Date of Award

Fall 2016

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Computer Engineering - (Ph.D.)

Department

Electrical and Computer Engineering

First Advisor

Abdallah Khreishah

Second Advisor

Nirwan Ansari

Third Advisor

Roberto Rojas-Cessa

Fourth Advisor

MengChu Zhou

Fifth Advisor

Ashutosh Dutta

Abstract

Content Caching at intermediate nodes, such that future requests can be served without going back to the origin of the content, is an effective way to optimize the operations of computer networks. Therefore, content caching reduces the delivery delay and improves the users’ Quality of Experience (QoE). The current literature either proposes offline algorithms that have complete knowledge of the request profile a priori, or proposes heuristics without provable performance. In this dissertation, online algorithms are presented for content caching in three different network settings: the current Internet Network, collaborative multi-cell coordinated network, and future Content Centric Networks (CCN). Due to the difficulty of obtaining a prior knowledge of contents’ popularities in real scenarios, an algorithm has to make a decision whether to cache a content or not when a request for the content is made, and without the knowledge of any future requests. The performance of the online algorithms is measured through a competitive ratio analysis, comparing the performance of the online algorithm to that of an omniscient optimal offline algorithm. Through theoretical analyses, it is shown that the proposed online algorithms achieve either the optimal or close to the optimal competitive ratio. Moreover, the algorithms have low complexity and can be implemented in a distributed way. The theoretical analyses are complemented with simulation-based experiments, and it is shown that the online algorithms have better performance compared to the state of the art caching schemes.

Share

COinS