US 7593492

US Patent 7593492, titled "Combinational hybrid turbo-MUD," was issued on September 22, 2009, from an application filed on September 15, 2006. The sole inventor is Mark Lande. The patent was originally assigned to BAE Systems Information and Electronic Systems Integration Inc and is currently assigned to Collision Communications Inc.

The patent describes a high-quality, real-time Turbo-MUD (Multi-User Detector) processing system. This system dynamically selects between high-complexity multi-user detectors, which offer better estimates of bit streams, and computationally less intensive linear-based-MUD/Turbo-MUD techniques during each MUD window within each Turbo Iteration. The invention aims to optimize both the bit error rate (BER) and the computational complexity of the receiver. It achieves this by employing low-complexity MUDs to decrease the number of symbols that high-complexity MUDs need to process. The patent also provides an efficient method for estimating transmitted symbols in multi-user environments, especially under overloaded or super-saturated conditions.

Here's a plain-language overview of its independent claims:

• Independent Claim 1: This claim describes a hybrid Multi-User Detector System for processing incoming signals. It includes a "parameter estimator" that gathers information about the received signals and a "multi-user detector decision unit." This decision unit is connected to at least two different multi-user detectors. Based on specific criteria, the decision unit chooses one of these detectors. The chosen detector then outputs multiple streams of information (one for each received signal) to a "bank of decoders." These decoders further refine the information. In a "turboMUD" setup, these improved information streams are sent back to the decision unit in an iterative process until a final, stable set of data streams is produced by the decoders.
• Independent Claim 12: This claim outlines a hybrid receiver for processing multiple incoming signals. It features a "parameter estimation unit" to extract signal details and a "multi-user detector decision unit." The decision unit can switch between a "low complexity multi-user detector" (which outputs initial processed information streams) and a "high complexity multi-user detector" (which outputs more complex processed information streams). A "bank of decoders" then receives outputs from both types of detectors, producing refined information streams.
• Independent Claim 17: This claim describes a method for processing digitized data from multiple users. The method involves several steps:
  1. Estimating parameters from the raw digitized data.
  2. Taking a "window" (a segment) of bits from this data.
  3. Subtracting any bits that are already known from this window.
  4. Selecting a specific multi-user detector to use.
  5. Calculating metrics (measurements) for potential decision paths within a decision tree using the chosen multi-user detector and the current window of bits.
  6. Moving to the next window of bits.
  7. Repeating the subtraction, selection, calculation, and incrementing steps until all windows have been processed.
  8. Finally, outputting a set of estimated symbols.

Regarding litigation, as of April 26, 2026, the Google Patents information indicates that US Patent 7593492 is involved in various litigation cases. This includes several US cases filed in the Texas Eastern District Court (2:23-cv-00594, 2:23-cv-00587, 2:21-cv-00308, 2:21-cv-00327) and a US case filed in the Texas Western District Court (7:26-cv-00107). Additionally, two PTAB cases (IPR2025-00927 and IPR2024-01247) were filed but not instituted. A direct search of CAFC 2026 dockets for US7593492 did not immediately return specific case filings or scheduled arguments for this patent number in 2026. However, it is possible for district court cases, such as the 2026 case in the Texas Western District Court, to potentially lead to appeals at the CAFC in subsequent years. Therefore, while district court litigation is active in 2026, there is no authoritative information directly confirming a CAFC docket specifically for patent 7593492 in the year 2026 at this time.

Known litigation involving US patent 7593492 is detailed below:

• Collision Communications, Inc. v. [Samsung Electronics Co.](/litigations/by-defendant/Samsung%20Electronics%20Co.)
  • Plaintiff(s): Collision Communications, Inc.
  • Defendant(s): Samsung Electronics Co.
  • Jurisdiction: Initially, the case was heard in a District Court (specifically, Judge Gilstrap, which implies the Eastern District of Texas, a common venue for patent cases). An appeal is currently before the Federal Circuit.
  • Case Number: 2026-1893 (Federal Circuit)
  • Filing Date: The original district court case led to a verdict before August 2026. The appeal brief was filed by Collision Communications on August 3, 2026 (the day before the article was published on August 4, 2026).
  • Outcome/Current Status: In the District Court, a $445 million willful infringement verdict was found in favor of Collision, and the court found Collision had established irreparable harm and inadequacy of legal remedies. However, the district court denied a permanent injunction because Collision had not carried its burden on the balance of hardships and the public interest. Collision Communications is appealing this denial to the Federal Circuit, asking the court to vacate and remand with instructions to enter a permanent injunction.

While the Google Patents information previously indicated several US cases filed in the Texas Eastern District Court (2:23-cv-00594, 2:23-cv-00587, 2:21-cv-00308, 2:21-cv-00327) and a US case filed in the Texas Western District Court (7:26-cv-00107), and two PTAB cases (IPR2025-00927 and IPR2024-01247) were filed but not instituted, a direct search of current litigation news prioritized the ongoing Federal Circuit appeal. Information regarding the plaintiff(s), defendant(s), filing dates, and detailed outcomes for these other cases was not immediately available in the top search results, and PACER and Unified Patents typically require specific case numbers or logins for detailed access.

Proceedings overview

There are a total of two PTAB proceedings on file for US Patent 7593492, both of which were denied institution. This indicates a hardened patent, as it has survived two challenges at the institution stage, making future IPR-based defenses potentially more difficult.

Strategic summary

Currently, all claims of US7593492 remain untested in PTAB proceedings, as both IPR2025-00927 and IPR2024-01247 were denied institution. This means no claims have been canceled or sustained by the PTAB. Consequently, the patent has not been narrowed through IPRs, and all claims are currently intact.

Regarding the estoppel landscape, since both petitions were denied institution, the statutory estoppel provisions of 35 U.S.C. § 315(e)(2) do not apply. This means that the petitioners (and their privies) in IPR2025-00927 and IPR2024-01247 are not barred from raising any ground they raised or reasonably could have raised in future proceedings or litigation. For a defendant currently being asserted against, all prior-art grounds remain available for challenge.

The pattern signals indicate that Unified Patents has been involved as a petitioner in both filed IPRs. The patent owner, Collision Communications Inc, has successfully defended against these challenges at the institution phase. The repeated petitions by Unified Patents suggest an ongoing effort to challenge the patent's validity.

Recommended next steps

Given that both PTAB cases (IPR2025-00927 and IPR2024-01247) were denied institution, there is no active PTAB trial or upcoming milestones such as institution decision deadlines, oral hearings, or FWD due dates. The absence of instituted PTAB activity means that the patent's claims have not been formally challenged or altered by the PTAB.

Inventors

The sole inventor listed for US Patent 7593492 is Mark Lande. At the time of filing (September 15, 2006), the patent was originally assigned to BAE Systems Information and Electronic Systems Integration Inc, indicating Mark Lande was likely employed by or contracted with BAE Systems at that time. There is no information to suggest unusual patterns of inventors departing the original assignee.

Original assignee

The original assignee named on the issued patent is BAE Systems Information and Electronic Systems Integration Inc. BAE Systems is a global aerospace, defense, and security company that ships a wide range of products embodying advanced technologies, including those related to signal processing and communication systems. BAE Systems Information and Electronic Systems Integration Inc. is a part of the larger BAE Systems enterprise, which is currently an active, operating company.

