Resource Optimization Of High Frequency Access Network In Large Area

In HF communication,the natural relay of the ionosphere is utilized,which provides many advantages,e.g.strong survivability,long transmission distance,simple and flexible realization,short construction time and low cost.Various new technologies,including adaptive HF link establishment,adaptive HF channel equalization,adaptive HF modulation and demodulation,wide-band HF technology,and network technology,have been developed and adopted to address the problems in HF communication,such as low transmission rate and poor transmission quality.HF communication remains widely used in the fields of military communications,emergency rescue,and maritime communications,despite the fact that satellite communications have been developing rapidly.With the development of broadband technology and networking technology in HF communication,the inadequacy of resources such as frequency are becoming more and more severe.We optimize the locations of stations for HF access networks in large areas with limited resources according to the change of ionospheric conditions,in order to optimize the assignment of frequency resources,and thus to improve the coverage of the area by HF networks,providing the mobile HF radio users with garenteed communication and access services at any time,in any case and anywhere in the area.This thesis focuses on the resource optimization HF access network in large area.The main contributions of this thesis are as follows:(1)In order to verify the reliability of ITS forecast data,the correlation analysis between forecast data based on ITS model and data from actual communication is carried out,the results show that the forecast data based on ITS model is credible,and further more the communication threshold of the forecast data is confirmed.Through experiments,we analyze the relationship of the cell side length,the leakage area and the computing time of the area,and appropriate cell size is determined.In order to solve the problem of large computation load due to the implementation of ITS prediction,which was the basic data for HF network frequency optimization,a method for rapid calculation of station coverage data based on dynamic overlay template is proposed by using the similarity of area covering shapes of adjacent HF stations.Firstly,through a number of representative stations coverage calculations,the general station coverage template of current channel conditions is obtained,then the template matching to station location is made,calculation of the regional coverage is defined in the new template region,reducing the data calculation,and increasing the calculation speed of HF stations coverage.(2)In order to improve the effectiveness and efficiency of HF frequency assignment,mutual information coefficient was defined for measuring similarity of coverage of adjacent frequency.Furthermore,the ant colony pheromone was released to the assigned frequencies of stations according to their coverage performance,and the ant colony pheromone was diffused to an unassigned frequency according to the coverage similarities of its adjacent assigned frequency,described by mutual information coefficient.The proposed method has the same effects as increasing the amount of ants to improve the probability of finding the optimal frequency assignment,leading to a better solution within shorter time.Experimental results show that the proposed algorithm achieves high effectiveness and efficiency in HF frequency assignment.(3)A preference ranking elimination non-dominant sorting genetic algorithm II(NSGAII)is proposed to deal with the time-consuming issue of the preference NSGAII algorithm in optimizing HF network frequency assignment in multi-areas outstanding coverage.The proposed algorithm sorts and eliminates solutions according to their preference evaluation prior to the non-dominate sorting.By eliminating solutions with low ranking,the number of solutions participates in non-dominate sorting is reduced.The calculation time and the probability of selecting low ranking individuals for crossover or mutation are both decreased.The proposed algorithm simultaneously achieves the best performance and least calculation time in 38 of 48 sets experiments.Constrained with the same iteration number,the proposed algorithm saves 27% of computation time against the preference NSGAII algorithm.Experimental results show that by adopting preference evaluation sorting,the proposed algorithm takes less time and obtains a better solution.(4)The problem is described as a bi-level optimization problem with super multiple objectives,to find an optimal station location scheme,which have the best coverage under all channel conditions.An improved bi-level non-dominant sorting genetic algorithm III-mutual information ant colony system(NSGAIII-MIACS)algorithm is used to solve the problem.In the process of solving the problem,first of all,using the orthogonal design method to reduce the dimension of the channel conditions;Then,a new local search mechanism of station location is proposed,when local search is carried out,the running times of the underlying algorithm can be reduced by using the lower layer solution information,so that the speed and effect of the algorithm can be improved.The information entropy of the population be used to control station selection and replacement to ensure the diversity of solutions and convergence of equilibrium;Once more,proposes an adaptive preference weighted NSGAIII algorithm to solve the problem of low efficiency,which due to the NSGAIII algorithm reference point too much and the Pareto front solution has a lot of decision makers will not consider.According to the weight of the preference,the reference points are eliminated and the number of reference points is reduced.Thus,the solution time of the algorithm is reduced and the penetration of the algorithm is enhanced.Finally,Different from the random selection of the crossover and mutation of NSGAIII algorithm,a generation mechanism is proposed to strengthen the cross population generation between the adjacent solutions,so as to enhance the penetration of the algorithm.The optimal station location scheme is given by the proposed Local Search-Adaptive Preference NSGAIII-MIACS(LS-AP-NSGAIII-MIACS)algorithm,compared with the optimal station location schemes given by NSGAIII-MIACS and NSGAII-MIACS algorithm,and the scheme of station uniform distribution,The coverage effect on multiple channel conditions is optimal of 25 channel conditions,and the algorithm in this paper is efficiency than the NSGAIII algorithm.Through the coverage optimization experiment with 25 additional channel conditions,the optimal location scheme selected by this algorithm has obvious advantages.It is shown that the channel condition represented by orthogonality can represent the whole distribution of channel condition.The method presented in this paper is practical and effective for the location optimization of HF network stations.