The cognitive radio devices operating in the digital terrestrial television (DTT) frequency band, known as white space devices (WSDs), use the free DTT channels under the condition that they do not cause interference above the limit set by the regulator. In 2015, the UK regulatory authority, Office of Communications (Ofcom), published the first European regulation allowing WSD systems to operate in the DTT band. It defines a methodology to calculate the maximum WSD transmission power based on single‐entry interference. It does not guarantee, however, that the co‐channel aggregate interference produced by the WSDs operating in the same region satisfies the intersystem interference limit. In addition, to define the maximum intersystem interference, the Ofcom methodology does not use the same criterion as that used in DTT system planning for the intrasystem interference. This work develops a new methodology for calculating the WSD transmission powers that solve the problems identified in the Ofcom regulation. Based on this methodology, a new WSD admission procedure that guarantees the protection of the DTT receivers from the WSD aggregate interference is proposed. It is compared to two other admission procedures based on the Ofcom regulation and Electronic Communications Committee report 186, respectively.

This study presents a probabilistic model to assess the co‐channel interference produced by multiple transmitters of a fixed wireless access network into a high altitude platform stations (HAPS) victim receiver. The authors propose an alternative probabilistic approach that avoids the worst case assumptions usually considered in traditional interference calculation methods, by considering the random modelling of some of the technical parameters involved. More specifically, the geographic positions of fixed wireless access system transmitters are modelled by a two‐dimensional point process and their antenna gains in the sidelobe region are modelled as gamma distributed random variables. The proposed method is validated by comparing the minimal operating distance resulting from the traditional deterministic interference calculation methods and that obtained with th proposed approach. Results corresponding to the four possible interfering scenarios indicate that the proposed probabilistic approach leads to a reduction in coordination distances.

The performance of modern communications transmitters utilising multi‐carrier modulation techniques is highly sensitive to non‐linear distortions arising mainly from the high‐power amplifier. The increase in symbol duration makes the system more sensitive to time‐varying channels. In addition, the wideband characteristics of multi‐carrier signals result in frequency‐dependent distortions, typically known as memory effects. This study presents analytical expressions that can be used to evaluate the impact of the distortions induced by time‐varying non‐linearities with memory on the bit error rate performance of a multi‐carrier signal such as the orthogonal frequency division multiplexing. The time‐varying non‐linearity with memory is modelled by a time‐varying Volterra series. The resulting mathematical expressions are applied to specific situations, allowing for a quantitative evaluation of the system performance degradation.

This paper presents a fast algorithm that performs a selective counting of the third-order intermodulation products generated when several carriers of various types are transmitted through a common nonlinear device. In deriving the algorithm, the carrier configuration and the frequency distribution of intermodulation products are both described by polynomials which are related to each other. The frequency distributions of third-order intermodulation products of the various kinds are then obtained with the aid of an FFT algorithm and a numerical method for solving a set of linear equations. When compared to a direct counting procedure, the proposed algorithm has presented a drastic reduction on the required Computational effort. This advantage is particularly important when the number of carriers sharing the nonlinearity is large. In an example with 1366 carriers of three different types sharing the nonlinearity, the CPU time was reduced from 7 h 46 min. down to 52 s.