US 5751356

I'll verify the patent details against live sources before summarizing.

US Patent 5,751,356 — Analyst Summary

Verification note: I searched for this exact number (no fuzzy matching, no auto-correction). Live results consistently return the same document, so the record below is grounded rather than inferred. I could not open USPTO PatentCenter's fee/ownership tab directly in this session, so the status/ownership entries come from the Google Patents family record plus corroborating third-party sources (noted where relevant).

Bibliographic data

Abstract (as issued)

A video/audio signal coding system including a video encoder that encodes a video signal into video data and attaches encoding information as a video header, an audio encoder that encodes an audio signal into encoded audio data, a multiplexer that multiplexes and outputs the video and audio data, and a controller for the multiplexer. The video encoder informs the controller of the video header and data indicative of the quantity of encoded video data corresponding to one picture; the controller judges a break in the picture of the encoded video data input to the multiplexer; and the multiplexer multiplexes the encoded audio and video data based on that judgement.

Plain-language overview of the independent claims

Claim 1 — system, two-memory architecture. A video encoder encodes video, attaches header information identifying the start of a video data unit, and also produces "data number information" (a count of the encoded video data for that unit). A first memory temporarily holds the header info and the data number info. A system header generator builds system headers from that stored information, and a second memory temporarily holds those system headers. A multiplexer combines three things: the encoded video, the encoded audio, and the stored system headers. A control means uses the stored header/data-number info to judge the boundary of the video data unit entering the multiplexer and thereby controls the multiplexing. In short: the encoder tells the mux where pictures end instead of the mux having to hunt for headers in the bitstream.

Claim 4 — method counterpart. Same substance in method form: encode video, attach header info, generate data-number info; encode audio; temporarily store the header info and data-number info; generate and temporarily store system headers from the stored information; judge the video-unit boundary from the stored information; and multiplex the video, audio, and stored system headers.

Claim 7 — system, consolidated-control variant. Same functional elements as claim 1, but the first memory, the system-header generation, and the second memory are consolidated into a single "first control means," with a separate "second control means" performing the boundary judgement and controlling the multiplexer.

Dependent claims add only two limitations: the predetermined video data unit is one picture (claims 2, 5, 8), and the encoders operate per MPEG-1 (ISO/IEC 11172-2) or MPEG-2 (ISO/IEC 13818-2) (claims 3, 6, 9).

Technical gist

The invention addresses a known MPEG-2 problem: at higher bit rates, per-picture data volume grows, so a processor that must pattern-scan the bitstream to find sequence/GOP/picture headers becomes a bottleneck. Here, the encoder's header creation circuit (12) already knows the bit length L between picture headers, so it hands that value and the headers to the multiplexer side (memory 3), where processor 4 builds system headers and steers data-select and byte-count signals (9, 10) to multiplexer circuit 6. The specification reports ~1/2.5 the processing time of the prior art because hardware data movement for one picture overlaps with software header processing for the next.

Litigation / docket posture (important caveat)

• No Court of Appeals for the Federal Circuit 2026 docket involving 5,751,356 was found. My searches for a 2026 CAFC appeal returned zero results. Given the patent expired 2016-08-06, a 2026 appeal would be unusual; I cannot rule one out, but I have no authoritative evidence of it and will not assert one.
• The only Federal Circuit matter I found touching this dispute is In re MediaTek, No. 07-M844 (Fed. Cir., petition filed 2007-02-07; opinion 2007-04-03), a mandamus proceeding naming MediaTek as petitioner and Sanyo Electric Co., Ltd. as respondent.
• District court cases: MediaTek, Inc. v. Sanyo Electric Co. (E.D. Tex. 6:05-cv-00323; complaint filed 2005-08-31, amended 2006-10-03 to add the '356 patent) and Sanyo Electric Co. Ltd. v. MediaTek Inc. (C.D. Cal. 2:06-cv-01225). The parties settled in 2007 with cross-licenses and dismissal with prejudice.
• Claim construction (E.D. Tex., Judge Davis, memorandum opinion 2007-03-26) is the most substantive judicial treatment of this patent. Rulings on the '356 claims:
  • "video encoding means" — governed by 35 U.S.C. § 112 ¶ 6; corresponding structure is a video signal encoder including an encoder circuit utilizing an MPEG standard, a header creation circuit, and a multiplexer circuit, and equivalents. The court rejected limiting the structure to MPEG-1/MPEG-2 only, reasoning MPEG is an evolving standard and dependent claim 3 separately adds MPEG-1/MPEG-2.
  • "data number information" — construed as "a value indicative of the bit length of a picture." Sanyo's proposed addition that it be not part of the encoded video data was rejected.
  • "predetermined video data unit" — construed as "a predetermined set of pictures" (not narrowed to GOP).
  • "header information" — construed as "data that carries information about video data."

Uncertainty flags

• I did not independently verify real-time maintenance-fee or ownership status in USPTO PatentCenter; Google Patents itself labels its legal-status and assignee listings as assumptions. The intervening re-assignments to United Module Corporation (a California entity) and then MediaTek are documented in the assignment records quoted in the family data, but I have not inspected the underlying assignment documents.
• No claim of an active 2026 CAFC matter should be attributed to this patent on the basis of my searches.

I'll research litigation involving US patent 5751356.

Let me search for details on the two specific cases.

Let me dig deeper into both cases and their outcomes.

Litigation Involving U.S. Patent No. 5,751,356

Patent identification (literal): U.S. Patent No. 5,751,356, "Video/Audio Signal Coding System and Method," issued May 12, 1998, from application US 08/689,268, filed August 6, 1996, priority date August 15, 1995 (JP 7-229704). Inventor Hiroshi Suzuki. Original assignee Nippon Steel Corp.; later assigned to United Module Corporation (Jan. 19, 2005) and then to MediaTek, Inc. (Jan. 21, 2005). Status: Expired – Lifetime (anticipated expiration Aug. 6, 2016).

Based on the sources reviewed, the '356 patent was asserted in one principal infringement action (E.D. Tex.), with a related C.D. Cal. action listed against the same patent family. Details below.

