US 8817805

Patent Overview: US 8,817,805 B2

Title: Apparatus, system and method for the transmission of data with different QoS attributes

Patent number: US 8,817,805 B2 (application US 14/102,120; publication US 2014/0140351 A1)

Inventor: Anthony Gerkis

Assignee/current assignee (as listed): Quarterhill Inc; original assignee WiLAN Inc.

Key dates: Priority date 2002-07-15 (assumed); filing date 2013-12-10; granted/publication date 2014-08-26; anticipated expiration 2023-07-11; legal status "Expired – Lifetime." Family has litigation and a PTAB case (IPR2020-00303, not instituted), but these are not analyzed here per instructions.

Purpose (from the supplied "Definitions" excerpt): The disclosure relates to an apparatus, system and method for transmitting data flows having different quality of service (QoS) attributes over a network link, and more specifically to providing and managing QoS for data flows transmitted over at least one link in a data network capable of transmitting data with different QoS requirements and/or attributes.

Plain-language scope (grounded in the supplied text): The patent concerns managing how data traffic with differing QoS needs is queued and transmitted over a network link. The excerpt's "Prior art keywords" include "logical channel," "data," "queues," "channel queues," and "quality," and the classifications cover traffic control, flow/congestion control, traffic shaping, queue scheduling (e.g., individual queue per QoS/priority), wireless resource allocation based on QoS, and negotiating QoS/service-level agreements. Together these indicate the subject matter involves segregating traffic into queues associated with logical channels and scheduling transmission according to QoS attributes.

Source: https://patents.google.com/patent/US8817805/en

What the supplied excerpt does not establish: It does not provide the claims, full specification, drawings, or detailed embodiments; therefore the precise claimed scope, elements, and limitations cannot be stated. Absent from the excerpt are the exact data structures, scheduling algorithms, and how logical channels map to queues. It also does not establish the real party in interest, the accuracy of the assignee listing, validity, or the content/outcome of any prior art. Dates are labeled as assumptions in the source. Claims of priority to related applications (e.g., US 14/292,380; US 14/506,510; US 15/229,869; US 15/456,218; US 16/025,504; US 16/692,729) appear in the family listing but their substance is not contained in the excerpt.

I'll research litigation involving US patent 8817805. Let me run several searches across litigation-tracking sources.

Let me dig deeper into specific cases and the litigation records referenced.

Litigation Involving U.S. Patent No. 8,817,805

I searched litigation-tracking and patent sources for the exact number 8,817,805 ("the '805 patent"), titled "Apparatus, system and method for the transmission of data with different QoS attributes," originally assigned to WiLAN Inc. (now Quarterhill Inc.), priority date 2002-07-15.

Important caveat up front: I was limited to a fixed number of search steps and hit the cap before I could fully verify outcomes on PACER/CourtListener and the Unified Patents litigation portal. The Google Patents family-history data for the '805 patent (which I treat as authoritative for the IDs) affirmatively flags two enforcement events, and I found corroborating case records for one district-court action and one PTAB proceeding. I have not independently confirmed a complete, exhaustive docket list. Treat the items below as the identified matters, with statuses flagged.

---

Identified matters

1. District court — Wi-LAN Inc. et al. v. Huizhou TCL Mobile Communication Co. Ltd. et al.

Notes: A July 24, 2019 NOTICE OF DISMISSAL and a simultaneously filed First Amended Complaint appear in the docket snippet, consistent with a party/claim adjustment (likely dropping/replacing certain TCL entities) rather than a full termination. The case is captioned in the docket as Wi-LAN Inc. et al v. Huizhou TCL Mobile Communication Co. Ltd. et al.

2. PTAB — Huizhou TCL Mobile Communication Co. Ltd. et al. v. WI-LAN Inc. et al.

This IPR is the administrative counterpart to the C.D. Cal. case above (same TCL petitioner group), indicating the defendants challenged the '805 patent at the PTAB while the district case proceeded.

---

What I could NOT confirm

• No other district-court cases naming the '805 patent were verified. The '805 patent is a member of a large Wi-LAN/Quarterhill family (U.S. 7,333,435 → 8,184,661 → 8,630,238 → 8,817,805 → 8,867,351 → 9,426,816 → 9,603,159 → 10,045,356 → 10,779,288 → 11,229,032, plus CA 2,393,373 and PCT/CA2003/001043). Wi-LAN has sued many parties (Apple, LG, HTC, Samsung, etc.) on related patents, but I did not find those cases expressly asserting the '805 patent specifically — do not attribute them to the '805 patent without verifying the asserted-patent list.
• The Google Patents family record also references a single "First worldwide family litigation" entry (Darts-IP family 30005550) and a C.D. Cal. case, which align with matter #1 above.
• The 2017 assignment chain (Wi-LAN → Quarterhill Inc. → Wi-LAN Inc.) and 2023 anticipated expiration are ownership/status events, not litigation.

Recommended next steps for a definitive list

1. PACER / CourtListener for C.D. Cal. 8:19-cv-00870 — pull the docket and the First Amended Complaint to confirm which of the '805 family patents were asserted and the final disposition.
2. Unified Patents litigation portal (https://portal.unifiedpatents.com/litigation/caselist) and RPX Insight for any additional defendants, since NPE/Quarterhill suits are often filed in parallel batches (e.g., E.D. Tex. and S.D. Cal.).
3. USPTO PTAB (PTAB E2E / PTAB API) for any further IPRs, PGRs, or ex parte reexams directed at the '805 patent beyond IPR2020-00303.

If you'd like, I can attempt targeted docket lookups for 8:19-cv-00870 and a PTAB number sweep for the '805 patent to close the open "status" gaps.

I'll start by checking the canonical proceedings block against what the patent page itself indicates, then fill in the details via web search.

I found a PTAB proceeding that the ODP block missed. Let me pull the details.

Let me check for any other IPRs against this patent and get the institution-decision reasoning.

⚠️ Threshold correction to the structured data

The "PTAB proceedings on file" block says ODP returns no AIA trial proceedings. That is wrong — or at minimum stale. Web search confirms IPR2020-00303 exists, was filed 2019-12-23, and was decided on institution 2020-08-05. The Google Patents page for this patent itself carries the banner "PTAB case IPR2020-00303 filed (Not Instituted - Merits)."

One docketing trap to flag up front: Docket Alarm indexes Paper 7 of IPR2020-00303 under the label "Institution Decision – Trial InstitutEd Document." That is a docket-label artifact. The document's own caption reads "DECISION Denying Institution of Inter Partes Review 35 U.S.C. § 314." Do not let anyone read the label as an institution. (Consistent with a denial, the record shows a fee refund request on 2023-07-13 and refund approval on 2023-07-18.)

---

Proceedings overview

There is one AIA trial proceeding on file against US 8,817,805: IPR2020-00303, institution denied (merits) — zero active, zero FWDs, zero claims canceled, zero claims sustained, zero settlements before the Board. No claim of the '805 patent has ever been tested at trial, so the bottom-line defensive posture is the opposite of "hardened": the patent is completely unlitigated on validity at the PTAB, which cuts both ways — the patent owner cannot point to a single PTAB win as validation, but a defendant likewise gets no § 315(e)(2) estoppel benefit from the denial, and the TCL art remains fully available to re-raise.

