|
One of the most important design issues in wireless sensor networks is energy efficiency. Data aggregation has
significant impact on the energy efficiency of the wireless sensor networks. With massive deployment of sensor nodes
and limited energy supply, data aggregation has been considered as an essential paradigm for data collection in sensor
networks. Recently, distributed source coding has been demonstrated to possess several advantages in data aggregation
for wireless sensor networks. Distributed source coding is able to encode sensor data with lower bit rate without direct
communication among sensor nodes. To ensure reliable and high throughput transmission with the aggregated data, we
proposed in this research a progressive transmission and decoding of Rate-Compatible Punctured Convolutional (RCPC)
coded data aggregation with distributed source coding. Our proposed 1/2 RSC codes with Viterbi algorithm for
distributed source coding are able to guarantee that, even without any correlation between the data, the decoder can
always decode the data correctly without wasting energy. The proposed approach achieves two aspects in adaptive data
aggregation for wireless sensor networks. First, the RCPC coding facilitates adaptive compression corresponding to the
correlation of the sensor data. When the data correlation is high, higher compression ration can be achieved. Otherwise,
lower compression ratio will be achieved. Second, the data aggregation is adaptively accumulated. There is no waste of
energy in the transmission; even there is no correlation among the data, the energy consumed is at the same level as raw
data collection. Experimental results have shown that the proposed distributed data aggregation based on RCPC is able to
achieve high throughput and low energy consumption data collection for wireless sensor networks
|
