In our simulation study, we only consider inaccuracy
of the TDoA measurement at the sensors (ti??™s), since ??ti??™s play
the same role. Such inaccuracy is modeled as a normal distribution in the
simulation.
We will evaluate the e?®ectiveness of TPSS. First, we want to study the
percentage of sensors whose locations can be resolved while varying the number
of initial beacons. We consider a network with 300 nodes. Figure 5 reports
the results for the first nine epochs.We can tell that the percentage of resolved
nodes increases as the number of the initial beacons increases. This also holds
true as the number of epochs increases. Another observation is that the more
the initial beacons are deployed, the less epochs TPSS will require to achieve
a high percentage of the resolved nodes in the network. Second, we test the
impact of network density on the localization process. Figure 6 illustrates the
percentage of resolved sensors when the number of initial beacons varies under
di?®erent network density. The number of epochs is set to 10. It shows that
as the network density increases, more and more sensors get localized. This
is reasonable. Given a fixed number of beacons deployed in a fixed-sized network,
the more sensors within the network, the more nodes can be covered by
a three-beacon group. Therefore an increase of the number of sensors will not
require the number of beacons to be increased as well, as long as the existent
186 Fang Liu et al.
Pages:
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315