Case 1 — MediaTek, Inc. v. Sanyo Electric Co., Ltd., Sanyo North America, Inc. and Sanyo North American Corporation

Grounding: MediaTek's Second Amended Complaint states that "[o]n May 12, 1998, the USPTO issued U.S. Patent No. 5,751,356 entitled 'Video/Audio Signal Coding System and Method' (hereinafter 'the '356 patent')" and identifies it as a patent-in-suit (Exhibit C). The E.D. Tex. court construed the '356 claim terms (means-plus-function "video encoding means," "multiplexing means," "system header," etc.) in its Memorandum Opinion and Order, Document 344, filed March 26, 2007, and a supplemental order of May 3, 2007 addressing Claim 4.

Case 2 — Sanyo Electric Co., Ltd. v. MediaTek Inc. (and MediaTek's related Sanyo actions)

Grounding: Google Patents lists a "US case filed in California Central District Court" under case 2:06-cv-01225 for this patent, and the Stanford NPE Litigation Database entry for patent 5751356 lists a single case — "Sanyo Electric Co Ltd v. Mediatek Inc," Civil Action 2:06-cv-01225, C.D. Cal. A press report (CDRinfo) confirms that Sanyo and MediaTek "amicably resolved the outstanding patent disputes … in the United States," covering Sanyo's claims against MediaTek and ASUS in California and MediaTek's claims against Sanyo in Texas, with both matters "dismissed with prejudice" and cross-licenses exchanged.

Related appellate matter

• In re MediaTek, Inc., No. 07-M844, U.S. Court of Appeals for the Federal Circuit (petition filed Feb. 7, 2007; decided April 3, 2007). This was an interlocutory/mandamus proceeding arising out of the MediaTek–Sanyo dispute. It is not a separate infringement suit but is part of the same litigation family.

Important caveats

1. Case-number/party attribution is imperfect in the databases. The '356 patent was originally a Nippon Steel patent, and MediaTek acquired it only in early 2005 (via United Module Corporation). It was MediaTek that asserted the '356 patent in the Texas action. The California case 2:06-cv-01225 is listed in the Google Patents/Darts-IP feed and in the Stanford NPE database as the California case associated with the '356 patent, even though Sanyo was the plaintiff there. I could not independently confirm from a primary court document that the '356 patent itself was pleaded as an affirmative claim in the California case; the association may reflect counterclaims, consolidation, or a database-classification artifact. Treat that linkage as uncertain.

2. No PTAB proceedings found. I found no IPR, PGR, or CBM review involving U.S. 5,751,356 in the sources reviewed.

3. No other litigation found. Beyond the MediaTek–Sanyo family above (and the related Federal Circuit mandamus), I did not identify any other suits asserting U.S. 5,751,356. The patent's term expired Aug. 6, 2016, making further assertion unlikely.

4. I did not have direct PACER access; the case details above are drawn from secondary/aggregated litigation databases (Stanford NPE Litigation Database, Google Patents litigation feed, Unified Patents, RPX Insight, and court documents surfaced via govinfo/CourtListener) and contemporaneous press and financial-report accounts. Filing dates for the specific docket entries should be verified against PACER for the two court dockets (E.D. Tex. 6:05-cv-00323 and C.D. Cal. 2:06-cv-01225) if an exact date is required.

Proceedings overview

Zero AIA trial proceedings (0 IPRs, 0 PGRs, 0 CBMs) have ever been filed against US 5,751,356 — the USPTO Open Data Portal trial list is empty, and my web searches surfaced no petition, institution decision, FWD, or Board document bearing this patent number — so at claim-level the scoreboard reads 0 canceled / 0 sustained / 0 settled / 0 institution-denied, and all 9 claims (1–9) UNTESTED at the PTAB; the bottom-line defensive posture is therefore not "hardened by IPR" but rather "never stress-tested at the Board, and now expired (anticipated expiration 2016-08-06) — there is no FWD to cite, but there is also no live patent to infringe."

Because there is no proceeding to itemize, the sections below substitute the only adjudicative record that exists on this patent (E.D. Tex. / C.D. Cal. claim construction) and flag where the record is thin.

No PTAB proceedings — what I checked and what I found

• Canonical source (USPTO ODP, via the structured block in this prompt): no AIA trial proceedings on file. I treat this as authoritative.
• Independent web sweep for "5751356" IPR, "5,751,356" inter partes review, "IPR2013/2015/2016" + 5751356, and CBM formulations: no petition or Board paper for this patent. Results returned this patent only in connection with district-court litigation and with an unrelated Japanese patent, JP5751356B2 (Canon, "Fixing apparatus and image forming apparatus") — a different patent that shares the numeric string. Do not conflate them.
• Negative-result caveat: an absence of hits is not the same as a certified "never filed." Pre-2012 Board papers and some older reexamination files are not fully indexed by commercial dockets. If you need a certifiable negative, pull the full file history from USPTO Patent Center and run the patent number through PTAB E2E (https://ptacts.uspto.gov/ptabweb/) directly. What I can say with high confidence: no AIA proceeding is on file in ODP, and nothing public contradicts that.

Fastest verification links:

• Google Patents record (the canonical text I was given): https://patents.google.com/patent/US5751356/en
• PTAB E2E search: https://ptacts.uspto.gov/ptabweb/
• Stanford NPE Litigation Database entry: https://npe.law.stanford.edu/patent/[5751356](/patent/5751356)

District-court record (substitute for a PTAB history) — MediaTek, Inc. v. Sanyo Electric Co., E.D. Tex. No. 6:05-cv-00323

