Research On Intelligent Decision Technology In Cognitive Radio

Cognitive radio (CR) is widely researched in the field of wireless communication.itis considered as a fundamental solution to waste of spectrum resource. Transformdomain communication system (TDCS) is developped from transform domainprocessing and spread spectrum technology. It is a new technology in the field ofwireless communication, and it is one of the options for cognitive radio transceivers.This thesis focuses on intelligent decision method of CR; physical technologysinclude TDCS, interference detection technology and some other data transmissiontechnologys. With the ability of sustaining sensing external environment such asjamming, business change, as well as both learning mechanism and parameteroptimization strategy, parameters of the intellilgent Cognitive Radio Engine would bedynamically adjusted, in order to balance system bit error rate (BER), signal power anddata throughout of the system.There are four parts in this thesis:The first part introduces physical technologies, and the most important one isOFDM-based TDCS. The BER formula of OFDM-based, difference coded TDCS underAWGN channel is deduced, and it is nicely supported by Matlab simulation results.Because frequency points with jamming are automaticly abandoned from TDCS system,so, as well as the spectrum Mask is right set, TDCS performance under jammingenvironment is very closed to performance under AWGN channel.This thesis alsointroduces a kind of jamming detection algorithm called FCME, and two other datatransmission technologies: difference OFDM technology and fast frequency hopping(FFH).The second part introduces three evolutionary algorithms for parameteroptimization: basic genetic algorithm (GA), simulated annealing based GA (SAGA) andbinary particle swarm algorithm (BPSO). This part describes the theory andimplementation steps of these three algorithms, and a group of tests for DeJong functionand TDCS system has been carried out. For DeJong function, all of three algorithmswill convergence to the best solution efficiently. Performance of basic GA In a good configuration is satisfactory, and performance of SAGA is more stable, but it is alsomore complex. BPSO runs fastest, but it is easily influenced by inertia weight and thelearning factor. For TDCS, There is a small probability that basic GA and SAGA can notconvergence to the optimal solution, and BPSO is outstanding: fewer convergenceiterative loop, shortter search time, and excellent global convergence performance.The third part proposes a rule-based intelligent CR system, which supports fourkinds of business: Instant message, file, audio and video transmission. One of the threedata transmission technologys, TDCS, FFH or OFDM would be choosed intelligently,according to the jamming and business types. Rule-based intelligent decision system iseasy to implement, but the effect of signal power is not taken into account.And it couldnot make the optimal decisions in some interference environment, because the systemdoes not have a learning engine.Based on above experiences, the forth part designs an intelligent CR system basedon BPSO and CBR, to adapt to the interference environment better. With Jammingdetection module calculating the jamming type and power, the system can dynamiclychoose a proper physical transmission technology, a proper signal power and a properanti-jammin methods, to balance three objectives followed: the minumize BER, theminimize power and the maximum data rate.The intelligent decision system for cognitive radio in this article has both referenceend Practical value. It provides some new ideas for further research in intelligentdecision system design.