---

IPR2020-00303 — Huizhou TCL Mobile Communication Co. Ltd., TCT Mobile (US) Inc. & TCL Mobile Communication (HK) Co., Ltd. v. Wi-LAN Inc.

• Type: Inter Partes Review (35 U.S.C. §§ 311–319); pre-AIA § 103 grounds (priority date 2002-07-15)
• Filed: 2019-12-23
• Status: Institution Denied (merits) — institution decision 2020-08-05. Not a settlement, not a discretionary denial; the Board reached the merits and found no reasonable likelihood of prevailing on any challenged claim.
• Judge panel: Garth D. Baer, Kamran Jivani, and Scott E. Bain, Administrative Patent Judges. Judge Jivani authored. (institution decision, Paper 7)
• Petition grounds: Challenged claims 1–19 (all three independent claims — 1, 11, 20 were not all reached; the challenged set is 1–19). Three grounds, all pre-AIA § 103:
  • Ground 1: claims 1–3, 6–7, 10–13, 16–19 obvious over WO 01/63855 A1 (Peisa)
  • Ground 2: claims 4–5, 8–9, 14–15 obvious over Peisa in view of US 7,016,366 (Kawarai)
  • Ground 3: claims 1–19 obvious over WO 2001/05098 A1 (Jorgensen) in view of US 5,831,971 (Bonomi)
  • Supporting expert: Declaration of Dr. Robert Akl (Ex. 1003). Petition is 90 pages. (Petition, Paper 2)
• Institution decision: Denied in full, 2020-08-05. The Board's operative sentence: "For the reasons explained below, we do not institute an inter partes review." The denial turned on the petition's failure to show the limitations as construed — per the preliminary response structure, the Board agreed that Petitioner misconstrued claim 11 and misapplied Peisa, failed to show that "available resources for data transmission" are determined "in a frame", and that Jorgensen does not disclose or suggest "selecting one of the . . . logical channel queues based on a first one of the quality of service attributes" or "packaging data from the selected one of the logical channel queues until . . . the available resources are used." I have not independently re-verified the Board's full verbatim reasoning paragraph-by-paragraph beyond the excerpt above; treat the claim-mapping rationale as sourced from the PO Preliminary Response argument the Board credited.
• Final Written Decision: None. Never issued. No claim was canceled, and no claim was adjudicated patentable. Claims 1–19 remain exactly as issued; claims 20–27 were never even challenged.
• Settlement / termination: Not a settlement. The proceeding ended by denial of institution. Patent owner was represented by Barry J. Bumgardner and John Murphy, Nelson Bumgardner Albritton P.C.; the firm's public announcement (2020-08-11) notes it defeated all three TCL petitions — IPR2020-00302, -00303, -00304 — and observes the FY20 institution rate exceeded 50%.
• Appeal: None, and none is legally available. A denial of institution is "final and nonappealable" under 35 U.S.C. § 314(d). There is no FWD to appeal. ([Federal Circuit practice on § 314(d): Thryv, Inc. v. Click-to-Call Techs., LP, 590 U.S. 45 (2020)])
• Defensive value: Weak as precedent — and that is the point to manage. The denial is not a validity holding; it reflects only that TCL's three-art combination ("Ground 1/2/3") cleared the § 314(a) threshold poorly, chiefly on claim-11 construction and the "in a frame" limitation. Because no FWD issued, no § 315(e)(2) estoppel attached to TCL, so TCL and its privies can re-assert Peisa, Kawarai, Jorgensen and Bonomi in district court, and any new petitioner can file the very same art tomorrow. For a defendant, the denial is a playbook, not a shield for the patent owner: it tells you which claim constructions the Board found persuasive for the patent owner and which reference mappings failed.

---

Related, non-AIA developments a defendant must know

Not AIA trials, but they belong in the threat picture:

• Ex parte reexamination, Control No. 90/014,588 — "Order Granting Request for Ex Parte Reexamination," mail date 2020-11-17. The requester is not identified in the public record I retrieved; the waiver of the 37 C.F.R. § 1.530 statement was filed 2020-11-20. This is a third-party validity attack on the same patent running outside the AIA-trial framework — it is not subject to § 315(b) time bar or § 315(e) estoppel, and it cannot be settled away by the requester. Verify the current status of 90/014,588 before relying on any claim-validity conclusion — the search results do not show me a reexamination certificate for the '805 patent, unlike the sibling '688 patent (certificate issued 2022-01-19 per the TCL complaint excerpt).
• District court assertion — Wi-LAN Inc. v. Huizhou TCL Mobile Communication Co. Ltd., No. 8:19-cv-00870-JVS-ADS (C.D. Cal.). Filed May 2019, asserting the '688, '805 and '577 patents. A stay motion was decided 2020-02-07.
• Sibling-patent PTAB activity (not this patent, but same portfolio/PO): LG Electronics filed six IPRs against Wi-LAN in 2018 (IPR2018-00673, -00674, -00704, -00705, -00709, -00710); Wi-LAN appealed the IPR2018-00705 FWD to the Federal Circuit (PO Notice of Appeal 2019-08-15). None of those six challenged the '805 patent.

---

Strategic summary

Claim status. Take the claim set as it stands today and bucket it honestly: CANCELED: none. SUSTAINED (by any PTAB/FWD): none. UNTESTED: all of them. Claims 1–19 were petitioned in IPR2020-00303 but never instituted, so they carry no PTAB validity adjudication whatsoever. Claims 20–27 (the "traffic shaping rate / allocating remaining capacity in priority order" method claims) were never petitioned at all. Any assertion letter from Quarterhill/Wi-LAN that cites claims 1–19 is citing claims that a petitioner failed to get over the § 314(a) threshold once — but there is no cancelation certificate, no estoppel, and no Board finding of patentability to hide behind. Do not overstate this either way.

Estoppel landscape. Because IPR2020-00303 produced no final written decision, § 315(e)(2) estoppel never attached to anyone. TCL and its privies are not barred from raising Peisa, Kawarai, Jorgensen, or Bonomi in the C.D. Cal. action or any later litigation — they already preserved that position. For a different defendant being asserted today, every prior-art ground you can find is structurally available, including the entire IPR2020-00303 art set; there is no IPR estoppel web to work around. The § 315(b) one-year bar runs from service of a complaint asserting the patent, so time your own petition filing accordingly. Caveat: the ex parte reexamination 90/014,588 could, if it produced a certificate canceling or amending claims, create intervening rights and claim-scope changes — pull the reexam file history before drafting invalidity contentions.

Pattern signals. No defensive aggregator is in the chain for the '805 patent — Unified Patents is not a petitioner here; the sole petitioner was TCL, an operating-company competitor defendant. Contrast that with Unified's history elsewhere (e.g., Unified's IPR against Dolby's U.S. 10,237,577, FWD 2022-06-15, Board found no challenged claim unpatentable). Wi-LAN/Quarterhill is a serial PTAB participant on both sides: it has defended IPRs on this family (defeating all three TCL petitions in 2020) and it has appealed adverse FWDs to the Federal Circuit (IPR2018-00705). The patent owner here has demonstrated it will fight at the Board on preliminary responses rather than settle, and it has a firm — Nelson Bumgardner Albritton — with a track record of winning institution denials on this exact patent family. Expect the same early-stage, POPR-heavy defense if you file.