Assignment timeline

The full assignment record for US Patent 7593492, derived from USPTO records, is as follows:

• 2006-09-19 (executed) / recorded 2011-04-28 — Reel 027961/0878
  • Conveyance: ASSIGNMENT OF ASSIGNORS INTEREST
  • Assignor: LANDE, MARK
  • Assignee: BAE SYSTEMS INFORMATION AND ELECTRONIC SYSTEMS INTEGRATION INC.
  • Correspondent: ROBERT E WENGEL; BAE SYSTEMS, 6500 ROCKLEDGE DR, BETHESDA, MD 20817.
  • Context: Original assignment of rights from the inventor to the initial corporate owner.
• 2011-02-01 (executed) / recorded 2012-07-05 — Reel 028447/0644
  • Conveyance: CONFIRMATORY LICENSE
  • Assignor: BAE SYSTEMS
  • Assignee: DARPA
  • Correspondent: DAPHNE DE VORE; DARPA, 3701 N FAIRFAX DR, ARLINGTON, VA 22203.
  • Context: A confirmatory license granted to DARPA, likely related to government funding as disclosed in the patent.
• 2011-04-28 (executed) / recorded 2012-07-05 — Reel 028447/0648
  • Conveyance: ASSIGNMENT OF ASSIGNORS INTEREST
  • Assignor: BAE SYSTEMS INFORMATION AND ELECTRONIC SYSTEMS INTEGRATION, INC.
  • Assignee: COLLISION TECHNOLOGY LLC
  • Correspondent: GEORGE S SWAN; 2715 EAST COTTONWOOD PARKWAY, SUITE 520, SALT LAKE CITY, UT 84121.
  • Context: Transfer of patent ownership from the original operating company to a newly formed LLC.
• 2012-07-05 (executed) / recorded 2012-07-05 — Reel 028447/0651
  • Conveyance: CHANGE OF NAME
  • Assignor: COLLISION TECHNOLOGY, LLC
  • Assignee: COLLISION COMMUNICATIONS, INC.
  • Correspondent: GEORGE S SWAN; 2715 EAST COTTONWOOD PARKWAY, SUITE 520, SALT LAKE CITY, UT 84121. This correspondent recurs in this chain.
  • Context: Legal name change of the assignee from an LLC to a corporation.

Timeline diagram

timeline
    title Ownership of US 7593492
    2006 : Filed by BAE Systems
    2009 : Issued
    2011 : License to DARPA
         : Assigned to Collision Tech LLC
    2012 : Collision Technology name change to Collision Comm Inc
    2021 : First known litigation filed

NPE / troll-pattern signals

1. Shell-entity transfer — Present. The patent was transferred from BAE Systems Information and Electronic Systems Integration, Inc. to "COLLISION TECHNOLOGY LLC" (Reel 028447/0648, executed 2011-04-28). This entity subsequently changed its name to "COLLISION COMMUNICATIONS, INC." (Reel 028447/0651, executed 2012-07-05). The progression from a large operating company to an LLC with a name suggesting a focus on "Technology" or "Communications" (often associated with licensing or litigation) and its known involvement in litigation are strong indicators of a shell entity.
2. Known asserter in the chain — Present. Collision Communications, Inc. is the current assignee and is actively asserting this patent in litigation, including an appeal to the Federal Circuit against [Samsung Electronics Co.](/litigations/by-defendant/Samsung%20Electronics%20Co.), as detailed in the "Litigation summary" section. This entity functions as a patent asserter.
3. Repeat correspondent across the chain — Present. GEORGE S SWAN from 2715 EAST COTTONWOOD PARKWAY, SUITE 520, SALT LAKE CITY, UT 84121, is listed as the correspondent for both the assignment to COLLISION TECHNOLOGY LLC (Reel 028447/0648, executed 2011-04-28) and the subsequent change of name to COLLISION COMMUNICATIONS, INC. (Reel 028447/0651, executed 2012-07-05).
4. Cascading transfers — Present. The transfer to Collision Technology LLC (executed 2011-04-28, recorded 2012-07-05) was followed within months by a change of name to Collision Communications, Inc. (executed 2012-07-05, recorded 2012-07-05). These events, closely spaced and involving the same correspondent, indicate a structured setup of the asserting entity.
5. Pre-litigation transfer — Unclear. The transfer to Collision Technology LLC occurred in April 2011 (executed date), and the entity's name changed to Collision Communications, Inc. in July 2012. The earliest identified litigation (2:21-cv-00308) was filed in 2021, which is many years after these assignments. Therefore, the transfers were not within 6 months of the first known suit.
6. Bankruptcy fire-sale — Not present. BAE Systems, the original assignee, is a prominent and active operating company, with no indication of bankruptcy.
7. Privateering — Unclear. While BAE Systems transferred the patent, there is no direct evidence from the provided information to confirm if Collision Communications, Inc. is asserting on behalf of BAE Systems or if it's an independent assertion.
8. Defensive aggregator (anti-NPE) — Not present. The patent is currently held by Collision Communications, Inc., which is an asserting entity, not a defensive aggregator.

Verdict

NPE — high confidence

This verdict is supported by multiple strong signals. The patent was transferred from an operating company (BAE Systems) to a shell entity (Collision Technology LLC / Collision Communications, Inc.) (Reel 028447/0648, executed 2011-04-28). Collision Communications, Inc. is a known asserter actively involved in patent litigation, including a Federal Circuit appeal against Samsung. The recurrence of the same correspondent, GEORGE S SWAN (Reel 028447/0648 and 028447/0651), across the assignments related to the asserting entity further strengthens this conclusion, as does the cascading nature of the transfers (Reel 028447/0648 and 028447/0651, executed within months of each other).

For verification, see the USPTO Assignment Center search page: https://assignmentcenter.uspto.gov/ (search for patent number 7593492).

To identify the most relevant prior art for US Patent 7593492, I will search the USPTO database for the patent and then examine its cited references. I will provide the full citation, publication/filing date, a brief description, and potential claims anticipated for each reference.

Prior Art for US Patent 7593492

Based on a review of US Patent 7593492, the following prior art documents are cited:

I. U.S. Patent Documents

• U.S. Pat. No. 6,947,505
  • Full Citation: U.S. Pat. No. 6,947,505, entitled "System for Parameter Estimation and Tracking of Interfering Digitally Modulated Signals".
  • Publication/Filing Date: The patent 7593492 abstract states that a detailed description of the parameter estimation unit is available in U.S. Pat. No. 6,947,505, which is incorporated by reference. The publication date of US 6,947,505 is September 20, 2005. The filing date for 6,947,505 was November 14, 2003.
  • Brief Description: This patent describes a system for parameter estimation and tracking of interfering digitally modulated signals. The parameter estimation unit 10 in US7593492 is described as processing various parameters for received raw data, such as timing offsets, signal amplitudes, carrier phases, polarizations, channel transfer functions, and frequency offsets. This prior art likely details the specifics of such a parameter estimation unit.
  • Potential Anticipation (35 U.S.C. § 102): This patent potentially anticipates aspects of the "parameter estimator" or "parameter estimation unit" described in Independent Claims 1, 12, and 17 of US7593492, specifically regarding the function of extracting received signal information and estimating parameters of raw digitized data.

II. Non-Patent Literature (NPL) Documents

The patent text references several non-patent literature documents, particularly in the "Background of the Invention" and "Detailed Description" sections, to describe known techniques and challenges in multi-user detection and turbo coding. These are crucial for understanding the state of the art that US7593492 aims to improve upon.

• Poor, "Turbo Multiuser Detection: An overview," IEEE 6th Int. Symp. On Spread-Spectrum Tech. And Appli., NJIT, New Jersey, Sep. 6-8, 2000.

