Clipping and Filtering

The initial stage applies a memoryless nonlinearity to the complex baseband signal. Any sample whose instantaneous amplitude exceeds a predefined clipping threshold is forcibly truncated to that limit, while the phase is preserved. This operation directly reduces the crest factor but introduces sharp discontinuities in the waveform, which act as broadband noise and cause severe spectral regrowth into adjacent channels.

Following clipping, a frequency-domain filter is applied to suppress the out-of-band emissions generated by the hard limit. This filter is typically a low-pass or band-pass filter designed to meet a specific spectral mask (e.g., 3GPP or ETSI requirements). The filtering process removes the high-frequency components of the clipping distortion, dramatically improving the Adjacent Channel Leakage Ratio (ACLR) and ensuring regulatory compliance.

A critical side effect of the filtering stage is peak regrowth. Removing out-of-band frequency components alters the time-domain waveform, causing previously suppressed amplitude peaks to re-emerge, often exceeding the original clipping threshold. This occurs because the filter's impulse response adds constructively with the remaining signal peaks, partially undoing the PAPR reduction achieved in the clipping stage.

To overcome peak regrowth, clipping and filtering are applied in a multi-stage cascade. Each subsequent stage uses a clipping threshold that is slightly lower than the target, compensating for the regrowth introduced by the following filter. Through repeated iterations, the signal's peak amplitude converges toward the desired limit while the out-of-band spectrum is progressively cleaned. The trade-off is increased computational latency.

While filtering controls out-of-band emissions, the clipping process itself introduces irreversible in-band distortion. This manifests as a displacement of the received constellation points from their ideal reference locations, quantified by an increase in Error Vector Magnitude (EVM). The aggressiveness of the clipping ratio directly correlates with EVM degradation, creating a fundamental trade-off between amplifier efficiency and modulation accuracy.

The Clipping Ratio (CR) is the primary design parameter, defined as the ratio of the maximum permitted amplitude to the RMS level of the unclipped signal. A lower CR achieves more aggressive PAPR reduction and higher power amplifier efficiency but results in greater EVM and requires more iterations to control spectral regrowth. System designers must balance these factors against the specific modulation scheme and error correction coding overhead.

The phenomenon where filtering a hard-clipped signal causes previously suppressed amplitude peaks to reappear above the target threshold. This occurs because the low-pass filter removes the high-frequency components introduced by the sharp clipping discontinuities, which alters the time-domain waveform envelope. Iterative clipping and filtering stages are required to progressively suppress regrown peaks until the target PAPR is achieved.

Unwanted out-of-band emissions generated by the nonlinear clipping operation. Hard clipping introduces sharp amplitude discontinuities that spread energy into adjacent frequency channels, degrading the Adjacent Channel Leakage Ratio (ACLR). The subsequent filtering stage is specifically designed to suppress this regrowth and ensure compliance with the spectral mask defined by standards bodies like 3GPP and ETSI.

A critical metric quantifying the in-band distortion introduced by the clipping process. EVM measures the deviation of the actual transmitted constellation points from their ideal reference positions. Aggressive clipping reduces PAPR but increases EVM, degrading modulation accuracy. System designers must balance Clipping Ratio (CR) against the maximum tolerable EVM specified by the wireless standard (e.g., 3.5% for 256-QAM in 5G NR).

Defined as the ratio of the maximum permitted signal amplitude after clipping to the RMS level of the unclipped signal. A lower CR indicates more aggressive clipping, yielding greater PAPR reduction gain but at the cost of increased EVM and spectral regrowth. CR is the primary design parameter controlling the trade-off between power amplifier efficiency and signal fidelity.

A cascaded architecture that applies successive stages of clipping and filtering with progressively tighter thresholds. Each stage clips the regrown peaks from the previous stage's filtering operation. This approach achieves aggressive PAPR reduction targets while distributing the distortion budget across multiple stages, resulting in better EVM performance than a single aggressive clipping stage.

The standard statistical tool for evaluating CFR performance. The CCDF curve shows the probability that a signal's instantaneous power exceeds a given threshold relative to its average power. Engineers use CCDF plots to measure PAPR reduction gain at specific probability points (e.g., 10⁻⁴) and verify that the clipped signal meets the design target before and after filtering.