic adjustment of packet size/packet rate from a measured traffic term (ρ* = R*/Ro) and a delay metric.
• § 102 assessment: Qualifies as §102(b) art by decades. It maps to claim 1's steps "obtaining a description of loading" and "selecting a block size … with reference to the description of link loading" — an adaptive packet-length response to load. It does not anticipate claim 1 as a whole: no wireless RNC→base-station link, no per-user selection from a discrete set that includes both a maximum and at least one reduced size applied to different users, no determination of how many users get reduced sizes, and no mapping of data to blocks sent to a base station. No claim is anticipated as an entire claim. Best characterized as §103 art going to the "vary packet/block size with load" genus.

2. US 5,649,299 A — "Apparatus and method for adapting a digital radiotelephone system to increased subscriber traffic" (Motorola)

• Citation: US 5,649,299 A; App. 08/144,166; filed 1993-10-27; issued 1997-07-15; inventors John W. Battin, Kevin M. Laird, Charles N. Lynk, Jr. Abstract confirmed via https://patents.google.com/patent/[US5649299A](/patent/US5649299A)/en
• Description: A digital radiotelephone system in which the fixed site measures system occupancy, compares it to a threshold (503), and commands "some or all" vocoders to switch to a lower data rate; higher rate is used by default for best quality. Fig. 5/7 logic includes "RANDOMLY SELECT M LOW [rate users]" and "SEARCH LIST FOR LOW PRIORITY SUBSCRIBER UNITS." Rates are restored when occupancy falls below threshold (Fig. 6/8).
• § 102 assessment: This is the closest single-reference (anticipation) candidate in the list. It discloses: a measured load/occupancy description; reducing a subset of users to a lower rate while others stay at the highest rate; the number of reduced users being chosen in view of occupancy; and priority differentiation (low-priority subscribers reduced first). Those elements line up with claim 1's "maximum block size for at least one user / reduced block size for at least one user" and "number of users that are to receive reduced block sizes … determined with reference to the description of link loading."
  • Missing for claim 1: the link is the RNC-to-Node-B transport link (in '299 it is the radio/capacity resource at the fixed site), and there is no "block size"/transport-format-block concept or mapping of data into blocks.
  • Claim 2: anticipates the idea of setting a number of users to a reduced rate as a function of load, but not the specific R_s = R*T/N / N_L computation. Weak § 102; strong § 103.
  • Claim 4 (priority class): '299's low-priority-first / high-priority-last logic is the best single-reference disclosure of the priority aspect — but since claim 4 depends from claim 1 and '299 lacks claim 1's link/block limitations, claim 4 is likewise not anticipated as a whole.
  • Verdict: probative § 102(a)/(b) reference; the "gist" of claims 1–2 and the substance of claim 4.

3. WO 1999/052307 A1 — "Flexible radio access and resource allocation in a universal mobile telephone system (UMTS)" (Ericsson)

• Citation: WO 1999/052307 A1; PCT/SE1999/000552; published 1999-10-14. https://patents.google.com/patent/WO1999052307A1/en. Inventors per the CN family record: Widegren, Willars, Wallentin.
• Description: A UTRAN (Node Bs + RNC) that, on a radio-access-bearer request, flexibly maps/allocates radio access bearers to physical transport resources (ATM) and radio channel resources based on QoS parameters and/or current traffic conditions, and can switch channel type mid-connection if parameters change.
• § 102 assessment: Supplies the claim-1 preamble environment (RNC/Node B, transport over ATM/AAL2, allocation based on current traffic conditions) and the concept of allocating based on "current traffic conditions." It is bearer-level resource allocation, not intra-bearer selection of a per-user block size from a discrete set containing a maximum and a reduced size, and it does not frame the control as reducing the rates of a determined number of users. No claim anticipated; relevant §103 art (and useful for construeing the claim preamble and "bearer rate").

4. WO 2001/063856 A1 — "Flow control between transmitter and receiver entities in a communications system" (Ericsson)

• Citation: WO 2001/063856 A1; PCT/SE2001/000407; filed 2001-02-23; published 2001-08-30. https://patents.google.com/patent/WO2001063856A1/ko
• Description: A UMTS RNC-internal flow-control scheme: a common-channel MAC-c entity and multiple dedicated-channel MAC-d entities; the MAC-c buffer fill level is regulated and each active MAC-d data flow receives a fair share of the data rate between MAC-c and MAC-d, with buffer space allocated sequentially/round-robin or by flow activity, based on MAC-c upper fill level and, in embodiments, MAC-d buffer levels.
• § 102 assessment: This is the closest reference to the fair-share rationale underlying the FRC algorithm, and it is expressly a UMTS/RNC mechanism. But the entity receiving the controlled data is an internal RNC MAC entity, not a Node B across the Iub link, and control is by credit/buffer-space allocation, not by selecting TFB/block sizes. No claim anticipated. Strong §103 art against the "fair share as a function of load" premise if combined with reference 2.

