نهان نگاری طرح احراز هویت برگشت پذیر برای شبکه های حسگر بی سیم

Authentication is a very important demand in wireless sensor networks (WSNs), especially in some critical applications. Previous watermarking-based approaches always make irreversible modifications in the watermarked data. In this paper, we propose a novel reversible watermarking authentication scheme for WSNs, which is based on prediction-error expansion. The sensor node groups the stream data, and two adjacent data groups compose the non-overlapping authentication group. The watermark bits are computed from the first data group and embedded into the second one before transmission. The sink does the verification and complete restoration. Compared with the previous work, our approach only buffers copy of current data element instead of the entire data group, which greatly reduces the delay. Analysis and experimental results show that our scheme losslessly authenticates the sensory stream data with low cost.