The honest bottom line: the '805 patent has attracted exactly one AIA trial and defeated it at the institution stage. That is a signal of a cash-rich, litigation-experienced owner and a technically contested claim set — not a signal of institutionalized patentability. A well-financed defendant that fixes the claim-construction and "in a frame" gaps the Board identified in 2020 is not foreclosed by anything that happened in IPR2020-00303.

Recommended next steps

1. Pull the full IPR2020-00303 record from PTAB E2E and read the Paper 7 institution decision end-to-end. It is the single highest-value document available on this patent — link: Docket Alarm/IPR2020-00303 record. Focus on (a) the Board's treatment of "available resources for data transmission . . . in a frame," and (b) the Jorgensen-based Ground 3, which attacked all of claims 1–19 and therefore is the ground with the broadest reach if you can cure the deficiencies.
2. Read the Patent Owner Preliminary Response (Paper 6, 2020-05-06) as your attack map in reverse — it is the patent owner's own statement of what it thinks the claims don't cover, and those constructions now bind it in district court under the Phillips standard applied at the PTAB.
3. Check the status of ex parte reexamination 90/014,588 (granted 2020-11-17) in Patent Center. No reexamination certificate for the '805 patent surfaced in my search — confirm whether prosecution is still open, whether claims were amended, and whether intervening rights are in play.
4. Do not cite IPR2020-00303 as an invalidity finding. It is a denial of institution under § 314(a). If you represent the defendant and need a merits vehicle, you need to file your own petition — and you have a clean slate to do it, because there is no § 315(e)(2) estoppel and no FWD on this patent.
5. If you represent the patent owner, the absence of any FWD means you cannot rely on Board validation, and the § 314(d) bar means there is no appeal avenue from the denial — but also no estoppel wall protecting you. Assume a re-filed petition.

Verification note: I found no AIA trial other than IPR2020-00303 for US 8,817,805, and the patent's own Google Patents record lists only that one PTAB case. My search was targeted, not exhaustive — before filing anything, run a PTAB E2E / Patent Center docket check on this patent number to confirm no 2024–2026 petition exists that neither ODP, Google Patents, nor the search results above have surfaced.

I'll start by verifying the ownership events against live sources before writing anything.

Ownership reconstruction — US 8,817,805 B2

Critical data caveat, stated up front: The authoritative source I was given (the Google Patents full-text record, fetched 2026‑09‑24) renders the legal events for this patent but does not expose reel/frame numbers or correspondent-of-record fields. My targeted searches for the underlying Assignment Center records returned no reel/frame data for this patent. I therefore report the conveyance events I can actually ground and explicitly mark reel/frame and correspondent as not retrievable. I am not going to invent reel numbers or attorney names to fill those slots — that would be fabrication, and in an ownership analysis it would be worse than useless. Verification path is given at the end.

---

Inventors

Pattern notes:

• Single-inventor patent — no co-inventor departure analysis is possible.
• I found no evidence of inventor departure, and the sole inventor does not appear as an assignor on any recorded conveyance in the chain. No inventor-assignment-to-employer record (the standard "assignment of inventors' interest" from Gerkis) surfaced in the available data; this is a data gap, not a finding.
• The patent is a five-generation continuation family (see below), which is itself relevant: the chain of title was refreshed at each continuation filing.

---

Original assignee

• Entity named on the issued patent: Wi‑LAN Inc. (Ottawa, Canada). Google Patents lists "Original Assignee: WiLAN Inc" and "Current Assignee: Quarterhill Inc" (with the disclaimer that Google has not analyzed assignee accuracy).
• Product embodiment: Yes, at least historically. Wi‑LAN was founded in 1992 as an operating company and shipped the AMOSPHERE wideband‑CDMA wireless broadband system (base stations + SOMAport™ CPEs), which the specification names as the embodiment. Wi‑LAN also commercialized early W‑OFDM/OFDMA technology. So the original assignee was, at the 2002 priority date, a genuine operating company with a product reading on the disclosure.
• Primary line of business today: Patent licensing only. Wi‑LAN transitioned to pure monetization and began litigating its portfolio in 2007. In April–June 2017 Wi‑LAN Inc. renamed itself Quarterhill Inc. and Wi‑LAN continued solely as "a key subsidiary of Quarterhill focusing on patent licensing." Per RPX, Quarterhill later sold ~90% of Wi‑LAN to a pair of New York–based investment firms, marking "the end of a pivot away from patent monetization."
• Current status: Operating (publicly traded Canadian entity, TSX/NASDAQ), but the patent-holding entity is a licensing vehicle, not a product company. Not dissolved, not in bankruptcy.

Family note (relevant to title): the priority filing is CA 2,393,373 (2002‑07‑15), international PCT/CA2003/001043 (2003‑07‑11), US national phase 10/521,581 → US 7,333,435, then continuations 12/028,365 → US 8,184,661, 13/468,925 → US 8,630,238, and this case 14/102,120 → US 8,817,805, with five further continuations issued afterward (US 8,867,351; 9,426,816; 9,603,159; 10,045,356; 10,779,288; 11,229,032; and US 2022/0110131). The same assignment events recur across all of them.

---

Assignment timeline

Reel/frame and correspondent fields are not present in the source record and could not be retrieved; every other field below is as rendered in the authoritative Google Patents legal-events data for US 8,817,805.

• Executed date not shown / recorded 2013‑12‑10 — Reel not retrievable

  • Conveyance: Assignment of Assignors' Interest ("ASSIGNMENT OF ASSIGNORS' INTEREST — SEE DOCUMENT FOR DETAILS")
  • Assignor: TURTLEBONES, INC.
  • Assignee: WI‑LAN, INC.
  • Correspondent: not retrievable from the source record
  • Context: Internal chain-of-title consolidation — the recording coincides exactly with the 2013‑12‑10 filing of continuation 14/102,120, i.e. a title clean-up by a related entity rather than a third-party sale. (TurtleBones, Inc. also appears in Unified Patents' data as the parent-company field for other Soma Networks/Wi‑LAN-family wireless patents, e.g. US 7,050,407.)

• Executed date not shown / recorded 2017‑06‑20 — Reel not retrievable

  • Conveyance: Merger and Change of Name ("MERGER AND CHANGE OF NAME — SEE DOCUMENT FOR DETAILS")
  • Assignor: WI‑LAN INC. (and QUARTERHILL INC.)
  • Assignee: QUARTERHILL INC.
  • Correspondent: not retrievable from the source record
  • Context: Change of name only — Wi‑LAN Inc. renamed itself Quarterhill Inc. (announced April 2017, name change effective ~June 1, 2017; record dated June 20, 2017). Corroborated by a PTAB filing of record in IPR2017‑01199 stating: "Wi‑LAN Inc., has changed its name to Quarterhill Inc." (dated June 21, 2017).

