Design And Implementation Of Bio-inspired Routing Mechanisms For ICN

Internet has become the basic infrastructure in information society.However,current TCP/IP architecture is facing many challenges,such as mobility,scalability and security,etc.As a kind of clean slate solutions,Information-Centric Networking(ICN)has become a research focus.Taking advantage of the biological feature,which is described as "performing local observations and interactions with others to collaboratively produce effective and efficient operational patterns toward the goals of the entire group",this thesis designs distributed bio-inspired routing mechanisms to solve the complex routing problems in ICN.The bio-inspired routing mechanisms put forward in this thesis are Cat Swarm Routing(CSR)and Cell Chemotaxis Routing(CCR).The main work of this thesis is described as follows:Firstly,we build a network model,including a link model and an intelligent node model which imitates biological behaviors.State perceiving and information exchanging of intelligent nodes are achieved according to the Reaction Diffusion,which is a kind of self-organization phenomena in living creature.Then,routing request is depicted and mapping relationship between Quality of Experience(QoE)and Quality of Service(QoS)is builded.Users' QoS satisfaction is calculated afterwards.Secondly,routing request is forwarded to find content for users based on the face chosing strategies derived from CSR and CCR.(1)CSR is a kind of multipath-alternative routing mechanism,in which forwarding probabilities of the faces are calculated based on users' QoS satisfaction of corresponding links.Then,forwarding faces are chosen respectively based on Tracing Mode Face Choosing(TMFC)and Seeking Mode Face Choosing(SMFC).After multiple content sources are found and multiple paths are generated,the path with the highest users' QoS satisfaction will be selected as the final path.In addition,a migration strategy to balance node load is designed in CSR.This strategy considers the occupancy rate of each node Content Store(CS)and the link utilization among the nodes.(2)CCR is a kind of single-path routing mechanism,imitating the phenomenon that each cell can detect local values of different morphogens to determine its location in the embryo development phase.Concentration of content hormone,residual capacity of links and node load are first defined,and further used to calculate the link concentration value.Forwarding face is then chosen based on Cell Chemotaxis Face Choosing(CCFC).In addition,a backtracking strategy is designed in CCR to improve the routing success rate.Finally,the designed routing mechanisms are simulated and evaluated on the topologies of CERNET and CERNET2.The results show that bio-inspired routing mechanisms are superior to the benchmark mechanism from the perspective of routing success rate,average routing hops and caching hit rate.Hence,mechanisms designed in this thesis are feasible and effective.