Crest Factor Reduction (CFR) Integration

The fundamental metric that CFR aims to reduce. It is the ratio of the signal's peak power to its average power, typically expressed in dB. High PAPR forces the power amplifier to operate at a large back-off from its saturation point, severely degrading efficiency. CFR Integration directly targets this metric to enable a higher average operating power without inducing excessive nonlinear distortion that the DPD must then correct.

A critical signal quality metric measuring the deviation of actual transmitted symbols from their ideal constellation points. The CFR-DPD co-design presents a direct trade-off: aggressive CFR reduces PAPR but introduces in-band distortion that degrades EVM. The integrated DPD must be designed to linearize the PA without over-correcting the intentional clipping artifacts from the CFR, balancing efficiency against modulation accuracy.

A common CFR technique that multiplies the signal peaks with a smooth window function (e.g., Gaussian, Kaiser, Hamming) to limit spectral regrowth compared to hard clipping. The windowed clipping noise is spectrally contained but temporally extended. Integration with DPD requires the predistorter model to account for the specific time-domain signature of the windowing function, as it creates a deterministic memory effect in the signal envelope.

A peak-cancelling CFR method that subtracts a scaled, spectrally-shaped cancellation pulse from detected peaks above a threshold. This technique offers precise control over the error spectrum. In a joint CFR-DPD architecture, the cancellation pulse can be co-designed with the DPD basis functions to ensure the subtracted energy falls into frequency bands where the PA's nonlinearity is most easily linearized, optimizing the overall system.

The amount by which the PA's average input power is reduced below its 1 dB compression point to meet linearity requirements. High back-off improves linearity but drastically reduces DC-to-RF efficiency. The primary goal of CFR Integration is to minimize this back-off by pre-conditioning the signal. The DPD then compensates for the nonlinearity at the new, higher operating point, enabling a net gain in transmitter efficiency.

The simplest CFR method, which directly saturates the signal magnitude at a defined threshold. While highly effective at reducing PAPR, it causes severe spectral regrowth and high EVM due to sharp discontinuities. In an integrated system, hard clipping is rarely used alone; instead, it is followed by filtering. The DPD must be robust enough to handle the strong nonlinear memory introduced by the subsequent filtering of the clipped signal.