Burst Leading Edge Slope

The burst leading edge slope is defined as the maximum first derivative of the signal envelope (dA/dt) during the turn-on transient. It is calculated by extracting the analytic signal's magnitude via the Hilbert transform and then computing the peak positive rate of change. This metric is directly proportional to the power amplifier's slew rate, which is the maximum speed at which its output can change, governed by internal current limits and parasitic capacitances.

This slope is not a software-defined parameter but a physical hardware fingerprint. It is determined by:

• Power amplifier (PA) biasing network: The resistor-capacitor (RC) time constants in the gate/base bias circuit dictate how quickly the transistor turns on.
• PA transistor physics: Electron mobility and channel formation speed in GaN, GaAs, or LDMOS technologies.
• Power supply regulation: The ability of the voltage regulator to supply instantaneous inrush current without sagging, which would otherwise limit the slew rate.

The leading edge slope is a highly discriminative feature for emitter identification because it captures the transient current inrush capability of the specific device. Two transmitters of the same make and model will exhibit statistically distinct slopes due to microscopic manufacturing variances in their capacitors, transistors, and bond wire inductances. This feature is particularly robust against channel effects like multipath, as the rapid amplitude change occurs over microseconds, often shorter than the channel's coherence time.

Accurate measurement requires high-fidelity equipment:

• Sampling Rate: Oscilloscopes or SDRs must sample at rates significantly higher than the Nyquist frequency to capture the fast edge without aliasing.
• Triggering Jitter: Precise burst onset detection is critical; timing errors in the trigger point will distort the calculated derivative.
• Noise Amplification: Differentiation amplifies high-frequency noise. Savitzky-Golay filtering or wavelet denoising is often applied to the envelope before calculating the derivative to prevent noise from masking the true slope.

The leading edge slope is a component of the broader amplitude ramp profile. It is closely related to:

• Overshoot Characterization: A steeper slope often correlates with a higher overshoot peak due to underdamped control loops.
• Rise-Time Variance: The statistical distribution of the 10%-90% rise time is inversely related to the slope; a faster slope yields a shorter, but potentially more variable, rise time.
• Transient Memory Effect: The slope can vary slightly depending on the transmitter's previous state (e.g., thermal condition), making it a dynamic rather than perfectly static fingerprint.

In a physical layer authentication system, the leading edge slope is extracted as a feature vector and fed into a one-class classifier or a Siamese neural network. During enrollment, the legitimate device's slope distribution is modeled. During verification, a new burst's slope is compared against this model. An adversarial device attempting to spoof the identity will fail to replicate the exact slew rate because it is an unclonable physical characteristic of the original hardware's power amplifier and power distribution network.

The maximum rate of voltage change the PA can produce at its output, typically expressed in V/µs. This physical limit is the direct cause of the measured burst leading edge slope. It is governed by the PA's bias current, transistor gate capacitance, and compensation networks. A slower slew rate produces a shallower slope, while a faster slew rate yields a steeper ramp. Key factors:

• Bias network charging time
• Gate-drain Miller capacitance
• Power supply current limiting
• Thermal effects at turn-on

The mathematical operation that extracts the instantaneous rate of amplitude change from the signal's envelope. Computed as dA(t)/dt where A(t) is the Hilbert transform magnitude. The peak value of this derivative during the ramp-up phase is the burst leading edge slope. This metric is highly sensitive to non-linearities in the PA's turn-on behavior and reveals inflection points caused by parasitic reactive components in the matching network.

The quantification of the transient amplitude excursion beyond the steady-state level during the ramp-up phase. This occurs when the PA control loop is underdamped, causing the envelope to exceed its target before settling. The overshoot percentage and subsequent ringing frequency are direct indicators of:

• Loop filter component values
• Phase margin of the control system
• Parasitic inductance in the supply path

Overshoot is often analyzed alongside the leading edge slope to fully characterize the attack profile.

The non-linear amplitude and phase distortion generated by the PA when driven through its non-linear region during the rapid ramp-up. As the PA transitions from cutoff to linear operation, the gain is not constant, causing the leading edge slope to deviate from an ideal linear ramp. This nonlinear trajectory is a rich source of fingerprinting features because it reflects:

• Transistor threshold voltage variations
• Gate oxide thickness inconsistencies
• Dopant concentration gradients in the semiconductor
• Thermal memory effects from prior transmissions