• Type: Not an AIA trial. Patent-infringement action (and mirror-image DJ/co-pending case, Sanyo Electric Co. Ltd. v. MediaTek Inc., C.D. Cal. No. 2:06-cv-01225).
• Filed: 2005 (Texas action); C.D. Cal. counterpart filed 2006.
• Status: Terminated — settled on the eve of trial ("The night before we were to give opening argument, the parties agreed to settle both suits."). Terms confidential.
• Judge panel: Not a PTAB panel. Judge Leonard E. Davis, E.D. Tex., Tyler Division.
• Claims at issue: All nine claims were in the case; the parties stipulated to give unconstrued claims their ordinary meaning, and claim 4 was carved out for separate treatment.
• Key ruling — Markman (2007-03-26), No. 6:05-cv-00323, Doc. 344 (https://www.govinfo.gov/content/pkg/USCOURTS-txed-6_05-cv-00323/pdf/USCOURTS-txed-6_05-cv-00323-0.pdf):
  • "Video encoding means" — agreed § 112 ¶ 6 means-plus-function. Function: "encoding a video signal into video data, attaching encoding information to said video data to generate encoded video data, and generating data number information indicative of a number of the encoded video data corresponding to a predetermined video data unit." Corresponding structure limited to "a video signal encoder including an encoder circuit utilizing an MPEG standard, a header creation circuit, and a multiplexer circuit; and equivalents." The court rejected MediaTek's broader "any encoder circuit" reading and rejected Sanyo's attempt to narrow to MPEG1/MPEG2 only ("The specification does not limit the structure to MPEG1 or MPEG2.").
  • "Data number information" construed as "a value indicative of the bit length of a picture" — and the court refused to add Sanyo's "which is not part of the encoded video data" gloss. This is the most defendant-unfavorable holding of the case.
  • "Predetermined video data unit" construed as "a predetermined set of pictures" — Sanyo's GOP-only narrowing rejected.
  • "Header information" construed as "data that carries information about video data" — again the narrowing rejected.
  • Claim 4 (2007-05-03 order): court noted the parties' agreement to ordinary meaning meant Sanyo had waived prosecution-history-estoppel and sole-embodiment arguments, but expressly preserved Sanyo's § 102 anticipation defense to claim 4, and observed a plain-meaning construction "is therefore likely to be broad and problematic for both parties."
• Appeal: No Federal Circuit appeal of a claim-construction judgment is on record; the case settled. (The 2015 Federal Circuit opinion in Silicon Storage Tech. v. United Module Corp. concerns different, WaferScale-origin patents held by the same UMC entity — not the '356 patent. Don't cite it as a '356 appellate ruling.)
• Defensive value: The most useful artifacts here are the § 112 ¶ 6 framing of every "means" limitation and the court's own remark that claim 4 was "likely to be broad and problematic for both parties." But note the settlement means no merits invalidity holding and no unenforceability holding exist.

Strategic summary

Claim status. All nine claims — 1, 2, 3, 4, 5, 6, 7, 8, 9 — are UNTESTED. Nothing is CANCELED and nothing has been SUSTAINED by the Board, because no claim has ever been before the Board. The only governmental narrowing of these claims is the E.D. Tex. constructions summarized above, which cut within the claims (limiting "video encoding means" to MPEG-based encoders) but did not invalidate anything. For completeness, the claim set divides into two independent system claims (1, 7), one independent method claim (4), and six dependents (2, 3, 5, 6, 8, 9); the dependent claims add only the "one picture" unit (2, 5, 8) and the MPEG1/MPEG2 limitation (3, 6, 9) — a very shallow claim tree with no meaningful fallback depth if the independent claims fall.

Estoppel landscape. There are no § 315(e)(2) estoppels to map, because there are no petitioners. That cuts both ways: no defendant is barred from running art at the Board or in court, but no defendant benefits from an earlier petitioner's work product or from a prior FWD's claim-construction findings. Equally, MediaTek (current assignee, per the reassignment chain Nippon Steel → United Module Corp. (2005-01-19) → MediaTek (2005-01-21)) has never faced IPR estoppel and has never had to defend these claims on the merits.

Pattern signals. No serial petitioner, no defensive aggregator involvement, and no PTAB appeal history — the "family has litigation" flag on this record traces to the 2005–2006 MediaTek/Sanyo campaign (Darts-IP family 16896397), not to any AIA trial. Note carefully: the appearance of portal.unifiedpatents.com in the litigation links is a data-source URL from Google Patents' UI, not evidence that Unified Patents filed anything against this patent. The one genuinely notable historical signal is procedural rather than adversarial: the patent lapsed for failure to pay a maintenance fee effective 2002-05-12 and was reinstated on 2004-05-17 after a petition and surcharge (37 CFR 1.378 practice), before the 2005 assignment to UMC and the 2005 assertion campaign. That gap is worth a footnote if you are building an equities narrative, but the Director's revival decision is generally final and the patent is long expired, so it is not a standalone defense.

The controlling fact. Anticipated expiration was 2016-08-06. Google Patents lists the legal status as "Expired – Lifetime." That is worth more than any IPR.

Recommended next steps

1. If you have received or are weighing a demand letter citing US 5,751,356, lead with expiration, not validity. The patent expired 2016-08-06 (20 years from the 1996-08-06 filing). Under 35 U.S.C. § 286, recovery is limited to infringement occurring within six years before the complaint is filed; a complaint filed after 2022-08-06 cannot reach any pre-expiration act, and an expired patent supports neither prospective injunctive relief nor ongoing royalties for post-expiration conduct. Practically: a 2026 assertion of this patent should have no damages runway and no injunction exposure, and a demand letter that ignores expiration is a candidacy problem for the sender. (Confirm the exact expiration/terminal-disclaimer position against the printed front page and the maintenance-fee record in Patent Center before you put this in a brief — I am working from the Google Patents status line, not a certified PTO transaction history.)
2. There is nothing to cite on the PTAB side — say so plainly. No IPR, PGR, or CBM was filed; there is no Final Written Decision to link or quote, and I will not manufacture one. If an adversary claims the patent "survived IPR," that is false; if they claim it "was invalidated," that is also false.
3. If any active dispute is nonetheless pending, the assertable record is the Markman history, not a Board record. Cite the 2007-03-26 E.D. Tex. claim construction (https://www.govinfo.gov/content/pkg/USCOURTS-txed-6_05-cv-00323/pdf/USCOURTS-txed-6_05-cv-00323-0.pdf) and the Federal Circuit/CourtListener opinion in MediaTek, Inc. v. Sanyo Electric Co. (https://www.courtlistener.com/opinion/[2502199](/patent/2502199)/mediatek-inc-v-sanyo-electric-co-ltd/) for the § 112 ¶ 6 treatment of "video encoding means" and the court's warning that claim 4 is "likely to be broad and problematic for both parties."
4. Close the verification gaps before relying on the negative. (a) Run the patent number through PTAB E2E and confirm no proceeding, including pre-2012-era papers; (b) pull the Patent Center file history to confirm expiration date, any terminal disclaimer, and the 2002 lapse / 2004 reinstatement; (c) check the foreign counterparts JPH0955935A (JP, status "Withdrawn"), KR970012600A ("Ceased"), and CN1149234A ("active — Pending") for foreign oppositions or nullity actions — I found no data on foreign challenges and am flagging that as explicitly unverified; (d) confirm the current recorded assignee, since the chain ended at MediaTek in 2005 and later transfers are not shown in the record provided.
5. Do not cite Silicon Storage Technology v. United Module Corp. (Fed. Cir. 2015) or the UMC/Keranos campaigns as '356 authority — those matters involved different patents, notwithstanding the shared UMC entity.

