Distributed average consensus for wireless sensor networks

摘要: Wireless sensor networks have emerged a few years ago, enabling large scale sensing at low cost. There are many interesting problems related to this new tool: designing robust and small hardware, defining adapted routing protocols, minimizing the energy consumption of each component, synchronizing sensors, etc. In thesis, we focus on processing sensed data within network itself. We study specific signal problem, called distributed average consensus. which connected in wireless network, need know all measurements network. Instead gathering central node, would compute broadcast it consensus algorithms offer solution averaging problem. By local message passing iterative computations only, nodes can learn measurements. More precisely, algorithms, iteratively weighted averages that conserve global estimates The node contract until they converge average. Many were designed literature is vast. This thesis starts by classifying existing algorithms. Then describes number useful techniques, handle analysis Preexisting as well revisited with these unifying simple techniques. addition, performance depends topology variety explored: graphs circles or trees, lattices, random geometric graphs, complete Finally, an extension voting derived. particular, show two bits memory reach finite time majority, when initial binary. Distributed majority for ternary quaternary signals also given.