In this paper, a distributed and autonomous technique for resource and power allocation in orthogonal frequency division multiple access (OFDMA) femto-cellular networks is presented. Here, resource blocks (RBs) and their corresponding transmit powers are assigned to the user(s) in each cell individually without explicit coordination between femto base stations (FBSs). The "allocatability" of each resource is determined utilising only locally available information of the following quantities: - the required rate of the user; - the quality (i.e., strength) of the desired signal; - the frequency-selective fading on each RB; and - the level of interference incident on each RB. Using a fuzzy logic system, the time-averaged values of each of these inputs are combined to determine which RBs are most suitable to be allocated in a particular cell, i.e., which resources can be allocated such that the user requested rate(s) in that cell are satisfied. Furthermore, link adaptation (LA) is included, enabling users to adjust to varying channel conditions. A comprehensive study of this system in a femto-cell environment is performed, yielding system performance improvements in terms of throughput, energy efficiency and coverage over state-of-the-art ICIC techniques.