Confidence statement: High confidence that no AIA trial proceeding is on file for US 5,751,356 (ODP is empty and public sources are consistent). Moderate-to-high confidence on the dates and status lines because they come from the authoritative patent record supplied to me. Low confidence / unverified: any foreign opposition history and the current chain of title.

I'll start by pulling the assignment records and cross-referencing the litigation and correspondent data.

Assignment & Ownership Analysis — US 5,751,356

Data caveat up front: The authoritative source I was given (the Google Patents full-text/legal-events page) exposes assignor, assignee, reel/frame numbers, and execution/recording dates, but it does not expose the correspondent of record for any of the three assignments, and I was unable to retrieve the correspondent fields or the litigation captions from live sources within this session. Every reel/frame below is taken literally from the patent record; wherever a required field (correspondent, plaintiff identity, UMC's corporate nature) could not be confirmed, I say so rather than fill it in.

Inventors

Unusual patterns: I found no evidence of inventor departure within 12 months of filing. To the contrary, the inventor's assignment to Nippon Steel was executed with an effective date of 1996-07-26 (recorded 1996-08-06), i.e. roughly 11 days before the US filing and 12 days after the JP priority filing — a routine employer-obligation assignment, not a depart-and-flip signal. This is a single-inventor case, so there is no co-inventor split to analyze.

Original assignee

Nippon Steel Corporation (Japan) — named on the issued patent and recorded as owner at grant (1998-05-12).

• Primary line of business: Integrated steel producer, but with a substantial Electronics/ semiconductor arm that developed MPEG codec and multiplexing technology. That business line is corroborated by Nippon Steel's own cited prior art, JP H0851623A ("Multiplex device of MPEG encoding device," published 1996-02-20, assigned to Nippon Steel Corp) — i.e., the same company was filing on MPEG multiplexer hardware in this exact window.
• Did they ship a product embodying the claims? Unclear. The patent is directed to a coder/multiplexer architecture (MPEG-1/2 TS multiplexing). Nippon Steel developed MPEG encoder/decoder LSI work, but I cannot confirm a commercial product from the sources at hand; do not treat this as either confirmed or disproven.
• Current status: Operating. Nippon Steel remains a going concern (merged with Sumitomo Metal in 2012, re-adopted the "Nippon Steel Corporation" name in 2019). It did not file bankruptcy in connection with this patent. Its semiconductor/IP holdings were divested (see timeline).

Assignment timeline

Three assignments are recorded. All three are ordinary "ASSIGNMENT OF ASSIGNORS INTEREST" conveyances; no security interests, licenses, mergers, or corrections appear.

• 1996-07-26 (executed) / recorded 1996-08-06 — Reel 008087/0505

  • Conveyance: Assignment (assignment of inventors' interest)
  • Assignor: Hiroshi Suzuki (inventor)
  • Assignee: Nippon Steel Corporation
  • Correspondent: Not exposed in the source of record I hold and not retrieved live. Flagging as a required-verification item.
  • Context: Routine inventor-to-employer obligation assignment, executed 20 days before the US filing (filing 1996-08-06).

• 2004-11-24 (executed) / recorded 2005-01-19 — Reel 015603/0829

  • Conveyance: Assignment
  • Assignor: Nippon Steel Corporation
  • Assignee: United Module Corporation (California)
  • Correspondent: Not exposed in the source of record I hold and not retrieved live. For a 2004-era Nippon Steel → California holding-company assignment this is the single most probative field; it must be pulled directly from Assignment Center.
  • Context: Divestiture of the MPEG patent by the operating steel/electronics company to a small California entity — a transfer to a non-practicing holder, on the evidence of the record (no product attribution for UMC).

• 2005-01-20 (executed) / recorded 2005-01-21 — Reel 015612/0472

  • Conveyance: Assignment
  • Assignor: United Module Corporation (California)
  • Assignee: MEDIATEK, INC. (Taiwan)
  • Correspondent: Not exposed in the source of record I hold and not retrieved live.
  • Context: Rapid flip — United Module held the patent for ~57 days (2004-11-24 exec. → 2005-01-20 exec.) before selling to MediaTek. Executed-to-recorded gap is one day, indicating a prepared, transactional transfer rather than an internal reorg.

Litigation (from the same record, "Family has litigation"):

• Texas Eastern District Court — case 6:05-cv-00323 (Unified Patents litigation data). Docket prefix 6:05 indicates a 2005 filing in the Marshall/Tyler division.
• California Central District Court — case 2:06-cv-01225 (Unified Patents litigation data; 2:06 indicates a 2006 filing).

I could not confirm the plaintiff in either docket from the sources available. This matters — see the verdict discussion.

Sibling family members (same priority 1995-08-15): JP H0955935A (JP 7-229704), KR 970012600A, CN 1149234A.

Timeline diagram

timeline
    title Ownership of US 5751356
    1995 : Priority filing JP 7-229704
    1996 : Filed by Hiroshi Suzuki
         : Assigned to Nippon Steel Corp
    1998 : US 5751356 issued
    2004 : Assigned to United Module Corporation
    2005 : Assigned to MediaTek Inc
         : Texas suit 6 05 cv 00323
    2006 : California suit 2 06 cv 01225
    2016 : Patent expired

NPE / troll-pattern signals

1. Shell-entity transfer — UNCLEAR (trending present).
   Concrete record: the patent left an operating company (Nippon Steel) for a California entity, United Module Corporation, by assignment executed 2004-11-24, recorded at reel 015603/0829, which then disposed of the patent 57 days later. That is a textbook non-practicing-intermediary shape. What is missing: the record I hold does not show UMC's business, address, or incorporation — the defining tells (no products, registered-agent address, DE/TX single-member LLC) are not in evidence, and "United Module Corporation" carries no IP/Licensing/Holdings suffix. Per your rule, a suggestive name is not a finding. Call: unclear, resolvable by pulling the reel 015603/0829 cover sheet for UMC's address and by verifying UMC product activity.