5. US 2003/0081628 A1 — "Throughput in multi-rate wireless networks using variable-length packets and other techniques" (Cognio)

• Citation: US 2003/0081628 A1; filed 2001-10-30 (provisional 60/330,755, also 2001-10-30); published 2003-05-01. https://patents.google.com/patent/US20030081628
• Description: A WLAN (802.11x) scheme that assigns a packet length to each user so all users transmit equal-duration packets, sets a user-rate-specific maximum packet size, uses shorter packets when possible, and dynamically changes packet lengths as network conditions / measured offered load change, conditionally engaging/disengaging the algorithm based on monitored conditions.
• § 102 assessment — important date point: Its publication (2003-05-01) is after US 7,031,254's 2002-01-25 filing date, so it cannot be §102(a)/(b) art. Its only §102 route is pre-AIA §102(e), as a U.S. published application "filed in the United States before the invention by the applicant" — here, filed 2001-10-30, three months before the '254 filing date. Even so it is not a clean anticipation of claim 1: it is a WLAN AP/STA system (not an RNC→Node B link), it does not derive the packet sizes from a discrete set applied to different users with a determined number of users receiving reduced sizes, and block boundaries are per-user rate rather than load-determined. It is, however, the closest reference to:
  • claim 1's "selecting a respective block size for each said user from a discrete set … including a maximum … and at least one reduced … size" and load-responsive adjustment; and
  • claim 5's "selecting a size of a transport format block … by changing the number … as a function of the size of transport format blocks of at least one other user" — since Cognio sets each user's max packet size by reference to the fastest user's rate, i.e., by reference to another user's parameters. §102(e) is arguable for claims 1 and 5 at the margins; §103 is the realistic posture.

---

Synthesis

1. No reference in this list anticipates claim 1 as a whole. Claim 1 requires, in combination, (i) the RNC→base-station link, (ii) per-user selection from a discrete set containing a maximum and at least one reduced size, (iii) mapping data to those blocks, (iv) both a max-size user and a reduced-size user, and (v) a determined number of users receiving reduced sizes as a function of the link-load description. References 1–5 collectively supply fragments of that combination, but no single one supplies all of it — principally because none of them controls transport-format-block size over the Iub transport link.
2. Ranked realistic §102 posture: US 5,649,299 (claims 1/2 concept, claim 4 priority aspect) > US 2003/0081628 A1 under §102(e) (claims 1, 5) > WO 1999/052307 (preamble only) > WO 2001/063856 (fair-share concept) > US 4,771,391 (adaptive packet sizing genus).
3. The likely real-world challenge is a §103 combination, e.g. US 5,649,299 (occupancy-triggered subset rate reduction, priority classes) + US 2003/0081628 A1 (per-user variable packet/block sizes selected responsively to network conditions) + WO 1999/052307 (the UMTS RNC/Node B/ATM transport environment), against claims 1, 2 and 5; and + WO 1999/052307 (bearer concept) + US 5,649,299 (priority) against claims 3 and 4.
4. Non-citations worth not treating as prior art: the Google Patents "Similar Documents" list for US 7,031,254 includes US 2003/0156580 A1 (the '254 pre-grant publication itself), EP 2 080 326 B1 ("System and method of load dependent rate control"), and the NPL item "Saraydar et al., Impact of rate control on the capacity of an Iub link: single service case (2003)." The Saraydar paper is the inventors' own and post-dates the 2002-01-25 filing, so it is not prior art against '254.

