OPTIMIZEDIMPLEMENTATION OF FFTPROCESSOR FOR OFDMSYSTEMS
Introduction Due to the advanced VLSI technology, Fast FourierTransform (FFT) and Inverse Fast FourierTransform (IFFT) has been applied to wide field ofdigital signal processing (DSP) applications. For modern communication systems, the FFT/IFFTis very important in the OFDM.
Cont…. Because of FFT/IFFT is the key computational blockto execute the baseband multicarrier demodulationand modulation in the OFDM system. The FastFourier Transform (FFT) and its inverse transform(IFFT) processor are key components in OFDMsystems. An optimized implementation of the 8-point FFTprocessor with radix-2 algorithm in R2MDCarchitecture is the processing element
OFDM . Orthogonal Frequency Division Multiplexing(OFDM) technology is an effective modulationscheme to meet the demand. This is an advanced modulation technique thateffectively expands channel utilization and reducesinter-symbol interference (ISI) and inter-carrierinterference (ICI) caused by multi-path effect.
Cont…. Multiple-input multiple-output (MIMO) signalprocessing technique has been utilized incombination with OFDM for next generationwireless communication system to enhance linkthroughput as well as the robustness of transmissionover frequency selective fading channel . FFT algorithms are based on the fundamentalprinciple of decomposing the computation of theDFT of a sequence of length N into successivelysmaller DFT using the symmetry, periodicity,compressibility and expansibility properties of WN
DFT Discrete Fourier Transform (DFT) is widely appliedto the analysis, design, and the implementation ofdiscrete-time signal processing algorithms andcommunications. The computational complexity of direct evaluation israther high so that it is hard to meet highperformance and low cost design goal.
FFT The fast Fourier transform (FFT) algorithm, togetherwith its many successful applications, represents oneof the most important advancements in scientificand engineering computing in this century. FFT algorithms are based on the fundamentalprinciple of decomposing the computation of thediscrete Fourier transform of a sequence of length Ninto successively smaller discrete Fourier transforms.
Cont… Decomposing is an important role in the FFTalgorithms. There are two decomposed types of the FFT algorithm. Oneis decimation-in-time (DIT), and the other is decimation-in-frequency (DIF) . The DIT algorithm means that the time sequence is decomposed intosmall subsequence, and the DIF algorithm decomposes the frequencysequence. In addition, there is no difference in computationalcomplexity between these two types of FFT algorithm.
R2MDC ARCHITECTURE One of the most straightforward approaches forpipeline implementation of radix-2 FFT algorithm isRadix-2 Multi-path Delay Commutator (R2MDC)architecture. It’s the simplest way to rearrange datafor the FFT/IFFT algorithm. The input data sequence are broken into two paralleldata stream flowing forward, with correct distancebetween data elements entering the butterflyscheduled by proper delays. 8-point FFT in R2MDCarchitecture
ADD AND SHIFT METHOD Another method proposed eliminates the non-trivialcomplex multiplication with the twiddle factors (W81, W8 3) and implements the processor withoutcomplex multiplication. . The proposed butterfly processor performs themultiplication with the trivial factor W8 2=-j byswitching from real to imaginary part and imaginaryto real part, with the factor W8 0 by a simple cable
Cont… With the non-trivial factors W1,W3 the processorrealize the multiplication by the factor 1/√2 usinghard wired shift-and-add operation as shown in Fig.5.
CONCLUSION . Our proposed variable-length FFT processor thatare suitable for various MIMO OFDM-basedcommunication systems, such as IEEE 802.11n, andIEEE 802.16 WiMAX, can perform256/128/64/16/8-point with 1-4 data sequences. The proposed 8-point FFT processor is used forIEEE 802.11n, and the proposed 8-point FFTprocessor is used for IEEE 802.11n and IEEE 802.16.