| Popis: |
Ad hoc wireless sensor networks are emerging as an important technology for applications such as environmental monitoring, battlefield surveillance and infrastructure security. Centralized processing in sensor networks works well for small-scale systems, but for large systems, it becomes time-intensive, inefficient, insecure and non-scalable. In contrast, in-field or in-network distributed processing results in faster detection of phenomena, faster query response, and a scalable, robust network. Since the computation and communication in distributed networks is local, the network becomes more energy efficient and the communication cost is also lower. While most research so far has focused on the network aspects of these systems (e.g., routing, scheduling, etc.), the capacity for scalable, in-field information processing is potentially their most important attribute. Networks that can infer the phenomenological structure of their environment can use this knowledge to improve both their sensing performance and their resource usage. These intelligent networks would require much less a priori design, and be truly autonomous. This thesis presents a distributed algorithm for inferring the global topological connectivity of an environment through a simple self-organization algorithm based on Hebbian learning. The application considers sensors distributed over an environment with a network of tracks. Vehicles of various types move on these tracks according to rules unknown to the sensor network. Each sensor node infers the local topology of the track network by comparing its observations with those from neighboring sensors, using a method similar to Hebbian learning in neural networks. The complete topology of the network emerges from the distributed fusion of these local views. The system’s performance is evaluated based on its similarity to the actual underlying network. Ongoing research focuses on using the inferred topology for intelligent scheduling of nodes to enhance network lifetime without loss of performance. |