Contradictions / uncertainty flags

• Contradiction with the prior "Verification note": the earlier section flagged an anomalous US 5744902 A string in the Google Patents bibliographic block and declined to fold it in. Consistent with your literal-interpretation rule, I confirm that US 5,749,902 does not appear anywhere in the '254 citation list; the cited patent is US 4,771,391. The US 5744902 A string therefore remains an unexplained metadata artifact (or an unrelated document) and is not prior art of record for US 7,031,254. I have not auto-corrected it; I am reporting only that it is absent from the citation table.
• WO 1999/052307 priority date discrepancy: Google Patents lists 1998-04-03; the CN 1,304,619 A family record shows priority 1998-04-02. Both pre-date the '254 priority date, so the discrepancy is immaterial to prior-art status; unresolved as to which is correct.
• Examiner/third-party markers on the citation table are probable but not verified from the rendered text.
• I could not complete direct USPTO PatentCenter/PTAB database queries (search step limit reached); the citation list of record is taken from the authoritative full text and corroborated against public mirrors. As noted previously, my searches found no 2026 CAFC docket activity for this patent — a search result, not a certified negative.
• I did not verify the granted-patent offspring of US 2003/0081628 A1 and will not speculate about its issue number.

I'll verify the prior-art references from the patent's citation list before analyzing them.

Now the two Ericsson PCT publications and the Cognio application:

Obviousness Analysis — U.S. Patent No. 7,031,254 B2 under 35 U.S.C. § 103

Analysis as of the current date. Builds on, and does not repeat, the bibliographic, claim-scope, and litigation sections already produced. One cross-check note: my system header lists today as 2026-09-30 while the task prompt lists 2026-04-26; nothing in this analysis turns on which is correct, but I flag the inconsistency rather than silently picking one.

---

1. Legal framework applied

I apply the Graham v. John Deere framework as refined by KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), and MPEP §§ 2141–2144:

1. Determine the scope and content of the prior art.
2. Ascertain the differences between the prior art and the claims.
3. Resolve the level of ordinary skill in the art.
4. Evaluate secondary considerations.

Under KSR, a combination is obvious where the prior art elements perform their known functions with predictable results; where a design need or market pressure provides a reason to combine known elements; where a simple substitution of one known element for another yields predictable results; or where the combination is "obvious to try" along a finite number of identified, predictable solutions. MPEP § 2144.04(iv) additionally allows treating a formula-based limitation as an obvious design choice where the result is a mere rearrangement of parts or a predictable optimization.

Limiting principle I apply throughout: the only independent claim is claim 1; claims 2–5 are dependent. A § 103 ground must therefore be evaluated against claim 1 first, with dependents treated as adding narrowing limitations that can defeat an otherwise-valid ground.

---

2. Scope and content of the prior art — the "Prior Art section" of the record

The record's prior-art section is the Patent Citations (5) list, which is the set relevant to this task. I verified each against live sources.

Two record-hygiene flags that matter to the prior-art set:

• The "Cited By" and "Families Citing" lists are not § 102/§ 103 prior art. They are later-published documents (e.g., US 2004/0166891 to Mahkonen, US 2005/0107090 to Hosein, US 7,478,158 Adobe, US 8,923,809 AT&T). Their publication dates post-date the '254 priority date. They cannot be used for anticipation or as combinable art; at most they evidence the state of the art indirectly, and I do not rely on them.
• Admitted prior art in the '254 specification itself. The specification states that in "current UMTS" practice a user's data is mapped to the largest TFB whenever the RLC buffer exceeds that TFB, and that "current systems would assign TFBs to a user in a bearer rate based simply on the amount of data the user had to transmit." These are applicant admissions usable under § 103 as representing the prior art (MPEP § 2129). They are important because they narrow the gap the cited art must bridge: the '254 does not purport to invent TFB fragmentation per se — only the link-load-conditioned selection of block sizes across users.

---

3. Level of ordinary skill in the art (POSITA)

A POSITA here would be an engineer with a bachelor's degree in electrical engineering or computer science and approximately 2–4 years' experience in wireless packet-data networking — specifically cellular radio-access transport (UMTS Release 99/UTRAN: RNC, Node B, Iub, RLC/MAC, transport format sets, AAL2/ATM framing) or an equivalent background in shared-medium packet scheduling. This is the level I apply to the motivation and predictability findings below.

---

4. What each reference teaches (verified content)

US 4,771,391 (IBM). "A distributed control apparatus and method … maintain[s] network stability in a packet switching local area network under changing traffic load conditions, by controlling the size of the packets transmitted by each node. The invention monitors the present information flow rate in the network and computes an average packet length which it then uses to adjust the size of the variable length data field in each message frame." The specification states the functional relationship between packet length and flow rate may be stored as "an arithmetic expression … or … a table of corresponding values," that a buffer "temporarily stores a block of the information," and that the node accesses a "sub-block" of the buffered block of the computed length for transmission. URL: https://patents.google.com/patent/US4771391A/en.

