The PAPR Optimization Algorithm And ICI Reduction Algorithm On Cognitive Radio

Wireless spectrum is a limited resource and the management frame of fixed assign spectrum can not satisfy the rapid development of wireless communications. Cognitive Radio (CR) which represents the concept of dynamic spectrum accessing is an efficient method to solve this problem. For designing the architecture of CR system, the centralized architecture based on OFDM has been widely accepted. Wavelet-based Orthogonal Frequency Division Multiplexing (WOFDM) integrates wavelet theory and OFDM. WOFDM can obtain the same perfect performance as OFDM. Furthermore, WOFDM can assign sub-channels more flexible and can eliminate the Inter-Symbol Interference (ISI) and impulse interference and selective interference more efficiently, and WOFDM can satisfy multi-rate signal transmission with different services and different service quality more conveniently. In addition, WOFDM has better bandwidth efficiency than that of OFDM because of no guard interval and pilot. According to these advantages, WOFDM can meet the flexible requirement of CR system much better. In the above-mentioned background, the problem of high Peak-to-Average Power Ratio (PAPR) which must be solved for using WOFDM in CR and the problem of the Inter-Carrier Interference (ICI) which influences the communication quality of licensed user in the Spectrum Pooling (SP) based on WOFDM are studied in this dissertation. The major achievements and results are outlined as follows:1. PAPR optimization algorithm based on assignment problem model is proposed.The efficiency of system radio can be decreased and the performance of system can be worse because WOFDM produces high PAPR. So, how to reduce the high PAPR is a practical problem for applying WOFDM. For the WOFDM system, infinite expanding the level of wavelet filter bank can increase the complexity of radio. If the level of wavelet filter bank is fixed, the number of wavelet packet basis which can be used in system is a constant and the number of sub-carriers can be determined in the system transmission end. It corresponds to linear rules.So, the PAPR optimization algorithm based on the assignment problem model of linear programming is proposed and a set of optimal wavelet packet basis is obtained by the algorithm. The PAPR of system can be reduced efficiently by using optimal wavelet packet basis to modulate signals. Numerical simulation results show not only the PAPR of system can be improved but the Bit Error Rate (BER) performance of system can be ensured and no any side information is required.2. The multi-thread optimizing PAPR parallel algorithm based on multi-core computing platform is proposed.In order to meet the requirement of practical application of CR systems, the PAPR optimization algorithm based on assignment problem model need to be improved. The multi-thread optimizing PAPR parallel algorithm based on multi-core computing platform and realized by the multi-thread concept is proposed to increase the execution speed of original serial optimization algorithm. For the PAPR optimization algorithm based on assignment problem model, if the number of wavelet packet tree layer is fixed, the iterative operation of searching efficient wavelet packet basis satisfies the parallel looping limitation that each iterative operation is independent. So, the new parallel algorithm can be designed with multi-thread theory. Because the new parallel algorithm based on multi-core processor platform executes with multi-thread parallel mode, it can execute faster than original serial algorithm. Numerical simulation results show the multi-thread optimizing PAPR parallel algorithm retains all merits of the PAPR optimization algorithm based on assignment problem model: it can reduce the PAPR of system efficiently and ensure the BER performance of system and need not any side information. Furthermore, the parallel algorithm has good degree of parallelism, good performance of speed-up ratio and parallel processing efficiency. So, the multi-thread optimizing PAPR parallel algorithm can reduce the PAPR of system much faster and efficiently and it is meaningful to practial applying WOFDM for CR systems.3. The spectrum pooling based on WOFDM is proposed and the optimal deactivating sub-channels guard band algorithm is proposed. Spectrum pooling is the most novel model of CR. For the spectrum pooling based on WOFDM, the spectrum of licensed users is separated into several parts and the bandwidth of one subband of the licensed system is matched with an integer multiple of the WOFDM carrier spacing used in the rental system. Furthermore, using the same level wavelet packet basis of wavelet packet tree to modulate sub-channels is needed in order to ensure the interval between sub-channels being same and the allocation vector used to represent the state of sub-channel (occupy/idle) is also needed. After the resource information is obtained by rental users, they transmit the data on the idle sub-channel and at same time transmit data symbol"0"on the occupied sub-channels. So, co-existence of licensed users and rental users in the spectrum pooling can be ensured and the spectrum pooling based on WOFDM can be realized.The prerequisite of applying CR system is the data transmission of licensed users can not be interfered. But the serious ICI occurs because of the WOFDM sub-channel signals of licensed users and rental users co-existing in the spectrum pooling. So, the optimal deactivating sub-channel guard band algorithm is proposed in the dissertation in order to mitigate the interference to licensed users of rental users efficiently. In the spectrum pooling based on WOFDM, after the ICI of licensed users and rental users is analyzed, the sub-channel of rental users which influences the ICI of licensed users most is located and this sub-channel is chosen as the optimal deactivating sub-channel. Then, the optimal deactivating sub-channel is flexibly selected to be deactivated according to the data transmission rate requirement of rental users. So, the ICI of licensed users can be mitigated. Numerical simulation results show the optimal deactivating sub-channel guard band algorithm can mitigate the average ICI energy of licensed users and improve the BER performance of licensed users. So, the new algorithm is meaningful to guarantee the communication quality of licensed users and realize the spectrum sharing efficiently.