Blind Equalization Algorithm Based On The Optimization Of RNA Genetic Algorithm

In underwater digital communication system, multiple-path effect and channel distortion are widespread. They are the main factors that cause inter symbol interference, and the presence of inter symbol interference makes communication quality drops greatly, and greatly reduces the data transmission rate and reliability of the underwater at the same time.Therefore, in order to eliminate inter symbol interference and improve the utilization rate of bandwidth, it’s very important to balance the signal at the receiving.Conventional adaptive equalization requires sending training sequences, which makes the poor bandwidth of the underwater more nervous. The blind equalize technology can balance the channel distortion without sending the training sequences.It can save the bandwidth of the underwater communication effectively, so as to improve the communication efficiency.Therefore, RNA genetic algorithm is used to optimize the blind equalization algorithm and explore the optimization algorithm.The main contributions are as follows:(1)Analyzing constant modulus blind equalization algorithm based on the optimization of RNA genetic algorithm.Because the searching method of the optimal weight vector in CMA is a gradient descent method. It’s easy to fall into local minimum.In order to overcome the defects, RNA genetic algorithm is introduced to CMA, using the global search of RNA genetic algorithm to search the globally optimal solution.RNA nucleotide chain coding method is used in this algorithm to represent the possible solutions of the equalizer optimal weight vector.Then the adaptive crossover and mutation operators is used on the encoded RNA chain to search the best individual in the population of RNA, and then decode the best individual, take the output vector decoding as the initial weight vector of CMA.(2) Analyzing Multi-Modulus blind equalization algorithm based on the optimization o f RNA genetic algorithm.The traditional constant modulus blind equalization algorithm just using the amplitude information of the equalizer output signal, without using the phase information.So it can’t eliminate the phase rotation caused by channel characteristics.For the un-constant modulus signal, it has a slow convergence rate and a big mean square error.But the multi-modulus blind equalization algorithm uses the amplitude and phase information of the equalizer output at the same time, solving the problem of phase rotation. But there are still slow convergence, the disadvantage of large steady-state error.So combine the RNA genetic algorithm with excellent global optimization ability with the traditional multi-modulus blind equalization algorithm to get the multi-modulus blind equalization algorithm based on RNA genetic algorithm. The new hybrid algorithm can effectively speed up the convergence rate, reduce the mean square error.(3) Analyzing weighted multi-modulus blind equalization algorithm based on the optimization of RNA genetic algorithm.Although the traditional multi-modulus blind equalization algorithm can effectively eliminate the phase rotation caused by the channel, and can reduce the steady-state crror.in the case of no noise, steady state error MMA algorithm is still exist.The weighted Multi-modulus blind equalization algorithm can solve this problem.The algorithm realize the adaptive revision of modulus value during the iteration process of factor of equalizer.To adjust the modulus in the cost function and have a further use of constellation prior information.The combination of RNA genetic algorithm and weighted Multi-modulus blind equalization algorithm formates a new hybrid algorithm, the algorithm uses the global search ability of RNA genetic algorithm and weighted Multi-modulus blind equalization algorithm performance can be adaptively adjust the modulus, the traditional blind equalization algorithm was further optimized, and the performance is greatly improved.(4) Analyzing frequency domain weighted multi-modulus blind equalization algorithm based on the optimization of RNA genetic algorithm.Frequency weighted Multi-modulus blind equalization algorithm is based on the traditional weighted multi-modulus blind equalization algorithm. The algorithm transform the algorithm from the time domain to the frequency domain with the using of the method of overlap save in the frequency constant modulus,it makes frequency weighted Multi-modulus blind equalization algorithm have a smaller calculation.But FWMMA still requires a continuous derivable cost function, and easy to fall into local optimum. RNA genetic algorithm with good global searching ability is used to iterate and search the optimal weighted vector of equalizer.lt can effectively improve the performance of blind equalization algorithm...