US 5,649,299 (Motorola). "Under most circumstances, the highest data rate is selected … When the system occupancy exceeds a particular threshold, as detected at the fixed site …, some or all of the vocoders are instructed to switch to a lower data rate to provide for an increase in the number of communication channels." The flowcharts show measuring occupancy, comparing to a detector level, then "RANDOMLY SELECT M LOW [priority] … USING HIGH RATES" and commands to reduce rates; alternative flows show priority-based correlation ("LOW PRIORITY SUBSCRIBER UNITS," "HIGH PRIORITY SUBSCRIBER UNITS") and rate-increase flows when occupancy falls. URL: https://patents.google.com/patent/US5649299A/en.

WO 99/52307 (Ericsson). Discloses the UMTS UTRAN architecture — RNC connected to plural Node Bs, an ATM connection between the nodes — and the dynamic mapping of radio access bearers to transport and radio resources: "the radio access bearers are dynamically assigned to UTRAN transport and radio channel resources"; "the mapping is based on one or more parameters including for example one or more quality of service parameters and/or current traffic conditions." It expressly discloses transport formats (TF/TFS) as the mechanism for varying a user's instantaneous rate on the transport connection. URL: https://patents.google.com/patent/WO1999052307A1/en.

WO 01/63856 (Ericsson). Flow control in the same UTRAN architecture. It describes a common MAC-c entity multiplexing data flows from multiple MAC-d entities and allocating "a fair share of the data rate between the Mac-d entities … and which also maintains the MAC-c buffer at or close to a desired fill level," using "available information pertaining to the buffer fill levels of the respective RLC/MAC-d entities"; a transmitter may transmit only on receiving credit from the receiver. It expressly frames the problem as "handling potential buffer overflows resulting from users overloading a communications system." URL: https://patents.google.com/patent/WO2001063856A1/en.

US 2003/0081628 (Cognio). "One technique is to assign packet lengths to each user in such a way as to ensure that all users transmit packets of equal duration. The packet duration for all users is set based on a maximum packet size at the fastest user data rate in the network." The user-rate-specific packet size "is set as a maximum packet size for that user", and "An enhancement … is to dynamically change the packet lengths of the network users to adjust for changing conditions." Also: "the AP monitors user throughput, offered network load, or other conditions … and … engages one of the variable packet length methods." URL: https://patents.google.com/patent/US20030081628A1/en.

---

5. Claim 1 — element-by-element mapping

Claim 1 elements (from the plain-language restatement already produced): (1a) obtain a description of link loading; (1b) select a block size per user from a discrete set of ≥2 sizes including a maximum and ≥1 reduced size; (1c) map data to blocks; (1d) send blocks over the link; with the wherein-clause requiring (1e) max size for ≥1 user and reduced size for ≥1 user, (1f) selection referenced to link loading, and (1g) the number of users receiving reduced sizes determined with reference to link loading.

Observation. Every limitation of claim 1 is disclosed, in relevant part, by the cited references. Limitation 1g — the most likely point of attack for the applicant, because it is the "number of users" concept most closely tied to the FRC algorithm — is squarely met by US 5,649,299's "RANDOMLY SELECT M" step, where M is precisely the number of users forced to the lower rate as a function of measured occupancy.

---

6. Grounds of rejection — combinations

Ground A — Primary: WO 99/52307 in view of WO 01/63856 (both Ericsson; same UTRAN architecture)

Proposed rejection of claim 1. WO '307 teaches the RNC↔Node B transport link, the mapping of each user's data onto transport formats/blocks, and doing that mapping as a function of "current traffic conditions." WO '856 supplies the missing per-user, load-conditioned distribution: it allocates MAC-c buffer space / rate among competing MAC-d flows based on buffer fill levels, and inherently reduces the share of some flows while others keep the maximum. Because WO '856 describes the identical UTRAN entity diagram (RNC 140 / Node B 150 / MAC-d / MAC-c / RLC) as WO '307, they are a priori combinable — indeed both are Ericsson filings aimed at the same Release-99 transport problem, and WO '307 is prior art that a POSITA working on WO '856's problem would necessarily consult.

Motivation: both references address "users overloading a communications system" (WO '856, first definition) and "efficiently allocating resources to support … services" (WO '307). Predictable result: keep aggregate offered traffic within transport capacity by reducing the offered rate of a subset of users.

