The Research On PAPR Reduction Of OFDM By Intelligent Algorithms

Orthogonal frequency division multiplexing (OFDM) is developed from themulti-carrier modulation (MCM) technology. OFDM symbol is the synthesis of themodulated subcarrier signals.When the phases of the modulated subcarrier signals aresame, the instantaneous power of the OFDM symbol will be far greater than the averagepower of the OFDM symbol, which resulting in high Peak to Average Power Ratio(PAPR).High PAPR is a main drawback of OFDM system, and this requries high lineardynamic range of the power amplifier of transmitter. If signals with high PAPR aretrapped into the nonlinear region of power amplifier, this may cause signals distortionand destroy the orthogonality of OFDM signals. It is needed to study effectivealgorithms to suppress the PAPR of OFDM.The methods used to reduce the PAPR of OFDM system currently can be dividedinto three categories: Signal pre-distortion technology, coding technology andprobabilistic technology. Probabilistic technology has no signal distortion but bringsless redundancy; it is the main direction of current study. Probabilistic technologyincludes the selected mapping scheme (SLM) and partial transmit sequence method(PTS), etc.PTS uses the right phase rotation factors to make the signal sequences have lowPAPR values. The main step of PTS is choosing the optimal or suboptimal method toobtain suitable phase factors. The reaearch results have shown that for the same pointsInverter Fast Fourier Transformation (IFFT), PTS makes the OFDM system has lowerPAPR than SLM. This paper is based on PTS and mainly concerned about twointelligent algorithms in searching phase factors of PTS. The main research works andachievements are as follows:1、 Researching genetic algorithm (GA) of intelligent algorithms and its applyingin PTS, then proposeing a new algorithm called phase factor optimal pair(PFOP). PFOP aims at selecting better phase factor combinations. Throughsimulating PAPR performance curves of GA-PTS and PFOP-PTS, we know new algorithm is better than GA-PTS0.2~0.7dB in the reduction of PAPR.2、 In order to solve the defect that PFOP is easy to fall into local optimumin at thelate iteration, two methods is proposeed to improve the PFOP: abandon/introducing mechanism and local search mechanism.3、 Particle Swarm Optimization (PSO) algorithm and PFOP are combined to beapplied in selecting phase factor combinations. Through the simulation resultswe find the joint method has better PAPR performance than using PSO orPFOP alone.4、 In traditional PTS, the detection and recovery of original signals at the receiverneed side information (SI). A detection method without side information isdesigned in this paper. By using the phase factors chose from the specificvalues, it is possible to achieve low bit error rate of the signal detection at thereceiving end.