This Master Thesis of “Applied Signal Processing and Implementation” specialization at Aalborg University is an investigation of FFT algorithms in OFDM receivers and the algorithms power usage on customizable platforms. The project focuses on mobile applications and cooperative radios, wherein only a part of the received frequency spectrum is needed. This can be exploited by special FFT algorithms to yield a lower operations count and intuitively a lower power consumption. However, what is not reflected in the operations count is the power-consumption of the controlling HW/SW. This thesis seeks to investigate the possibilities and tradeoffs, with regards to power usage, when computing a subset of the frequency spectrum, as opposed to the full spectrum. Initially, the concept of cooperative radio and a signal model for OFDM is defined. Afterwards, two Fourier transform algorithms - a full Split-Radix FFT and an FFT algorithm computing only a subset of the spectrum (SFFT) - are examined and mapped to a Cyclone III FPGA architecture. Next, the power performance of each implementation is examined and an investigation into possible improvements is performed. In conclusion the algorithms are compared to a performance measure of computational complexity traditionally used to theoretically evaluate FFT algorithms. The test results shows that the SFFT is not feasible with regards to power usage, without further improvements. These improvements include, among others, an enhanced power-off mechanism when subsystems are not in use. If a power-off state is introduced it is predicted that the SFFT becomes feasible and that computational complexity corresponds to the power usage for this implementation.