Beruflich Dokumente
Kultur Dokumente
directional, multi-directional, omnidirectional. Coverage of each sensor is determined by the kind of sensing. Omni directional sensors - acoustic or seismic the coverage can be assumed as a 2D-Gaussian envelope.
minimal energy consumption. The sensor placement is in-deterministic The sensor however are not dynamic enough to assume a deterministic position to assume maximum coverage. Thus the problem now works around three issues of sensor field under-covered, aptly covered, over covered. Each case redundancies still exist due to placement.
theoretic techniques Voronoi Diagrams and graph search algorithms Centralized approach Assumes a central command centre. Optimal polynomial time algorithm for coverage in sensor network Converts continuous geometric problem into discrete graph problem
Algorithm
nodes into mutually exclusive sets where the members of each of these sets of nodes completely monitors the given area. Only one such set is active at any moment and consumes power the other set is activated when the first one is deactivated. Assumes isotropic circular sensing systems.
sensor resource management Represents sensor field as grid (2 or 3dimensional) and works on deterministic placement of the senor nodes. The algorithm places each sensor on a grid point, one sensor at a time greedy heuristic. Comparison is done between random placement Vs their deterministic PLACE_SENSORS algorithm
Discussion
The Voronoi Tessellation and Simulated
annealing will provide good result but will have to little to offer in the context of distributed self-organized networks. Computational ability is also a concern. This requires a more real time and distributed algorithm for coverage issue.
capability leading to a 2D-Gaussian. Establish a cluster head and allow each node initially to beacon its location obtained from the GPS to the cluster head Produce a image map at the cluster head to represent the deployed sensor fields Gaussians and look for black patches and bright patches on the Image. Obtain the maximum likelihood between sensors based on the probability density function. Fix a threshold ( p(x,y) > 0.70 ) to shutdown the sensor since the sensors are likely to cover the same area of the sensor field.
Under-represented Coverage
Over-represented Coverage
sensor field coverage is shown. The coverage map technique promises to decrease redundancy. Different sensor modalities should be considered and subsequently correlation factor should be observed. Efficient physical level node scheduling scheme for energy consumption
References