• Executed date not shown / recorded 2017‑07‑12 — Reel not retrievable

  • Conveyance: Assignment of Assignors' Interest
  • Assignor: QUARTERHILL INC.
  • Assignee: WI‑LAN INC. (the newly formed Delaware subsidiary of Quarterhill)
  • Correspondent: not retrievable from the source record
  • Context: Internal reorganization — assets spun from the renamed public parent down into a purpose-formed, licensing-only subsidiary. Confirmed in the same IPR2017‑01199 record: "Quarterhill Inc., formed a new, wholly‑owned subsidiary, Wi‑LAN Inc., and transferred certain assets from Quarterhill Inc., to Wi‑LAN Inc."

Family litigation / assertion events (not assignments, listed for the NPE analysis):

• 2019 — Case filed, U.S. District Court, Central District of California, 8:19‑cv‑00870 (per Unified Patents litigation data).
• 2019‑12‑23 — IPR2020‑00303 petitioned by Huizhou TCL Mobile Communication Co. Ltd. against Wi‑LAN INC. on this patent (application 14/102,120). Google Patents labels the outcome "Not Instituted – Merits"; IPVerse reports "Institution Denied" at the Aug. 5, 2020 decision date. A refund was requested and approved (2023).
• The patent is therefore an asserted asset, and Wi‑LAN/Quarterhill is tracked by RPX and appears in Unified Patents' PTAB dataset as a repeated respondent.

---

Timeline diagram

timeline
    title Ownership of US 8817805
    2002 : Priority application filed in Canada
    2003 : PCT and US national phase filed
    2013 : Continuation filed with WiFiLan
         : TurtleBones assigns interest to WiLan
    2014 : Patent issued to WiLan
    2017 : WiLan renamed Quarterhill
         : Assets assigned to new WiLan subsidiary
    2019 : Suit filed in California Central District
         : IPR petition filed by TCL
    2020 : IPR institution denied

---

NPE / troll-pattern signals

1. Shell-entity transfer — present (qualified).
The chain terminates inside an entity purpose-built for asset holding: on 2017‑07‑12 Quarterhill assigned the asset to a newly formed, wholly-owned Delaware subsidiary named Wi‑LAN Inc., which exists solely to run "patent licensing." That is a textbook transfer into a licensing-only vehicle (recorded 2017‑07‑12, same family litigated by "Wi‑LAN Inc." as Patent Owner). Qualification: Wi‑LAN is not an anonymous single-member shell of the MPHJ type — it is a large, publicly disclosed, 10,000‑asset licensing operation with a named corporate parent. The "shell/registered‑agent address" sub‑tells (Delaware single‑member LLC at a registered-agent address) are not evidenced for this chain, so I mark the signal present on the licensing-only-entity criterion only, not on the anonymity criterion. I could not verify the registered address because the correspondent field was unavailable.

2. Known asserter in the chain — present (strong).
The task's own NPE list explicitly names Wi-LAN. Corroboration beyond the list itself: RPX describes Quarterhill as a publicly traded entity that "began litigating its patents in 2007"; IAM describes Wi‑LAN as "the Canadian NPE" and a repeat user of "privateering type deals"; Wi‑LAN appears as Patent Owner/respondent across numerous PTAB proceedings (e.g. IPR2017‑01199, IPR2018‑00705) and as appellant in Wi‑LAN v. Apple, Fed. Cir. 15‑1256. Assignee identity is confirmed in the 2013‑12‑10, 2017‑06‑20 and 2017‑07‑12 recordings.

3. Repeat correspondent across the chain — insufficient data.
The source record does not expose the correspondent of record for any of the three recordings, and my searches did not surface it. I am declining to name an attorney or firm. This is the single highest‑value field in the requested analysis and it is the one field I cannot ground — treat it as an open work item, not as a negative finding.

4. Cascading transfers — present (weak-to-moderate).
Two consecutive rounds of recorded title movement inside a ~22-day window (2017‑06‑20 merger/name change; 2017‑07‑12 assignment into the new subsidiary), on top of a 2013 confirmatory assignment — three recordings total. They share a single corporate principal (Quarterhill), which is consistent with chained intra-group transfers rather than a genuine market cascade. Multiple transfer steps, single economic owner.

5. Pre-litigation transfer — not present.
The last recorded transfer is 2017‑07‑12. The first suit naming this patent that I could confirm is the 2019 California Central District case (8:19‑cv‑00870), and the IPR campaign is 2019‑12‑23. That is roughly 24 months, well outside the 6‑month window. (Caveat: the '805‑family patents '040 and '640 were also asserted against Apple in S.D. Cal. 13‑cv‑0798, decided 2013–2015, but that suit predates issuance of the '805 patent and I found no confirmed '805 count earlier than 2019. If an earlier 2014–2016 '805 assertion exists, this signal would need re‑running.)

6. Bankruptcy fire-sale — not present.
Quarterhill/Wi‑LAN divested its licensing arm to investment firms (per RPX) and restructured by name change and internal spin‑down — but there is no Chapter 7/11 proceeding for the original assignee in any record I retrieved. Divestiture ≠ insolvency.

7. Privateering — unclear (chain-specific: not present).
Wi‑LAN's corporate model is documented by IAM as including privateering-type deals "where the seller receives a portion of any revenues." However, for this patent the chain is entirely intra‑group (TurtleBones → Wi‑LAN → Quarterhill → Wi‑LAN) — there is no operating-company‑to‑NPE transfer of this asset, and no evidence this patent was asserted on a third party's behalf. The privateering characterization applies to Wi‑LAN generally, not to this chain.

8. Defensive aggregator — not present (inverse confirms assertion).
The chain does not terminate at RPX, AST, LOT, Unified Patents or OIN. The opposite is true: Unified Patents' data tracks this patent through a TCL‑filed IPR (IPR2020‑00303), i.e. a defensive aggregator challenged the patent rather than acquiring it. The asset remains in the asserter's hands.

---

Verdict

NPE — high confidence.

Two strong signals plus one supporting signal, each grounded in a dated recording or filing rather than in naming: (a) the chain terminates in a licensing-only entity created for that purpose — Quarterhill's 2017‑07‑12 assignment of the asset to the newly formed Wi‑LAN Inc. subsidiary, corroborated in the IPR2017‑01199 record and in RPX/IAM coverage; and (b) the assignee is Wi-LAN, a named entry on the NPE list supplied in this task, an entity RPX dates to litigating since 2007 and IAM labels "the Canadian NPE." Supporting: (c) cascading intra-group transfers on 2017‑06‑20 and 2017‑07‑12, and an asserted posture (2019 C.D. Cal. suit; IPR2020‑00303 filed by TCL against Wi‑LAN on 2019‑12‑23, institution denied Aug. 5, 2020 per Google Patents/IpVerse).

Honest limits on this verdict: the correspondent-of-record field was unavailable, so signal #3 — the strongest structural tell in this methodology — is unresolved, and the pre‑litigation-transfer window (+6 mo.) is not met because the last recording (2017‑07‑12) predates the confirmed first assertion (2019) by roughly two years. Neither gap changes the direction of the evidence, but both would need closing before relying on this in a contentious filing.

