Abstract
In the internet, a worm is usually propagated in a random multi-hop contact manner. However, the attacker will not likely select this random multi-hop propagation approach in a mobile sensor network. This is because multi-hop worm route paths to random vulnerable targets can be often breached due to node mobility, leading to failure of fast worm spread under this strategy. Therefore, an appropriate propagation strategy is needed for mobile sensor worms. To meet this need, we discuss a hop-by-hop worm propagation model in mobile sensor networks. In a hop-by-hop worm propagation model, benign nodes are infected by worm in neighbor-to-neighbor spread manner. Since worm infection occurs in hop-by-hop contact, it is not substantially affected by a route breach incurred by node mobility. We also propose the carryover epidemic model to deal with the worm infection quota deficiency that might occur when employing an epidemic model in a mobile sensor network. We analyze worm infection capability under the carryover epidemic model. Moreover, we simulate hop-by-hop worm propagation with carryover epidemic model by using an ns-2 simulator. The simulation results demonstrate that infection quota carryovers are seldom observed where a node’s maximum speed is no less than 20 m/s.
Highlights
Worm attacks against sensor motes are realistic in sensor networks, as demonstrated by [1,2] In particular, they show that a worm could be propagated by exploiting code injection vulnerabilities in Computers 2015, 4 target sensor motes
We introduce a hop-by-hop worm infection with carryover epidemic model strategy that is well-suited for mobile sensor networks
We derive the Chernoff bounds on the number of time slots required for the entire network infection
Summary
Worm attacks against sensor motes are realistic in sensor networks, as demonstrated by [1,2] In particular, they show that a worm could be propagated by exploiting code injection vulnerabilities in Computers 2015, 4 target sensor motes. In the sense that epidemic model [3] is good at modeling the spread of epidemic disease, researchers adapted it to worm propagation modeling in the internet and static sensor network [4,5,6,7] These previous works do not consider node mobility which can greatly affect the worm infection capability and they are not appropriate for worm propagation in a mobile sensor network. The simulation results show that a worm infection quota deficit rarely occurs under the situation where a node’s maximum speed is at least 20 m/s and worm infection is not delayed but completed in time This indicates that node mobility makes it possible for a mobile node to have enough neighbors, leading to expedition of worm infection.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
