AFWERX / SpaceWERX CSO Phase I Offeror // C++ & FPGA IP Cores

Deterministic Fractional-Calculus DSP Cores for Contested RF & SATCOM Environments

We eliminate the "Amplitude Trap" below 0 dB SNR. Our software-defined baseband pre-filter dynamically excises swept CW jamming and enforces phase coherence across tactical satellite downlinks and terrestrial 5G O-RAN—with zero AI hallucination and sub-microsecond L1 FPGA latency.

Request Azure HPC Sandbox Access Explore Mathematical Architecture
-0.9 dB
QPSK Phase-Lock SNR Threshold
6× MAD
Outlier-Immune Excision Gate
<1.5%
FPGA LUT Resource Footprint
1 CLK
Attractor LUT Coercion Timing

The Tactical Dead-End of Probabilistic & Power-Threshold DSP

Why legacy Automatic Gain Control (AGC) and black-box AI demodulators fail when electronic warfare (EW) pushes the battlespace below the noise floor.

The Negative-SNR "Amplitude Trap"

Legacy RF filters and modems operate on rigid power thresholds. When dynamic atmospheric clutter or swept continuous wave (CW) jamming pushes signal amplitude below the thermal noise floor, legacy receivers suffer catastrophic failure.

  • Viterbi & Viterbi Cycle Slips: Below 0 dB SNR, standard phase trackers exponentially amplify noise, spinning into an unrecoverable "donut" phase rotation.
  • AGC Collapse: Swept CW and barrage jammers desensitize front-end gain control, causing total packet error rate (PER) spikes across command and control links.
  • Black-Box AI/ML Fragility: Neural network demodulators hallucinate fake symbols and require millions of MACs—violating sub-microsecond L1 PHY timing budgets.
The MatterMath Pre-Filtering Solution

MatterMath abandons amplitude thresholding. Positioned digitally between the terminal ADC and the modem, our deterministic C++ pre-filter conditions raw I/Q feeds before standard channel estimation.

  • Sub-Threshold Phase Persistence: Measures differential phase alignment vectors to detect coherent micro-structures assembling inside the thermal noise floor down to -0.9 dB SNR.
  • Surgical Jammer Excision: Punches spectral holes exactly where swept CW jammers reside while preserving Protected Tactical Waveform (PTW) processing gain.
  • 100% Mathematically Auditable: Zero neural network weights, zero offline retraining, and strict compliance with DoD aerospace software verification standards.

The 3-Pillar Deterministic Signal Recovery Stack

Authored in hardware-agnostic C++ and synthesized via High-Level Synthesis (HLS) into SOSA/MORA-compliant FPGA bitstreams.

Pillar 01 // Temporal Memory
Grünwald–Letnikov Fractional Baseline

Replaces stationary moving averages with discrete Grünwald–Letnikov fractional integro-differential memory operators.

Fractional Fading Memory Buffer W_GL = [0.68, 0.22, 0.10]
3-Tap Lattice IIR (2 MACs / bin)

Inherits rolling power-law fading memory across frames, preventing swept CW jammers from desensitizing thermal noise baseline calculations.

Pillar 02 // Manifold Excision
Robust MAD & Wiener Gating

Replaces standard deviation (σ) with Median Absolute Deviation (MAD) statistics to eliminate outlier sensitivity from high-power interference.

Statistical Excision & Soft Gate MAD = median(|X_i - median(X)|)
Gate: Intensity > (6 × MAD) = 0.05x
Wiener Gate: x² / (x² + 1)

Surgically suppresses hostile CW jamming spikes while passing clean sub-threshold spread-spectrum signal bins uninhibited.

Pillar 03 // Phase Coherence
Phase-Space "Gravity Well" Attractor

Constructs an artificial potential well across QPSK phase-space, defining ideal diagonal attractor basins (±π/4, ±3π/4).

Sinusoidal Relaxation Coercion θ_rel = θ_raw + α·sin(4(θ_tgt - θ_raw))
Attractor α = 0.35 | 256-Entry LUT

Pre-computed 256-entry lookup table executes attractor coercion in 1 clock cycle, locking phase before Viterbi demodulation.

