Social-context based routing and security in delay tolerant networks

Delay Tolerant Networks (DTNs) were originally intended for interplanetary communications and have been applied to a series of difficult environments: wireless sensor networks, unmanned aerial vehicles, and short-range personal communications. There is a class of such environments in which nodes follow semi-predictable social patterns, such as wildlife tracking or personal devices. This work introduces a series of algorithms designed to identify the social patterns present in these environments and apply this data to difficult problems, such as efficient message routing and content distribution. Security is also difficult in a mobile environment. This is especially the case in the event that a large portion of the network is unreliable, or simply unknown. As the network size increases nodes have difficulty in securely distributing keys, especially using low powered nodes with limited keyspace. A series of multi-party security algorithms were designed to securely transmit a message in the event that the sender does not have access to the destinations public key. Messages are routed through a series of nodes, each of which partially decrypts the message. By encrypting for several proxies, the message can only be intercepted if all those nodes have been compromised. Even a highly compromised network has increased security using this algorithm, with a trade-off of reduced delivery ratio and increased delivery time -- Abstract, page iv