Transforming the distribution of OFDM signals for peak‐to‐average power ratio reduction

Abstract

SUMMARYLarge peak-to-average power ratio (PAPR) is the main drawback of orthogonal frequency division multiplexing (OFDM) systems. In this paper, a novel nonlinear scheme is proposed to reduce the PAPR. This scheme transforms the amplitude of original OFDM signals from Rayleigh distribution into beta distribution. Moreover, given different parameters, this scheme has different transformation expressions. It is demonstrated that the well-known companding schemes based on exponential function and error function are actually the special cases of the proposed scheme. Further, we derive another simpler expression of the proposed scheme. Simulation results show that the new transformation can achieve a better tradeoff between PAPR reduction and bit-error rate (BER) performance than the exponential function and error function companding schemes.

On Minimizing the Cubic Metric of OFDM Signals Using Convex Optimization

“…In problem (12), the functions (12a) and (12c) ensure the minimization of E |x n | 6 ; the constraint (12d) remains the average power nondecreasing; (12e) and (12f) constrain the modified signal to satisfy the EVM and spectral mask requirements, respectively; finally, (12g) keeps the pilot signals unchanged. (Note that for the optimized signal, S (X − X) = ε holds, so (12d) can be used to equivalently replace the nonconvex constraint (9).…”

“…Now the problem (12) can be transformed to an equivalent problem with variablesX (1) and c as follows…”

“…In general, these PAPR-reduction techniques can be divided into two categories: techniques with distortion and without distortion. The former includes clipping/filtering [5]- [10] and companding [11], [12] techniques, while the latter includes probabilistic techniques [13], [14] and tone reservation (TR) [15]. On the other hand, as mentioned above, CM is a more accurate metric for envelope fluctuations and thus reducing CM has great significance.…”

Cubic metric reduction in OFDM systems by using convex optimization

“…The techniques without distortion include coding [5], selected mapping (SLM) [6] and partial transmit sequences (PTS) [7]. The techniques with distortion include repeated clipping and filtering [8] and companding [9]. Compared with the techniques without distortion, the techniques with distortion can be easily applied and the PAPR-reduction performance is very significant.…”

A model for OFDM signals with applications

“…In , it is shown that under some conditions on a k and for N → ∞ , the continuous time domain OFDM signal distribution converges to the correlated complex Gaussian distribution. A different approach for modelling the discrete time domain signal is invoking the central limit theorem (CLT) . Applying the CLT to yields that the marginal distribution of x is complex normally distributed, that is, for N large enough and .…”