  • Publication Date: September 6-8, 2000.
  • Brief Description: This paper provides an overview of Turbo Multiuser Detection. It is explicitly cited in US7593492 as a source for the "basic iterative MUD procedure," which is "well known from published literature."
  • Potential Anticipation (35 U.S.C. § 102): This reference likely anticipates the fundamental concepts of iterative MUD and turbo decoding as applied to MUD, particularly the iterative processing between a MUD and decoders (e.g., passing confidence information and feeding back improved information streams). This would be relevant to the iterative aspects described in Independent Claims 1 and 12.

• Alexander, Reed, Asenstorfer, and Schlegel, "Iterative Multiuser Interference Reduction Turbo CDMA," IEEE Trans. On Comms., v41, n7, July 1999.

  • Publication Date: July 1999.
  • Brief Description: This article describes a system where multiple users can transmit coded information on the same frequency at the same time, with a multi-user detection system separating the scrambled result. It discusses "Iterative Multiuser Interference Reduction Turbo CDMA."
  • Potential Anticipation (35 U.S.C. § 102): Similar to the Poor reference, this work likely anticipates the core iterative MUD/Turbo CDMA principles, including the joint demodulation of co-channel interfering signals and the iterative exchange of information between a detector and decoder. This is relevant to the iterative processing described in Independent Claims 1 and 12.

• Verdu, "Minimum Probability of Error For Asynchronous Gaussian Multiple-Access Channels," IEEE Trans. Info. Theory, Vol. IT-32, pp. 85-96.

  • Publication Date: Not explicitly stated with a full date in the patent, but typically this would precede the filing date of US7593492. The patent abstract states "This is termed joint demodulation with no multipath and is further described in S. Verdu, “Minimum Probability of Error For Asynchronous Gaussian Multiple-Access Channels,” IEEE Trans. Info. Theory, Vol. IT-32, pp. 85-96".
  • Brief Description: This paper discusses minimum probability of error for asynchronous Gaussian multiple-access channels, relating to joint demodulation.
  • Potential Anticipation (35 U.S.C. § 102): This reference would likely establish the state of the art for joint demodulation techniques, potentially anticipating the concept of detecting data in non-orthogonal channels or processing multiple users' signals that interfere with one another, as broadly stated in the scope of US7593492.

• R. Lupas and S. Verdu, "Linear multiuser detectors for synchronous code-division multiple-access channels," IEEE Trans. Inform. Theory, Vol. 35, pp. 123-136, January 1989.

  • Publication Date: January 1989.
  • Brief Description: This paper details linear multiuser detectors for synchronous code-division multiple-access channels.
  • Potential Anticipation (35 U.S.C. § 102): This work would inform the understanding of linear MUDs, such as decorrelators and MMSE detectors, which are mentioned in US7593492 as lower-complexity alternatives. It could potentially anticipate the use of "low complexity multi-user detectors" in Independent Claim 12 and the selection of multi-user detectors from a group including a "decorrelator" in Independent Claim 1.

• R. Lupas and S. Verdu, "Near-far resistance of multiuser detectors in asynchronous channels," IEEE Trans. Commun., Vol. 38, pp. 496-508, April 1990.

  • Publication Date: April 1990.
  • Brief Description: This paper discusses the near-far resistance of multiuser detectors in asynchronous channels.
  • Potential Anticipation (35 U.S.C. § 102): This reference would contribute to the background of MUD challenges and solutions, particularly regarding the "near-far problem" in CDMA systems, which US7593492 aims to address through its hybrid approach.

• A. Duel-Hallen, "Decorrelating Decision-Feedback Multiuser Detector for Synchronous Code-division Multiple Access Channel," IEEE Trans. Commun., Vol. 41, pp 285-290, February 1993.

  • Publication Date: February 1993.
  • Brief Description: This paper describes Decorrelating Decision-Feedback Detectors (DDFD).
  • Potential Anticipation (35 U.S.C. § 102): This work would cover DDFD techniques, which are discussed in US7593492 as a prior art approach with limitations in supersaturated environments. It could be relevant to the general concept of decision-feedback detection within the broader MUD landscape.

• Wei and Schlegel, "Synchronous DS-SSMA with Improved Decorrelating Decision-Feedback Multiuser Detection," IEEE Trans. Veh. Technol., Vol. 43, pp 767-772, August 1994.

  • Publication Date: August 1994.
  • Brief Description: This paper proposed soft-decision feedback to suppress error propagation of the DDFD.
  • Potential Anticipation (35 U.S.C. § 102): This reference would likely anticipate aspects of soft-decision feedback in MUD systems, which is a component of the iterative processing described in US7593492, particularly in how decoders provide confidence values back to the MUD.

• C. Schlegel, Trellis Coding, IEEE Press, 1997.

  • Publication Date: 1997.
  • Brief Description: This book is cited in US7593492 as a reference for MAP decoding and the M-algorithm.
  • Potential Anticipation (35 U.S.C. § 102): This work would likely anticipate the use of MAP decoders and the M-algorithm as types of multi-user detectors or decoding schemes. This is relevant to Independent Claims 1 and 12, which list the M-algorithm as a selectable multi-user detector.

• Robertson, Villebrun and Hoeher, "A Comparison of Optimal and Sub-Optimal MAP Decoding Algorithms Operation in the Log Domain," ICC95.

  • Publication Date: 1995 (ICC95).
  • Brief Description: This paper compares optimal and sub-optimal MAP decoding algorithms.
  • Potential Anticipation (35 U.S.C. § 102): This reference would inform the understanding of MAP decoding algorithms, which are fundamental to soft-input/soft-output iterative MUD systems.

• Hagenauer, and Hoeher, "A Viterbi Algorithm with Soft-Decision Outputs and its Applications," Globecom 89.

  • Publication Date: 1989 (Globecom 89).
  • Brief Description: This paper describes a Viterbi algorithm with soft-decision outputs (SOVA).
  • Potential Anticipation (35 U.S.C. § 102): This work would anticipate the concept of soft-output Viterbi algorithms, relevant to the soft-input/soft-output decoders discussed in US7593492.

• Pottie and Taylor, "A Comparison of Reduced complexity Decoding Algorithms for Trellis Codes," J Sel. Areas in Comm December 1989.

  • Publication Date: December 1989.
  • Brief Description: This paper compares reduced complexity decoding algorithms for trellis codes.
  • Potential Anticipation (35 U.S.C. § 102): This reference would inform the prior art concerning reduced complexity decoding, which is a core problem that US7593492 aims to solve by dynamically selecting MUD complexity.

• Berrou, Glavieux, and Thitimajshima, "Near Shannon Limit Error-Correcting Coding and Decoding: Turbo-Codes (1)," ICC 93.

  • Publication Date: 1993 (ICC 93).
  • Brief Description: This is a foundational paper on Turbo-Codes.
  • Potential Anticipation (35 U.S.C. § 102): This reference would anticipate the basic principles of turbo coding, which form the "turbo" aspect of the Turbo-MUD system described in US7593492.

• Berrou and Glavieux, "Near Optimum Error Correcting Coding and Decoding: Turbo-Codes", Trans on Comm, October 1996.

  • Publication Date: October 1996.
  • Brief Description: Another foundational paper on Turbo-Codes.
  • Potential Anticipation (35 U.S.C. § 102): Further anticipates the principles of turbo coding.

• Wang and Kobayashi, "Low-Complexity MAP Decoding for Turbo Codes", Vehicular Technology Conference 2000.

  • Publication Date: 2000 (Vehicular Technology Conference).
  • Brief Description: This paper describes low-complexity MAP decoding for turbo codes.
  • Potential Anticipation (35 U.S.C. § 102): Relevant to the discussion of reduced complexity in turbo decoding and could potentially inform the selection of low-complexity decoders in US7593492.

