||Recently, due to the rise of smart home systems, the need for border battles and also with|
the low cost and convenience of sensors, the wireless sensing network becomes an indispensable
system. However, since the sensor is exposed in the open space, the possibility of being attacked
or malfunctioning is greatly increased. Furthermore, because sensors are less likely to be checked
and repaired individually across different regions, the fusion center must have the ability to
correctly judge information when the sensor has already been attacked.
This thesis basically considers an under-attack wireless sensor network system. To reduce
the system error rate, a new method is proposed. First, the likelihood ratio test is used to
analyze the transmission signals of each sensor. Meanwhile, the transmission signals of other
sensors are summed up and fusion center judges which signals were transmitted. The two
transmission signals are then compared to check whether if they are matched. If both signals
are identical, then determine the sensor as a legitimate sensor; otherwise, determine the sensor
as an attacker. Finally, the fusion center only considers the transmission signal of the legitimate
sensor and uses the majority decision to make the final decision.