In wireless sensor networks, the resilient data fusion is an effective method to deal with the malicious attack while it is always energy exhausting, and does not consider the fusion function and the attack detection synthetically, which may affect the accuracy of fusion result. A novel information utility measures based resilient data fusion algorithm is proposed in this paper trying to solve this problem. Based on the prevalent clustering method, the algorithm firstly detects whether there is an attack in networks or not, and then computes the information utility measures from which we can evaluate the under-attack level of each cluster. Finally it uses these measures as parameters jointly to calculate the fusion output. Compared with the existing techniques, the new algorithm is more resilient and can upgrade the accuracy and applicability of the fusion function. Simulation results show the effectiveness of the presented algorithm.
[1]
Levente Buttyán,et al.
RANBAR: RANSAC-based resilient aggregation in sensor networks
,
2006,
SASN '06.
[2]
Francesco Flammini,et al.
Wireless Sensor Data Fusion for Critical Infrastructure Security
,
2008,
CISIS.
[3]
Insup Lee,et al.
Resilient multidimensional sensor fusion using measurement history
,
2014,
HiCoNS.
[4]
Pan Quan.
Performance Analysis on Information Utility Function in Wireless Sensor Networks
,
2007
.
[5]
Jemal H. Abawajy,et al.
A Data Fusion Method in Wireless Sensor Networks
,
2015,
Sensors.