Burst Trailing Edge Slope

The burst trailing edge slope is the maximum negative rate of amplitude change (dV/dt or dP/dt) measured during the ramp-down phase of a signal burst. This slope is fundamentally determined by the discharge time constants of the transmitter's power supply decoupling capacitors, the bias network bleed resistors, and the charge carrier recombination rates in the power amplifier's semiconductor junctions. Unlike the leading edge, which reflects charging dynamics, the trailing edge reveals how quickly energy storage elements can be depleted, making it a distinct and unclonable hardware fingerprint.

The trailing edge slope is sensitive to specific component tolerances and parasitic effects:

• Power Supply Holdup Capacitance: Larger or aging capacitors with higher equivalent series resistance (ESR) slow the discharge rate, reducing the slope magnitude.
• Bias Tee Inductance: The inductor in the DC bias network resists rapid current changes, creating a characteristic exponential decay profile.
• Gate/Base Bleed Resistors: The value of resistors that discharge the power amplifier's input capacitance directly controls the turn-off time constant.
• Charge Trapping in GaN/SiGe Devices: Semiconductor defects can cause a slow tail in the discharge curve, creating a unique transient memory effect.

While related, burst trailing edge slope and fall-time variance capture different aspects of the ramp-down:

• Trailing Edge Slope is the instantaneous maximum rate of change (a dynamic, derivative-based feature).
• Fall-Time Variance is the statistical distribution of the total 90%-10% transition duration across multiple bursts (a temporal, interval-based feature). A device with a consistent fall time may still exhibit a unique slope profile if the decay is non-linear, containing inflection points caused by parasitic LC resonances in the power distribution network.

The trailing edge slope is generally more robust to multipath channel distortion than phase-dependent features because it is derived from the amplitude envelope. However, challenges remain:

• Channel Fading: Deep fades can obscure the true amplitude trajectory, requiring channel equalization prior to slope extraction.
• Doppler Spread: In high-mobility scenarios, the envelope may be distorted, necessitating Doppler compensation.
• Amplifier Non-Linearity at Capture: The receiver's own automatic gain control (AGC) must be frozen or de-embedded to avoid compressing the true slope. Contrastive learning techniques are often employed to train models that are invariant to these channel-induced distortions.

In physical layer authentication systems, the trailing edge slope serves as a hard-to-spoof feature because it is dictated by the physical discharge path, which cannot be easily modified by software. It is combined with other transient features in a multi-dimensional fingerprint vector for emitter identification. For example, a rogue device attempting to mimic a legitimate transmitter's ramp-up signature will likely fail to replicate the exact trailing edge slope and its associated undershoot characterization, as these are governed by different physical circuits (charging vs. discharging paths).