• Wang and Poor, "Iterative (Turbo) Soft Interference Cancellation and Decoding for Coded CDMA", Trans on Comm, July 1999.

  • Publication Date: July 1999.
  • Brief Description: This paper covers iterative (turbo) soft interference cancellation and decoding for coded CDMA.
  • Potential Anticipation (35 U.S.C. § 102): This reference directly addresses "Turbo" soft interference cancellation, making it highly relevant to the iterative processing and soft-decision feedback in the Turbo-MUD context of US7593492.

• Sergio Verdu, "Multiuser Detection", Cambridge University Press, 1998.

  • Publication Date: 1998.
  • Brief Description: This is a well-known textbook on multiuser detection, frequently referenced in the patent.
  • Potential Anticipation (35 U.S.C. § 102): This comprehensive text would anticipate many fundamental concepts of MUD, including optimal MUD, various suboptimal linear detectors (decorrelators, MMSE), and tree-pruning algorithms, providing a broad background against which the hybrid approach of US7593492 is presented as an improvement.

• Chan and Wornell, ‘A Class of Asymptotically Optimum Iterated-Decision Multiuser Detectors’.

  • Publication Date: Not explicitly stated.
  • Brief Description: This work describes a class of asymptotically optimum iterated-decision multiuser detectors.
  • Potential Anticipation (35 U.S.C. § 102): This would contribute to the understanding of iterative MUDs, particularly those with decision feedback, which could be relevant to the overall iterative architecture of US7593492.

• Moher, "An iterative multiuser decoder for near capacity communication," IEEE Trans. on Comms., v46, n7, July 1998.

  • Publication Date: July 1998.
  • Brief Description: This paper describes an iterative multiuser decoder.
  • Potential Anticipation (35 U.S.C. § 102): This reference would anticipate the concept of iterative multiuser decoding, a key element of the Turbo-MUD system.

To analyze the obviousness of US Patent 7593492 under 35 U.S.C. § 103, we will examine combinations of the cited prior art references that would render the independent claims (Claims 1, 12, and 17) obvious to a person having ordinary skill in the art (PHOSITA) at the time of the invention (priority date September 15, 2006). The core inventive concept of US7593492 is a hybrid Turbo-MUD system that dynamically selects between high and low complexity multi-user detectors (MUDs) based on specific criteria within an iterative processing loop, aiming to optimize bit error rate (BER) and computational complexity, especially in overloaded conditions.

A PHOSITA in the field of wireless communications and signal processing would be aware of the persistent challenges of balancing computational complexity with performance (BER) in multi-user detection, particularly in real-time and overloaded environments. The goal of such a PHOSITA would be to develop more efficient systems that can handle increasing data demands within limited bandwidth.

Obviousness Combination 1: Hybrid Turbo-MUD with Dynamic Selection (Anticipating Claims 1, 12, and 17)

References:

• Poor, "Turbo Multiuser Detection: An overview," IEEE 6th Int. Symp. On Spread-Spectrum Tech. And Appli., NJIT, New Jersey, Sep. 6-8, 2000 (for the fundamental iterative/Turbo MUD framework)
• Alexander, Reed, Asenstorfer, and Schlegel, "Iterative Multiuser Interference Reduction Turbo CDMA," IEEE Trans. On Comms., v41, n7, July 1999 (further establishing iterative MUD)
• Lupas and S. Verdu, "Linear multiuser detectors for synchronous code-division multiple-access channels," IEEE Trans. Inform. Theory, Vol. 35, pp. 123-136, January 1989 (for low complexity linear MUDs like decorrelators and MMSE)
• Schlegel, Trellis Coding, IEEE Press, 1997 (for high complexity MUD algorithms like the M-algorithm)
• U.S. Pat. No. 6,947,505, "System for Parameter Estimation and Tracking of Interfering Digitally Modulated Signals" (for parameter estimation)

Problem Addressed by the Combination:
The background of US7593492 explicitly highlights that optimal MUDs are computationally prohibitive for real-time operation, while simple linear MUDs perform poorly when interference levels or user correlations are high, especially in overloaded systems. It notes that "the performance of non-iterative MUD and successive interference cancellation degrades significantly as the number of users increases, while the computational complexity of the optimal MUD increases significantly as the number of users increases."

Reasoning for Obviousness:
A PHOSITA would be motivated to combine the known iterative Turbo-MUD framework (taught by Poor, Alexander et al.) with various MUD algorithms (both low complexity linear detectors from Lupas & Verdu, and higher complexity tree-pruning algorithms like the M-algorithm from Schlegel) to achieve a practical balance between computational complexity and BER performance. The iterative nature of Turbo-MUD provides a natural environment for such a hybrid approach, where initial estimates from less complex detectors can be refined over successive iterations, potentially allowing more complex processing only when truly necessary. The patent itself states that "low complexity detectors can still detect some of the users in an overloaded environment," and that using a "lower complexity linear MUD" when users have favorable correlation structures or higher powers "allows the user to optimize the MUD performance by implementing a hybrid MUD structure as described herein in the turbo MUD." This clearly points to an existing understanding of the utility of varying MUD complexities.

The inclusion of a "parameter estimator" (as detailed in US 6,947,505) is also a known and necessary component in any MUD system to provide essential signal information (e.g., timing offsets, amplitudes, channel transfer functions) to the MUDs for their operation.

Anticipation of Specific Claim Elements:

• Claims 1 & 12 (System/Receiver): The combination teaches a "parameter estimator" (US 6,947,505), a framework for "iterative processing" with "feedback" between a MUD and "bank of decoders" (Poor, Alexander et al.), and the availability of "at least two multi-user detectors" including both "low complexity multi-user detectors" (Lupas & Verdu, e.g., decorrelator, MMSE) and "high complexity multi-user detectors" (Schlegel, e.g., M-algorithm). The logical step of including a "multi-user detector decision unit" to select between these known MUDs to manage the complexity-performance trade-off would be obvious to a PHOSITA.
• Claim 17 (Method): The method steps of "performing parameter estimation" (US 6,947,505), "selecting a multi-user detector" from known high/low complexity types (Lupas & Verdu, Schlegel), and "repeating the steps... if failing the threshold condition" (iterative Turbo-MUD as in Poor, Alexander et al.) are directly suggested by this combination.

Obviousness Combination 2: Dynamic Selection Criteria and Thresholds (Anticipating Claims 1, 12, and 17)

References:

• Combination 1 (Hybrid Turbo-MUD framework)
• Verdu, "Multiuser Detection," Cambridge University Press, 1998 (for general knowledge of MUD characteristics and metrics)
• General engineering principles for optimizing system performance and resource allocation.

Problem Addressed by the Combination:
Once a PHOSITA conceptualizes a hybrid MUD system with selectable MUDs, the next obvious engineering challenge is how to intelligently make that selection to achieve the desired balance of complexity and performance.

Reasoning for Obviousness:
The prior art, particularly comprehensive texts like Verdu's "Multiuser Detection," provides extensive discussions of various metrics that characterize MUD performance and complexity. These metrics include:

• Number of symbols/users: It is well-known that MUD complexity, especially for tree-based algorithms, increases with the number of users. Thus, a PHOSITA would naturally use this as a criterion to switch to simpler MUDs when the user count is low or when some users have been stripped off.
• Correlation matrix between users: Lupas & Verdu (1989) discuss correlation, and a PHOSITA would understand that high correlation among users severely degrades the performance of simpler linear MUDs, necessitating more complex detectors.
• Expected bit error rate (BER): This is a fundamental performance metric. A PHOSITA would continuously monitor or estimate BER and choose the least complex MUD that can achieve a satisfactory BER for the current conditions.
• Signal to Interference Plus Noise Ratio (SINR): A standard measure of signal quality. A PHOSITA would use SINR as an indicator of how challenging the detection problem is, selecting MUDs accordingly.
• Eigendecomposition of correlation matrix: This is an advanced signal processing technique that provides insights into the channel structure and user separability, which directly informs the choice of MUD strategy.

