Study on Corporate Strategy in Information Communication Industry by Dynamic Game Approach

随着网络业务模块化的推进，信息通信行业的市场进入变得更加容易，业务竞争更加激烈。在信息通信行业，商业竞争在几个主要的企业集团之间进行，这是很常见的。但是，除了群体竞争之外，我们还可以发现不同公司分布在同一地区或同一国家的分支机构或子公司之间也存在着业务竞争，本研究利用多参数奇异摄动系统模型定义了一个群体动态博弈概念来描述上述现象。该模型由N个低级快速(决策速度)子系统(分支)和一个高层慢子系统(母公司)组成，基于多参数奇异摄动系统模型，从求均衡解的角度研究了(1)零和博弈、(2)纳什博弈和(3)合作博弈。为了找到平衡解，首先需要发展数值方法来求解各种广义交叉耦合多参数代数Riccati方程和Lyapunov方程。我们不断地改进算法，包括递归算法、迭代算法和牛顿法，来计算这些方程的解。因此，我们开发了几种高效的算法来计算具有二次收敛速度的解，并将上述结果应用于研究信息通信行业中的商业竞争，例如网络投资竞争和定价策略。

As the advancement of network business modularity, the market entry in information communication industry becomes easier, and business becomes more competitive. It is common in the information communication industry that business competitions take place among a few major corporate groups. But, besides the group competition, we can also find that business competitions take place among branches or subsidiaries which are distributed in the same area or in the same country by different companies.In this study, we have defined a group dynamic game concept to describe the phenomena described above by using the system model of multi-parameter singularly perturbed systems. This model is composed of N lower lever fast (decision speed) subsystems (branches) and a higher level slow subsystem (parent company).Based on the system model of multi-parameter singularly perturbed systems, we have studied various dynamic games which include (1) zero-sum game, (2) Nash game and (3) cooperative game from the viewpoint of finding equilibrium solutions. In order to find the equilibrium solutions, we need, at first, to develop the numerical approach to solve various kinds of generalized cross-coupled multi-parameter algebraic Riccati and Lyapunov equations. We have continuously improved the algorithms, which include the recursive algorithm, the iterative algorithm and the Newton's method, to calculate the solutions of these equations. As the result, we have developed several efficient algorithms to compute the solutions with quadratic convergence rates.We will apply the results obtained above to study business competitions, for example, network investment competition and pricing strategies, in the information communication industry.