US 5991389

I'll run several searches in parallel to gather authoritative information on this specific patent number.

Let me search specifically for the claim text and any 2026 litigation activity.

US Patent 5,991,389 — Analyst Summary

Caveat on scope of search: The USPTO/CAFC-specific queries I ran for litigation activity around this number returned no 2026 Federal Circuit docket. I therefore report what the patent record itself shows, and flag the litigation gap explicitly below rather than inferring a 2026 appeal exists. (Also note your prompt says April 26, 2026 while my operating date is 2026‑09‑28; the record snapshot I'm using was fetched 2026‑09‑28.)

Bibliographic data

Uncertainty: I could not retrieve a CAFC 2026 docket for 5,991,389, and given the patent expired 2017‑05‑30, any 2026 appellate activity would be unusual (and any damages window would have closed pre‑expiration). Treat "no 2026 CAFC docket found" as a search result, not a certified absence.

Abstract

An apparatus and method for providing a telephone operating company with the ability to rapidly deploy advanced services into a public switched telephone network includes a programmable switch matrix, a service control unit (SCU), and a media resource unit (MRU). The call processing of a call (associated with one or more ports on the programmable switch matrix) is controlled externally by the SCU when particular triggering criteria is met (i.e., the call requires or desires control by the SCU). Call control processing is achieved through a high-speed communications link between the programmable switch matrix and the SCU using a communications protocol defining a comprehensive set of primitives (instructions) for call manipulation and control at the programmable switch matrix. The SCU executes different service application software programs that operate within the SCU for different types of service calls that are under the control of the SCU. The MRU interconnects between the SCU and the programmable switch matrix to provide voice processing and message capabilities for connection to a service call via the programmable switch matrix.

(Abstract text as published; consistent with the patent's own summary section and reproduced by secondary aggregators.)

Technical setting (brief, since prior sections cover it)

The patent is a distributed, externally programmable alternative to classic IN/SCP-SSP architecture. A DMS‑250‑class switch (the "programmable switch matrix") contains a Programmable Service Node (PSN) implemented in the switch core. When a trigger in a switch‑resident trigger database fires, in‑switch call processing is suspended for that call and a New Call event notification is sent over an Ethernet link to an external Service Control Unit (SCU), which runs service application programs. The SCU then drives the switch by sending primitives (instructions), and the switch reports back event notifications. A Media Resource Unit (MRU) provides voice/announcement/IVR resources, controlled by the SCU but bridged into calls by the switch over DS1 trunks. Control is per‑"agent" (port/trunk member), not per‑call, with single‑ended supervision so call halves can be controlled independently. The AIN/SCP path can co‑exist and can even redirect a call to SCU control.

Independent claims — plain-language overview

The record identifies at least five independent claims (1, 6, 11, 14, 16) plus further independent claims directed to the interface protocol and the method of communicating over it. Claim numbering below reflects what I could verify; the last two are flagged as approximate.

Claim 1 — Apparatus (switch-side, claim text partly reconstructed from patent summary).
An apparatus with a programmable switch matrix having at least two ports (first and second), plus one or more predetermined triggers that detect when a service call on the first port should be controlled externally by a service control unit. A service node in the switch emits an event notification to the external SCU when the trigger fires and receives primitives back. Call‑processing circuitry in the switch then connects the first port to the second port in response to those primitives. The SCU sits outside the switch and controls it over a communications link.

Claim 6 — Apparatus (system-level "means‑for" claim).
A system comprising an SCU with service application programs, a programmable switch matrix with plural ports, and means for: generating a service‑call request when external control is desired; generating and outputting an event notification (including call information) to the SCU; receiving primitives; and connecting the first port to the second port under those primitives' control. It adds three links: a first data link (switch↔SCU) carrying primitives and event notifications; a media resource unit that outputs messages to a port; a second data link (MRU↔switch) carrying that message to the port; and a third data link (SCU↔MRU) letting the SCU control the MRU's message output.

Claim 11 — Method of controlling/processing a service call.
Receive a call on a first switch‑matrix port; determine from call information that the call should be handled by a service application program external to the switch; send a New Call event notification to the SCU identifying that the call is taken over; the SCU (or its app) executes the service application program to control the call; transmit one or more primitives from the SCU to the switch; and connect the first port to a second port in response to those primitives. (Dependent claims 12, 13, 15 add refinements.)

Claim 14 — Method (return‑address variant).
Same take‑over flow as claim 11, but adds that in response to the New Call event notification the SCU sends a New Call Accepted primitive back to the switch, that primitive containing (a) data defining the first port and (b) data defining a return address within the SCU to which subsequent event notifications from the switch will be sent. This is the "call steering / arbitration" mechanism — the arbitrator tells the switch where to route future events.

Claim 16 — Method (trigger‑database variant).
Receive a call on a first port with call information; perform in‑switch call processing until a trigger is detected; trigger when the call information meets one or more predetermined trigger criteria in a trigger database; then send a first data message to the external SCU containing (a) data defining the call type and (b) data defining the address of the first port. The SCU replies with a second data message indicating it will control the call; the SCU selects a service application program based on the call‑type data; executes it to generate instructions; and sends them to the switch to control/connect the call.

Claims directed to the Service Programming Interface (SPI) protocol — independent claim number approximate. An independent claim covers a service programming interface protocol between a switch matrix and an external SCU, comprising primitives sent SCU→switch — Bridge (conference specified ports, optionally bridging a message), Collect Digits & Report (collect a specified number of DTMF/multifrequency digits at a port and report them), Connect (connect a predetermined first port to another port), Disconnect (disconnect a first port from another), and New Call Accepted (tell the switch the SCU is taking control) — and event notifications sent switch→SCU: Digits Collected, Instruction Completed, New Call, Off‑Hook, and On‑Hook.

Method‑of‑communicating claim — independent claim number approximate. A method of communicating over a link between the switch matrix and the external SCU, comprising instructing the switch (via primitives) to (1) bridge ports, (2) collect and transmit digits, (3) connect a first port to another port, (4) disconnect a first port from another port, and to accept SCU control; and informing the SCU (via event notifications) of digits collected, instruction completion, a New Call trigger, off‑hook, and on‑hook.

Uncertainty to flag: The published claim set extends beyond claim 16, and I could not verify with full confidence the exact independent claim numbers, the total claim count, or the precise wording of every limitation (multiple web sources truncate the claims; the Google Patents full text I was given stops mid‑"In Service event notification" and does not include the claims section). The overviews above are faithful in substance to the claim language I retrieved from the RPX patent record and cross‑checked against the specification, but exact claim numbering for the SPI‑protocol and method‑of‑communicating independents should be confirmed against the USPTO PatentCenter/Patent Public Search claim listing before being relied on for infringement or validity analysis.

Sources consulted

• Google Patents record and full text: https://patents.google.com/patent/US5991389A/en (bibliographic data, assignments, family, status, description)
• RPX/Insight patent record (claims text snippets): https://insight.rpxcorp.com/patent/US5991389A
• Justia Patents: https://patents.justia.com/patent/5991389
• Secondary aggregator reproducing the published abstract and inventor/assignee fields: https://radaris.com/f/Geetha/Ram/Inventor
• Sibling/priority family documents: WO 1997/048240 A2, EP 0906705 A2, AU 719309 B2, and related US 6,028,924 (same 1996‑06‑13 priority)

I'll search for litigation involving US patent 5991389 across patent litigation databases.