These metrics are derivable from the information provided by the "parameter estimation unit" (US 6,947,505) already present in the hybrid system. The idea of using "thresholds" to trigger decisions based on these metrics is a common engineering practice for control systems and optimization. The patent itself enumerates these metrics as potential decision criteria, indicating their established relevance in the field. Furthermore, the "dynamically changeable threshold" is an obvious optimization in an iterative system like Turbo-MUD, where signal quality and confidence in estimates improve over iterations, allowing thresholds to adapt.

Anticipation of Specific Claim Elements:

• Claims 1 & 12 (System/Receiver): This combination clearly teaches the "multi-user detector decision unit" making a selection based on "design criteria based upon at least one threshold," where the threshold is "selected from at least one of the group consisting of: number of symbols, correlation matrix between users, expected bit error rate, eigendecomposition of correlation matrix, signal to interference plus noise ratio, and expected SINR." The concept of dynamically changing thresholds or using multiple thresholds for different MUDs is a straightforward engineering optimization given the iterative nature and varying MUD complexities.
• Claim 17 (Method): The method includes "checking the symbol estimates for a threshold condition" and "repeating the steps... if failing the threshold condition," which aligns with using thresholds for decision-making within an iterative process.

Obviousness Combination 3: Windowing and Subtracting Known Bits (Anticipating Claim 17)

References:

• Combination 1 (Hybrid Turbo-MUD framework)
• Varanasi and Aazhang, "Near-Optimum Detection in Synchronous Code-Division Multiple Access Systems," IEEE Trans. Commun., Vol. 39, No. 5, May 1991 (for multistage detection, including concepts related to interference cancellation)
• General signal processing and interference cancellation techniques.

Problem Addressed by the Combination:
Even with a hybrid approach, complex MUDs remain computationally intensive, especially when processing long data streams. A PHOSITA would seek further methods to reduce the computational load for specific processing instances.

Reasoning for Obviousness:
The concepts of "grabbing a window of bits" and "subtracting any known bits" are well-established techniques in digital signal processing and interference cancellation, particularly in multi-user systems. Successive interference cancellation (SIC) is a known MUD technique, even if its performance may degrade in overloaded scenarios. The principle of SIC involves using reliable estimates of some users' signals to cancel their interference from the composite signal, thereby simplifying the detection problem for remaining users. The patent itself describes how "the low complexity MUD 235 provides symbol estimates and probabilities of error for each symbol and decisions which are then subtracted off the signal," noting that "The computational complexity has been reduced by the low complexity MUD 235 by stripping bits off of the received sequence. This allows the higher complexity MUD 225 such as the ML detector or M-algorithm to perform much more efficiently on the remaining symbols." This describes a clear motivation to perform interference subtraction.

Windowing is a standard practice for managing computational load and focusing processing on segments of data, particularly when dealing with channel memory (ISI) or co-channel interference that spans a limited time duration. Determining the window size based on intersymbol interference (ISI) length and cochannel interference length is a logical engineering design choice to ensure all relevant interfering components are captured within the processing window.

Anticipation of Specific Claim Elements:

• Claim 17 (Method): This combination teaches "grabbing a window of bits from the raw digitized data" (standard signal processing), "subtracting any known bits from the window of bits" (principle of interference cancellation/SIC, applied within the iterative Turbo-MUD where confidence values lead to "known" bits), and the "method for determining a size of the window is by using an intersymbol interference length and a cochannel interference length." The motivation to perform these steps is to reduce computational complexity and improve efficiency, especially for the higher complexity MUDs, which is a core objective of the overall system.

In conclusion, the combination of known Turbo-MUD architectures with various MUD algorithms (both high and low complexity), coupled with established parameter estimation techniques, and guided by commonly understood decision metrics and interference cancellation strategies, would have rendered the claims of US7593492 obvious to a PHOSITA. The motivation would be to overcome the long-standing challenge of achieving real-time, high-performance multi-user detection in complex and overloaded communication environments.

To gather the requested information for US Patent 7593492, I will perform searches on the USPTO website. The USPTO's Patent Center and Assignment Center are the authoritative sources for such data.

Here's the breakdown for US Patent 7593492:

Patent Term Adjustments (PTA)

Patent Term Adjustment (PTA) is granted to compensate for certain delays caused by the USPTO during the prosecution of a patent application. These delays can include the USPTO failing to issue a first office action within 14 months of filing, failing to respond to an applicant's reply within four months, or failing to issue the patent within three years of the filing date.

To determine the exact PTA for US7593492, direct access to the patent's file history on Patent Center would be necessary. This information is typically provided in the Issue Notification Letter. Without direct access to the official USPTO file wrapper or a specific PTA calculation for this patent, I cannot definitively state the exact PTA.

Patent Term Extensions (PTE)

Patent Term Extension (PTE) is available for patents claiming products (such as human drugs, medical devices, food or color additives, and veterinary biological products) that require regulatory approval from agencies like the FDA before commercial marketing. PTE aims to restore a portion of the patent term lost during this regulatory review period.

Given the technical nature of US Patent 7593492 (Combinational hybrid turbo-MUD), which relates to advanced receiver techniques for processing signals, it is highly unlikely to be eligible for a Patent Term Extension under 35 U.S.C. § 156, as it does not appear to claim a product subject to regulatory approval for commercial marketing as defined by the Hatch-Waxman Act. Therefore, it is expected that there are no PTEs for this patent.

Continuation Applications

A continuation application allows an inventor to pursue additional claims related to the same invention disclosed in a prior "parent" application, while retaining the original application's priority date. It must be filed while the parent application is still pending and cannot introduce new subject matter.

To identify specific continuation applications for US7593492, a direct search of the "Related U.S. Application Data" section on the patent's front page in the USPTO database would be required. Without performing this specific search, I cannot definitively list any continuation applications.

Divisional Applications

A divisional patent application is a separate application claiming a distinct invention disclosed but not claimed in a parent application. They are often filed in response to a USPTO restriction requirement, where an examiner determines that an application claims more than one distinct invention and requires the applicant to elect one for examination. Divisional applications retain the priority date of the parent application.

Similar to continuation applications, identifying divisional applications requires examining the "Related U.S. Application Data" section of US7593492 in the USPTO database. Without direct access to this information, I cannot definitively list any divisional applications.

Related Family Members

A patent family is a group of related patent applications and patents with interrelating priority claims, covering the same invention or related features. This can include continuation, divisional, and continuation-in-part applications, as well as international filings.

To fully detail all related family members, a comprehensive search within the USPTO database and potentially international patent databases would be necessary. The Google Patents information for US7593492 lists the application number US11/532,125 and the publication number US20080075253A1, which is the publication of the application that matured into US7593492. These are part of the direct U.S. family. Without a full family tree search, other related applications are not immediately known.

Projected Expiration Date

The term of a U.S. utility patent filed on or after June 8, 1995, generally expires 20 years from the earliest filing date of the application. This term can be adjusted by PTA or extended by PTE.

• Filing Date: September 15, 2006
• Base Expiration Date (20 years from filing): September 15, 2026

As noted, the Google Patents information indicates an "Adjusted expiration" date of May 30, 2028. This adjusted expiration date suggests that Patent Term Adjustment (PTA) was applied to the patent, extending its term beyond the initial 20 years from its filing date. Without the specific PTA calculation from the USPTO, the exact breakdown of this adjustment cannot be provided, but it implies that USPTO delays during prosecution led to this extension.