2. Known asserter in the chain — NOT PRESENT.
   None of the three assignees (Nippon Steel Corporation, United Module Corporation, MediaTek Inc.) appears on the referenced NPE lists (Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation, Spangenberg entities). MediaTek in particular is a large operating semiconductor designer. No RPX/Unified high-frequency-plaintiff match surfaced for any link.

3. Repeat correspondent across the chain — UNCLEAR (data not retrieved).
   This is the signal you specifically care about and it is the one I cannot substantiate: the correspondent of record for reel 008087/0505, reel 015603/0829, and reel 015612/0472 is not exposed in the source I hold, and my live lookups did not return it. I will not guess a name or firm. Action item: query Assignment Center by patent number 5751356 and open each of the three reel/frame entries; the recurrence test only counts if the same attorney/firm appears on ≥2 links.

4. Cascading transfers — PRESENT (moderate).
   Two consecutive arm's-length assignments inside a ~2-month execution window (2004-11-24 at reel 015603/0829 → 2005-01-20 at reel 015612/0472), with a one-day record lag on the second. Chain length is only two hops, and the second hop lands at an operating company, so this is suggestive of a broker/intermediary step rather than a long NPE daisy-chain.

5. Pre-litigation transfer — LIKELY PRESENT (date-unconfirmed).
   The ED Texas docket 6:05-cv-00323 is a 2005 filing; the Nippon Steel → United Module transfer was executed 2004-11-24 (reel 015603/0829), and the United Module → MediaTek transfer executed 2005-01-20 (reel 015612/0472). A 2005 filing therefore falls within roughly 2–6 months of the 2004-11-24 transfer. However, because MediaTek already owned the patent by 2005-01-21, the plaintiff could be either entity — so I mark this likely present but not confirmed. Exact filing date and plaintiff caption are required to make it a firm finding.

6. Bankruptcy fire-sale — NOT PRESENT.
   Nippon Steel did not enter Chapter 7/11 around this divestiture; there is no recorded bankruptcy sale, no court-sale reel, and no trace of the Kodak/Nortel/Polaroid pattern. The 2004 divestiture is a negotiated assignment, not a bankruptcy disposition.

7. Privateering — UNCLEAR.
   The shape (operating company → small entity → possible assertion → sale to an operating company) is consistent with privateering, but there is no SEC 8-K/10-K disclosure, no Patent Progress/EFF coverage, and no proof that Nippon Steel benefited from or directed any assertion. Insufficient evidence.

8. Defensive aggregator (anti-NPE) — NOT PRESENT.
   The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at MediaTek, Inc. (assignment effective 2005-01-20, reel 015612/0472), an operating semiconductor company. Note the patent is now expired (anticipated expiration 2016-08-06; status "Expired – Lifetime"), so any assertion value is historical.

Verdict

NPE — moderate confidence.

The verdict rests on two record entries: the operating steel/electronics company Nippon Steel assigning to the small California entity United Module Corporation at reel 015603/0829 (executed 2004-11-24), followed by United Module disposing of the patent 57 days later to MediaTek at reel 015612/0472 (executed 2005-01-20) — with an ED Texas suit (6:05-cv-00323) falling within months of the first transfer. That is the classic short-hold, non-practicing intermediary plus litigation-timing pattern. Confidence is capped at moderate because (a) I could not retrieve the correspondent of record on any of the three links — the recurrence test that most often nails these chains is therefore untested, and (b) no assignee matches a published NPE list, no shell-entity tells (address, no-products, single-member LLC) are in evidence for United Module Corporation, and MediaTek — the terminal owner — is an operating company. If the 2005/2006 plaintiffs turn out to be MediaTek rather than United Module, this verdict should be re-run as "operating-company assertion."

Verify the three reel/frame entries and, critically, the correspondent fields here:

• https://assignment.uspto.gov/patent/index.html (search patent number 5751356)
• https://assignmentcenter.uspto.gov/

Cross-check the dockets at Unified Patents' litigation portal for 6:05-cv-00323 and 2:06-cv-01225 to fix the plaintiff identity, which is the pivot for the final call.

I'll start by searching the USPTO database for the specific patent number 5751356.

I have the patent and its citation list. Now let me research each cited prior-art reference in detail.

Prior-Art Analysis — U.S. Patent No. 5,751,356

Scope note: The following is a technical patent-landscape analysis based on the U.S. patent record and the six references cited on the face of the patent. It is not a legal opinion; §102 anticipation requires that every element of a claim, arranged as in the claim, be disclosed in a single reference. Where the retrieved record was incomplete, I say so explicitly rather than guessing.

1. Confirmed bibliographic record (USPTO / Google Patents / FPO)

Because the application was filed in 1996 (before the AIA first-to-file date of 2013‑03‑16), pre-AIA §102 governs. The effective prior-art date for the disclosure is the foreign priority date, 1995‑08‑15.

2. The claims at a glance (what must be met for §102)

All independent claims share a common architecture:

• Claim 1 (system): (a) video encoding means that attaches header info and generates data number information (number/bit-length of encoded video data per predetermined video unit); (b) audio encoding means; (c) first memory storing header info + data number info; (d) system-header generating means; (e) second memory storing the system headers; (f) multiplexer for encoded video + encoded audio + system headers; (g) control means that judges a picture boundary from the header info + data number info.
• Claim 4 (method): method counterpart of claim 1.
• Claim 7 (system variant): same elements recast as a "first control means" (store header/data-number info + generate & store system headers) and a "second control means" (boundary judgment / multiplex control).
• Claims 2, 5, 8: unit = one picture.
• Claims 3, 6, 9: encoding per MPEG‑1 (ISO/IEC 11172‑2) or MPEG‑2 (ISO/IEC 13818‑2).

The alleged point of novelty is that the encoder hands the multiplexer's controller the header and the per-picture bit length directly, so the multiplexer need not parse the bit stream to locate picture headers.

