基于网络编码的分布式无线光网络在线优化

最近, 一种集成骨干光传输网络、无源光网络和无线接入网的混合宽带无线光网络被提出。这种网络具有大带宽、低费用和无处不在的信息接入等特点。考虑在这种网络中的基于网络编码的多播会话的设计问题, 使得网络效用最大化, 而布网的费用最小化。这个问题被转换为一个混合的整数非线性规划问题, 精确求解极其困难。为了使得问题简化, 采用了一种两步优化方法进行求解, 交替地为多播会话选择光网络单元和网关。在每一次迭代过程中需要解决两个问题：光网络的网络编码设计问题和无线网络的用户和带宽分配问题。前者通过基于拉格朗日对偶分解的分布式方法实现; 后者通过广义Benders分解实现。通过仿真验证了所采用的方法的有效性。

Distributed wireless optical network online optimization based on network coding

Recently, the hybrid wireless-optical broadband network integrating optical backbone networks, passive optical networks, and wireless access networks have been proposed to provide the high-bandwidth, low-cost, and ubiquitous communication connections. This paper considered the design of network coding-based multicast applications in such networks with the objective of maximizing the total network utility and minimizing the deployment cost. The problem was formulated as a mixed integer nonlinear programming problem and the exact solution is prohibitively complex. In order to make the problem more tractable, it developed a two-step optimization procedure that iteratively selected the optical network unit and gateways for the multicast sessions. During each iteration, it solved two subproblems :a network coding design problem for the optical network and a user assignment and bandwidth allocation problem for the wireless network. The former was solved in a distributed way based on the Lagrangian-dual decomposition; the latter was solved based on the generalized bender decomposition. Simulation results illustrate the effectiveness of the proposed solutions.