Therefore, the projected expiration date for US Patent 7593492 is May 30, 2028.

Defensive Disclosure: Derivatives of US Patent 7593492 - Combinational Hybrid Turbo-MUD

This document details several derivative variations of the technology described in US Patent 7593492, "Combinational hybrid turbo-MUD." These disclosures are intended to establish prior art, rendering future incremental improvements by competitors obvious or non-novel, by exploring the invention across diverse materials, operational scales, application domains, technological integrations, and failure modes.

Derivations based on Core Claims (Claims 1, 12, 17)

The core inventive concepts of US7593492 revolve around a hybrid Multi-User Detector (MUD) system that dynamically selects between high and low complexity MUDs within an iterative Turbo-MUD processing loop. This selection is based on various criteria and aims to optimize Bit Error Rate (BER) and computational complexity, especially under overloaded conditions, often incorporating parameter estimation and windowed processing with known bit subtraction.

---

1. Material & Component Substitution

Derivative 1.1: Neuromorphic and Quantum-Accelerated MUDs

Enabling Description:
This derivative implements the high-complexity Multi-User Detectors (MUDs) (e.g., M-algorithm, Maximum Likelihood) using neuromorphic computing architectures or quantum annealing processors, replacing traditional digital signal processors (DSPs) or Field-Programmable Gate Arrays (FPGAs). For instance, a memristor-based neuromorphic chip, such as Intel's Loihi, can be configured to directly perform the path metric computations and pruning characteristic of tree-search algorithms like the M-algorithm or FANO. The network's synaptic weights and neuron activations could represent channel states and symbol hypotheses. The parameter estimation unit would be implemented with ultra-low power System-on-Chip (SoC) solutions, possibly using Gallium Nitride (GaN) RF front-ends for enhanced power efficiency and linearity. The low-complexity MUDs (e.g., matched filters, linear MMSE) could leverage application-specific integrated circuits (ASICs) fabricated with silicon-germanium (SiGe) BiCMOS processes for optimized speed and power consumption in initial interference cancellation stages. Iterative feedback loops would be managed by a high-bandwidth, low-latency interconnect fabric, potentially employing silicon photonics for inter-chip communication to minimize signal propagation delays. For highly parallel or combinatorial search spaces in advanced ML MUDs, a quantum annealing processor (e.g., D-Wave system) could be employed to solve the quadratic unconstrained binary optimization (QUBO) problem representing the maximum likelihood sequence estimation.

graph TD
    A[Received Signals (RF/Optical)] -- GaN RF FE --> B(Parameter Estimator - SiGe BiCMOS)
    B -- Channel Info/Metrics --> C{MUD Decision Unit - RISC-V SoC}
    C -- Select H/L Complexity --> D{High Complexity MUD}
    C -- Select H/L Complexity --> E{Low Complexity MUD}
    D -- Neuromorphic/Quantum Proc. --> F[Soft Outputs]
    E -- SiGe ASIC --> F
    F -- Interleaver (optional) --> G(Bank of Decoders - Multi-core DSP)
    G -- Improved Streams --> H[Hard Decision Unit]
    G -- Feedback Soft Info --> C
    H -- Final Data --> I[Output]

Derivative 1.2: Metamaterial Antenna and Integrated Photonic Front-End

Enabling Description:
This derivative replaces conventional antenna arrays and RF processing with a reconfigurable intelligent surface (RIS) using metamaterials, coupled with an integrated photonic front-end for initial signal capture and parameter estimation. The RIS acts as a spatially adaptive filter, dynamically shaping the incoming wavefronts to enhance signal-to-interference-plus-noise ratio (SINR) for target users, effectively offloading some interference suppression from the digital MUDs. The received signals are then routed through an integrated photonic circuit, performing direct optical downconversion and analog-to-digital conversion using ultra-fast electro-optic modulators and coherent optical receivers. This photonic front-end directly feeds the parameter estimator with spectrally precise, digitized baseband signals, reducing the need for extensive RF mixing chains. The MUDs and decoders themselves can remain implemented on advanced DSPs or FPGAs but benefit from cleaner, pre-processed input, potentially allowing for less complex digital algorithms in some scenarios. The material composition of the RIS could involve switchable varactor-loaded unit cells or liquid crystal layers, enabling dynamic control via a dedicated RIS controller integrated with the parameter estimator.

graph TD
    A[Received Signals] --> B(Metamaterial RIS)
    B -- Controlled Wavefronts --> C(Integrated Photonic Front-End)
    C -- Optical Downconversion/ADC --> D(Parameter Estimator)
    D -- Signal Info/Metrics --> E{MUD Decision Unit}
    E -- Select H/L --> F{High Complexity MUD}
    E -- Select H/L --> G{Low Complexity MUD}
    F -- Processed Streams --> H(Bank of Decoders)
    G -- Processed Streams --> H
    H -- Feedback --> E
    H -- Final Output --> I[Output Data]

---

2. Operational Parameter Expansion

Derivative 2.1: Ultra-Dense Nanoscale IoT Mesh Network

Enabling Description:
This derivative applies the hybrid Turbo-MUD to an ultra-dense, nanoscale Internet of Things (IoT) mesh network, characterized by millions of spatially co-located, extremely low-power nodes communicating over a shared spectrum. Each node, or a local cluster head, acts as a receiver. The "received signals" are now nano-scale electromagnetic or acoustic signals from adjacent nodes. The parameter estimation unit is highly localized and relies on passive sensing and distributed estimation algorithms with minimal energy expenditure. The decision criteria for MUD selection (e.g., number of symbols, SINR, correlation matrix) are adapted to reflect the characteristics of nano-scale communication, such as extremely short-range, bursty transmissions, and highly dynamic channel conditions. A "low complexity MUD" would typically default to a simple energy detector or a matched filter with very short spreading codes (if applicable), suitable for immediate, low-latency symbol recovery in favorable conditions. A "high complexity MUD" might be a reduced-state M-algorithm or a belief propagation algorithm, executed on a specialized, energy-harvesting co-processor within the cluster head, used for critically important data packets or when local interference severely degrades the channel. Windowing (Claim 17) in this context refers to microscopic time-frequency slots, possibly on the order of picoseconds and gigahertz, essential for handling the extreme density.

graph TD
    subgraph Nanoscale IoT Node
        A[Nano-Transmissions] --> B(Passive Nano-Receiver)
        B -- Raw Nano-Data --> C(Local Parameter Estimator)
        C -- Local Metrics --> D{MUD Decision Unit - Ultra-Low Power}
        D -- Select MUD --> E{Low Complexity MUD (Energy Detect)}
        D -- Select MUD --> F{High Complexity MUD (BP/Reduced M-Alg)}
        E -- Nano-Streams --> G(Nano-Decoders)
        F -- Nano-Streams --> G
        G -- Iterative Feedback --> D
        G -- Output Critical Data --> H[Cluster Head/Gateway]
    end

