Font Size: a A A

Research Of Content Placement In Satellite Cache Networks

Posted on:2020-07-15Degree:DoctorType:Dissertation
Country:ChinaCandidate:E B WangFull Text:PDF
GTID:1368330602963889Subject:Military communications science
Abstract/Summary:PDF Full Text Request
In the ever-changing information age,with the spread of mobile communication devices such as smartphones and tablets,hundreds of millions of end users have become more and more demanding for network access anytime,anywhere,and at your fingertips.However,the current network infrastructure supporting user access is mainly composed of the ground equipment,which can be easily damaged in natural disasters such as fires,typhoons and earthquakes.Fortunately,with the continuous development of satellite communication technology,the satellite network characteristics,such as widely coverage,high rate,and high reliability and so on,have become increasingly prominent.Therefore,in order to satisfy the service demands of multi-user and multi-service,the construction of satellite terrestrial network(the convergence of satellite network and terrestrial network)has become the mainstream trend of the future network development.Currently,many Internet companies such as Google and Facebook have or plan to launch their own satellites for Internet access.Since satellites can provide a wider range of coverage,satellite terrestrial networks have great potential for delivering services in terms of content(mainly considering application attributes,i.e.,the popularity of content).Different from the traditional terrestrial networks,both the throughput and delay performance of satellite networks are affected by the high dynamic satellite network topology,the intermittently connected satellite links,and the limited satellite transceivers restrict.In fact,in the terrestrial networks such as the Internet,content distribution network(CDN)and information center network(ICN),the efficient utilization of network cache resources can significantly improve network throughput and reduce end-to-end network latency.Note that,network cache means each network element(such as satellite nodes,gateways,and routers)has the ability to cache data.Therefore,with the development of storage technology and the enhancement of on-board processing capabilities,the development and utilization of onboard cache resources has gradually become the key to improve the performance of satellite terrestrial networks.However,there are still many problems of effectively utilizing the satellite storage resources to comprehensively improve the satellite network performances,such as,1.It is difficult to accurately estimate the cache occupancy by using the existing queuing theory methods.2.Due to the high speed mobility of nodes in the satellite cache network,the ground cache network load balancing solution does not apply to satellite cache networks.3.How to design a content placement strategy in a cognitive satellite caching network.In order to solve the above problems,this paper utilizes the theoretical tools such as Martingale process and game theory to study the related strategies and methods of satellite cache network content placement.? Aiming at the problem that the propagation delay of the satellite network is large and the cache occupancy information cannot be transmitted frequently,an encapsulationbased cache occupancy estimation method is proposed.According to the data generation rate,channel access protocol,transmission rate and queuing mode,this method first constructs the corresponding Martingale encapsulation function,and then gives the equivalent arrival Martingale and service Martingale through the system cascading strategy,and finally estimates cache occupancy via the equivalent arrival Martingale and service Martingale.In order to evaluate the estimated performance of the proposed method,the statistical box plot is compared with the upper bound of cache occupancy.The results show that the median of the statistical box plot is almost on the upper bound curve of the cache,hence the changes in the cache occupancy of satellite nodes can be better reflected by the method that using the Martingale encapsulation function to estimate the node cache.? Aiming at the problem of multi-layer satellite cache network load balance,a Stackelberg game model based on low-orbit satellite cache congestion degree and its solving algorithm are proposed.For solving this non-convex game problem,this paper first assumes that everyone participates in the game,thus the non-convex problem is transformed into a convex optimization problem,and then solves the transformed optimization problem.In order to evaluate the load balancing performance of the proposed method,we compare the proposed method with the existing satellite network load balancing method.The results show that the proposed method can significantly reduce the packet loss rate and can upload data(content)to the upper node.To reduce the occupancy of the lower node cache,the lower nodes will have more storage space to store more popular content.The proposed method not only improves the load balancing performance of low-layer nodes,but also increases the efficiency of cache usage.? Aiming at the problem that users download content is delayed in cognitive satellite cache network,a utility function-based content placement method is proposed.Based on the situation that the cognitive satellite cache network presence sensing technology restricts the spectrum sharing effect,and combined with the distribution of the satellite users in the ground,two different transmission methods are designed to improve the cache efficiency of satellite nodes and reduce the download delay of satellite users.In addition,for further improving the satellite nodes' cache efficiency,a content placement method is proposed.According to the content success download probability,this method can determine the optimization ratio of the most popular content MPC and general popular content GPC.Besides,in the sparse user area,this method can optimize the utility function based on the forwarding cost and the storage cost to obtain the proportion of the slice stored on the satellites.
Keywords/Search Tags:cache backlog estimate, node cache congestion control, content placement, satellite cache network
PDF Full Text Request
Related items