Abstract
This study introduces a new concept of using data and storage management techniques to boost the reliability of challenged ad hoc and sensor networks by mitigating the effects of link and graph connectivity outages. Novel distributed uniform and multi-rate storage management approaches are proposed and through proper modeling of the stochastic disconnectivity patterns, network heterogeneity degrees, data resolution thresholds and various network design parameters, this work shows how these techniques can improve the overall network performance. Both analytical and simulation results are presented and various sensitivities are thoroughly investigated to give the network designer solid guidelines.