%0 Conference Paper
%A Cello, Marco
%A Gnecco, Giorgio
%A Marchese, Mario
%A Sanguineti, Marcello
%B 44th Conference of Italian Operational Research Society (AIRO 2014)
%C Como, Italy
%D 2014
%F eprints:3134
%K Intermittently-connected networks; Store-and-forward mechanism; IP networks; Markov chains.
%P 102
%T Average Packet Delivery Delay in Intermittently-Connected Networks
%U http://eprints.imtlucca.it/3134/
%X Delay/Disruption-Tolerant Networking (DTN) is addressed. It is a communication paradigm that enables  the communication over Intermittently-Connected Networks (ICNs), which are characterized by unpredictable  or scheduled contacts among nodes, high latency, and high bit error rates. DTN exploits storeand-forward  techniques in order to cope with intermittent link issues. A model is proposed, to compute the  average packet delivery delay in ICNs. We assume that the inter-meeting time as well as the contact time  between any two nodes is an exponentially-distributed random-variable. As a consequence, the behavior of  the communication between each pair of nodes can be modeled by a two-state Continuous-Time Markov  Chain (CTMC). It is assumed that the packet generation process at the source node follows a Poisson process,  so in the analysis one can exploit the Poisson Arrivals See Time Averages (PASTA) property. Both  the IP-like paradigm used by traditional TCP/IP protocols and DTN are considered. Numerical results and  simulations are presented.