Security Services in Wireless Sensor Networks Using Sparse Random Coding

• Main Authors: Delgosha, F., Ayday, E., Chan, K., Fekri, F.
• Format: Conference Proceeding
• Language: English
• Subjects: Availability; Collaboration; Computer networks; Data security; Military computing; Peer to peer computing; Protocols; Routing; Sensor phenomena and characterization; Wireless sensor networks
• ISSN: 2155-5486
• DOI: 10.1109/SAHCN.2006.288407

The task of providing security services for wireless sensor networks is not trivial due to the resource constraints of the sensor nodes. An adversary may launch a wide range of attacks including eavesdropping, message forgery, packet dropping, and noise injection. In this paper, we propose random coding security (RCS) that provides protection against all the aforementioned attacks. For this purpose, the proposed protocol makes extensive use of node collaboration and data redundancy. Moreover, using location information, we both localize adversarial activities to the area under attack and enhance routing the data toward the sink. The objectives of using the novel idea of sparse random coding in RCS are twofold. First, every node generates correlated data by calculating random linear combinations of the received packets. Hence, the availability of the data at the receiver is guaranteed with a high probability. The second advantage is the feasibility of implementing the RCS in the real case scenario in which the communication media between the sensors is usually modeled as the erasure channel. The existing protocols cannot be trivially modified to suit this realistic situation. In the overall, RCS provides many security services with computation and communication overheads comparable with other schemes