作者: Koji Nuida
DOI: 10.1007/978-3-642-04846-3_10
关键词:
摘要: In recent research on collusion-secure fingerprint codes, some relaxation of the conventional security assumption (Marking Assumption) have been introduced from a viewpoint reality in practical situations, and several codes proposed under those assumptions. this article, we consider such relaxed give an extension short 2-secure (under Marking recently by Nuida et al. (IEICE Trans. A, 2009) to our assumption. We perform theoretical numerical evaluation required code lengths. For example, bound error probability 0.01% for 10,000 users, 162-bit, 220-bit 329-bit lengths are sufficient even if each bit codeword is either flipped (in addition other collusion attacks) with probabilities 1%, 2.5% 5%, respectively, or erased 2%, 5% 10%, respectively.