Methods to increase the robustness of the frequency selection and power allocation processes for HF communications have been a part of two research and development projects funded by the Swedish Armed Forces. This work started in 2012 and have continued until 2015 and will end with this report, which summarizes the work that have been performed from the start until now. The main application of this research have been naval tactical communications. Therefore, many examples and traffic scenarios are for naval situations. The results are, however, general so that they are applicable for other scenarios as well. The focus of the work has been on improving the resource allocation, in terms of frequency selection and power allocation, by including information of the local interference at the receivers. The first step to including such information is to measure and characterize the interference environment at the receivers. In the simplest case, only the average power of the interference is considered. The average power is sufficient to characterize the interference if it is assumed to be white Gaussian. One extension is to consider impulse interference. Then, the average power as well as the impulsiveness of the interference have to be measured and characterized to better allocate the available resources. Methods for frequency selection and power allocation, that exploit different levels of information about the local interference, are proposed and analyzed in this report. Both individual transmission links as well as multiple channels such as multicast and broadcast are considered. The concept of interference-based frequency selection in HF communications is presented and some implementation aspects are discussed. Several methods for interference based frequency selection are proposed and analyzed, taking both Gaussian and impulse interference into account. An analysis of the tradeoff between overhead transmission cost and performance gain of using additional information of the receiver interference environment is also provided. Furthermore, the report considers power allocation on a set of parallel channels affected by impulse interference. The optimal power allocation that maximizes channel capacity is derived. The special case of frequency selection, where all of the available power is allocated to a single channel is also considered. In addition, power allocation of co-located transceivers is considered, and some analysis and interpretations of the strategy that minimizes the total transmit power under SNR constraints are shown. In all applications studied in this report, it is seen that including knowledge of the local interference in the resource allocation procedure can be beneficial. Moreover, if the noise is impulsive, inclusion of the impulse characteristics of the noise can also improve system performance. However, distribution of local information in the network is crucial and has to be done properly.