Let me search the specific litigation reference found in the patent record and dig deeper.

I found a strong lead: US 5,991,389 appears in Delaware case 1:07-cv-00507-GMS. Let me verify the details and search for the outcome.

Litigation involving U.S. Patent No. 5,991,389

Based on my searches of patent-litigation sources (Unified Patents litigation portal reference embedded in the Google Patents record, the District of Delaware docket/forms, and related pleadings), I identified one litigation in which U.S. Patent No. 5,991,389 was asserted.

Case: Vonage v. Nortel Networks (D. Del.)

How the '389 patent fits in: The case began as a Vonage filing, and Nortel answered and counterclaimed for patent infringement. In the Answer and Counterclaim filed December 14, 2007, Nortel asserted infringement against Vonage on multiple patents, including specifically:

"CLAIM 11: INFRINGEMENT OF U.S. PATENT NO. 5,991,389 … Vonage has and continues to infringe, contribute to the infringement of, and/or induce infringement of the '389 Patent … Vonage's continued infringement of the '389 Patent is willful."

Other patents Nortel asserted in that same counterclaim included U.S. Patent Nos. 6,091,808; 6,445,695; 7,050,861; 6,823,370; 7,035,390; 6,934,279; 6,337,858; 6,798,786; and 6,799,210. The case was a Nortel–Vonage cross-licensing/infringement dispute in the Delaware district court.

Corroborating source: The Google Patents record for US5991389 itself carries a litigation flag ("US case filed in Delaware District Court") pointing to the same docket, via the Unified Patents litigation dataset: https://portal.unifiedpatents.com/litigation/Delaware%20District%20Court/case/1%3A07-cv-00507.

Important caveats and exclusions

1. No other litigation found. Apart from the Delaware action above, I did not find any additional suit asserting U.S. Patent No. 5,991,389.

2. Excluded look‑alike patent — U.S. Patent No. 6,845,389. Several Rockstar/Nortel portfolio cases (e.g., the Charter/Google–Rockstar declaratory-judgment actions and the Rockstar v. Google/ZTE Android suits) reference a "'389 patent," but that reference is to U.S. 6,845,389, a different patent. It is not U.S. 5,991,389 and I have excluded it per your instruction not to conflate similar numbers.

3. Ownership history (not litigation, but relevant). The '389 patent was originally assigned to Northern Telecom Ltd./Nortel and passed through the Nortel bankruptcy portfolio to Rockstar Bidco, LP (2011) → Rockstar Consortium US LP (2014) → RPX Clearinghouse LLC (2015), with a subsequent security interest to JPMorgan Chase (2016) and a release in 2018. Ownership transfers do not themselves constitute litigation, and I found no case in which Rockstar/RPX asserted the '389 patent.

4. Patent status. US 5,991,389 is listed as "Expired – Lifetime" with an anticipated expiration of 2017-05-30 (priority date 1996-06-13), which limits the possibility of later-era suits.

5. Unverified item. I could not confirm the final disposition (settlement, dismissal, judgment, or license) of the Vonage v. Nortel action within the searches available. I am explicitly flagging this as unverified rather than guessing. To confirm the outcome, the primary sources to pull would be the PACER docket for D. Del. 1:07-cv-00507-GMS and the Unified Patents litigation page for that case number.

Sources:

• Google Patents record for US5991389 (litigation flag to Delaware case 1:07-cv-00507): https://patents.google.com/patent/US5991389/en
• Unified Patents litigation entry: https://portal.unifiedpatents.com/litigation/Delaware%20District%20Court/case/1%3A07-cv-00507
• Nortel Answer & Counterclaim, D. Del. 1:07-cv-00507-GMS, D.I. 28 (Dec. 14, 2007) — asserts "U.S. PATENT NO. 5,991,389" (Claim 11)
• D. Del. AO 120 Report on the Filing of an Action, 1:07-cv-00507, filed Aug. 20, 2007 (Vonage Holding Inc. v. Nortel Networks Inc./Ltd.)

If you want, I can attempt to retrieve the final judgment/settlement documents for D. Del. 1:07-cv-00507 to fill in the outcome field.

I'll verify the structured block's "no proceedings" default against public sources before writing anything up.

Let me run a couple more targeted checks to be thorough about older/unindexed proceedings.

Let me check the sibling/related patents and confirm no AIA challenges exist on this family.

Proceedings overview

Total AIA trial proceedings on file for US 5,991,389: 0. Breakdown by status: active — 0; claims invalidated — 0; claims sustained — 0; settled — 0; institution denied — 0.

The structured "PTAB proceedings on file" block (USPTO Open Data Portal, most recent ingest) returns no AIA trial proceedings, and my independent web checks surfaced nothing to contradict it — no IPR, PGR, or CBM docket naming US 5,991,389 as the challenged patent, and no Federal Circuit appeal flowing from such a trial. I searched for IPRs against the number itself and against the sibling patent in the same Nortel family (US 6,028,924, "Apparatus and method for controlling processing of a service call," filed 1997-05-30 from app. 08/865,692) and found nothing.

The defensive posture this gives a defendant is therefore not "the patent has survived two IPRs and is hardened," and it is also not "claims 1–5 have been canceled." It is a third, more useful posture: there is no PTAB record to fight about, and the patent is expired. US 5,991,389 expired on 2017-05-30 (20 years from its 1997-05-30 filing date), and Google Patents records the legal status as "Expired - Lifetime." An IPR is not available against an expired patent for prospective relief, and any assertion today would be limited to past damages, with a § 286 six-year lookback that — for a complaint filed after 2023-05-30 — reaches back no earlier than a period in which the patent was already gone. Note the ownership chain ends in the name of a defensive aggregator, RPX Clearinghouse LLC, since 2015-02-09.

Because there are zero proceedings, there are no per-proceeding sections to write. I will not manufacture proceeding numbers, panels, or claim dispositions to fill the template. What follows is what the absence means and the traps to avoid.

A critical identification trap — read this before citing anything

My searches returned a large volume of hits referring to "the '389 patent" in PTAB and district court records — Cisco, Time Warner Cable, Rockstar Consortium, Spherix, and the Nortel bankruptcy docket in Delaware. Those are almost all U.S. Patent No. 6,845,389, a different patent, also from the Nortel portfolio. For example, the Time Warner Cable declaratory judgment counterclaim in the Nortel bankruptcy lists "the '499, '389, '048, '879, '917, and '299 patents" — where '389 is 6,845,389, not 5,991,389. The same string confusion appears in Rockstar/Constellation dockets and in the Cisco counterclaims.

Do not import claim-level outcomes, FWD language, estoppel positions, or cancellation findings from any of those matters into an analysis of 5,991,389. They are unrelated patents with an unfortunate shorthand collision. If opposing counsel hands you a brief that says "the '389 patent was invalidated in IPR," verify the patent number in the caption before you build a defense on it.

The one litigation data point in the structured record

The Google Patents "litigation" field for US 5,991,389 lists a single case:

• Delaware District Court, case 1:07-cv-00507, source: Unified Patents Litigation Data, tag "Critical."