3. Reference‑by‑reference §102 analysis

Reference 1 — U.S. Pat. No. 5,122,875

• Full citation: Raychaudhuri, D.; Zdepski, J. W.; Reitmeier, G. A.; Wine, C. M., "An HDTV compression system," General Electric Co., Appl. No. 07/661,993, filed 1991‑02‑27, issued 1992‑06‑16 (US5122875A).
• Date status: Issued >1 yr before the 1995‑08‑15 priority → §102(b) art.
• Description: Compression circuit produces hierarchically layered video codewords; a priority-selection circuit parses them into high/low-priority sequences; a transport processor forms transport blocks each containing header data + codewords + error-check bits; audio codewords are formed into mutually exclusive transport blocks and interleaved with the video transport-block sequences.
• Potential §102 target claims: Discloses the video+audio multiplexing-with-headers element (claim 1(f), 4, 7) and MPEG-like coding (relevant to claims 3, 6, 9). It does not disclose a first memory holding data-number information per picture or a system-header generator feeding a second memory, nor a controller that derives picture boundaries from that data-number information.
• Assessment: Weak standalone anticipation. Potentially anticipates only a bare video+audio/transport-block multiplexing claim; better characterized as an obviousness (§103) reference against claims 1/4/7 and as art on claims 3/6/9.

Reference 2 — U.S. Pat. No. 5,168,356

• Full citation: Acampora, A. A.; Siracusa, R. J., "Apparatus for segmenting encoded video signal for transmission," General Electric Co., issued 1992‑12‑01 (US5168356A). Date caveat: Google Patents lists 1991‑02‑27 for this record; Justia lists filed Dec. 20, 1991 (https://patents.justia.com/examiner/michael-l-lee). Treat the 1991 date as the priority/filing reference.
• Date status: Issued 1992‑12‑01, before priority → §102(b) art.
• Description: Segments compressed video codewords into transport blocks; a transport control circuit (responsive to TYPE codewords) extracts header indicia to form transport-block headers, temporarily stores the headers, holds a corresponding quantity of video data in a buffer memory, and emits a sequence of transport headers followed by video data — maximizing re-entry points via secondary transport headers.
• Potential §102 target claims: Maps to the "memory + control + header-driven multiplexing" architecture of claim 1 (elements c/f/g) and the method/variant claims 4/7.
• Assessment: Closest of the two GE patents on the multiplexing/segmentation mechanism, but it is video-only — it lacks the audio encoding element and system-header generation, so it does not fully anticipate independent claim 1/4/7. Strong §103 candidate when combined with an audio-multiplexing reference.

Reference 3 — U.S. Pat. No. 5,583,562

• Full citation: "System and method for transmitting a plurality of digital services including imaging services," Scientific-Atlanta, Inc., filed 1993‑12‑03, issued 1996‑12‑10 (US5583562A). (Inventor not confirmed in the retrieved snippet; do not rely on an inventor name without verification.)
• Date status: Filed 1993‑12‑03 (before invention) but issued 1996‑12‑10 (after priority) → §102(e) art.
• Description: Multiplexes video, audio, and low-/high-speed data services; generates a "multiplex map"/multiplex structure control word specifying the size of the transport region and the location of individual packets, permitting the packet arrangement to be varied per field; expressly supports compression "per ISO 11171 or ISO 13818 M.P.E.G."
• Potential §102 target claims: Discloses multiplexing of MPEG video + audio + data into a structured transport with control/header tables (claim 1(f), 4, 7; and the MPEG limitation of claims 3, 6, 9).
• Assessment: Relevant to the multiplex-control and MPEG elements but does not disclose the encoder-supplied per-picture data-number information or the boundary-judging controller. Not a standalone anticipator of claims 1/4/7; useful as §103 art, and directly on point for claims 3/6/9's MPEG framing.

Reference 4 — U.S. Pat. No. 5,515,107

• Full citation: "Method of encoding a stream of motion picture data," Sigma Designs, Inc., filed 1994‑03‑30, issued 1996‑05‑07 (US5515107A).
• Date status: Filed before invention / issued after priority → §102(e) art.
• Description: A method for encoding a stream of motion-picture (MPEG-type I/P/B) data.
• Potential §102 target claims: On the strength of the title/bibliographic record alone, this reference concerns the video-encoding side — potentially relevant to the "video encoding means attaching headers" element of claim 1/4/7 and to the MPEG claims 3/6/9. I could not retrieve its full text within the search budget, so I am not asserting a specific claim-by-claim anticipation mapping.
• Assessment: Insufficient data to assert anticipation. Needs full-text review before any §102 conclusion.

Reference 5 — JP H08‑51623 (JP 8‑051623 A)

• Full citation: "Multiplex device of MPEG encoding device" (MPEG符号化装置の多重化装置), Nippon Steel Corp., filed 1994‑08‑05, laid open 1996‑02‑20 (JPH0851623A). Same original assignee as the patent under review.
• Date status: As a foreign published application laid open after the 1995‑08‑15 priority date, it is not §102(a)/(b) art (not "published before"), and pre-AIA §102(e) does not reach foreign publications. It may still be citable only if an earlier invention/§102(a) theory or §103 basis applies.
• Description: A multiplexing device for an MPEG encoder — i.e., the same problem space (multiplexing MPEG-encoded video/audio) as the patent under review and filed by the same company roughly one year earlier.
• Potential §102 target claims: Most relevant reference on subject matter — squarely aimed at the multiplexing function of claims 1/4/7 and the MPEG context of claims 3/6/9. However, because of its publication date, its §102 status is questionable and it functions primarily as §103 art (or as evidence of the assignee's own prior work / a §102(a) reference if the invention date is later than 1996‑02‑20).
• Assessment: Closest art in subject matter; do not treat as clean §102 anticipation without resolving the date issue.

Reference 6 — U.S. Pat. No. 5,600,366

• Full citation: "Methods and apparatus for digital advertisement insertion in video programming," NPB Partners, Ltd., filed 1995‑03‑22, issued 1997‑02‑04 (US5600366A).
• Date status: Filed 1995‑03‑22 (before priority 1995‑08‑15) / issued 1997‑02‑04 (after) → §102(e) art (borderline — only ~5 months before priority).
• Description: Digital advertisement insertion into video programming (local insertion / splicing of content into a video program).
• Potential §102 target claims: Peripheral to the claimed encoding/multiplexing architecture; it may touch on program/content boundary handling in a video stream but does not disclose the encoder→multiplexer header/bit-length data path.
• Assessment: Very weak §102 relevance. At most a secondary §103 / background reference; no credible standalone anticipation of any of claims 1–9.

