Font Size: a A A

On low power self-organizing sensor networks

Posted on:2001-05-19Degree:Ph.DType:Dissertation
University:University of California, Los AngelesCandidate:Sohrabi, KatayounFull Text:PDF
GTID:1468390014452188Subject:Engineering
Abstract/Summary:
This dissertation presents the problem of building a wireless sensor network. Low radio range, potentially high node density, and limited energy reservoirs at each sensor node, in addition to the need for small size, impose stringent requirements for low energy and low complexity on algorithmic and hardware design.; Designing efficient wireless communications for sensor networks requires understanding of the propagation medium. In a sensor network, antenna heights are typically low, since sensor nodes are most likely deployed near surfaces, i.e. on the ground or floor. We present the first characterization of near ground RF channel at the 800 MHz–1GHz band. Our experiment indicates that a fourth power law is a suitable model for characterizing the path loss.; Then we propose a method for network link layer self-organization. The Self-organizing Medium Access Control for Sensor-nets (SMACS) enables wireless nodes to form a connected multi-hop network. To reduce energy consumption, radio communication between the nodes must be scheduled, and not contention based. However, traditional methods of scheduling node activities, such as the various well known “link activation” and “node activation” algorithms require some form of network-wide time synchronization. They also require nodes to have information about radio topology of the network in order to avoid packet collisions. The key innovative mechanism of SMACS is that, by use of a hybrid TDMA/FH method which we call non-synchronous scheduled communication, it enables links to be formed and scheduled concurrently throughout the network without the need for costly exchange of global connectivity information or time synchronization.; We also introduce a mechanism for code synchronization on point to point links in wireless sensor networks. The Transmitter Aided Code Acquisition (TACA) method, enables frequency hopping for a packet network, where packet lengths are short. It uses randomly assigned long codes, specific to transmitters, to control frequency hopping on links. The transmitter inserts a sequence of tones between packets, conveying in coded form, the value of the hopping pattern's phase. Only the intended receiver(s) will be able to decode this information. Simulation and analysis results for performance of TACA will be given.
Keywords/Search Tags:Sensor, Network, Low, Wireless, Node
Related items