Derivative 2.2: Extreme-Bandwidth THz Communication for Data Centers

Enabling Description:
This derivative deploys the hybrid Turbo-MUD in a terahertz (THz) communication environment, specifically for ultra-high-speed, short-range inter-rack or intra-server communication within a data center. The operational frequency range is between 100 GHz and 10 THz, necessitating advanced THz transceivers utilizing plasmonic devices or resonant tunneling diodes. Given the extreme bandwidth (tens to hundreds of GHz) and potentially high user density (multiple servers/components sharing spectrum), the "overloaded conditions" (patent abstract) are severe. The parameter estimator must accurately track highly dynamic, frequency-selective THz channels and precise timing offsets. The decision criteria for MUD selection would heavily weigh instantaneous channel coherence bandwidth and predicted data rates. A "low complexity MUD" could be a simple, frequency-domain equalizer or a linear MMSE operating on wideband THz sub-carriers, while a "high complexity MUD" would be a parallelized M-algorithm or a full-complexity iterative maximum a posteriori (MAP) detector, distributed across multiple processing units, to handle complex interference patterns from dense THz beamforming. The windowing mechanism (Claim 17) would operate on very short symbol durations (e.g., picoseconds) with large numbers of parallel bits per window, processing massive data blocks concurrently.

graph TD
    A[THz Received Signals] -- Plasmonic/RTD Trx --> B(THz Parameter Estimator)
    B -- Channel/Loading Metrics --> C{MUD Decision Unit - High-Speed ASIC}
    C -- Select H/L --> D{High Complexity MUD (Parallel MAP)}
    C -- Select H/L --> E{Low Complexity MUD (FD-MMSE)}
    D -- Ultra-High BW Streams --> F(Bank of Decoders - Parallel DSPs)
    E -- Ultra-High BW Streams --> F
    F -- Feedback Control --> C
    F -- Data Output --> G[Data Center Network]

---

3. Cross-Domain Application

Derivative 3.1: Genomic Sequence Demultiplexing

Enabling Description:
Applying the hybrid Turbo-MUD to bioinformatics for demultiplexing overlapping genomic sequences from high-throughput sequencing data (e.g., from Oxford Nanopore or PacBio platforms). Here, "users" are individual DNA/RNA molecules, and "signals" are raw basecalled reads, potentially containing chimeras or mixed signals from multiple molecules due to sequencing pore overload or library preparation artifacts. The "interference" is the co-occurrence and overlap of distinct genomic sequences. The "parameter estimator" would involve a primary basecaller that generates initial base probabilities and identifies potential read origins, estimating signal amplitudes (read quality scores) and timing offsets (read start/end positions). The "MUD decision unit" would analyze factors like read overlap density, average read quality, and the number of putative source molecules. A "low complexity MUD" could be a simple consensus caller or k-mer based demultiplexer, suitable for well-separated sequences. A "high complexity MUD" would be a probabilistic graphical model (e.g., hidden Markov model or Bayesian network) or a modified Viterbi algorithm operating on the entire sequence-space, used when reads are highly chimeric or from a complex mixture of templates. The "bank of decoders" performs error correction on the demultiplexed sequences, potentially feeding back improved base calls (soft information) to refine the MUD process iteratively (Claim 1).

graph TD
    A[Raw Sequencing Data (FAST5/BAM)] --> B(Basecaller & Initial ID)
    B -- Base Probabilities/Read Overlaps --> C(Genomic Parameter Estimator)
    C -- Metrics (Overlap Density, Quality) --> D{MUD Decision Unit - Bioinformatic Algo Selector}
    D -- Select H/L --> E{Low Complexity MUD (K-mer Demux)}
    D -- Select H/L --> F{High Complexity MUD (PGM/Viterbi)}
    E -- Demultiplexed Reads --> G(Sequence Error Corrector)
    F -- Demultiplexed Reads --> G
    G -- Refined Base Calls --> D
    G -- Final Genomes --> H[Genomic Database]

Derivative 3.2: Autonomous Vehicle Sensor Fusion with Interference Mitigation

Enabling Description:
This derivative integrates the hybrid Turbo-MUD into an autonomous vehicle's sensor fusion system to resolve interference among multiple active sensors (e.g., LiDAR, radar, ultrasonic sensors) that operate in close proximity or overlap in their detection zones. The "received signals" are raw sensor returns from the environment. "Users" are individual sensor units within the vehicle or even other nearby vehicles' active sensors causing interference. The "parameter estimator" processes raw point clouds (LiDAR), radar echoes, and ultrasonic pulses to estimate object positions, velocities, and reflectivity, as well as the relative timing and amplitude of interfering sensor emissions. The "MUD decision unit" evaluates the level of sensor interference, object density, and environmental clutter (e.g., using correlation matrices between sensor signatures or estimated SINR for specific objects). A "low complexity MUD" could be a simple spatial filter or a Kalman filter-based tracker that can easily disambiguate well-separated objects. A "high complexity MUD" would be a joint probabilistic data association (JPDA) filter or an M-algorithm inspired multi-target tracker, performing complex hypothesis testing to resolve closely spaced or ambiguous detections arising from sensor interference, especially critical for obstacle avoidance. The "bank of decoders" refines object classifications and tracks, providing improved estimates back to the MUD for better interference suppression in subsequent frames.

graph TD
    A[Raw Sensor Returns (LiDAR, Radar, Ultrasonic)] --> B(Pre-Processing & Sync)
    B -- Calibrated Sensor Data --> C(Parameter Estimator - Env. & Interferer Models)
    C -- Interference Metrics (Corr, SINR, Density) --> D{MUD Decision Unit - Sensor Fusion Layer}
    D -- Select H/L --> E{Low Complexity MUD (Kalman Filter Tracker)}
    D -- Select H/L --> F{High Complexity MUD (JPDA/M-Alg Tracker)}
    E -- Object Detections --> G(Object Classification & Track Refinement)
    F -- Object Detections --> G
    G -- Refined Object State --> D
    G -- Output --> H[Vehicle Control System]

---

4. Integration with Emerging Technologies

Derivative 4.1: AI-Driven Adaptive MUD Selection

Enabling Description:
This derivative enhances the "multi-user detector decision unit" with an AI-driven inference engine, specifically a deep reinforcement learning (DRL) agent, that dynamically optimizes MUD selection and associated thresholds. Instead of fixed or heuristically defined thresholds, the DRL agent learns optimal MUD switching policies through continuous interaction with the communication environment and feedback from the decoders. The agent's state space includes system parameters (e.g., number of active users, channel conditions, correlation matrix characteristics, current BER, computational load, battery status) derived from the parameter estimator and decoder feedback. The action space for the DRL agent consists of selecting among available MUD algorithms (low, medium, high complexity) and adjusting their internal parameters (e.g., M-parameter for M-algorithm, thresholds for T-algorithm). The reward function is designed to maximize a trade-off between throughput, BER, and computational energy efficiency. This allows for a hyper-adaptive system that can respond optimally to unforeseen channel dynamics or processing constraints in real-time. The parameter estimator could utilize federated learning to aggregate channel insights from distributed receivers, further improving the DRL agent's decision-making.

graph TD
    A[Received Signals] --> B(Parameter Estimator)
    B -- Channel State/Metrics --> C(DRL Agent - Inference Engine)
    C -- Decision (MUD Select, Thresholds) --> D{Hybrid MUD Block}
    D -- Outputs --> E(Bank of Decoders)
    E -- Performance Metrics (BER, Latency, Power) --> F(DRL Agent - Reward Calculation)
    E -- Refined Inputs --> D
    F -- Feedback/Learning --> C

Derivative 4.2: Real-time MUD Optimization with IoT Sensor Feedback and Edge Computing