4. Summary table

5. Bottom line and caveats

• No single cited reference appears to disclose all elements of independent claim 1 (or 4/7) — specifically the combination of a first memory holding encoder-supplied per-picture data-number information, a system-header generator writing to a second memory, and a controller that derives picture boundaries from that data-number information. The cited art is therefore best understood as §103 (obviousness) art, or as §102 art directed at narrower sub-combinations (transport-block formation, MPEG video+audio multiplexing).
• Claims 3, 6, 9 (MPEG‑1/MPEG‑2) are the most likely to be met by the art, since References 1–3 and 5 all operate in the MPEG/multiplexing environment.
• References 1 and 2 (both GE) are the strongest §102(b) documents because they clearly pre-date the 1995‑08‑15 priority date.
• Reference 5 (JP H08‑51623, same assignee) is the closest subject-matter match, but its post-priority publication date makes its §102 reach questionable.
• Unresolved items I am flagging rather than guessing: (i) full text/claims of US 5,515,107 were not retrieved, so I do not assert a claim-specific mapping for it; (ii) a filing-date discrepancy exists for US 5,168,356 (Google lists 1991‑02‑27; Justia lists Dec. 20, 1991); (iii) the inventor of US 5,583,562 was not confirmed in the retrieved sources.

Sources used: https://patents.google.com/patent/US5751356 ; https://www.freepatentsonline.com/5751356.html ; https://npe.law.stanford.edu/patent/5751356 ; https://patents.google.com/patent/US5122875A ; http://www.everypatent.com/comp/pat5122875.html ; https://patents.google.com/patent/US5168356 ; https://uspto.report/patent/grant/5168356 ; https://patents.justia.com/examiner/michael-l-lee ; https://patents.google.com/patent/US5583562 ; http://www.everypatent.com/comp/pat5583562.html

I'll verify the cited prior-art references against live sources before building the § 103 analysis.

Good detail on the citations. Now let me check the two remaining cited references and whether the Japanese citation has a US counterpart (relevant to its §102 date).

I hit the tool-call limit mid-verification, so the analysis below is grounded in the prior-art set I did retrieve (five of the six cited references in full or in abstract, one by title only) plus the file and claim-construction record already established in the prior sections. I flag where verification is incomplete.

Obviousness Analysis — US 5,751,356 under 35 U.S.C. § 103

1. Framework and threshold caveats

Governing law. The application was filed 1996-08-06 with a 1995-08-15 JP priority date, so pre-AIA § 103 governs (AIA § 3(n)(1)). The analysis follows Graham v. John Deere (scope/content of art; differences; PHOSITA level; secondary considerations) as refined by KSR Int'l v. Teleflex, 550 U.S. 398 (2007) — i.e., a flexible, expansive obviousness inquiry where the claimed elements are known, each performs its known function, and the combination yields no more than predictable results. A teaching-suggestion-motivation in the references is not required; a design need or market pressure plus predictable solutions suffices.

Claim construction drives the scope. Per the E.D. Tex. memorandum opinion (2007-03-26) summarized earlier:

• "video encoding means" — § 112 ¶ 6, structure = a video signal encoder with an MPEG-based encoder circuit, a header creation circuit, and a multiplexer circuit, plus equivalents; the court expressly declined to limit structure to MPEG-1/MPEG-2.
• "data number information" — "a value indicative of the bit length of a picture."
• "predetermined video data unit" — "a predetermined set of pictures" (not narrowed to GOP).
• "header information" — "data that carries information about video data" (broad).

These constructions broaden the claims relative to the specification and therefore make the § 103 case easier, not harder. The § 112 ¶ 6 framing also means a reference disclosing an equivalent structure suffices — the patent cannot avoid art by pointing to its particular block-diagram partition.

Practical note. The patent expired 2016-08-06, so any §103 conclusion has no live enforcement consequence; this is an art-analysis exercise and a reconstruction of what the 2005–07 litigation would have contested.

2. The reference set (all six are on the patent's face as examiner citations)

Two provenance problems I must flag rather than paper over:

1. JPH0851623A's publication date (1996-02-20) postdates the claimed priority date (1995-08-15). It therefore cannot be §102(a) art against the 1995 invention date, and it is not §102(b) art (published less than one year before the 1996-08-06 filing). It is only available as §103 art if (a) it has a US counterpart qualifying under §102(e) by its 1994-08-05 JP filing, or (b) the actual invention date is later than 1996-02-20. I could not verify a US counterpart (my search for one was cut off by the tool limit). Treat the following JP '623-based combinations as contingent on that predicate.
2. Pre-AIA § 103(c) common ownership. If JP '623 qualifies only under §102(e)/(f)/(g), it is disqualified as §103 art because both it and the '356 patent were owned by Nippon Steel Corporation at the time of invention. This is a serious, independent obstacle to any JP '623-centered obviousness theory. The examiner's citation of it is not evidence it was available as §103 art.

3. Element mapping — claim 1 (the two-memory system claim)

Claims 4 and 7 are the same subject matter in method form and in consolidated-first control means form respectively. Claim 7's reorganization (merging E3/E4/E5 into one functional block) is not a substantive difference: it is the same hardware described at a different level of aggregation, and a "first control means" holding the video info, generating system headers, and buffering them reads directly on US 5,125,875's transport processor. Expect claim 7 to be the easiest of the three independents to invalidate, because the two-memory separation the specification touts as the point of novelty is absent from claim 7.

4. The obviousness combinations and the motivation for each

Combination A (strongest) — GE pair: US 5,168,356 + US 5,125,875

Both are Acampora/Raychaudhuri-era General Electric patents in the same HDTV compression/transport system, with overlapping inventors and a shared disclosure of MPEG-like layered coding (the family description expressly names MPEG and ISO/IEC DIS 11172). Motivation to combine is at its zenith: same assignee, same system, one reference supplying the segmenting/header-generation logic ('356) and the other the transport-block header construction and quantity-derived parsing ('875). What results: video is encoded; segments are counted/measured and headers identifying group heads are generated; headers are generated and staged; headers are multiplexed with video and other service data. The only meaningful gap is audio encoding (E2) — supplied by US 5,512,507 or US 5,583,562. Under KSR, adding an audio elementary stream to an MPEG-like multiplex is the paradigm of a predictable combination of known elements.