The structured data gives no parties, no claim outcome, and no disposition — I could not verify any of those from independent sources, so I will not assert them. What matters procedurally: this case predates AIA trials entirely. The AIA trial regime (IPR/PGR/CBM) began for petitions filed on or after 2012-09-16. A 2007 suit could only have been met with inter partes reexamination (available from 1999) or ex parte reexamination — neither of which is an AIA trial proceeding and neither of which would appear in the ODP AIA proceedings list. If you need the pre-AIA reexamination history, that has to be pulled from the file wrapper at USPTO PatentCenter, not from the AIA trial docket. I have not verified whether any reexamination was requested and decline to speculate.

Strategic summary

Claim status: entirely UNTESTED at the PTAB. No claim of 5,991,389 has been canceled, confirmed, or even challenged in an AIA trial. There is no surviving-claims list to report because no claims were ever at risk. The claim set stands as issued (subject to any ex parte or inter partes reexamination history I could not confirm — that is the one open verification item). Independent claims would include the programmable-switch-matrix apparatus claims, the service-control-unit apparatus claims, the method claims reciting in-switch call processing to a detected trigger, and the service programming interface protocol claims reciting the bridge / collect-digits-and-report / connect / disconnect / new-call-accepted primitives plus the digits-collected / instruction-completed / new-call / off-hook / on-hook event notifications. Anyone representing that this claim set has been narrowed by IPR is wrong.

Estoppel landscape: empty. § 315(e)(2) estoppel only attaches to a petitioner that reached a final written decision. With zero FWDs, no petitioner and no privy is estopped from raising any ground — including § 102 and § 103 art on the very primitives and event-notification structures claimed. But this cuts both ways and is largely academic: because no IPR has ever been filed, there is also no favorable PTAB record to borrow, no claim-construction ruling from the Board, and no Board credibility finding on any prior-art reference. A defendant's invalidity case would start from zero at the PTAB and at the district court.

Pattern signals. The patent family here — 5,991,389 and its sibling 6,028,924, both from the 1996-06-13 Nortel priority date — shows no repeated-petitioner pattern and no defensive-aggregator challenge pattern. What it does show is a distinctive ownership pattern: Northern Telecom Ltd → Nortel Networks Corporation (1999-12-23) → Nortel Networks Limited (2000-08-30) → Rockstar Bidco, LP (2011-10-28, the Nortel bankruptcy asset sale) → Rockstar Consortium US LP (2014-03-05) → RPX Clearinghouse LLC (2015-02-09). RPX is a defensive patent aggregator whose business model is to take patents out of assertion, not to litigate them. Ownership landing at RPX in 2015, followed by expiration in 2017, is the single most informative fact about this patent's assertion risk profile. The patent owner did not pursue PTAB appeals aggressively — there were no PTAB proceedings to appeal from.

Recommended next steps

If a demand letter or complaint cites US 5,991,389:

1. Check the expiration date first, on the face of the document. The patent expired 2017-05-30. Under 35 U.S.C. § 286, damages are limited to six years before the complaint's filing; for any complaint filed on or after 2023-05-30, the entire recoverable window postdates — or is entirely consumed by — the patent's expiry. Put the filing date and the 2017-05-30 date side by side in your first response and let the arithmetic do the work.
2. Demand the current owner's standing documentation. The § 282 burden-shifting and Rule 11 posture both turn on who actually holds title. The recorded chain ends at RPX Clearinghouse LLC (2015-02-09); confirm whether any later assignment exists.
3. Ask for the file wrapper. Since I could not confirm the reexamination history on this patent, this is the one genuine evidentiary gap. Pull the PatentCenter file wrapper and check for any ex parte or inter partes reexamination certificate, which would narrow or confirm the currently-in-force claim scope.
4. If you want a PTAB record and the patent were somehow still live, note that none exists. There is no FWD to link to and quote, because no FWD was ever issued. Any statement to the contrary — by an adversary or in a licensing pitch — is a fabrication, and it is worth calling out as such.

If instead you are evaluating this patent as part of a broader Nortel-portfolio diligence: be extremely careful with "'389" shorthand. US 5,991,389 and US 6,845,389 have both been called "the '389 patent" in real filings across this portfolio, and the latter — not the former — is the one with the Rockstar / Constellation / Cisco / Time Warner Cable assertion history. Confirm the full seven-digit number in the caption of every document before relying on it.

Verification note. Everything in the "no proceedings" finding rests on (a) the structured ODP block in this prompt and (b) my searches, which returned no contrary proceeding. If a recently-filed petition exists that neither source has indexed, I did not find it and cannot rule it out — but the patent's 2017-05-30 expiration makes a new petition impractical for prospective relief, and RPX's ownership since 2015 makes it unlikely.

I'll search for the recorded assignment data (reel/frame, conveyances, correspondents) for US 5,991,389, plus corroborating sources for the assignees.

Let me dig for the specific reel/frame records and correspondents for this patent number.

Assignment & Ownership Analysis — US 5,991,389

Data-integrity note up front. The assignmentcenter.uspto.gov / assignment.uspto.gov record pages are not directly retrievable by search engine, and my tool calls for the exact '389 reel/frame strips were cut off at the step limit. What follows separates (a) facts on the '389 record itself (Google Patents legal-events feed, which mirrors the USPTO assignment data, and the USPTO-sourced release reel 038041/0001) from (b) reel/frame numbers verified only on sibling Nortel/Rockstar patents in the same bulk transfers. I label each accordingly and do not present sibling reels as confirmed '389 reels.

Inventors

All four are Dallas–Fort Worth suburbs surrounding Nortel's Richardson, Texas facility, and all four appear as assignors on the 1997-12-22 assignment of interest to Northern Telecom Limited — i.e., straightforward employee invention assignments, not founder-held rights.

Unusual-pattern check — "all inventors departing within 12 months of filing": NOT PRESENT. Both Ram and Thompson continued inventing for the Nortel entity well after filing:

• Ram is a named inventor on US 6,625,258 ("System And Method For Providing Unified Communication Services Support"), filed 1999-12-27, assigned to Nortel Networks Ltd (St. Laurent).
• Thompson is credited with 11 US patents while at Nortel Networks Limited, including '389.

So there is no evidence of an inventor exodus that would precede a portfolio fire-sale. The departures that mattered here were corporate (bankruptcy), not individual.

Original assignee

Northern Telecom Limited (Quebec, Canada) — named on the face of the issued patent (Justia Patent History: Assignee: Northern Telecom Limited (Quebec); prosecution firm of record Carr & Storm, L.L.P.; Primary Examiner Scott Wolinsky).

• Primary line of business: public telecommunications switching and transmission equipment. The specification states the programmable switch matrix "may be based on a DMS-250 programmable switch matrix designed and manufactured by Nortel, Inc." — i.e., a carrier tandem/toll switch.
• Product embodying the claims: Substantially yes, on the record's own terms. The '389 architecture (Programmable Service Node / Service Control Unit / Media Resource Unit, driven by the SPI primitive protocol) is described as an implementation on the DMS-250 platform, Nortel's commercial product line. I did not find an independent product brochure confirming a shipped "PSN" SKU, so treat "shipped a product embodying the claims" as supported by the spec, not independently corroborated.
• Current status: Bankrupt / wound down. Nortel filed for creditor protection in June 2009 (US Chapter 11, Bankr. D. Del. No. 09-10138; Canadian CCAA). Its ~10,000-patent portfolio was auctioned; the winning $4.5B bid (Asset Sale Agreement dated 2011-06-30) went to Rockstar Bidco, LP. Nortel's remaining estate was allocated in the joint US/Ontario proceeding (Nortel Networks Corporation (Re), 2015 ONSC 2987). The patent's original assignee no longer exists as an operating entity.
• Assignment chain of corporate name (per the '389 legal-events record): Northern Telecom Limited → Nortel Networks Corporation (change of name, recorded 1999-12-23) → Nortel Networks Limited (change of name, recorded 2000-08-30).