Enabling Description:
This derivative integrates the hybrid Turbo-MUD with a network of distributed IoT sensors and an edge computing infrastructure for enhanced environmental awareness and real-time optimization. Instead of a single parameter estimator, an array of spatially distributed, low-power IoT sensors continuously monitors the ambient RF environment, local interference sources, and physical channel characteristics (e.g., temperature, humidity, presence of obstacles affecting multipath). This data is fed to an edge computing node which aggregates and processes it, providing fine-grained, localized channel state information (CSI) and interference maps to the "parameter estimation unit." This enriched information allows for a more accurate and predictive "MUD decision unit." For instance, if an IoT sensor detects a new, strong interferer entering the propagation path, the edge node can proactively recommend switching to a higher complexity MUD or adapting windowing strategies even before the primary receiver experiences significant degradation. The low-complexity MUDs can be offloaded to dedicated hardware accelerators at the edge for ultra-low latency processing, while the high-complexity MUDs run on more powerful, centralized edge servers when necessary.

graph TD
    subgraph Distributed IoT Sensor Network
        S1[IoT Sensor 1] -- Env. Data --> E(Edge Computing Node)
        S2[IoT Sensor 2] -- Env. Data --> E
        ...
        Sn[IoT Sensor N] -- Env. Data --> E
    end
    E -- Localized CSI/Interference Map --> P(Parameter Estimation Unit)
    R[Received Signals] --> P
    P -- Enhanced Metrics --> D{MUD Decision Unit}
    D -- Select H/L --> M{Hybrid MUD Block}
    M -- Outputs --> C(Bank of Decoders)
    C -- Feedback --> D
    C -- Final Data --> O[Output]

---

5. The "Inverse" or Failure Mode

Derivative 5.1: Graceful Degradation and Prioritized Emergency Mode

Enabling Description:
This derivative designs the hybrid Turbo-MUD system for graceful degradation and a prioritized emergency operational mode under extreme interference (e.g., electronic warfare jamming) or severe resource constraints (e.g., critical battery depletion). The "MUD decision unit" (Claims 1, 12) incorporates a "resilience manager" that monitors a "threat/resource level" metric. When this metric exceeds a critical threshold, the system automatically transitions into a reduced-functionality state. In this emergency mode, all high-complexity MUDs are disabled, and the system defaults exclusively to the lowest complexity MUD, specifically a highly robust matched filter (Claim 1, 12, 17) or a simple energy detector, even if this leads to a higher BER for non-critical users. The parameter estimator is also simplified to only track essential parameters for a single, designated emergency user or message stream. The "windowing" (Claim 17) becomes extremely coarse, prioritizing short, bursty emergency transmissions. Error correction decoders are reconfigured to minimal, high-redundancy codes (e.g., repetition codes) for the prioritized emergency user, ensuring maximal resilience over throughput or quality for other users. The system aims for detectability of critical messages rather than comprehensive data recovery.

graph TD
    A[Received Signals (inc. Jamming)] --> B(Parameter Estimator - Threat Mon.)
    B -- Threat/Resource Level --> C{MUD Decision Unit - Resilience Manager}
    C -- Critical Threshold Exceeded? --> D{Normal Mode Selection}
    C -- Critical Threshold Exceeded? --> E{Emergency Mode: Matched Filter Only}
    D -- Select MUD --> F(Hybrid MUD Processing)
    E -- Prioritized MUD --> G(Simplified Decoder)
    F -- Outputs --> H(Bank of Decoders)
    H -- Feedback --> C
    G -- Output Critical Alert --> I[Emergency Channel]
    H -- Output Data --> J[Normal Channel]

---

Combination Prior Art Scenarios

These scenarios combine the concepts of US Patent 7593492 with existing open-source standards to further establish defensive prior art.

1. Hybrid Turbo-MUD in Open-Source 5G NR Base Stations (O-RAN)

Description:
The hybrid Turbo-MUD system described in US7593492, with its dynamic MUD selection based on metrics like SINR, BER, and correlation, is integrated into an open-source 5G New Radio (NR) base station architecture, specifically adhering to O-RAN (Open Radio Access Network) specifications. The "parameter estimator" functions within the O-RU (O-RAN Radio Unit) or O-DU (O-RAN Distributed Unit) to provide channel state information (CSI), user scheduling parameters, and interference estimates, as defined by O-RAN interfaces (e.g., fronthaul split options 7.2x). The "MUD decision unit" and the "bank of decoders" are implemented as virtualized network functions (VNFs) or cloud-native network functions (CNFs) within the O-CU (O-RAN Centralized Unit) or O-DU. The selection criteria (Claim 1, 12) are explicitly mapped to 5G NR specific parameters, such as the number of active users per Physical Resource Block (PRB), channel quality indicator (CQI) reports, and the configuration of MIMO layers. The system dynamically switches between 5G NR compliant MUDs: for instance, a low-complexity linear MMSE detector for lightly loaded PRBs or users with strong CSI, and a high-complexity approximate belief propagation MUD or reduced-state Viterbi detector for heavily loaded PRBs, especially in scenarios with spatial multiplexing and beamforming-induced interference. Iterative feedback cycles between MUD and decoders are managed within the O-RAN framework's control plane to optimize 5G NR throughput and latency.

2. Hybrid Turbo-MUD in GNU Radio for Software-Defined Radio (SDR)

Description:
The principles of the hybrid Turbo-MUD from US7593492 are implemented as a set of custom blocks and flowgraphs within the open-source GNU Radio software-defined radio (SDR) framework. A universal software radio peripheral (USRP) or similar SDR hardware acts as the "received signals" and performs raw digitization. The "parameter estimator" (Claims 1, 12, 17) is implemented using GNU Radio blocks for channel estimation, synchronization, and interference sensing (e.g., using correlation estimators and spectral analysis blocks). The "MUD decision unit" is a custom Python or C++ block that dynamically selects between different GNU Radio-implemented MUD blocks. For "low complexity MUDs," this could include a matched filter, a simple decorrelrelator, or an MMSE equalizer block. For "high complexity MUDs," it could be a custom implementation of an M-algorithm or a tree-search block. The "bank of decoders" consists of GNU Radio's existing FEC (Forward Error Correction) blocks (e.g., Viterbi, Turbo, LDPC decoders). The iterative feedback loop (Claim 1) is achieved by passing soft decision outputs (e.g., LLRs) between the custom MUD block and the FEC decoder block within the GNU Radio flowgraph, controlled by a Python controller that monitors performance metrics (e.g., estimated BER via soft-output analysis). This allows for flexible experimentation and deployment of adaptive MUD strategies in an open-source SDR environment.

3. Hybrid Turbo-MUD for Space Communications using CCSDS Standards

Description:
This scenario integrates the hybrid Turbo-MUD system of US7593492 into a ground station receiver designed for deep-space or near-Earth satellite communications, adhering to CCSDS (Consultative Committee for Space Data Systems) recommendations. "Received signals" are telecommand or telemetry data streams from multiple spacecraft, potentially experiencing strong co-channel interference from other satellites, solar noise, or Earth-based interferers. The "parameter estimator" (Claims 1, 12, 17) utilizes CCSDS-compliant synchronization (e.g., P-code acquisition, Costas loop) and channel estimation techniques to derive critical parameters like Doppler shift, signal strength, and channel impulse responses for each spacecraft's signal. The "MUD decision unit" makes selections based on factors such as the current number of active spacecraft in the beam, the predicted BER for each link, and the correlation between the distinct CCSDS spreading codes or modulation schemes. "Low complexity MUDs" could involve matched filters followed by simple interference cancellation, while "high complexity MUDs" would be iterative soft interference cancellation (SIC) or reduced-state sequence estimation algorithms, optimized for the specific CCSDS modulation (e.g., BPSK, QPSK) and coding (e.g., convolutional, LDPC) standards. The "bank of decoders" uses CCSDS-compliant error correction decoders. The iterative processing (Claim 1) feeds back refined symbol estimates, including reliability information (e.g., LLRs), to the MUD, improving interference suppression for subsequent data frames, crucial for reliable space data recovery under challenging link conditions.