Verify at:

• USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search application 14/102,120 or patent 8,817,805; request the correspondent/correspondent‑address fields, which are the missing piece here)
• USPTO Assignment legacy index — https://assignment.uspto.gov/patent/index.html
• Google Patents legal events for this record — https://patents.google.com/patent/US8817805/en
• PTAB — IPR2020‑00303: https://portal.unifiedpatents.com/ptab/case/IPR2020-00303
• District court docket: https://portal.unifiedpatents.com/litigation/California%20Central%20District%20Court/case/8%3A19-cv-00870

I'll research the prior-art citations for US 8,817,805. Let me run targeted searches.

Let me drill into the citation and PTAB records for this specific patent number.

I found the PTAB petition's prior-art set. Let me now retrieve the patent's own front-page citations and the reexamination references to be thorough.

Prior Art for US 8,817,805 — Analyst Findings

Scope and sourcing note (read first)

I searched for this specific number (8,817,805) and its citation/prior-art records. Two important limitations:

1. The authoritative full text supplied in this task does not contain the patent's own front‑page "References Cited" / "Patent Citations" listing. The Google Patents fetch that was provided is truncated after the abstract/description/claims and before the citation tables. I therefore cannot enumerate the examiner‑cited U.S. patent references from the primary source, and I will not invent them. I report that gap explicitly rather than filling it.
2. What I could retrieve is the fully documented prior‑art set used against this exact patent in the PTAB proceeding IPR2020‑00303 — i.e., the art that a petitioner and the Board actually treated as the most relevant prior art to the '805 patent. That is the substance of Section B/C below.

Critical framing caveat on § 102: the task asks which claims each reference "potentially anticipates under 35 U.S.C. § 102." In the actual record, the petitioner asserted these references as § 103 obviousness combinations, not as § 102 anticipation grounds. I flag this so the § 102 characterizations below are not mistaken for holdings. Where a reference is used alone it is still framed as § 103 obviousness, not anticipation.

---

A. The patent's own citation record

• The supplied authoritative text does not include the "References Cited" section for US 8,817,805, so I cannot list the front‑page patent citations the examiner considered. Gap, not verified.
• The only "references" visible in the body of the patent are background/technical documents the specification itself discusses (Section E below).

---

B. Documented prior‑art set — PTAB IPR2020‑00303 (TCL v. Wi-LAN), filed 2019‑12‑23

Petitioner: Huizhou TCL Mobile Communication Co. Ltd., TCT Mobile (US) Inc., and TCL Mobile Communication (HK) Co., Ltd. Patent Owner: Wi‑LAN Inc. Outcome: Institution Denied (decision 2020‑08‑05). Challenged claims: 1–19.

Exhibits (as captioned in the Petition/Schedule):

(Non‑prior‑art exhibits: TCL‑1002 prosecution file history of App. 14/102,120; TCL‑1003 Akl declaration; TCL‑1004 Bright declaration; TCL‑1009 claim‑construction order for US 8,867,351, Wi‑LAN v. LG, S.D. Cal., May 28, 2019.)

Grounds as pleaded (per Patent Owner's Preliminary Response):

• Ground 1 — § 103 over Peisa alone: claims 1–3, 6–7, 10–13, 16–19.
• Ground 2 — § 103 over Peisa in view of Kawarai: certain dependent claims (the traffic‑shaping‑rate claims — e.g., 4–5, 8–9, 14–15).
• Ground 3 — § 103 over Jorgensen in view of Bonomi: claims 1–19.

The Board denied institution on 2020‑08‑05, so none of these references was adjudicated invalidating.

---

C. Reference‑by‑reference (citation, dates, description, claimed relevance)

Dates below marked (unverified) were not confirmed in my retrieved sources; I give them only where confident and flag otherwise. All are pre‑AIA art relative to the 2002‑07‑15 priority date.

1. WO 01/63855 A1 — Janne Peisa ("Peisa")

• Full citation: International (PCT) Patent Publication WO 01/63855 A1, inventor Janne Peisa.
• Date: Published in 2001 (WO "01/…" series). (Exact publication day unverified.) Because its number is high in the 2001 sequence, its publication likely falls after 2001‑07‑15 — i.e., less than one year before the 2002‑07‑15 priority date — which would place it under pre‑AIA § 102(a) (and potentially § 102(e) as a PCT designating the US). It is not more than one year before priority, so it is an unlikely § 102(b) reference. (Treat the statutory subsection as conditional on the exact date.)
• Description (role in the record): Peisa is the primary/lead reference in Grounds 1 and 2. Per the petition structure, Peisa was relied on for the core queue‑selection/packaging workflow — logical channels/queues with QoS attributes, selecting a queue and transmitting its data — i.e., the subject matter of independent claims 1 and 11. I did not retrieve Peisa's full text, so I do not quote it; the characterization is from the ground mapping, not from the reference itself.
• Claims it was asserted against: 1–3, 6–7, 10–13, 16–19 (Ground 1); combined with Kawarai for 4–5, 8–9, 14–15 (Ground 2).
• "§ 102" note: Peisa was asserted alone but under § 103, not § 102. If Peisa in fact discloses, in a single reference, determine‑resources → select‑by‑priority → package‑until‑stop‑condition, then claims 1–3, 6–7, 10–13, 16–19 are the claims most susceptible to a § 102 challenge — but that is my inference from the mapping, not a record finding.

2. U.S. Patent No. 7,016,366 — ("Kawarai")

• Full citation: U.S. Patent No. 7,016,366.
• Date: Granted 2006‑03‑21 (from memory; grant date not confirmed in retrieved sources); filed on an earlier application. Pre‑AIA § 102(e) is the most likely statutory basis (US patent granted on an application filed before the '805 invention). (Verify filing date before relying on § 102(e).)
• Description (role): Secondary reference combined with Peisa in Ground 2, apparently to supply the traffic‑shaping data‑rate limitation (the "second QoS attribute," and the leaky‑bucket/prioritized‑bit‑rate features of dependent claims 4–5, 8–9, 14–15 and claim 21).
• Claims it was asserted against: the traffic‑shaping‑rate dependents (4–5, 8–9, 14–15).
• "§ 102" note: Kawarai was used only in combination; it was not asserted to anticipate any claim standing alone. Any § 102 case would require a fully self‑contained disclosure, which the record does not assert.

3. WO 2001/05098 A1 — Jacob Jorgensen ("Jorgensen")

• Full citation: International (PCT) Patent Publication WO 2001/05098 A1, inventor Jacob Jorgensen.
• Date: Published 2001 in the early‑year "01/05098" sequence — likely January 2001 (exact day unverified). If published on/before 2001‑07‑15, it predates the priority date by more than one year and qualifies under pre‑AIA § 102(b).
• Description (role): Lead reference in Ground 3, combined with Bonomi, against claims 1–19. Patent Owner's Preliminary Response confirms the dispute focused on whether Jorgensen teaches "selecting one of the … queues" and "packaging data from the selected one … until …" — i.e., the core limitations of claims 1 and 11.
• Claims it was asserted against: 1–19 (in view of Bonomi).
• "§ 102" note: Asserted as a § 103 combination, not § 102. Of the four, Jorgensen is the most plausible standalone § 102(b) candidate on date grounds (early‑2001 publication, PCT designating the US), if its disclosure is as broad as the petitioner claimed — which Patent Owner contested.

