ABOUT NOEMA
NOEMA is a signal-domain compression and reconstruction system built for raw radar, RF, SAR, and I/Q data.
It is designed for Level-0 signal pipelines where storage, transfer, and long-term archival of raw data are the dominant operational constraint.
NOEMA reduces raw signal data volume while preserving reconstructability, enabling compressed archives that remain usable in downstream processing workflows.
Built for signal data, not generic data
Raw radar and RF datasets are structurally different from general scientific datasets.
They are generated from physical sensing systems and contain properties such as:
I/Q amplitude distributions
quantization structure from ADC systems
noise-corrupted signal dynamics
local stationarity over time windows
structured redundancy across channels and segments
These properties are not optimized for generic compression methods.
NOEMA operates directly on these signal characteristics rather than treating data as generic byte sequences.
Compression at the signal layer
NOEMA applies compression based on statistical structure in signal space.
Instead of relying on general-purpose redundancy detection, NOEMA models:
entropy variation across signal windows
distribution shifts in I/Q streams
segment-level statistical structure
redundancy across time-frequency regions
Compression is adaptive and dataset-dependent.
Performance varies based on sensor type, acquisition mode, and signal complexity.
Reconstructable archives
Every NOEMA output is a working archive, not a static compressed file.
Each archive preserves:
reconstruction capability for full datasets
selective reconstruction of signal regions
deterministic decode mapping through indexed metadata
configuration context used during compression
Reconstruction is validated using signal-level metrics, including:
signal-to-noise ratio (SNR)
root mean square error (RMSE / NRMSE)
correlation consistency across reconstructed outputs
Indexed signal structure
Compressed data is paired with a structured index that enables operational use without full decompression.
The index supports:
region-level retrieval
time-window extraction
segment mapping
reconstruction parameter resolution
This enables targeted access to signal subsets without requiring full archive expansion.
System behavior
NOEMA is designed for predictable integration into existing signal pipelines.
It supports:
batch processing of archived datasets
streaming ingestion for continuous capture systems
on-prem and air-gapped deployment
cloud-based archive processing
It does not require changes to upstream acquisition systems.
Role in signal infrastructure
NOEMA operates at the archival layer of signal processing systems.
It sits between raw data generation and downstream analytics pipelines, reducing storage and transfer load while preserving reconstructability and analytical usability.
As signal datasets scale in size and duration, this layer becomes a primary constraint in system design.
NOEMA is built to address that constraint directly.
Summary
NOEMA is a signal-domain compression system for Level-0 radar, RF, SAR, and I/Q data.
It reduces raw signal storage requirements while preserving:
reconstructability
fidelity measurability
indexed access to signal regions
compatibility with downstream workflows
It is designed for environments where signal data volume has become a structural limitation.