Study On OFDM Channel Estimation And Adaptive Resource Allocation Algorithm

Due to its good spectrum efficiency and anti-interference property, Orthogonal Frequency Division Multiplexing(OFDM) becomes hot spot of communication. Channel estimation is one of the key technologies. The contradiction of estimation property and estimation simplicity is the estimation algorithm's key issue to be solved. Meanwhile, existing OFDM system's operation mode in which every subcarrier shares same power, carries equal number of bit, result in low working efficiency. In this paper, the linear minimum mean-squared error(LMMSE) algorithm based on singular value decomposition(SVD) is systematically analyzed and a new low-rank order choose criterion is raised by applying the signal/noise subspace theory. Based on channel estimation, adaptive subcarrier power and bit allocation are studied. Second best approximative power allocation algorithm and bit allocation algorithm based on SNR threshold are researched respectively. All the algorithms are tested by matlab stimulation.The results show that: (1) Because of the Rhh's characteristic, only a few eigenvalue of Rhh obviously bigger than zero, which make the LMMSE algorithm simplified greatly. Meanwhile, due to the utilization of self correlation, the algorithm can take complexity and property into account concurrently. (2) The key factor of LMMSE algorithm has the same essence as signal/noise subspace mode. Taking the dimension of signal subspace as low-rank order choose criterion can improve original algorithm. With the change of the environment, the modified algorithm can adjust its operation parameters to ensure the estimation accuracy. (3) Second best power allocation is the approximation of best allocation. It has the same trend as best allocation, besides that only a few channels with bad quality are allocated with no power. The second best algorithm is proved to have better property than equal-power and equal-SNR algorithm. For a fixed subcarrier modulation system, second best algorithm can minimize the system bit-error-rate under fixed transfer power. (4) Bit allocation algorithm based on SNR threshold chooses the modulation order according to the SNR of each modulation mode under a certain bit-error-rate. That means the channel with good quality would transfer more information. On the basis of each subcarrier's channel quality, the algorithm can realize the system bit-error-rate minimization under the fixed transfer rate. The modified low-rank order choosing criterion strengthens the LMMSE algorithm's environment adaptability. Since that the algorithm take both property and complexity into consideration, it is suit for application. On the basis of accurate channel estimation, adaptive subcarrier power and bit allocation algorithm can improve spectrum efficiency, allocate the resource reasonably and minimize the system bit-error-rate.