Assignment timeline

Reel/frame marked [SIBLING-VERIFIED] was confirmed on other Nortel/Rockstar patents in the same bulk recording, not on the '389 record itself. Reel 038041/0001 [389-VERIFIED] is taken directly from the '389 legal-events entry.

1. 1997-12-22 (recorded) — Reel not retrieved for '389 (a Nortel-family original-assignment reel seen on sibling US 6,778,653 is 010504/0344 [SIBLING-VERIFIED])

   • Conveyance: Assignment of assignors' interest ("ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)")
   • Assignors: Kent W. Smith; Keith W. Landau; Geetha R. Ram; Roger L. Thompson
   • Assignee: Northern Telecom Limited
   • Correspondent: not retrievable from available sources
   • Context: Routine employee-to-employer assignment of invention rights — not an acquisition.

2. 1999-12-23 (recorded) — Reel not retrieved

   • Conveyance: Change of Name
   • Assignor: Northern Telecom Limited → Assignee: Nortel Networks Corporation
   • Correspondent: not retrievable
   • Context: Internal corporate renaming only; no change in beneficial ownership.

3. 2000-08-30 (recorded) — Reel not retrieved

   • Conveyance: Change of Name
   • Assignor: Nortel Networks Corporation → Assignee: Nortel Networks Limited
   • Correspondent: not retrievable
   • Context: Internal corporate renaming/reorganization; no new consideration.

4. 2011-10-28 (recorded); effective/executed ≈2011-07-29 per sibling INPADOC — Reel 027164/0356 [SIBLING-VERIFIED on US 6,309,117 and cited as the Nortel→Rockstar link in the IPR2014-01167 chain-of-title for US 6,778,653]

   • Conveyance: Assignment of assignors' interest
   • Assignor: Nortel Networks Limited → Assignee: Rockstar Bidco, LP (New York)
   • Correspondent: not retrievable (deal counsel for the auction were Paul Weiss for Ericsson and Canadian counsel LaBarge Weinstein for the consortium; neither is confirmed as the recorded correspondent)
   • Context: Bankruptcy fire-sale — the $4.5B Nortel patent auction, June 30, 2011 Asset Sale Agreement.

5. 2014-03-05 (recorded); effective 2012-05-09 per sibling INPADOC — Reel 032422/0919 [SIBLING-VERIFIED on US 6,309,117] and Reel 031760/0848 [SIBLING-VERIFIED on US 6,778,653] — two reels appear for the same transfer on different family members; treat the paired reels as evidence of more than one recording of this step.

   • Conveyance: Assignment of assignors' interest
   • Assignor: Rockstar Bidco, LP → Assignee: Rockstar Consortium US LP (Texas)
   • Correspondent: not retrievable
   • Context: Internal consortium transfer to the dedicated licensing/assertion vehicle.

6. 2015-02-09 (recorded); effective 2015-01-28 per sibling INPADOC — Reel 034924/0779 [SIBLING-VERIFIED on US 6,309,117]

   • Conveyance: Assignment of assignors' interest
   • Assignors: Rockstar Consortium US LP; Rockstar Consortium LLC; Bockstar Technology LLC; Constellation Technologies LLC; MobileStar Technologies LLC; NetStar Technologies LLC → Assignee: RPX Clearinghouse LLC (California)
   • Correspondent: not retrievable (Skadden, Arps advised RPX on the deal; not confirmed as recorded correspondent)
   • Context: Defensive aggregation / clearinghouse — the $900M Rockstar→RPX transaction announced 2014-12-23, closed 2015; syndicate of 30+ licensees (incl. Cisco, Google); FRAND licensing commitment.

7. 2016-03-09 (recorded) — Reel not retrieved

   • Conveyance: Security Agreement
   • Assignors: RPX Clearinghouse LLC and RPX Corporation → Assignee (collateral agent): JPMorgan Chase Bank, N.A.
   • Correspondent: not retrievable
   • Context: Securitization — collateral pledge of the RPX patent portfolio to a lender; not a change in beneficial ownership.

8. 2018-01-02 (recorded) — Reel 038041/0001 [389-VERIFIED] — note the record itself references "REEL 038041 / FRAME 0001"

   • Conveyance: Release
   • Assignor: JPMORGAN CHASE BANK, N.A. → Assignee: RPX CORPORATION / RPX CLEARINGHOUSE LLC
   • Correspondent: not retrievable
   • Context: Release of the 2016 security interest; confirms RPX Clearinghouse LLC as unencumbered owner of record.

No post-2018 assignment appears on the '389 record. The patent expired (anticipated expiration 2017-05-30), so no further transfers are expected.

Timeline diagram

timeline
    title Ownership of US 5991389
    1996 : Priority date June 13
    1997 : Filed May 30
         : Inventors assign to Northern Telecom Limited
    1999 : Patent issued November 23
         : Name change to Nortel Networks Corporation
    2000 : Name change to Nortel Networks Limited
    2009 : Nortel files for creditor protection
    2011 : Nortel portfolio sold to Rockstar Bidco LP
    2014 : Transferred to Rockstar Consortium US LP
    2015 : Acquired by RPX Clearinghouse LLC
    2016 : Security interest pledged to JPMorgan Chase
    2017 : Patent expires
    2018 : Security interest released

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT. The patent left the operating company (Nortel Networks Limited) for Rockstar Bidco, LP (reel 027164/0356, recorded 2011-10-28) and then for Rockstar Consortium US LP (reel 032422/0919, recorded 2014-03-05). Rockstar Consortium is not merely a holding name: it is a licensing-only entity with no products in commerce, chartered to "license [patents] to other companies on a royalty-bearing basis" (Rockstar press release, 2012-03-12). Where I must be careful: Rockstar's initial 2011 assignee, Rockstar Bidco LP, was a consortium vehicle for operating companies (Apple, Microsoft, RIM, Ericsson, Sony, EMC) rather than a classic single-purpose Delaware shell — but the 2012/2014 transfer into Rockstar Consortium US LP is the pure assertion vehicle.

2. Known asserter in the chain — PRESENT (with a naming caveat). Rockstar Consortium does not appear on the enumerated list you supplied (Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, DGC, Spangenberg), so I cannot call it a listed NPE. It is nonetheless a documented mass-assertion entity: it sued Google, Cisco, Samsung, HTC and others (2013 onward), and a party filing in the Nortel bankruptcy stated that "In the Texas Case, Constellation asserts that TWC infringes the '649, '389, '048, '879, '917, and '299 patents" (Bankr. D. Del. 09-10138, D.I. 14559). Constellation Technologies LLC is one of the six assignors named on the '389→RPX recording (reel 034924/0779). This is a contradiction with the earlier litigation section of this analysis, which concluded "no case in which Rockstar/RPX asserted the '389 patent" — flagging explicitly below.

3. Repeat correspondent across the chain — UNCLEAR / NOT RETRIEVABLE. I could not obtain the recorded correspondent (attorney/firm/address) for any of the eight recordings; the Assignment Center detail pages that expose that field were not reachable. I therefore make no recurrence finding. Do not treat the deal-counsel names (Skadden for RPX; LaBarge Weinstein for Rockstar; Paul Weiss for Ericsson) as recorded correspondents — they are transaction counsel, a different thing.

