Probabilistic signal shaping in MIMO OFDM via successive candidate generation

In this paper, the recently introduced method of channel-coding-based peak-to-average power ratio (PAR) reduction is further developed. On the one hand, not only PAR is considered, but a broad class of signal parameters to be optimised is treated by defining a respective metric to be minimised. On the other hand, the candidate generation and assessment is done successively. If a given tolerable metric limit is met the procedure stops. Analytical expressions for the average number of tested candidates and the achieved performance are derived. Critical metric values, as generalisation of the ‘critical’ PAR value ξ = log(D), where D is the number of carriers, are defined. The theoretical statements are covered by means of numerical simulations using either PAR or cubic metric as desired signal parameter. Copyright © 2010 John Wiley & Sons, Ltd.