Protecting from Inside Attacks in Wireless Sensor Networks

Wireless sensor networks (WSNs) are widely used. They are easily and rapid deployed, low cost, low power, self-organized, cooperatively collect the environmental information and realize the integration of the physical world and communication network. However, for most WSNss the open nature of the wireless medium an adversary can easily eavesdrop and replay or inject fabricated messages. Different cryptographic methods can be used to defend against some such attacks. But the inside attacks are not detectable with only the classic cryptographic techniques. These attacks mainly include node compromise which is another major problem for WSN security. This allows an adversary to enter inside the security perimeter of the network which raises a serious challenge for WSNs. This paper focusses on investigating inside attacks of WSNs. We show our novel method works well for security under some fixed parameters designed by the network designer. We can also reasonabley predict the highest signal noise ratio (S/N). This method controls the sinker node with period sleeping time period for protecting inside attacks in WSNss. Therefore we may allow a sinker to be open only around a particular time period to receive the signals from the sources while the other time slots are in "sleeping state" to ignore any signals, including the inside attacking signal. In particularly, we found the highest S/N timing can be controlled by the sending rating for a fixed network. We can manipulate the sending rate to control the time when the highest S/N is occurring to protect from "inside attacks". The simulation results from our novel algorithm are promising.