4. Cascading transfers — PRESENT (moderate). Three ownership changes inside ~44 months: Nortel Networks Ltd → Rockstar Bidco LP (2011-10-28) → Rockstar Consortium US LP (recorded 2014-03-05) → RPX Clearinghouse LLC (recorded 2015-02-09). Two of the three are within a 12-month window (2014-03 → 2015-02). However, these were bulk portfolio transfers with distinct economic rationales (bankruptcy sale, consortium reorg, syndicated buy-out), not the classic pattern of a patent ping-ponging through identically-corresponded shell LLCs. Rate it as a weak-to-moderate cascade signal.

5. Pre-litigation transfer — NOT PRESENT on the 6-month test. The first Rockstar-era infringement activity dates from October 2013; the '389 was acquired by Rockstar ~26 months earlier (2011) and was moved to RPX after the litigation (2015). No assignment falls within six months before a first suit naming this patent.

6. Bankruptcy fire-sale — PRESENT (strong). Direct hit. Nortel's insolvency (June 2009) led to the June 30, 2011 Asset Sale Agreement and the $4.5B auction to Rockstar Bidco; the recorded transfer (reel 027164/0356, recorded 2011-10-28) is precisely the Kodak/Nortel/Polaroid genre of sale-out-of-bankruptcy.

7. Privateering — PRESENT (strong). This is the defining feature of the chain. Five/six operating companies (Apple, Microsoft, RIM/BlackBerry, Ericsson, Sony, EMC) bought a failed rival's portfolio through an intermediary (Rockstar Bidco LP → Rockstar Consortium US LP) and then asserted it against competitors — Google, Cisco, Samsung, HTC, and (per the bankruptcy filing) Time Warner Cable via Constellation Technologies LLC. The DOJ antitrust review and the 2014 RPX syndication deal (30+ licensees, incl. Cisco and Google, paying $900M to end the litigation) are both on the public record.

8. Defensive aggregator — PRESENT (chain terminus). The chain terminates at RPX Clearinghouse LLC (reel 034924/0779, recorded 2015-02-09), a subsidiary of RPX Corporation — one of the defensive aggregators you enumerate. RPX publicly pledged not to assert the acquired patents and licensed them to a 30+ company syndicate on FRAND terms; the 2018 release (reel 038041/0001) cleared the JPMorgan security interest. Inverse signal: this patent has been neutralized since 2015.

Verdict

Defensive / non-asserting (chain terminates at a defensive aggregator).

The current and final assignee of record is RPX Clearinghouse LLC, acquired by assignment recorded 2015-02-09 at reel 034924/0779 [SIBLING-VERIFIED] and unencumbered by the 2018 release at reel 038041/0001 [389-VERIFIED] — a defensive aggregator on your enumerated list that pledged not to assert the portfolio. That said, the verdict describes only the terminal state: this is emphatically not a "clean" operating-company patent. Signals 1, 6 and 7 (shell-entity transfer, bankruptcy fire-sale, privateering) are all present, and the 2011–2015 Rockstar holding interval is a textbook privateering arrangement under which the former-Nortel portfolio — plausibly including this patent — was asserted against Google, Cisco, Samsung, HTC and (via Constellation Technologies LLC) Time Warner Cable. Anyone reading "defensive/non-asserting" as "never asserted" would be wrong; the correct reading is "asserted by a privateer, then bought out and neutralized by a defensive aggregator."

Verification links:

• USPTO Assignment Center (search patent 5,991,389 or reel 034924/0779 and 038041/0001): https://assignmentcenter.uspto.gov/
• Legacy Assignment Search UI: https://assignment.uspto.gov/patent/index.html
• Google Patents legal events for '389 (source of the eight dated entries above): https://patents.google.com/patent/US5991389/en
• RPX/Rockstar transaction (SEC 8-K, Dec. 22 2014): http://ir.rpxcorp.com/static-files/a9d14a7b-4dba-4d1f-9358-d15d9e412100

⚠️ Explicit contradiction with the earlier litigation section

The previously generated Litigation summary states: "No other litigation found… I found no case in which Rockstar/RPX asserted the '389 patent." Two pieces of evidence found during this assignment analysis cut against that conclusion, and I am surfacing them rather than silently harmonizing:

1. Constellation Technologies LLC v. Time Warner Cable (E.D. Tex.) — a filing in the Nortel bankruptcy (Bankr. D. Del. 09-10138, D.I. 14559) states that "In the Texas Case, Constellation asserts that TWC infringes the '649, '389, '048, '879, '917, and '299 patents." Constellation Technologies LLC is one of the six assignors named on this patent's RPX recording. Caveat: the filing abbreviates the patent as "'389" and I could not independently confirm it is US 5,991,389 rather than a different '389-numbered patent in the Constellation/TWC set. This is a strong lead, not a confirmed hit.
2. The 2007 D. Del. action (1:07-cv-00507) is unaffected by the above — it predates Rockstar ownership and is consistent with Nortel-era ownership.

Recommendation: pull the Constellation v. TWC complaint (E.D. Tex.) patent list and the PACER docket for D. Del. 1:07-cv-00507 to (a) confirm whether the asserted '389 is this patent and (b) fill the still-unverified outcome field in the litigation section. Until then, the earlier section's "no other litigation" statement should be treated as incomplete, not authoritative.

I'll search the USPTO and related databases specifically for the prior-art references cited on US 5,991,389.

I found the non-patent citations and need the complete "Patent Citations" list for US 5,991,389 itself (careful — one result's "Patent Citations (23)" belonged to a different patent, US 2006/0008069). Let me pull the '389 front-page references directly.

Prior-Art Analysis — US 5,991,389 (Ram et al.)

Search executed: I queried for the literal number 5991389 / US5991389 / 5,991,389 against Google Patents' USPTO mirror and the citing/cited tables, restricted to that number. I did not substitute or auto-correct any number.

⚠️ Two honesty flags before the analysis

1. I could not retrieve the verified front-page list of U.S. Patent Documents cited on US 5,991,389 within this session's search budget. When a search result showed a "Patent Citations (23)" table containing US 4,713,806; US 5,157,390; US 5,551,035; US 5,168,515; US 5,323,452; US 5,185,782; US 5,475,817, etc., I traced that table to the page for US 2006/0008069 ("Method and apparatus for providing real time execution of specific communications services in an intelligent network") — a later document that cites the '389 patent. Those references are that patent's citations, not the '389 patent's. I am deliberately not attributing them to US 5,991,389. Fabricating that list would be worse than reporting the gap.

2. The non-patent citations below were retrieved from a page that also rendered the US5991389A row, so attribution should be confirmed on the '389 front page itself. Their subject matter (AIN service-execution/service-creation environments, 1990–1993) is exactly the profile of an examiner's "Other Publications" list for this patent, but I flag it as retrieved-but-not-fully-confirmed.

A. Verified prior-art references (the "Other Publications" / Non-Patent Citations, 5 items)

These are printed publications cited on the '389 record. All pre-date both the 1996-06-13 priority date and the 1997-05-30 U.S. filing date, so each qualifies as prior art under pre-AIA § 102(b) (more than one year before the U.S. filing date).

