Effective Hybrid Compressive Sensors Using Wireless Networks in Clustering Methods

Compressive sensing (CS) reduces the number of data transmissions and also balances the traffic load throughout the networks. However, the total number of transmissions for data collection by using pure compressive sensing is still large. The hybrid method of using compressive sensing technique was proposed to decrease the number of transmissions in the sensor networks. The previous works also used the Compressive Sensing technique on the routing trees. Proposed a clustering method which uses hybrid compressive sensing for the wireless sensor networks. The sensor nodes are arranged into clusters which means the group of nodes. Inside the cluster, nodes transmit the data to cluster head (CH) without using compressive sensing technique. The two levels of transmission in clustering method using hybrid CS technique are: Intra cluster transmission that do not use CS technique and Inter cluster that uses CS technique. The data size is same in both the methods. Reducing the number of data transmission can decrease the energy consumption of the sensor nodes. The sensor nodes are independently and uniformly distributed in sensor field. The separation between sensor hubs in the sensor field is controlled by Euclidian separation which is in correspondence range. Sensor hubs gather the information intermittently and transmit to sink through multibounce with less number of transmissions utilizing grouping technique. Sensor information in the se nsor systems has transient or spatial connection. The corresponded information is as wavelet space or Fourier change area. Proposed model explains the relationship among the size of clusters and number of data transmissions in the hybrid CS method which aims of calculating the optimal size of clusters that can lead to minimum number of data transmissions. Proposed a centralized clustering algorithm based on the results obtained from the analytical model. The proposed methodology aims at using information of the node distribution and node location to enhance a clustering method which uses hybrid CS for sensor network.