The effect of detection and restoration times for error recovery in communication networks

Abstract

Detection and restoration times are often ignored when modeling network reliability. In this paper, we develop Markov Regenerative Reward M o dels (MRRM) to capture the e ects of detection and restoration phases of network recovery. States of the MRRM represent conditions of network resources while state transitions represent occurrences of failure, repair, detection and restoration. Reward r ates, assigned to states of the MRRM are c omputed b ased o n a p erformance m o del that accounts for contention. We compare our model with ones that ignore these parameters and show signi cant di erences, in particular for transient measures.

Distributions on level-sets with applications to approximation algorithms

“…Our problem with arbitrarỳ i , in addition to its intrinsic interest, is motivated by the following optical networking problem. Given their high data rates (Gigabits/second and Terabits/second), a key requirement in optical networks is fast restoration from node/edge failures [18,25,10,24,9]. Many restoration strategies have been studied/deployed.…”

Modeling the time varying behavior of mobile ad-hoc networks

“…A weakness of most of the literature on the performance of mobile ad-hoc networks is that steady state analysis techniques are used even though transient or nonstationary periods will occur in the network, especially after a link or node failure. The importance of this transient behavior after failures has been illustrated in several wired network technologies including circuit switched [10] networks, packet switched data networks [11] and in packet based signaling networks [12]. This work taken together shows that the dominant factor on network performance after a failure is this transient or nonstationary congestion period.…”

A time dependent performance model for multihop wireless networks with CBR traffic

“…The importance of this transient behavior after link/node failures has been illustrated in several network technologies including circuit switched networks [4], packet switched data networks [5], packet based signaling networks [6], cellular networks [7], and MANETs [8]. This work, taken together, demonstrates that the dominant factor on network performance after a dynamic event such as a link failure is the transient or nonstationary congestion period.…”