Orthogonal Frequency-Division Multiplexing (OFDM) is a key modulation technique in modern communications technology due to its robustness against frequency-selective transmission channels. However, a major problem is the high peak-to-average power ratio (PAPR), which results in nonlinear distortions due to the limited linearity range of analog components. While this can be largely compensated for in the unambiguous part of the transmission characteristics using suitable pre-distortion methods, this is no longer possible at higher input levels because there is no reversible relationship between the input and output signals anymore. This phenomenon, in which the signal is hard limited, is known as clipping and occurs, for example, in the saturation range of high-frequency power amplifiers or below the laser threshold in optical transmission systems. In current communication standards, this problem is typically addressed by PAPR reduction methods or by reducing the average signal power. However, these approaches have disadvantages, such as reduced data rate or degraded signal quality at the receiver. The aim of this work is to first analyze and mathematically describe clipping and its introduced distortion. Subsequently, concepts that optimize the transmission performance of clipped OFDM systems, considering the obtained results, are developed. The investigations are carried out exemplary on radio and optical free-space transmission systems, since the signals have different statistical properties and the results thus cover most of the relevant application scenarios in which OFDM is used. First, realistic models of the two transmission systems are introduced and, based on these, the statistical and spectral properties of clipping distortion are calculated analytically and verified by simulations. Next, the aspect of digital baseband clipping, which is often regarded as a PAPR reduction method, is examined. However, this only shifts the problem from analog to digital domain, resulting in the need to digitally process a non-band-limited signal. In the next part of the work, optimization approaches are developed and examined using simulations. First, the transmission side is considered and the optimal transmit power that leads to maximum transmission quality is determined as a function of the system parameters. Furthermore, it will be shown that a significant gain can be achieved by correctly considering clipping interference in adaptive modulation methods. Subsequently, receiver-side optimizations are investigated, with a particular focus on decision-based iterative methods that successively remove clipping distortion in the received signal. Finally, the influence of different propagation times in the individual transmitter-receiver links on channel capacity in optical free-space systems with multiple transmitters and receivers (multiple-input multiple-output, MIMO) is investigated. Thereby, it is demonstrated that, despite very strong line-of-sight dominance, significant gains can be achieved by using MIMO techniques, particularly at high bandwidths.
Alexander Frömming (2026) studied this question.