Document details

Content storage and retrieval mechanisms for vehicular delay-tolerant networks

Author(s): Silva, Bruno Miguel Correia

Date: 2010

Persistent ID:

Origin: uBibliorum

Subject(s): Content storage and retrieval; Cache; Web caching; Mobile ad-hoc networks caching; Vehicular ad-hoc networks caching; Delay-tolerant networks caching; Vehicular Delay-Tolerant networks


Vehicular delay-tolerant networks (VDTNs) were proposed as a novel disruptive network concept based on the delay tolerant networking (DTN) paradigm. VDTN architecture uses vehicles to relay messages, enabling network connectivity in challenging scenarios. Due to intermittent connectivity, network nodes carry messages in their buffers, relaying them only when a proper contact opportunity occurs. Thus, the storage capacity and message retrieving of intermediate nodes directly affects the network performance. Therefore, efficient and robust caching and forwarding mechanisms are needed. This dissertation proposes a content storage and retrieval (CSR) solution for VDTN networks. This solution consists on storage and retrieval control labels, attached to every data bundle of aggregated network traffic. These labels define cacheable contents, and apply cachecontrol and forwarding restrictions on data bundles. The presented mechanisms gathered several contributions from cache based technologies such as Web cache schemes, ad-hoc and DTN networks. This solution is fully automated, providing a fast, safe, and reliable data transfer and storage management, while improves the applicability and performance of VDTN networks significantly. This work presents the performance evaluation and validation of CSR mechanisms through a VDTN testbed. Furthermore it presents several network performance evaluations and results using the well-known DTN routing protocols, Epidemic and Spray and Wait (including its binary variant). The comparison of the network behavior and performance on both protocols, with and without CSR mechanisms, proves that CSR mechanisms improve significantly the overall network performance.

Document Type Master thesis
Language English
Advisor(s) Rodrigues, Joel José Puga Coelho
Contributor(s) Silva, Bruno Miguel Correia
facebook logo  linkedin logo  twitter logo 
mendeley logo