4. U.S. Patent No. 5,831,971 — ("Bonomi")

• Full citation: U.S. Patent No. 5,831,971.
• Date: Issued 1998‑11‑03 (exact grant date unverified) — comfortably before the 2002 priority date; potential § 102(b) art on date grounds.
• Description (role): Secondary reference in Ground 3 (Jorgensen + Bonomi). The petition title/ToC confirms Bonomi is used to supply one or more of the QoS‑attribute/queue‑handling limitations missing from Jorgensen.
• Claims it was asserted against: claims 1–19, only as part of the Jorgensen combination.
• "§ 102" note: Combination‑only; not asserted to anticipate alone.

5. U.S. Publication No. 2003/0067903 A1 — Jacob Jorgensen ("Jorgensen II")

• Full citation: U.S. Patent Application Publication No. 2003/0067903 A1, inventor Jacob Jorgensen.
• Date: Published 2003‑04‑10 (from the 2003/… series; day unverified). NOTE — date problem: a publication dated 2003‑04‑10 is after the 2002‑07‑15 priority date and therefore is not prior art to the '805 patent on its face (pre‑AIA § 102(e) could still reach it via its earlier effective filing date, if any). (Verify its filing/priority date before treating it as art.)
• Description (role): Listed as an exhibit (TCL‑1010), likely the US counterpart/related disclosure to Jorgensen used for support.
• "§ 102" note: Cannot anticipate on publication date alone; only relevant if its § 102(e) filing date predates the invention.

6. U.S. Provisional Application No. 60/185,005

• Date: Provisional filed on/around 2000 (60/185,xxx series, ~Feb–Mar 2000) (unverified).
• Role: Likely the priority/benefit document behind Peisa or Jorgensen (TCL‑1011). A provisional is not itself § 102 prior art as a printed publication; it matters only via the § 102(e)/(b) date it confers on the non‑provisional publication that claims its benefit.
• Claims: none directly.

---

D. Anticipation (§ 102) assessment — with explicit caveat

• No reference in the record was asserted to anticipate any claim under § 102. Each ground was pleaded under pre‑AIA 35 U.S.C. § 103 (obviousness). The Board denied institution, so there is no adjudicated § 102 finding for US 8,817,805.
• On the face of the grounds, the claims most exposed to a single‑reference § 102 theory would be the independent claims 1 and 11 (and their closest dependents 2–3, 6–7, 10, 12–13, 16–19), with Peisa (WO 01/63855) as the candidate under Ground 1 and Jorgensen (WO 2001/05098) as the candidate under Ground 3 — provided either reference alone discloses the full determine‑resources → select‑by‑priority → package‑until‑second‑attribute/stop‑condition loop. I have not verified that either does, and Patent Owner affirmatively argued they do not.
• The traffic‑shaping‑rate limitations (claims 4, 5, 8, 9, 14, 15, 20, 21, 26) were never tied to a single reference; they were supplied only by the Peisa + Kawarai and Jorgensen + Bonomi combinations.

---

E. Other references discussed in the patent body (context, not examiner citations)

The specification itself cites, as background/technical support:

• IETF RFC 791 (1981) — IPv4 TOS byte; RFC 1600 (Reynolds & Postel, 1992) and RFC 1609 (Almquist, 1992) — suggested TOS settings. These are descriptive background, not asserted prior art.
• Published PCT application WO 01/91407 (to Mantha), assigned to the same assignee — described as an embodiment of a suitable broadcast data channel (BDCH). This is the one substantive co‑owned publication referenced in the description.

These are "references cited by the applicant" in the narrative sense; they were not part of the IPR grounds.

---

F. Related administrative record

• Ex parte reexamination, Control No. 90/014,588, requested against US 8,817,805; Order Granting Request for Ex Parte Reexamination, mail date Nov. 17, 2020 (per the Justia family page for US 11,229,032). The reexam references were not retrievable in my searches — gap, flagged.
• IPR2020‑00303 — Institution Denied (2020‑08‑05).
• District court: Wi‑LAN v. Huizhou TCL et al., No. 8:19‑cv‑00870‑JVS‑ADS (C.D. Cal.) (order re motion to stay, Feb. 7, 2020).

---

G. Explicit uncertainties

1. The '805 front‑page citation list is not in the supplied authoritative text, and my searches did not surface it. I therefore do not represent that the references in Section B are the examiner‑cited references; they are the petitioner‑cited references from IPR2020‑00303.
2. Publication/grant dates for Peisa, Jorgensen, Kawarai, and Bonomi are given from the numbering conventions and general knowledge; only the ones you can independently confirm should be relied on for a § 102(b) vs § 102(a)/(e) determination. I flagged each.
3. Titles/abstracts of WO 01/63855, WO 2001/05098, US 7,016,366, and US 5,831,971 were not retrieved; I did not invent them. Descriptions above reflect each reference's role in the grounds, not its text.
4. Jorgensen II (US 2003/0067903) has a publication date that appears later than the priority date; treat its prior‑art status as unconfirmed until its filing/priority date is checked.
5. US Provisional 60/185,005 is a benefit document, not itself printed‑publication art.

If you can supply the front‑page "References Cited" block (the examiner‑cited US patents and any foreign/NPL references), I can complete the reference‑by‑reference § 102 mapping for the art‑of‑record as distinct from the IPR art above.

I'll research the prior art landscape for this patent before analyzing obviousness. Let me run several targeted searches.

Excellent — I've found the actual IPR2020-00303 petition and its grounds. Let me pin down the exact prior-art patent numbers and the reexamination outcome.

Obviousness Analysis — U.S. Patent No. 8,817,805 under 35 U.S.C. § 103

A note on the inputs I was given

The task instructs me to "use the results from the Prior Art section of this page." No Prior Art section was actually supplied in the material handed to me — I received the patent's own full text (claim set, abstract, description), plus previously generated summary and litigation sections. In place of a supplied prior-art section, I reconstructed the operative prior-art record from the primary documents surfaced by search: the IPR2020-00303 petition and the accompanying Declaration of Dr. Robert Akl (Ex. 1003) filed by Huizhou TCL Mobile Communication Co. Ltd., TCT Mobile (US) Inc., and TCL Mobile Communication (HK) Co., Ltd. against Wi-LAN Inc.

I flag this explicitly because the analysis below is grounded in TCL's asserted grounds and references, not in a pre-packaged "Prior Art section." Where I could not confirm a reference's exact identifier (I hit the search-step cap before resolving every patent number), I say so rather than inventing one.

---

1. Governing framework

Under Graham v. John Deere Co., 383 U.S. 1 (1966), obviousness turns on (1) the scope and content of the prior art, (2) the differences between the prior art and the claims, (3) the level of ordinary skill, and (4) objective indicia. Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), a combination is obvious where the references are analogous art, the combination yields predictable results, and there existed a "design incentive" or "market demand" to combine, or the solution was "obvious to try" among "a finite number of identified, predictable solutions."

