Order Qam Fast Digital Demodulation And Circuit Design

To solve the contradiction of vast communication data and the limited frequency bandwidth, a variety of modern modulation/demodulation techniques are seeking stable capability of high-speed rate. Quadrature Amplitude Modulation (QAM) technique has the advantages of high bandwidth efficiency and noise resistance so that it could meet the requirements of the communication systems.Generally, the transmitted high-order QAM signals would be affected by the multi-path problem. To eliminate the disturbance of multi-path to the signal transmission, the demodulator needs a self-adapt equalizer. When the local clocks of transmitter and receiver are not matched, there is a phase excursion or even a phase rotation. To resolve the unmatched local clocks, the demodulator needs to recovery the carrier wave of the received signals. The traditional self-adapt equalization algorithms that based on the fixed constellation phase can not converge in the conditions of a strong phase jitter. So it is necessary to find an algorithm that is of little correlation of phase. The most common algorithm is Constant Modulus Algorithm (CMA), but its MSE is relatively large after convergence. CMA is only used in the primary equalization. In the same time, before the carrier wave recovery module works stably, the equalizer needs to eliminate the interference (i.e. multi-path). Since the equalization has a large error, the carrier recovery module will spend a long time in converging, or even can not converge. Therefore, it is hard to find a perfect holistic scheme because of the contradiction.Based on researching some demodulation algorithms to solve the problem, the article proposed the CMA-HCMA (HCMA, Hierarchical Constant Modulus Algorithm) dual self-adapt equalization algorithm to solve the above problems, and improved the carrier recovery algorithm. The signals' moment is independent of the signal phase, so the CMA-HCMA equalization algorithm uses the second-order statistics of high-order QAM demodulation signals to compute the equalization coefficients. The algorithm can converge before carrier recovery, and its MSE is small. For the 64QAM, the carrier recovery algorithm in this article can capture the frequency difference of 11 % of the baud rate in 5000 symbol periods. Based on the algorithm's improvement, the system circuit structure could be adjusted as follows: The first circuit module is the symbol synchrocyclotron module, which has strong independence. The next is the self-adapt equalizer module, and the last is the carrier recovery module. Obviously, this circuit structure is clearer and simpler than before. In the condition that the 64QAM signals have passed the cable channel model and have the frequency difference of 11% of the baud rate, the matlab simulation proves that this structure finishes the demodulation in 12500 symbol periods.To prove the above algorithm, the article uses the Verilog HDL to complete the RTL design. At last, a hardware circuit has been designed based on the FPGA of EP2C8Q208C8N produced by Altera. In the platform, the modulation and demodulation algorithms are realized primarily, which proves the algorithms correct.