Font Size: a A A

Research On Key Technologies Of In-network Caching-based Cooperative Transport Issues

Posted on:2017-03-14Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y Y GaoFull Text:PDF
GTID:1108330482987049Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
With Internet scale spreading out, various new applications and requirements emerge, the location-based end-to-end communication model could not meet these new requirements. For example, in the content/information-centric network, the traditional five-tuple end-to-end transport model could not meet the new architecture. In content/information centric network, named content replace the Internet protocol address and in-network caching among the universal network nodes has changed the traditional end-to-end model. Efficient transport schemes are the fundamentals of network performance, traffic management and quality of experience for users. This dissertation focuses on researching in-network caching-based transport schemes. The primary contributions are as following:1. In-network caching-based end-to-end model is constructed. This mathematical model could shed a light on the relationship between transport success possibility and network parameters. Five-tuple transport model is analyzed as a comparison, the simulation results show that the in-network caching end-to-end transport is better than the traditional one in the aspects of transport success possibility, network traffic load, node responses. Simulation on the different traffic groups provides the basis of designing and improving in-network caching cooperative transport schemes.2. A hierarchical sensing algorithm-based cooperative transport scheme is proposed. Firstly, we define a new set of caching node metrics, which carries the topology-aware information for caching placement and replacement. Node metrics include neighboring metric, centric metric and weight metric. Transporting content based on the node metrics would make the corresponding faster, improve the network resource utility efficiency and extend the caching time-to-live. Caching hit ratio, network traffic load and latency for content access is evaluated for this transport scheme. Secondly, based on the proposed sensing algorithm, a mini-set caching method is designed to support the cooperative transport scheme. Simulation results show that the mini-set caching transport scheme could greatly improve network performance.3. A Flash crowd events prediction-based cooperative transport control scheme is proposed. Firstly, we propose a Pending interest table assisted prediction algorithm, which could tell the possibility of flash crowd events previously to rapidly adjust the content caching and transport-related metrics. Simulation results show that the prediction could provide better performance in aspects of cache hit ratio, delay and traffic load. Based on the prediction algorithm, a first-hop determine congestion control method is proposed to avoid network congestion, it is based on the cooperation of the first node control and caching node adjustment. Simulation results show that this method could improve the network robust and improve the network performance within flash crowd events occur in the network.4. A connection identifier mapping cache-based cooperative transport scheme is proposed. Connection identifier is brought from the universal identifier network, which could conceal the five-tuple connection detail in the backbone network, thereby enhancing the network privacy and security. The CID mapping caching and mapping management cooperatively support the CID mapping-based transport scheme, which provides a distributed and rapid content transport. Simulation results show that the connection identifier mapping-based cooperative transport scheme is feasible and effective in the aspects of identifier uniqueness and efficiency.
Keywords/Search Tags:Content Centric Networking, Universal Identifier Network, In-network Caching, Transport Model, Cooperative Transport Schemes, Transport Control
PDF Full Text Request
Related items