A reinforcement learning-based routing for delay tolerant networks

Enabling wireless cooperation in user provided networks

“…Although the Delay Tolerant Reinforcement-Based (DTRB) [Rolla and Curado, 2013b] routing solution is explained in detail in Chapter 4, an introduction to this routing solution is given here for the sake of comparison with the routing solutions presented in Subsection 2.1.2. DTRB enables device to device data exchange without the support of any pre-existing network infrastructure.…”

“…DTRB [Rolla and Curado, 2013b] Flooding-based The nodes that recently gossip about the destination of a given UPN data message are more likely to deliver the message PRoPHET Table 2.1: Summary of delay tolerant routing solutions.…”

“…These routing solutions use a store-carry-forward approach to adeptly deliver the message to the destination. Examples of distinguished delay tolerant routing solutions are: PRoPHET [Lindgren et al, 2003], Delay Tolerant Reinforcement-Based (DTRB) [Rolla and Curado, 2013b], and Spray and Wait (SnW) [Spyropoulos et al, 2005].…”

“…The rewarding process is executed by a learning algorithm based on the distances between the nodes, which are calculated as a function of time from the last meetings. DTRB [Rolla and Curado, 2013b] is a flooding-based delay tolerant routing solution. The simulation results show that DTRB can deliver more messages than a traditional delay tolerant routing solution does in densely populated areas, with similar end-to-end delay and lower network overhead.…”

A near-optimal green control for probabilistic routing protocols in delay-tolerant networks

“…Burns et al (2005), while assuming that movements of humans and vehicles are periodic, propose a new predictability concept based on the recorded movement of the node during the last t rounds. In Rolla and Curado (2013), the authors present the delay tolerant reinforcement-based (DTRB) in order to learn about routes in the network and forward the messages that produce the best reward. The rewarding process is executed by a learning algorithm based on the distances between the nodes, which are calculated as a function of time from the last meetings.…”