Strategic Alignment: Space Force & Air Force Transition

Designed for zero-spaceflight-risk insertion into existing terrestrial military satellite ground hubs and joint tactical testbeds.

Space Systems Command (SSC)

System Delta 88 (SYD 88) & Protected Tactical SATCOM

Directly supports SYD 88's "commercial-first" mandate under Col. A.J. Ashby by providing ground-only anti-jam resilience upgrades for military satellite communications.

  • Protected Anti-Jam Tactical SATCOM (PATS): Enhances PTW over WGS (PTWoW) and Commercial (PTWoC) downlinks against peer EW jamming.
  • PTS-G ($4B IDIQ): Licensed as an algorithmic FPGA core to prime contractors to fortify commercial baseline ground segments ahead of the 2028 launch wave.
  • PTES Joint Hubs: Integrates digitally ahead of Air Force-Army Anti-Jam Modems (A3M) and Navy WAMS to lower effective J/S interference ratios.
AFRL Sensors Directorate & Army DEVCOM

MUSTER BAA, REFLECT EW Call & C4ISR Testbed

Aligned with AFRL/RY (Wright-Patterson AFB) under Acting Director Col. Rodrick A. Koch for RF sensing, spectrum warfare, and avionics EW survivability.

  • AFRL MUSTER (`FA8650-21-S-1180`): Targeted for autonomous SDR enhancement and resilient tactical L1 physical layer processing.
  • REFLECT Call & TAAP-DEW: Provides auditable ELINT signal countermeasure simulation and airborne threat exploitation.
  • Army DEVCOM Armaments: Formatted for over-the-air SATCOM emulation across UHF to X bands at the Fort Huachuca C4ISR Testbed (DREN).

Commercial Dual-Use: Terrestrial 5G/6G & LEO Ground Stations

Dual-use commercial software (EAR99) engineered for high-throughput telecommunications infrastructure.

5G/6G O-RAN Inline L1 PHY Acceleration

In terrestrial Open Radio Access Network (O-RAN) architectures, combating dense urban co-channel interference requires sub-millisecond physical layer control.

  • 7.2x Fronthaul Split Integration: Bypasses slow Near-RT RIC loops (>10 ms) to operate directly within the O-DU (Open Distributed Unit) as an L1 Inline Accelerator.
  • HLS eASIC / FPGA Execution: Compiles directly onto commercial carrier tower accelerator cards to clean raw user-plane I/Q data before OFDM demodulation.
LEO Earth Observation Downlink Recovery

Low Earth Orbit (LEO) Earth Observation and remote sensing satellites suffer severe packet drops during critical 10-minute downlink windows due to rain fade and thermal noise.

  • Terrestrial Server Ingest Plugin: Deployed as a containerized Linux C++ plugin directly onto commercial ground station servers (e.g., Kratos, Amergint).
  • 25% Daily Data Yield Increase: Recovers sub-threshold image telemetry packets at the ground level with zero spaceflight hardware qualification required.
Empirical Evaluation Protocol // Bring Your Own I/Q

The Azure HPC Sandbox for DoD TPOCs & Prime Integrators

Evaluating deep-tech DSP algorithms requires empirical, real-world data validation. To facilitate frictionless Proof-of-Concept (PoC) trials for DAF, Space Force, and prime contractor engineers, MatterMath hosts a secure Microsoft Azure High-Performance Computing (HPC) Sandbox under mutual NDA.

Step 01 // Ingest
Upload Contested I/Q

Partners upload their own raw, corrupted I/Q sample datasets (from problematic SATCOM downlinks, swept CW jamming emulators, or flight testbeds).

Step 02 // Execute
3-Pillar Pre-Filtering

We route the raw file through the Grünwald–Letnikov, MAD excision, and Gravity Well attractor blocks at target FPGA sample rates.

Step 03 // Audit
Empirical BER Delivery

Receive the restored I/Q dataset alongside comparative telemetry proving exact Bit Error Rate (BER) reduction and constellation phase lock.

Request Sandbox NDA & Technical Briefing