Font Size: a A A

Design And Analysis Of Content Delivery System At The Mobile Edge Network

Posted on:2019-02-07Degree:DoctorType:Dissertation
Country:ChinaCandidate:J P JiaoFull Text:PDF
GTID:1488305705486244Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
The surge of content-oriented communications leads to an unprecedented growth of data traffic and imposes great challenges to existing communication networks,particularly the mobile wireless communications network.Unlike conconventional content delivery solutions that mainly focused on the core network,content delivery solutions based on the mobile edge network has recently emerged,as a promising solution.The mobile edge network consists of the radio access network and the backhaul network.It is a widely distributed communications infrastructure that can be embedded with extra computing and caching resources.This dissertation adapts a new design paradigm that aims to jointly optimize the usage of the 3C(communications,computing,and caching)resources embedded in the mobile edge network to address the problem of massive content delivery.The main contributions of this dissertation are as follows.1)The problem of optimal caching facility placement over the mobile edge network is investigated.The problem is formulated as a(p+1)-median problem,which is a non-trivial extenstion of the classic p-median problem.The problem is NP-hard,so that two heuristic caching facility placement algorithms are propoposed to find the close-to-optimal solution.The first algorithm is suitable for edge networks with unidirectional-links and its design considers the homogeneity of user demand distribution.The second algorithm is suitable for edge networks with bidirectional links and adapts a tabu-search-based strategy.Both algorithms are shown to give near-optimal results with polynomial-time complexities.2)A content delivery scheme that jointly leverages multicasting and proactive caching is proposed.By taking the predicted user demand as prior information,the scheme determines how to optimially cache content at two levels:the user terminal level and network node level,so that the total communication cost of content delivery can be minimized.A novel analytical framework is proposed to theoretically characterize the tradeoff among cache,computing,and communication resources.The framework defines cache resource as the number of content replicas,computing resource as the scale and accuracy of user demand predicion,and communicaiton resource as the number of communication hops.It is showed that this framework yields a closed-form formula for the 3C resource trade-off relationship in mobile edge networks with a symmetric tree topology.Such a closed-form formula can serve as the bounds for mobile edge networks with general topologies.3)A joint content caching and streaming scheme over a communication network with two service tiers is proposed.The scheme jointly leverages the buffering,proactive caching,and communication service-tier scheduling methods to minimize the total cost of content streaming.Optimal algorithms are derived for the joint buffer-scheduling policy and proactive caching policy.The former yields an algorithm with closed-form solution,while the latter is shown to be an extended water-filling algorithm.Simulation results show that the proposed scheme can effectively reduce the cost of content streaming with medium-to-low traffic loads.In summary,this dissertation proposes three effective content delivery schemes that jointly leverage the 3C resources in mobile edge networks for performance enhancement and cost reduction.Our research results reveal the theoretical performance bounds of the proposed schemes and provide useful guidelines for practical system designs.
Keywords/Search Tags:Mobile edge network, content delivery, content streaming, proactive caching
PDF Full Text Request
Related items