A crowd cooperative defense model for mitigating DDoS attacks in Mobile Crowdsensing networks

Abstract

Mobile Crowdsensing (MCS) is rapidly growing up while encountering severe security threats simultaneously. DDoS attacks tend to disrupt large-scale sensing tasks to the MCS and reduce the availability of MCS. However, the existing centralized defense technology has the problem of high cost and the existing research lacks the application of distributed cooperative characteristics of MCS. A crowd cooperative defense model for mitigating DDoS attacks in MCS is proposed based on these. Firstly, the crowdsensing networks are formally described. The cooperative mechanism of sensing nodes participating in overlapping sensing coalition is generalized, which breaks the independence between the distributed sensing nodes and enables the sensing nodes to cooperate against DDoS attacks. Then, the crowd cooperative defense problem is transformed into a multi-objective combinatorial optimization problem. A multi-objective discrete Harris Hawk optimization (MODHHO) algorithm is designed. The algorithm applies non-dominated sorting and crowding strategy to evaluate the advantages and disadvantages of each cooperative defense solution on the cooperative defense objectives. In addition, the global exploration, local exploitation, and besiege phases are used to continuously update the cooperative defense solutions, and finally obtains the global optimal cooperative defense solution after a certain number of iterations. The simulation results show that the proposed model can minimize the defense cost compared to non-cooperative and non-overlapping sensing coalition methods. Compared with centralized defense, the loss cost of the proposed model is reduced by 61.8%. In addition, the average throughput of the sensing network is increased by 35.5%, which enhances the availability of the sensing network.