The number of users of Ka band multiple spot beam communication satellite system reaching millions is tremendous relative to C/Ku VSAT system. And the bandwidth efficiency for the system is very low if we used the method for VSAT system. Even though there are new algorithms such as convex programming, 0-1 programming , artificial neural network, genetic algorithm, taboo search, dynamic programming, analytic hierarchy process, gray relative analysis method, dominant strategy equilibrium and strategic equilibrium, cooperative and non-cooperative game etc., wc focus on the communication network. Although these algorithms with different names, the idea is the same: first we build model based on communication network, then we obtain the objective function with the highest rate of convergence or top efficiency or other objective, last we solve the objective function obtain the max or min value. In most cases the value is local optimal solution, and the convergence speed , convergence property are also challenges. In this paper we focus on the users driving force for the network throughout. For the million users the behavior habit of using network can be prediction and this is the idea of big data. According to the idea wc adopt Markov chain to model the prediction model. With the number of users increasing, the performance of the model is optimizing relative to other traditional algorithms. However, the idea of big data made a great success in many field, and it is a worth a try in satellite communicate resources optimization and scheduling.
展开▼
机译：相对于C / Ku VSAT系统，Ka波段多点波束通信卫星系统的用户数量达到数百万。如果使用VSAT系统方法，则系统的带宽效率非常低。即使有新算法，例如凸规划，0-1规划，人工神经网络，遗传算法，禁忌搜索，动态规划，层次分析法，灰色相对分析方法，优势策略均衡与策略均衡，合作与非合作游戏等，wc专注于通信网络。尽管这些算法的名称不同，但是思想是相同的：首先，我们基于通信网络建立模型，然后获得收敛速度最高或具有最高效率的目标函数或其他目标，最后我们求解目标函数以获得最大值或最小值。在大多数情况下，该值是局部最优解，并且收敛速度，收敛性质也是挑战。在本文中，我们将重点放在整个网络的用户驱动力上。对于数百万用户而言，使用网络的行为习惯是可以预测的，这就是大数据的想法。根据这个想法，wc采用马尔可夫链对预测模型进行建模。随着用户数量的增加，该模型的性能相对于其他传统算法正在优化。但是，大数据的概念在许多领域都取得了巨大的成功，值得在卫星通信资源的优化和调度中进行尝试。
展开▼
