DANCER: The routing algorithm in delay tolerant networks based on dynamic and polymorphic combination of dimensions and energy consideration

摘要

Be differ from traditional networks, the nodes in delay tolerant networks are connected intermittently and they are mobile continuously, which result in high transmission delay and low delivery ratio. In addition, the energy of the mobile node is finite. Due to these constraints, efficient routing in delay tolerant networks is a tough task. In order to achieve higher routing efficiency in delay tolerant networks, we propose DANCER in this article, which utilizes a binary topology for more accurate and energy-efficient routing in delay tolerant networks. First, inspired by the concept of dimension in general relativity; the distance is diverse in different dimensions. Each node builds its own topology and the nodes in each topology are ranked in the top s in multiple dimensions. The link weight, which reflects the packet delivery ability between the two nodes in the topology, is defined by the reciprocals of ranking values of the two nodes in multiple dimensions and the reciprocals of important coefficients of each dimension. Second, due to each node has limited energy, Nash equilibrium solution is utilized to simplify the topology to determine which links are useful for a specific packet relaying. After simplification, each node obtains the binary topology which consists of links where flag bits are respectively 1 or 0. Finally, based on the binary topology, each node could decide the suitable routing for every specific packet sending. The result of the simulation demonstrates that DANCER outperforms benchmark routing algorithms in the respect of average delay, packets hit rate, and routing cost.