ECPC: Preserve Downtime Data Persistence in Disruptive Sensor Networks
Mingsen Xu, Wen‐Zhan Song, Deukhyoun Heo, Jonghoon Kim, Byeong-Sam Kim · 2013
In challenging environments, sensor networks may become disruptive even unavailable (downtime), resulting in data losses. In-situ encoding and caching data in a distributed fashion can preserve data persistence during downtime. However, the existing approaches are mainly based on random walks, which incurs significant communication overhead and may itself fail due to network disruptions. In this paper, we present a distributed Erasure Coding with randomized Power Control (ECPC) mechanism to preserve downtime data persistence in disruptive sensor networks. ECPC only requires each node to perform a single broadcast at each of its several randomly selected power levels. Thus it incurs low communication overhead. Moreover the storage space requirement is lower and uniform across the network. It applies randomized power control in localized data broadcast to achieve expected code degree distribution and ensure pseudo-global randomness for erasure coding. Performance comparisons between ECPC and other existing approaches show that ECPC mechanism can reach higher data reliability under varying node failure probabilities. In addition, our ECPC approach is scalable with network sizes.