Level of ordinary skill (my assessment): a POSITA here would have a bachelor's degree in electrical/computer engineering (or equivalent) and roughly 2–4 years of experience in packet-switched wireless MAC-layer scheduling and QoS — consistent with the person described in Dr. Akl's declaration and with the petition's treatment of the field. All asserted references are either in the same field (MAC/network-layer packet scheduling and QoS) or are reasonably pertinent to the problem the '805 patent addresses (providing differentiated QoS over a shared, variable-capacity radio link). That makes them analogous art under In re Bigio / In re Clay principles.

---

2. The prior-art record (as asserted in IPR2020-00303)

TCL advanced three grounds, all under § 103(a), against claims 1–19:

The references (as they appear in the petition record):

• Peisa — a U.S. patent (cited in the Akl declaration as "TCL-1011" and cross-referenced as Ex. 1013 in related briefing). Discloses a UMTS MAC-entity scheduling method "for providing bandwidth to entities in a communications system," in which logical channel queues are mapped to transport channels, multiple transport channels are mapped to a single physical channel, and the MAC entity determines available bandwidth (via the Transport Format Combination Set/Indicator, TFCS/TFCI) and then distributes it among logical channels selected in priority order. Attributes include guaranteed-rate, weighted-fair-queuing (WFQ) rates, and QoS class, with backlog counters used to address congestion. (I could not confirm Peisa's exact patent number within my search budget — treat the name/Exhibit citation as the identifier for now.)
• Kawarai — U.S. Patent No. 7,016,366, filed 2001-03-20, published 2001-10-25 (prior art under § 102(e)/(a)). Discloses improving WFQ to control priorities of a guaranteed-bandwidth class, with per-QoS-class buffers and contention control / drop control per output buffer and QoS class.
• Jorgensen — cited for a MAC-hardware scheduler and segmentation/reassembly (SAR) that runs an "advance reservation algorithm" which "can assign the slots from highest priority to lowest priority, exhausting the number of available slots" in frames, and "flushes" per-IP-flow class queues. (Exact patent number not confirmed within my search budget.)
• Bonomi — cited for a combined leaky-bucket traffic-shaping and fair-queueing shaper in which a cell is assigned "a virtual finishing time indicative of the priority," a cell is "non-conforming" if it violates the connection's leaky-bucket parameters, and transmission of a selected cell is interrupted when the selected cell is non-conforming; the shaper also checks whether the queue is empty. (Exact patent number not confirmed within my search budget.)

Procedural outcome to keep in view: the PTAB denied institution on August 5, 2020 ("Institution Denied — on the merits"). A denial of institution is not a holding of validity; it means TCL did not establish a reasonable likelihood of prevailing on the petition as filed. Separately, the patent family record shows an ex parte reexamination (Control No. 90/014,588), granted Nov. 17, 2020, and the C.D. Cal. tentative Markman order (case 8:19-cv-00870) states in a footnote that claim 11 was cancelled during the EPR. Treat the claim-11 cancellation as reported-but-needs-confirmation. This materially affects the enforceability of claim 11 even though it does not change the § 103 analysis of the original claims.

---

3. Independent claim 1 — method (determine capacity → select by first QoS attribute → package until stop condition → repeat)

Claim 1 elements mapped to the art:

Assessment — Ground 3 (Jorgensen + Bonomi) is the strongest § 103 theory for claim 1. Jorgensen supplies the priority-ordered, frame-based slot-exhaustion loop and the "queue emptied/last portion" stop condition; Bonomi supplies the missing "second QoS attribute" — the leaky-bucket/traffic-shaping rate that halts packaging for a queue before its data is exhausted. Because the claim's three stop conditions are written disjunctively ("one of…") — a point both the petitioner and the Board's claim-construction discussion treat as "or," not "and" — a POSITA reading Jorgensen alone arguably reaches the resource-exhaustion and queue-empty prongs, with Bonomi supplying the shaping-rate prong.

Motivation to combine (KSR):

• Same field / same problem. Both references address congestion management and fair bandwidth distribution among competing flows in packet networks; Bonomi's "combined leaky bucket traffic shaping and fair queueing shaper" is, as the declaration puts it, "simply a variation and an improvement of the WFQ scheme" already used in the scheduling art.
• Predictable result. Layering a per-flow leaky-bucket policer over a priority scheduler is a conventional combination that "yields predictable results and provides the expected functionality" — i.e., enforcing per-flow bandwidth limits while preserving priority order.
• Finite, identified solutions. Traffic-shaping via leaky/token buckets and priority/fair scheduling were among a small set of known, predictable tools for QoS differentiation.

Assessment — Ground 1 (Peisa alone) is a credible single-reference obviousness theory, since Peisa expressly discloses (a) logical channel queues, (b) frame/scheduling-interval determination of available bandwidth, (c) priority-ordered selection of logical channels, and (d) per-channel guaranteed/WFQ rate attributes that function as the "second QoS attribute," continuing "until the whole bandwidth has been allocated." A POSITA would have found claim 1 obvious over Peisa because Peisa is directed to exactly the same problem (QoS-differentiated bandwidth allocation over a shared variable-capacity wireless channel) and supplies every recited mechanism.

---

4. Independent claim 11 — wireless device / "link controller"

Claim 11 is the apparatus counterpart, requiring "a plurality of logical channel queues" and "a link controller adapted to" determine/select/package/repeat. In Grounds 1 and 3, the asserted "link controller" is the MAC hardware / MAC scheduler:

• Peisa: the MAC entity maps logical channels to transport channels and performs the scheduling — a "link controller" in the claim's functional sense.
• Jorgensen: expressly a MAC-hardware scheduler with segmentation/reassembly (SAR) — structurally identical to the patent's "link controller + segmentation cache" architecture.

The device framing (a wireless terminal or access point, claims 18/19) is substantively met by the UE/Node-B entities in Peisa and the MAC hardware in Jorgensen; the patent itself concedes that a UE and its transmit/receive circuitry were routine.

Assessment: claim 11 would likewise be obvious over Peisa, and over Jorgensen + Bonomi, for the same reasons as claim 1. Caveat: because claim 11 was reportedly cancelled in EPR 90/014,588, this analysis is of historical/family value for that claim; the live claim scope (if any) that descended from it must be re-checked against the reexamination certificate.

---

5. Dependent claims 2–10, 12–19

These add limitations that are each separately taught or conventional:

Motivation for Ground 2 (Peisa + Kawarai): the claims 4/8/14/5/9/15 family is directed to the "second QoS attribute = traffic-shaping data rate (leaky bucket)." Peisa already teaches rate-based QoS attributes (guaranteed rate, WFQ rate) applied per logical channel; Kawarai teaches drop control for a guaranteed-bandwidth class keyed to QoS class. A POSITA would combine them to add per-flow rate enforcement (including leaky bucket) to Peisa's priority scheduler — the references are in the same field, and the combination is a predictable application of known rate-limiting to known queues.

---

6. Independent claim 20 (and claims 21–27) — the two-phase allocation claim not challenged in the IPR

Important scope point: IPR2020-00303 challenged only claims 1–19. Claim 20 and its dependents (21–27) were therefore not before the Board and are not covered by the institution-denial. Their validity was not tested by TCL's petition.

