Research On Danger Signal Perception Method And Response Strategy For Internet Of Things Perception Layer

With the development of the Internet of Things(IoT), security issues graduallybecome revealed and become a bottleneck of restricting the development of internetof things. The security issues share great similarity with biological immune system(Biological Immune System, BIS) when it comes to the problems. That is, both needto maintain the stability of the system in a constantly changing environment.Immune danger theory, as the latest research in the immunity doctrine, focuses ondamage while the traditional immunity doctrine immune cells are only concernedwith foreign bodies. As a result, it can be activated by which the danger signalemitted from the damaged tissue. Danger signal perception method and activeresponse strategy for IoT is studied based on immune danger theory in this paper.Firstly, danger signal perception method is proposed according to thecharacteristics of IoT perception layer dangers of discrete, that using numericaldifferential method to extract the danger signal. Danger signal perception model isbuilded in this paper, that regarding the main performance index change rate whensuffering attacks as a measure against the size of the risk, the model has two modules:immune tolerance module, risk perception and accumulated module. Immunetolerance module’s duty is screening detector which is not matching with autologouscollection at the beginning of the sensing layer node deployment, it also constantlyadjust the current detector according to the working environment in process ofperception. The accumulation of risk perception and module is responsible for thedanger signal recognition and accumulation, it works as follow: detector set which isgenerated by immune tolerance module is used to determine whether the currentsignal danger signal at first, if abnormal signal is detected, accumulates the dangersignal value, and reduces the danger signal value if not. Hazard warnings whendanger signal value is greater than the setting of the relief valve, which suggests thatsuffered from foreign attack. Experimental results show that the dangerous signal perception model proposed in this paper can effectively detect the danger with a lowrate of false positives.Secondly, active response strategy based on cost-sensitive is put forward in thispaper aim at the perception layer node characteristic which is energy limited andhard to supplement, complicated work environment. In this strategy potential costand response cost are calculated according to the attack information. Response costis related to node unit energy consumption, node sleep time, node identity, potentialcost is related to node energy consumption per unit time during risk accumulation,node sleep time, node proportion of residual energy. Appropriate response methodwill be taken according to the results of comparing potential cost and response cost.The experimental results show that in the case of foreign attack, the responsestrategy can effectively reduce loss of IoT perception layer, it also has goodadaptability that could adjust constantly according to the work environment andnode proportion of residual energy.