§ 102 caveat that matters here. None of these five is, on the record I retrieved, a single-reference anticipation of the '389 claim set as a whole. Anticipation requires one reference disclosing every limitation as arranged in the claim. On their face these teach the architecture (external/distributed service logic, a resource-control API) but the references do not appear to disclose the specific combination the '389 independents recite — trigger database in-switch + New Call event notification carrying call-type and port-address data + New Call Accepted returning an SCU return address + the enumerated primitive/event set (claim 16 and claim 14 in particular). Realistically these were § 103 combination art, which is consistent with the examiner citing them as a group rather than alone. Applicant/agent of record: Baker & Botts, L.L.P. (per the sibling record).

(I did not retrieve the examiner's actual rejection or the file wrapper, so I cannot say which of these was applied with a § 103 rationale or in what combination. Flagging that explicitly.)

B. Documents citing the '389 patent — not prior art against it

Search results repeatedly surfaced these; I list them to prevent a common counting error. They post-date (or are contemporaneous with) the '389 filing and therefore cannot anticipate it:

• US 6,028,924 A — Ram et al., "Apparatus and method for controlling processing of a service call," issued 2000-02-22, Northern Telecom. This is the co-pending sibling expressly incorporated by reference in the '389 family (Ser. No. 08/866,229 and 08/865,887 are the two concurrent cases). Sibling/§ 102(e)-era same-family matter — not citable as prior art against '389.
• WO 1997/048238 A2/A3 and AU 727,007 B2 — "Programmable service node for call control processing," Northern Telecom, same 1996-06-13 priority family.
• WO 1997/036430 A1 — "Service logic portability based on interface definition of execution environment in an intelligent network," Northern Telecom, 1997-10-02.

These are family/later documents, and the correct label for them is "cited by" or "same-family," not "prior art."

C. What is still needed to complete this task properly

To close the one real gap (Section A's U.S. patent-document references), the authoritative primary sources are:

1. USPTO Patent Public Search (https://ppubs.uspto.gov) — retrieve the '389 full document and read the (56) References Cited block.
2. USPTO PatentCenter for application 08/866,229 — includes the examiner's citations and any PTO-892.
3. Google Patents front page https://patents.google.com/patent/US5991389A/en — the "Patent Citations" table (I retrieved the surrounding page but not that table's rows).

Given that this patent expired 2017-05-30 (term from 1997-05-30 filing; listed "Expired – Lifetime," status Critical/Current per the record), § 102 analysis at this point is historical validity context, not an enforcement posture — any challenge window (IPR/PGR) and damages window have closed, and the previously documented litigation (D. Del. 1:07-cv-00507-GMS, Vonage v. Nortel) predates expiration by a decade.

Bottom line: The five references I could verify are all 1990–1993 IN service-execution/API papers, qualifying as § 102(b) prior art, individually most probative against claims 11 and 16 (external service-application execution) and against the SPI-protocol / method-of-communicating independents (Sage in particular, as an IN resource-control API). I am not certifying the U.S.-patent-document citation list for '389, because the only such list I retrieved belonged to US 2006/0008069.

I'll build on the prior sections by first pulling the actual prior-art references cited on the face of this patent, since the "Prior Art section" material wasn't in the text I was given.

Let me pull the actual front-page "Patent Citations" list for US 5,991,389 and the sibling Nortel applications.

Let me get the examiner-cited prior art from the family's international/EP search reports, which is the most reliable "Prior Art section" for this family.

I hit the tool-step limit, so I'll write the analysis now. Note up front that I could only partially retrieve the face-of-patent "Prior Art" tables before the limit; I say exactly what I did and did not verify, and I do not fill the gaps with invented reference numbers.

Obviousness Analysis — U.S. Patent No. 5,991,389 under 35 U.S.C. § 103

0. Evidence base for the "Prior Art section" (and its limits)

Your instruction was to use the Prior Art section of the page. The Google Patents record for US5991389 (the authoritative full text supplied) exposes that section in four pieces, which I used:

1. "Prior art keywords" (auto-generated, not examiner-selected): switch matrix, port, call, service, control unit.
2. "Prior art date": 1996-06-13 (this is actually the priority date, per the record's own assumption statement).
3. Patent Citations / Cited By tables — of which I could retrieve only a partial list before the tool limit.
4. Family search-report art (EP 0906705 / WO 1997/048240 / AU 719309 family) — I could not retrieve the examiner's category X/Y/A listing for this family.

Explicitly flagged retrieval gap (consistent with the caveat already in the Patent summary section): I was not able to pull the complete front-page "(56) References Cited" list for US5991389, nor the WO 97/048240 / EP 0906705 search report. What I did retrieve of the citation tables produced these references, which I treat below as Category A (verified in the record):

Category B — applicant-admitted prior art (highest-value, and not subject to any retrieval doubt). The Specification itself is an extended admission of the prior art under MPEP § 2129 / In re Fout: it states that Intelligent Networking "is driven by standards that have defined (and include) components such as Service Control Points (SCP) and Service Switching Points (SSP) communicating via a common channel signaling system 7 (CCS7) infrastructure"; that IN "decouple[s] service development from internal switch implementation"; and that "stand-alone programmable switching matrices" with external programmability were known (with the admitted deficiencies of poor reliability/capacity and duplicate resource use). These admissions alone supply most of the preamble-level elements of every independent claim.

Category C — well-known field art I rely on with a confidence caveat (public before 1996-06-13; I did not retrieve documents for these in this session): Bellcore AIN 0.1/AIN 0.2 / AIN Release 1 (SSP call model, trigger tables, TCAP query to SCP); ITU-T Q.1201/Q.1204/Q.1211/Q.1214/Q.1218 (IN CS-1, INAP operations, BCSM, SIBs, SRF/assist operations); ITU-T Q.761–Q.764 (ISUP); Q.931 (PRI); and the published service-creation literature that appears in sibling-family search reports (Zolzettich, "Customized service creation," 1992; Ebert et al., "Application of a service-independent architecture," ISS 1990; Fujioka et al., IEEE Comm. Mag. 1991; Bosco et al., 1992; Lauer, IEEE Network 1994). I flag these as Category C precisely because I could not re-verify them this session — they should be document-pulled before being relied on in a filing.

Category D — not usable, and why. The sibling/companion applications named in the specification (Ser. No. 08/866,229 itself; Ser. No. 08/865,887; and US 6,028,924 "Apparatus and method for controlling processing of a service call"; WO 97/048238 "Programmable service node for call control processing") are same-family, same-inventor, commonly owned. Even where they might qualify as pre-AIA § 102(e) art, they are disqualified for § 103 purposes by pre-AIA § 103(c)(1). I do not rely on them.

Legal framework note: The application was filed 1997-05-30 with a 1996-06-13 priority date; the patent expired 2017-05-30. Pre-AIA § 102/§ 103 governs. The controlling obviousness standard is Graham v. John Deere as refined by KSR Int'l v. Teleflex, 550 U.S. 398 (2007) — i.e., the TSM test is not the sole route; a combination may be obvious where the elements are known, the combination is of the "familiar … try[ing] … a finite number of identified, predictable solutions," each element performs its known function (Sakraida), or a design/market incentive existed. MPEP §§ 2143, 2144.

