What algorithm is chosen for a particular system will depend on
such factors as the frequency of wakeup events, the node density and practical
constraints such as the availability of a two-radio sensor node platform or of
low-power paging technology. In addition, these factors might be dynamic and
non-uniform in space, which may require an adaptive solution.
214
Chapter 8 Wakeup Strategies in Wireless Sensor Networks
In principle if no paging channel is available, a sender-based asynchronous
solution with a wakeup tone has the least amount of overhead for a sleeping
node. Therefore it is expected to result in the lowest average energy consumption
for a given wakeup delay constraint, if wakeup events are very infrequent.
However, this conclusion only considers a single link. In multi-hop
networks, the total end-to-end setup latency of a path should be considered
rather than that of one hop alone. Staggering the wakeup of consecutive links
has been proposed to reduce this latency [8] [29], and is potentially more e?®ective
in synchronized scenarios. In addition, tight integration with routing [34]
or topology management [13] can further lower the average power consumption.
Finally, the wakeup delay of each link in the network does not need to
be equal. Since some nodes are more energy-critical than others, it would be
beneficial to have them in a deeper sleep state, while ensuring the networkwide
end-to-end latency is still bounded.
Pages:
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355