Claim 20 recites a two-phase structure over an uplink: (Phase 1) allocate a portion of the frame's uplink capacity to queues selected in priority order, applying a traffic-shaping process based on each queue's traffic-shaping rate, "while the data transmission capacity is not exhausted"; (Phase 2) "thereafter, if the data transmission capacity is not exhausted, allocat[e] the remaining portion … to one or more of the logical channel queues, selected in priority order."

Prior art for claim 20:

1. Peisa is highly material. Its scheduling is expressly multi-stage: a first level sizes the transport-channel bandwidth using guaranteed-rate/WFQ/QoS-class attributes, and the second level sizes the logical channels using priorities / guaranteed rates / WFQ rates — i.e., a shaped, priority-ordered allocation with subsequent redistribution of remaining bandwidth. That maps onto claim 20's "shape in priority order, then fill leftover" structure.

2. The Holma & Toskala WCDMA for UMTS textbook (a standard reference for the ordinary artisan, surfaced in a companion Wi-LAN/UMTS invalidity record) is directly on point for the two-phase idea. It describes both:

   • the "simplest algorithm … where the higher priority bearers will always get the capacity before the lower priority bearers," and
   • a second "approach … to give a minimum bit rate … for all the users and then allocate the remaining capacity for the high priority users."

   The second algorithm is essentially claim 20's Phase 1 (minimum/shaped allocation) followed by Phase 2 (leftover to priority order). A POSITA would have been motivated to implement this textbook approach on a UMTS uplink to balance guaranteed service against efficient residual-capacity use — precisely the "market demand"/design-incentive rationale of KSR.

3. Claims 21, 26 (prioritized bit rate + leaky-bucket size) track the 3GPP prioritized bit rate (PBR) and bucket-size bearer parameters and the Bonomi leaky-bucket shaper; claims 23–25 (uplink radio resources, pool of dedicated data channels, dynamic allocation) track the UMTS/E-DCH dedicated-channel model Peisa describes and the patent itself calls conventional.

Assessment: claims 20–27 are, in my view, the strongest § 103 targets for a future challenge because (a) they were never adjudicated in the IPR, (b) their "shape-first-then-fill" structure is squarely taught by Peisa's multi-stage scheduler and the Holma textbook, and (c) their dependent features (PBR, leaky bucket, dedicated uplink channels, segmentation) are all standard UMTS/E-DCH art. A combination of Peisa + Holma (and + Bonomi for the leaky-bucket claims) would support an obviousness case, with the motivation being the well-known desire to guarantee minimum rates to all bearers while harvesting unused capacity for higher-priority bearers.

---

7. Recurring "motivation to combine" themes (why a POSITA would combine)

Across all grounds, the KSR rationales repeat consistently:

1. Same field of endeavor and same problem. Every reference is directed to allocating bandwidth / enforcing QoS among competing packet flows over a shared network resource — the exact problem of the '805 patent. All are analogous art.
2. Predictable results. Combining a priority scheduler (Peisa/Jorgensen) with a rate-based shaper (Bonomi/Kawarai) is a routine engineering composition; the combination does no more than the parts would be expected to do (priority ordering + per-flow rate limiting + fill residual capacity).
3. Finite, identified, predictable solutions. Priority scheduling, WFQ, guaranteed-rate allocation, and leaky/token-bucket shaping were the small, well-known toolkit for QoS differentiation by the 2002 priority date — "obvious to try."
4. Express teachings in the references themselves. Peisa's own goal ("provide best quality of service to flows having the highest QoS class(es), while still providing a minimum level of service to all flows") supplies the incentive to add per-flow shaping policies; Kawarai and Bonomi expressly address the same congestion/fairness objectives.

---

8. Overall conclusion and confidence

• Claims 1–10 (method) and 12–19: For the reasons above, there is a strong, articulable § 103 case that these claims are obvious over Peisa alone, and independently over Jorgensen + Bonomi (with Peisa + Kawarai covering the traffic-shaping/leaky-bucket dependents). This is essentially TCL's Grounds 1–3. Confidence in the legal framework and mapping: high. Confidence that a fact-finder would have instituted and prevailed: moderate — the Board denied institution on the merits, so the mapping did not clear the "reasonable likelihood" threshold as presented (the Patent Owner's Preliminary Response attacked, inter alia, the "in a frame" and "first QoS attribute" limitations; see § 9).
• Claim 11 (device): same analysis as claim 1, but note the reported EPR cancellation of claim 11 — verify against the reexamination certificate before relying on it. Confidence: moderate (identity of claims post-EPR needs confirmation).
• Claims 20–27: not challenged in IPR2020-00303; my independent assessment is that they are obvious over Peisa in view of the Holma WCDMA for UMTS textbook (and Bonomi for the leaky-bucket claims), with the two-phase "guaranteed/shaped allocation, then leftover-by-priority" structure expressly taught. Confidence in the reference mapping for claim 20 specifically: moderate-to-high, but this has never been adjudicated and would need a full claim chart.

---

9. Key uncertainties and contradictions to flag

1. No "Prior Art section" was provided in the prompt; I reconstructed the prior-art record from the IPR2020-00303 petition/declaration documents surfaced by search. If a specific prior-art section was intended, please supply it so I can re-anchor.
2. Reference patent numbers unconfirmed for Peisa, Jorgensen, and Bonomi. I confirmed Kawarai = U.S. 7,016,366, but I hit the search-step cap before confirming the others' numbers. I have not invented numbers for them; they are identified by party name and by the IPR exhibit labels (e.g., "TCL-1011"/Ex. 1003 Akl declaration).
3. Litigation/reexam facts to reconcile with the earlier sections: the previously generated litigation summary did not mention the ex parte reexamination (90/014,588, granted 2020-11-17) or the reported cancellation of claim 11. Those come from the C.D. Cal. tentative Markman order and the IPR exhibit list. I report them as stated and flag them as needing confirmation against the reexamination certificate — this is a potential contradiction with the "27 issued claims" premise carried in the earlier summary.
4. Denial is not validity. The institution denial (2020-08-05) means the petitioner failed to show a reasonable likelihood of prevailing as presented; it is not a holding that the claims are non-obvious. My analysis above states where I think the art is strong (claims 20–27, never tested) versus where the PTAB was unpersuaded (the claims 1–19 grounds).
5. CAFC activity: nothing I retrieved establishes any Federal Circuit appeal naming US 8,817,805, consistent with the earlier section's conclusion.

Bottom line: On the record as it exists, the strongest § 103 combinations are (i) Peisa alone for claims 1–3, 6–7, 10–13, 16–19; (ii) Peisa + Kawarai for the traffic-shaping/leaky-bucket dependents (claims 4–5, 8–9, 14–15); (iii) Jorgensen + Bonomi as an independent route to all of claims 1–19; and — my own addition beyond the petition — (iv) Peisa + the Holma WCDMA for UMTS textbook (and + Bonomi for the leaky-bucket features) for the untested claims 20–27, where the textbook's "give a minimum bit rate to all users, then allocate the remaining capacity to the high-priority users" teaching maps almost verbatim onto claim 20's two-phase allocation. In every combination, the motivation is the same: these are analogous references addressing the identical problem of QoS-differentiated bandwidth allocation over a shared, variable-capacity radio link, and their combination yields only predictable, expected results.