Semi-Markov model for delivery delay prediction in multi-copy opportunistic networks with heterogeneous pairwise encounter rates

This work proposes an analytical model based on absorbing semi-Markov process for predicting the expected pairwise delivery delay for single and multi-copy opportunistic forwarding schemes. The framework can be applied to any inter-meeting times distribution. For the exponential case, it only requires the number of nodes and the pairwise encounter rates, which can be completely heterogeneous. Results show that the model predictions are quite accurate, with error values being less than 1% in some cases.