1. Level of ordinary skill and claim-construction anchors

POSITA: a telecommunications switching engineer with a B.S. in electrical engineering/computer science plus ~3–5 years of experience in stored-program-controlled (SPC) switch call processing, and familiarity with CCS7/TCAP, SS7 ISUP, ISDN Q.931, and either IN/AIN or computer-telephony (host-controlled switch) architectures as of June 1996. This is the skill level the patent itself assumes (its Background is written for such a reader).

Constructions that matter for § 103 (the full claim set was not in the text I was given — see the standing caveat in Patent summary — so these are keyed to the specification and to the substance of the independents as previously reconstructed):

• "primitive" = an instruction message sent SCU→switch; the specification is explicit: "a primitive is an instruction that controls the programmability of the programmable switch matrix 24 and controls the call or agent processing." This is the functional equivalent of an INAP/TCAP operation (e.g., Connect, ReleaseCall, CollectUserInformation, PlayAnnouncement, AssistRequestInstructions).
• "event notification" = a status message switch→SCU reporting a port event, sent either in response to a primitive or as a peripheral (asynchronous) event — functionally the BCSM event report / TCAP response.
• "macro" = multiple primitives in one message — a message-packing design choice identical in concept to multiple TCAP/INAP components in one dialogue.
• "trigger"/"trigger database" = match on call information to decide whether external control is invoked — the AIN trigger table / detection point.
• "service programming interface protocol" = a set of named messages. This is where § 103 analysis must confront In re Swinehart, 439 F.2d 210 (CCPA 1971), and the printed-matter cases: the names of the primitives carry no patentable weight; only their functional content does, and that content is the conventional IN operation set.
• "agent"/per-port control — the specification's own gloss for "port or trunk member," with "no concept of an originator or terminator." This is the one construct with genuine room for a nonobviousness argument (§ 6 below).

2. Ground A (the primary ground): Bellcore AIN / ITU-T IN CS-1 in view of a stored-program-controlled switch, plus the applicant's own admissions

This is the strongest ground because the patent's Background concedes every architectural element.

What AIN already taught before 1996-06-13 (Category C, and mirrored by the Category B admissions):

The only arguable delta is the transport: CCS7/TCAP vs. the claimed "communications link" (a 10 Mbit/s Ethernet 10BaseT LAN in the preferred embodiment). That is a textbook KSR combination: substituting a known general-purpose data link for a known special-purpose signaling link, with each performing its known function. The specification itself supplies the motivation: it complains that CCS7/TCAP "leaves little room for differentiation of service offerings" and that IN "complexity is high thus requiring a phased implementation." The artisan in June 1996 — with the IN objectives already articulated but unmet, and with ubiquitous Ethernet/TCP-IP LANs driving down cost — had both explicit reason (overcome TCAP's expressiveness limits; use commodity LANs) and reasonable expectation of success (the switch already used Ethernet/TCP-IP for OA&M, as the specification admits by placing the EIUs in the link peripheral processor and by describing "Ethernet (IEEE 802.3) with TCP/IP and X.25/X.75"). KSR squarely covers this: "if a technique has been used to improve one device, and a person of ordinary skill … would recognize that it would improve similar devices in the same way, using the technique is obvious."

Motivation-to-combine summary (Ground A):

• (i) Same field, same purpose, same problem — both AIN and the '389 invention exist to decouple service development from switch-vendor release cycles (specification: "services can be written independently from the switch, implemented by third-parties").
• (ii) KSR "predictable variation" — replacing one message transport with another, where both are message-based, procedural, and both had already been used in the same switch.
• (iii) "Obvious to try" — the artisan had a finite set of signaling transports (CCS7/TCAP, X.25, Ethernet/TCP-IP) and understood their trade-offs; the specification's background is essentially a recitation of those known trade-offs.
• (iv) Market/competitive pressure — the Background frames the problem as competitive service differentiation, a classic KSR market-force motivation.

Result: Claims 1, 11, and 16 would have been obvious over AIN/CS-1 + a programmable SPC switch + the applicant's admissions. Claim 16's "call type data + address of the first port → select service application" is, in substance, the AIN service-key computation; Claim 14's "return address" is the SCP point code / transaction identifier correlation already inherent in TCAP dialogues.

3. Ground B: Host-controlled / external-logic switching (Category A refs)

This ground attacks the "programmability comes from outside the switch" concept without needing IN at all, and it is anchored in references I actually retrieved.

US 5,168,515 (Unifi, "Switchless automatic call distribution system," 1992) is a particularly strong § 103 anchor for claim 6 and the MRU limitations: it discloses a host computer that establishes and tears down connections among ports and adjunct resources (voice-response/announcement units) on a per-port basis, i.e., the external processor is the master and the switching resource the slave — exactly the "master–slave relationship … SCU is the master and the programmable switch matrix 24 is the slave" language in the '389 specification. Substitute a switch fabric with a programmable matrix (admitted known; the DMS‑250 is named) and the claimed system reads on the combination.

US 5,157,390 (Hitachi, "Method for providing network service in a communication system," 1992) adds the network-service-provisioning-in-a-communication-system teaching (external determination of service handling).

US 4,713,806 (AT&T Bell Labs, "Communication system control arrangement," 1987) supplies the long-standing teaching that a central control arrangement can direct distributed switching elements — useful for the "call processing circuitry … connects the first port to the second port in response to [] primitives received" limitation.

US 5,551,035 (Lucent, "Inter-object communication in an object-oriented program controlled system," 1996) supplies the mechanism and the motivation for message/packet-transported instructions between separately located software objects — directly supporting the "primitives over a data link" and "SPI protocol handler" aspects, and giving the artisan a reason to carry them over a general-purpose link (the reference is expressly about inter-object messaging, not CCS7).

US 5,185,782 (AT&T, 1993) — ACD arrangement — corroborates the automated-distribution/agent-status reporting environment in which "agent" status events are per-port (supporting the Off-Hook/On-Hook/Instruction-Completed notification claims).

Motivation to combine (Ground B): all four references are in the same field of endeavor (telephony call control) and address the same problem (getting service/call logic out of the switch and into a reprogrammable external processor). KSR factor (i) ("design incentives … in the same field") and (iii) ("one of the known options") are satisfied. No reference teaches away; US 5,168,515's "switchless" approach is not a teaching away from adding a programmable matrix, because the specification itself concedes programmable matrices are known and merely criticizes their reliability/capacity — which is precisely the problem solved by putting the matrix inside a carrier-grade SPC switch (the DMS‑250). That is a combination of known elements each performing its known function → Sakraida.

4. Ground C: Service-creation/service-logic portability art supplies the "sequence of primitives" and the "select a service application by call type"

Two of the retrieved references are directed at external definition of call-processing logic:

• US 5,323,452 (Bellcore, "Visual programming of telephone network call processing logic," 1994) — teaching graphical/off-switch authoring of call-processing logic, i.e., the artisan was already taught to express service behavior as a sequence of primitive call-processing steps generated outside the switch. This is the strongest Category-A support for Claim 16's "executing the selected service application program, the selected service application program generating one or more instructions," and for the entire family of "primitives" claims.
• The service-independent-architecture literature (Category C: Ebert 1990; Fujioka 1991; Zolzettich 1992; Bosco 1992; Lauer 1994) teaches SIBs — standard, reusable, service-independent functional blocks with standardized inputs/outputs — which is the textbook definition of an INAP-style primitive catalog. The artisan looking at "Bridge, Collect Digits & Report, Connect, Disconnect, New Call Accepted" would recognize nothing more than a domain-specific instantiation of a known, standardized service-independent building-block set.

