Radiometric Detection

A radiometer squares the magnitude of the received signal and integrates this energy over a fixed observation interval T and bandwidth W. The resulting test statistic is proportional to the total received energy. If a signal is present, this energy will statistically exceed the integrated noise floor. The time-bandwidth product (TW) is the critical design parameter—larger products improve detection sensitivity by averaging out noise variance, but increase observation latency.

The integrated energy is compared to a pre-calculated threshold λ to declare signal presence or absence. This threshold is derived from the desired Probability of False Alarm (PFA)—the rate at which noise fluctuations alone trigger a detection. Under the noise-only hypothesis, the test statistic follows a chi-squared distribution with 2TW degrees of freedom. Setting the threshold requires precise knowledge or estimation of the noise power spectral density N₀.

Radiometers suffer from a fundamental sensitivity limit known as the SNR Wall. Below a certain input SNR, no amount of integration time can reliably distinguish the signal from noise uncertainty. This occurs because the estimator's variance in the noise floor σ² creates an irreducible ambiguity. For a radiometer with noise uncertainty of x dB, signals weaker than this uncertainty floor become undetectable regardless of the time-bandwidth product.

The Achilles' heel of radiometric detection is noise power uncertainty. In practical receivers, the thermal noise floor fluctuates due to temperature changes, component aging, and automatic gain control (AGC) drift. Even a 1-2 dB uncertainty in the noise estimate can catastrophically degrade performance. This limitation motivated the development of more robust blind detectors, such as eigenvalue-based and cyclostationary methods, which do not require explicit noise floor knowledge.

To monitor broad spectrum ranges, a single radiometer is insufficient. A channelized radiometer splits the input bandwidth into parallel narrowband channels using a filter bank or FFT-based polyphase architecture. Each channel independently integrates energy, enabling simultaneous detection and coarse frequency estimation of multiple signals. This architecture is essential for detecting frequency-hopping spread spectrum (FHSS) signals, where energy appears transiently in different channels.

Radiometric detection is classified as a non-coherent and blind method. It requires no prior knowledge of the signal's modulation, spreading code, carrier phase, or timing synchronization. This makes it the universal first-stage detector in electronic warfare and spectrum monitoring systems. However, this generality comes at a cost: it cannot classify modulation type, identify specific emitters, or separate co-channel signals—it only declares energy presence.

A blind spectrum sensing method that computes the eigenvalues of the received signal's sample covariance matrix to detect the presence of a spread spectrum signal. Unlike radiometric detection, this technique does not require knowledge of the noise floor, making it robust in uncertain noise environments. Common algorithms include the Maximum-Minimum Eigenvalue (MME) detector and the Energy with Minimum Eigenvalue (EME) detector, which exploit the fact that signal-plus-noise covariance matrices have distinct eigenvalue distributions compared to noise-only matrices.

A unique periodic pattern embedded in a signal's spectral correlation function, intentionally generated by modulating the spreading code to enable robust signal identification. Unlike energy-based methods, cyclostationary analysis exploits the hidden periodicities in modulated signals that are absent in stationary noise. Key discriminators include the cyclic frequency and spectral frequency parameters, which form a distinctive two-dimensional fingerprint for each modulation and spreading scheme.

A detection architecture that splits a wide bandwidth into parallel narrowband channels, integrating energy in each to detect and characterize frequency-hopping signals in real time. By employing a filter bank or FFT-based polyphase decomposition, this approach simultaneously monitors multiple frequency bins. When combined with time-frequency analysis, it enables the reconstruction of hop timing and hop set identification, providing a practical bridge between simple energy detection and full cyclostationary processing.

A signal acquisition framework that reconstructs sparse wideband spread spectrum signals from sub-Nyquist rate samples by exploiting their inherent structure in a dictionary basis. This technique is particularly valuable for intercepting frequency-hopping signals across wide bandwidths where traditional Nyquist-rate sampling is impractical. By solving an l1-minimization problem, compressive sensing recovers the signal's time-frequency occupancy map, enabling simultaneous detection and parameter estimation with dramatically reduced hardware requirements.

The identification of short-duration, intermittent spread spectrum emissions in time-domain energy profiles or spectrograms, often used to counter low-probability-of-detection (LPD) tactics. Key challenges include distinguishing burst signals from impulsive noise and determining precise start/stop times for subsequent despreading. Techniques include:

• Adaptive thresholding with constant false alarm rate (CFAR)
• Sequential change-point detection algorithms
• Short-time Fourier transform (STFT) energy integration

A non-coherent detection architecture that multiplies a received DSSS signal by a delayed version of itself to generate a spectral line at the chip rate for estimation. This technique exploits the fact that the product of a signal with its time-shifted copy produces a sinusoidal component at the chip rate when the delay is less than the code period. The resulting spectral line can be detected with a narrowband filter or FFT, enabling chip rate estimation without prior knowledge of the spreading code.