Combination B — US 5,168,356 / US 5,125,875 + US 5,600,366

US 5,600,366 supplies the precise MPEG-2 systems vocabulary the claims inhabit (188-byte TS packets, PID, PAT/PMT, PES), and — crucially — teaches computing a quantity of data to reach a frame or packet boundary (claims 22–23) and the necessity of identifying packet/frame boundaries because such streams "cannot be switched at arbitrary points." Motivation: US 5,600,366 is directed to ad-insertion switching, but it states the same problem the '356 identifies — the mux/formatter must know where the picture ends. A POSITA seeking to remove the '356's admitted bottleneck (pattern-scanning the bitstream) would naturally adopt the boundary-identification technique of '366. This is the combination that most directly reaches E7.

Combination C (contingent) — JP H8-51623 as primary, + US 5,168,356 or US 5,600,366

JP '623 is by far the closest art: same assignee, same "MPEG encoding device" multiplex context, same stated problem (reduce software multiplexing load, enable high-rate encoding), and it already discloses a header-detection unit whose output drives a header-flag creation unit consumed by software, with the flags then used by the multiplex unit. Adding the '356's two refinements is a short, predictable step:

• supplying the bit-length/quantity value from the encoder (which already knows it — as does US 5,168,356's segmenting means and US 5,600,366's frame-boundary calculation), and
• staging system headers in a second memory so hardware data movement overlaps software header processing.

But this combination is presumptively dead on arrival for the § 103(c)/§ 102(e) reasons in § 2 above. If a US counterpart exists with a pre-1995-08-15 § 102(e) date, § 103(c) common ownership removes JP '623 from the § 103 art. If no counterpart exists, JP '623 is not prior art at all. A litigant should not build its case on JP '623 — and a defendant relying on it must first resolve the date/ownership predicates.

Combination D — US 5,515,107 + US 5,168,356 (or US 5,600,366)

US 5,515,107 supplies MPEG-1 (ISO/IEC DIS 11172) video and audio encoding with real-time multiplexing of the encoded streams — directly covering E1 (in part), E2 and E6, and covering the MPEG-1/MPEG-2 dependent claims by express incorporation of the CD 11172 draft. Motivation to combine with the GE transport art: both address packing/formatting variable-length MPEG data into a deliverable multiplex; a POSITA needing to know picture boundaries and to generate system-layer headers would look to the transport-formatting art. Weaker on E7 (no express boundary-judgement mechanism), so it is best as a secondary reference.

Combination E — anything above + US 5,583,562 (US 5,583,562 as the audio/plural-services reference)

Only useful to shore up E2/E6 if the primary combination lacks audio. I could not read this reference, so I do not assert specific disclosures beyond its title.

5. Dependent claims

• Claims 2, 5, 8 ("one picture") — obvious. The court read "predetermined video data unit" as "a predetermined set of pictures," so the narrowing to one picture adds only a choice among MPEG's three standard units (picture / GOP / sequence). The '356 specification itself concedes that the reference quantity for the multiplexing operation is the "bit length corresponding to one picture." US 5,600,366 claims 22–23 calculate quantities to the end of a current frame; JP '623's FIG. 3-style layering likewise operates per picture within a GOP. Strong § 103.
• Claims 3, 6, 9 (MPEG-1 ISO/IEC 11172-2 / MPEG-2 ISO/IEC 13818-2) — obvious and arguably entitled to little weight. JP '623 is expressly an "MPEG encoding device" multiplex device; US 5,515,107 incorporates ISO/IEC CD 11172; US 5,600,366 describes MPEG-2 TS packets, PES, PAT/PMT and PIDs. Note the doctrinal wrinkle the E.D. Tex. court already identified: it refused to import MPEG-1/MPEG-2 into the structure of video encoding means precisely because dependent claim 3 separately recites them — i.e., the court treated the standards recitation as an add-on limitation rather than a structural one, which makes it a weak hook for non-obviousness.

6. The best non-obviousness argument, and how it holds up

The single genuinely distinguishing feature across the art is direction of information flow: in JP '623 and in the GE references' transport processors, the header/length determination sits downstream of, and reads from, the encoded bitstream; in the '356, the video encoder pushes the header info and the picture bit-length to the mux side so the mux never searches the bitstream. That is the asserted advantage ("about 1/2.5 the processing time").

Assessment: this is a weak non-obviousness position under KSR. Reasons:

1. Every constituent element is known: generating headers from encoded video (JP '623, GE pair); computing a data quantity to reach a picture/packet boundary (US 5,600,366 cl. 22–23); computing a ratio of compressed video information and using it to control block formation (US 5,125,875); generating location/length information about group beginnings (US 5,168,356).
2. The change is a predictable design choice to eliminate redundant computation: the header creation circuit already determines the bit length L (the '356's own admission), so passing that number to the mux rather than re-deriving it by pattern scan is the definition of an obvious engineering optimization.
3. KSR squarely covers "improving a known device's throughput by a known technique" and the use of "a design need or market pressure" — and the '356 admits the market pressure verbatim ("as the coding rate increases…the time necessary for detecting the header information is undesirably increased").
4. The patent's own Background is a § 103 admission about the prior art (pattern-comparing all video data to find headers) — the very practice JP '623 characterizes as "conventional."

Where a patentee would push back: (a) no reference teaches the encoder-to-multiplexer signalling, and (b) the two-memory arrangement enabling hardware/software pipelining is not shown. Counter: claim 7 collapses those memories, and the '356 specification says the overlap is why it is faster — an argument of degree, not of kind; no comparative data beyond a bare "1/2.5" assertion appears in the record I retrieved.

7. Anticipation (§ 102) — likely not, and that is why § 103 is the fight

No single reference appears to anticipate claim 1. JP '623 is the closest but (i) detects header information from the video data rather than receiving it from the encoder, and (ii) discloses no data number information indicative of picture bit-length. The GE pair lacks audio encoding and the audio-data-unit signalling. So the examiner's allowance is explicable on § 102 grounds, and § 103 is the only realistic invalidity theory — which matches the litigation posture (the patent survived to the 2007 E.D. Tex. claim-construction stage, with the parties settling that year).

8. Objective evidence / secondary considerations

I found no objective indicia in the retrieved record: no reexamination or IPR, no licensee/industry-praise evidence, no evidence of copying or a long-felt unmet need, and no nexus evidence tying the alleged 1/2.5 speedup to the claim elements rather than to the general hardware/software pipelining described. Post-expiration, there is no commercial-success record to point to. Absent such evidence, the secondary-consideration prong of Graham does not rescue the claims from the Combination A/B case.

9. Confidence and open items