Critical point for the protocol claims (the SPI-protocol independent and the method-of-communicating independent): these claims are drawn to a menu of named messages. Under In re Swinehart and the printed-matter line, the names ("bridge primitive," "new call accepted primitive") are non-functional descriptive matter; only the functional semantics are compared to the art — and each semantics is a known IN/switch operation. Further, selecting the membership of a protocol message set (whether to include a "Bridge" primitive) is the epitome of "a finite number of identified, predictable solutions" (KSR). A claim whose only distinction is which conventional call-control operations were chosen to be enumerated is obvious as a matter of law absent evidence that the selection produced an unexpected technical result.

5. Ground D: Conferencing/bridging art for the "Bridge" and "Stop Message / Play Message" limitations

The Bridge primitive ("connect a plurality of predetermined ports to each other within the switch matrix," optionally bridging a message/announcement) is a conference-circuit / meet-me bridge function, and the specification concedes the switch already has "a plurality of six port (or more) conference circuits."

• US 5,903,637 (Linkusa Corp., priority 1994-06-08 — verify date/assignee before reliance), "Teleconferencing bridge with edgepoint mixing / System and method for call conferencing," is directly on point for externally-controlled multiparty conferencing: a conferencing bridge that mixes/bridges multiple endpoints under external control. Combined with Ground A or B, it renders the Bridge primitive and the "bridged ports no longer hear the message" behavior of the Stop Message primitive obvious (and the specification's own Stop-Message description is just ordinary conference algebra).
• Announcement/Play-Message/Play-Prompt limitations are met by the admitted "announcement machine, such as Nortel's Enhanced Digital Recorded Announcement Machine (EDRAM)" plus the admitted known "interactive voice response" and VRU capabilities of the MRU. In IN terms, this is the Specialized Resource Function (SRF)/Intelligent Peripheral, whose assist/connect-to-resource operations and announcement priming were standardized before the priority date (IN CS-1, Q.1214/Q.1218).

Motivation (Ground D): the artisan had (a) a switch with conference circuits, (b) a known announcement machine, and (c) an external controller already tasked with call manipulation. Exposing the conference circuit and announcement machine as controller-invocable primitives requires no new architecture — only new message definitions, which Swinehart removes from the § 103 calculus.

6. Where the patent has its best (but still vulnerable) nonobviousness arguments

I flag these honestly rather than pretending the whole claim set falls cleanly:

1. Per-"agent" control with single-ended supervision ("no concept of an originator or terminator"; each call half controllable independently). This is the strongest candidate for a genuine § 103 distinction, and it is notable that the Litigation summary shows claim 11 was the asserted claim in Vonage v. Nortel — so this is the claim whose validity would actually be litigated.
   • Rebuttal: the BCSM is already split into O_BCSM (originating) and T_BCSM (terminating) halves with independently reported events and separate detection points (IN CS-1). Once the artisan adopts an IN-style model, treating each half/port as an independent controllable entity is a predictable implementation choice, not an inventive leap. The specification offers no unexpected-result data for single-ended supervision — no latency, capacity, or call-completion statistics. Under KSR, an assertion of a difference without evidence of a resulting difference in function is insufficient.
2. "New Call Accepted" carrying a return address ("port service information," e.g., an IP address) so subsequent events go to a specific SCU process.
   • Rebuttal: routing subsequent messages to a designated address is the ordinary function of a transaction identifier/point code in TCAP and of any client-server session protocol; the artisan also knew IP-addressed messaging from the switch's own OA&M Ethernet. Motivation: load distribution across multiple SCUs (which the specification itself describes via the "arbitrator"/CDM) is an admitted design goal.
3. Co-existence of AIN triggers and SCU triggers, with AIN able to redirect a call to the SCU.
   • Rebuttal: AIN trigger tables are per-dialed-digit/per-call-type provisioned data; adding a trigger type that points to a new service platform rather than the SCP is a routine datafill change in a table the artisan already administered. MPEP § 2144.04 (obviousness of a change in a database/table when the data itself is a design choice). No teaching away.
4. The specific FSM (FIGS. 12a–12c) and the layered SCU software architecture (data-comms/middleware/service layers with CDM/CCM/DBM/TRK-RM).
   • Rebuttal: these are described in the specification but largely not claimed; where they are claimed (e.g., "arbitrator" IP-address functions), they map onto the SCP-side service-logic management of IN, plus the object-oriented inter-object messaging of US 5,551,035. A claim to a software layering that yields no asserted unexpected performance benefit is obvious under KSR's "predictable variation."

No reference retrieved teaches away. The specification's criticisms of IN (slow standards, complexity) and of standalone programmable switches (reliability, capacity, duplication) are known trade-offs, and using them as the motivation to combine is precisely the KSR analysis (identify the problem the inventor was trying to solve — expedite deployment of advanced services on a carrier-grade switch — and ask whether the prior art elements, combined, were predictably capable of solving it; they were).

7. Bottom line per claim (with confidence levels)

Recommended primary § 103 ground for a challenge: Bellcore AIN / ITU-T IN CS-1 (SSP-SCP, trigger table, BCSM event reports, INAP operations) in view of a stored-program-controlled programmable switch and a general-purpose data link (Ethernet/TCP-IP), further in view of US 5,168,515 and US 5,323,452, and taking official notice of the applicant's own admissions in the Background (MPEP § 2129). Secondary grounds: US 5,168,515 + US 5,157,390 + US 4,713,806 + US 5,551,035; and US 5,323,452 + US 5,903,637 for the protocol claims.

8. Caveats and verification to-dos (do not skip these)

1. Retrieval gap (material): I did not obtain the complete "(56) References Cited" list for US5991389 nor the WO 97/048240 / EP 0906705 search report, due to a tool-step limit. Any conclusion above resting on Category C art must be re-grounded by pulling those two documents (they will show exactly which references the examiner treated as closest, and any category-X art would strengthen, not weaken, the § 103 position).
2. Claim-numbering caveat (carried forward and still unresolved): the authoritative text supplied here contains a summary of the claims, not the numbered claim set. The independent-claim numbers for the SPI-protocol and method-of-communicating claims remain approximate; the number "3" for one independent claim in the user message, and my earlier note of "1, 6, 11, 14, 16," cannot be reconciled without the claim listing. Confirm against USPTO Patent Public Search before relying on numbering.
3. US 5,903,637 was retrieved only as a citation-table snippet; verify its priority date (stated 1994-06-08), assignee, and content before using it against the Bridge limitation.
4. Family art excluded by law. US 6,028,924, WO 97/048238 and Ser. No. 08/865,887 are commonly owned same-family documents; pre-AIA § 103(c)(1) disqualifies them for obviousness. Do not cite them in a § 103 ground.
5. Posture note. The patent expired 2017-05-30, so § 103 matters now only for a validity challenge (e.g., ex parte reexamination of the enforceability-period claims, or as a defense). Given the Litigation summary shows claim 11 was the claim Nortel asserted against Vonage in D. Del. 1:07-cv-00507-GMS, claim 11 is the claim whose § 103 status has the most practical significance — and it is also the claim the IN/INAP ground attacks most cleanly.

If you want, the two highest-value next retrievals are (a) the full "(56) References Cited" table for US5991389 and (b) the WO 97/048240 international search report — either of which could convert several of the "Med-High" rows above into "High" and would let me replace the Category C placeholders with specific, citable documents.