SASHIMI: Secure Aggregation via Successively Hierarchical Inspecting of Message Integrity on WSN

Aggregation schemes for reducing transmission cost have been proposed for wireless sensor networks for a long time. Aggregated results can be easily altered by adver- saries since sensors are prone to being captured in a harsh environment. Hence, several secure data aggregation schemes have been proposed to solve this problem. Many schemes ensure data integrity during aggregation procedures, but most of them are post-active since integrity can only be confirmed after the data reaches the base station. Another limitation is that the network topology is assumed to be fixed. However, this assumption violates the characteristic of sensor networks. In this paper, we present a secure data aggrega- tion scheme called SASHIMI. SASHIMI utilizes successively hierarchical inspecting of message integrity during aggregation. If attacks arise during aggregation, attacks can be detected within two levels of the hierarchal tree structure. In other words, penalty and overhead caused by attacks can be reduced. In average, SASHIMI incurs only O(n) communication cost where n is the number of nodes. In the case of attacks, SASHIMI performs better than existing schemes. Moreover, SASHIMI supports dynamic network topology. Finally, a comprehensive analysis demonstrates that SASHIMI is more secure and efficient than other schemes. keyword: Wireless Sensor Network; Data aggregation; Data integrity