【摘要】 提出了一种基于马尔可夫链的轻量级机会路由转发策略(MOR),该方法将网络运行划分为多个等距的时间周期,利用活跃度描述节点每个周期的随机相遇状态,多个连续周期的状态序列构成一个离散的马尔可夫链,将节点活跃度评估转换为状态预测问题,利用马尔可夫模型预测相遇节点后续周期的状态,有效地提高了节点活跃度评估的准确性。构建了基于节点活跃度和平均相遇间隔的综合效用评估方法,作为消息转发决策的依据,节点只需维护前一周期的状态和状态转移概率矩阵,以及记录与其他节点平均相遇间隔的向量表,算法简单高效,时空复杂度低。建立了多指标的最优消息副本设置方法,有效地均衡了网络资源的使用率。实验结果表明,与已有算法相比,MOR算法能够有效地提高消息传递的成功率,降低平均转发时延,且具有较低的网络交付代价。更多还原

【Abstract】 A lightweight opportunistic routing forwarding strategy(MOR) was proposed based on Markov chain. In the scheme, the execute process of network was divided into a plurality of equal time period, and the random encounter state of node in each time period was represented by activity degree. The state sequence of a plurality of continuous time period constitutes a discrete Markov chain. The activity degree of encounter node was estimated by Markov model to predict its state of future time period, which can enhance the accuracy of activity degree estimation. Then, the method of comprehensive evaluating forwarding utility was designed based on the activity degree of node and the average encounter interval. MOR used the utility of node for making a routing forwarding decision. Each node only maintained a state of last time period and a state transition probability matrix, and a vector recording the average encounter interval of nodes. So, the routing forwarding decision algorithm was simple and efficient, low time and space complexity. Furthermore, the method was proposed to set optimal number of the message copy based on multiple factors, which can effectively balance the utilization of network resources. Results show that compared with existing algorithms, MOR algorithm can effectively increase the delivery ratio and reduce the delivery delay, and lower routing overhead ratio.更多还原