Ground B — Primary: US 4,771,391 in view of US 5,649,299 (and optionally WO 99/52307 for the UMTS context)

This is the strongest combination because it maps cleanly onto the two concepts claim 1 actually requires: (i) adaptively sizing a transport unit as a function of measured load, and (ii) reducing the number of active users' rates to fit capacity on a shared medium.

Proposed rejection. US '391 teaches measuring the flow rate on a shared medium and changing the transmitted packet size according to a stored relationship, buffering data and transmitting a sub-block of the computed size. US '299 teaches, in a cellular shared-radio context, that when occupancy exceeds a threshold, some or all users are commanded to a lower rate, that the selection may be random and may follow priority classes, and that rates are restored when occupancy falls.

Motivation / KSR rationales:

• Same field and same problem: both address throughput/capacity under variable load on a shared medium; US '299 is expressly a cellular (CDMA) system, and the '254 concerns a cellular (UMTS/CDMA) link.
• Known technique, predictable result: variable-length/variable-size transport units to trade per-user rate against aggregate capacity is the express teaching of US '391; applying it to a radio access link (US '299's environment) yields no more than the expected result.
• Telegraphing in the art: US '299's own flowchart ("RANDOMLY SELECT M") reaches the claim-1 "number of users" concept directly; WO '307 (if added) supplies the UTRAN transport-format vocabulary ("block size") so the combination lands exactly on the claim language.

Ground C — WO 99/52307 (or WO 01/63856) in view of US 2003/0081628 (Cognio)

Proposed rejection. WO '307/'856 provide the UMTS transport-format link and load-conditioned allocation; Cognio provides per-user packet-length assignment where the length for a user is computed from the data rate of other users and dynamically re-computed as conditions change, with lengths capped at a per-user maximum. Cognio is not restricted to 802.11 — its rationale (matching the transmission unit to the user's rate to prevent low-rate users from starving the medium) is a general shared-medium insight a POSITA would apply to any multi-user link, including Iub. This ground is especially strong against claim 5 (see § 8).

Ground D — US 5,649,299 alone (for certain limitations) combined with the specification's admitted prior art

Because the '254 specification admits that TFB mapping from a TFS and largest-TFB selection were known UMTS practice, a rejection can rest on the admission plus US 5,649,299 for the load-conditioned, count-based reduction of some users — without needing a second reference to teach "block size." Under KSR, the remaining increment (conditioning the number of reduced users on load) is disclosed by US '299's occupancy-threshold/random-M teaching.

---

7. Why a POSITA would have combined — motivation, consolidated

1. Common problem, common solution space. All five references confront the same failure mode: aggregate offered load on a shared transport exceeds capacity, degrading all users. The solution space is finite and identified: (a) shrink the transmission unit per user (US '391, Cognio), and/or (b) cut the rate of a subset of users (US '299, WO '856). Combining (a) and (b) is the predictable union of two known remedies.
2. Same architecture, same parties. WO '307 and WO '856 are Ericsson filings on the identical UTRAN entity model; combination is not hindsight — it is the design space the references themselves define.
3. Design incentive (KSR factor). Reducing offered rate to fit capacity while preserving service differentiation is a "design need" driven by the QoS classes (conversational → background) already standardized in UMTS.
4. Predictability. Selecting a discrete unit size from an enumerated set (TFS) and throttling a computed number of users to hold aggregate rate below capacity is a bounded, predictable optimization with no unexpected result.
5. "Obvious to try." Where the art discloses both load measurement and unit-size selection and rate reduction, the only remaining variable is how many users to reduce — a value a POSITA would determine by routine calculation, which is exactly claim 2's arithmetic.

---

8. Dependent claims

Claim 2 (sustainable rate R_s; reduced rate R_TFj; number N_L). This is the narrowest and most defensible claim. No cited reference literally discloses the formula R_s = R·T/N, the selection of R_TFj ≤ R_s, or solving for the minimum N_L. However, it is still vulnerable: (i) US '391 discloses storing the packet-length-vs-flow-rate relationship as a "table of corresponding values" or "an arithmetic expression" — i.e., a rate/load-to-size mapping, the genus of which claim 2 is a species; (ii) US '299 discloses threshold comparison and selecting a number M of users to reduce; (iii) deriving N_L from a conservation-of-rate equation is a routine algebraic design choice (§ 103 "design choice" / KSR "predictable variation"). Expect the applicant/examiner to litigate claim 2 on the ground that the particular closed-form optimization is not taught — and expect a § 103 rejection to survive only if supported by a declaration or a third reference teaching rate-proportional allocation. Confidence that a two-reference ground invalidates claim 2: moderate.

