Secure combinatorial key predistribution scheme for sensor networks by regulating frequencies: magneto optic sensors

摘要

Low cost, decentralized architecture and ad hoc nature are a few desirable properties of wireless sensor networks that make them well suited for gathering sensitive information in hazardous deployment grounds. Existing security protocols exploit various cryptographic tools to strengthen their security. Not many works focus on constraints faced by an adversary. One such constraint being practical difficulties to trace a particular frequency band from a large range of unknown frequencies, specially in unharmonious geographical locations. Our research capitalizes on this weakness encountered by an adversary and preassigns nodes with multiple frequency bands from a wide range of frequencies allocated to the network. Deployed nodes can discover these bands during set network setup phase by a simple trick. Nodes are to internally switch their frequency bands depending on parametric variation that are caused by an (external) impulse. Experiments conducted using magneto optic sensors confirm that any variation of their parameter affects frequencies of emergent waves. Similar behavior is expected from application specific sensors. Frequency regulation (FR) concept is applied to combinatorial key predistribution schemes (KPS) having (regular) degree r. Depending on r and each node's capability to switch internal frequencies, nodes are preallocated with n frequency bands. This naturally partitions the (distributed) network and results in improved resilience. Combining our FR concept to a KPS where the number of shared key between a pair of nodes is at most one (𛾠≤ 1) may yield best case scenario of an ideally resilient key predistribution. Our analysis of systems that combines FR idea with KPS where 𛾠> 1 leads to an optimized key-band distribution argument anddrastic resilience improvements.Results of simulations conducted assuming real-life scenario ascertain our analysis and establishes superior performance of our protocols as compared to prominent ones.