The Channel Utilization Related Algorithm Based On The GSM Mobile Communication System

Efficient use of available frequency resources in GSM mobile communicationsystem, calculating the channel utilization and estimating a reasonable level is one ofthe key issues of the current mobile operators to solve. Utilization related algorithmwas studied in this paper, the main algorithms are: the fixed channel allocationalgorithm and busy detection algorithm in the mobile data network.The recent years, Algorithms to solve the problem of channel allocation aremainly heuristic algorithm and swarm intelligence algorithm, the improvedartificial bee colony algorithm was adopted to solve the problem of channelallocation in Chapter four, artificial bee colony algorithm also belongs to swarmintelligence algorithm.The first improved channel allocation algorithm is the improved artificial beecolony algorithm, the step of this algorithm used gradually reduced dynamic step tobalance local and global search capability; the introduction of the selective mutationincreased population diversity and the convergence speed. Simulation results showthat: the improved artificial bee colony algorithm can be better to solve the wirelesschannel allocation, improve the convergence rate and convergence speed ofalgorithm.The second improved channel allocation algorithm is the cooperative hybridartificial bee colony algorithm, the dynamic step was used to increase the searchaccuracy of the algorithm, the dynamic step was cosine rule with the increasingtimes of iteration; The artificial bee colony were divided into doublesubgroups,every subgroup adopted different evolutionary strategy to co-evolve, itwas easy to jump out of local optimal solution in this way; The introduction of theadaptive selective mutation increased diversity of population and the speed of convergence. Simulation results show that: the cooperative hybrid artificial beecolony algorithm can be better to solve the more complex classical channelallocation problem, namely Benchmark problem.Another algorithm of the related utilization is the busy detection algorithm fordata services in the mobile network. The traditional busy detection methods mostlyused statistical methods to estimate the busy time of the voice, after the introductionof data services, busy time of data traffics and voice traffics is different, if we stilladopt traditional methods, not only waste time but also estimation was not accurate.it was not based on the times of observations to use the principle of relativecondition entropy to calculate the value of relative conditional entropy on EDGEnetwork,and then to analyze data, lastly gained busy time. Result showed that: thebusy detection algorithm in the mobile data network can better manage and detectbusy time, and provide theoretical basis for the network optimization.