Claim 3 (reduced rate within a bearer rate). WO '307's bearer/service model plus its transport-format mapping within a bearer directly supports this; the '254 specification itself concedes bearer rates are held constant for a connection. Claim 3 adds little. Confidence: high.

Claim 4 (within bearer rate and priority class). US 5,649,299 is a direct hit: its alternative flowcharts reduce low-priority subscribers first, and distinguish "HIGH PRIORITY SUBSCRIBER UNITS" from "LOW PRIORITY." Combined with WO '307's bearer/rate classes, claim 4 is the weakest dependent. (Note the '254's own multi-priority extension is disclosed only as Table 7 in the specification — a claim drawn to it would face a written-description/enablement question as well, but that is a § 112 issue, not § 103.) Confidence: high.

Claim 5 (TFB formed from transport blocks; number of transport blocks changed as a function of the TFB sizes of at least one other user). This is the claim where Cognio (US 2003/0081628) is unusually strong: its fixed-duration scheme derives each user's packet length from the fastest user's data rate — i.e., from at least one other user — and dynamically re-computes it as users come and go. Combined with WO '307's transport-format vocabulary (a TFB formed of one or more transport blocks), claim 5 is well within reach. Confidence: high.

Net claim-level conclusion: claim 1 is the broadest and the most exposed; claims 3–5 are also vulnerable on the cited art; claim 2 is the residual redoubt and likely the only claim where the cited-art set alone may be insufficient without an additional teaching or a design-choice rationale.

---

9. Where the applicant would push back, and the rebuttals

• Argument: references are from different fields (LAN, 802.11 WLAN, computer-telephony). Rebuttal: US '391 and US '299 are structurally analogous art — both control transmission-unit size/rate on a shared communications medium under load. WO '307 and WO '856 are in the very field of the '254 (UMTS UTRAN). Under KSR, "familiar elements … according to known methods" is sufficient; and analogous art is judged by the problem to be solved, not the label of the network.
• Argument: no reference teaches conditioning the number of users reduced on load (limitation 1g). Rebuttal: US '299's "RANDOMLY SELECT M LOW … USING HIGH RATES," where M varies with occupancy, is that teaching.
• Argument: the '254's FRC rejects the "always largest TFB" practice. Rebuttal: that practice is admitted prior art, and the departure from it — load-conditioned sizing — is precisely what US '391 and Cognio teach.
• Argument: secondary considerations. The record contains no evidence of unexpected results, long-felt but unmet need, industry praise, licensing-due-to-merit, or commercial success with nexus. The patent is expired (adjusted expiration 2024-08-22), and — per the litigation section already produced — no assertion, IPR/PGR/CBM, or 2026 CAFC activity was found. Absent objective indicia, there is nothing to rebut the KSR case.

---

10. Caveats and confidence

• Source depth. My mapping rests on the abstracts, "Definitions"/summary passages, and figure legends of the five cited references as surfaced by live search (URLs inline in § 4). I did not perform a full-text, column-and-line, claim-chart export of each reference; individual element citations should be confirmed against the reference PDFs before filing a rejection or an invalidity contention. Element mappings marked ⚠️ in § 5 are implicit/inherent rather than express, and are correspondingly weaker.
• Antedating. US 2003/0081628 qualifies only via its 2001-10-30 provisional (60/330,755); if that priority were successfully challenged, it would fall out as pre-priority art and Ground C would need to be rebuilt on WO '307/'856 plus US '391/'299.
• The US 5744902 A string appearing in the record's metadata next to application US10/057,449 remains flagged as an unverified parse artifact (per the earlier verification note) and is not treated as prior art here.
• Bottom line: on the cited-art set, claim 1 is very likely obvious (strongest grounds: US '391 + US '299; or WO '307 + WO '856); claims 3, 4, and 5 are likely obvious; claim 2 is the hardest to reach with only these five references and would most likely require an additional reference or a design-choice/routine-optimization rationale to fall.

If useful, I can convert Ground B into a formal claim chart with reference-specific paragraph/page cites, or draft an examiner-style § 103 rejection with the KSR rationales enumerated per MPEP § 2143.