Research On The Disguise Method Of Test Questions Without Carrier Based On Improved Search

Search-based coverless information hiding(SCIH)is the typical way of the latest proposed coverless information hiding(CIH).The main idea of SCIH is finding a set of natural unmodified carriers to express secret information from a big database to resist steganalysis.But there are some problems in SCIH:1)Natural unmodified carrier has poor ability to express irrelevant secret information,and it results in the low hidden capacity of single carrier.2)In order to search appropriate carriers which contain appointed secret vectors,a large database consisted of enough natural image or text carriers is needed.Therefore,the expensive cost of database building and searching is inevitable.Even if multi-level inverted index is constructed,the cost for storing and maintaining this index is also very large.3)Although the modification trace can’t be detected easily due to the natural single unmodified carrier,SCIH can also attract attacker’s attentions because of the massive transmission of carriers.4)The selection range of location tags in some methods is small,and the key dependence of these location tags is low.In addition,SCIH lack the authentication and self-repairing strategy.Therefore it can’t authenticate the restored information correctness and then recover the attacked restored information.To address these problems,the main research works of this paper are summarized as following:1)A coverless test paper disguise method combined with non-direct transmission and random codebook was proposed.Firstly,to avoid direct transmission of secret information,secret information was coded by random codebook which had been scrambled by pseudo-random sequence.Secondly,the coded secret information was then converted into 32-decimal digit sequence and further expressed by 24-decimal and 9-decimal digit sequences.Finally,each 24-decimal and 9-decimal digits in these two sequences were disguised by the key based random offsets relative to sort arrangement order in choice question and answer of blank-filling question respectively in a random generated test paper.By Comparing with the existed methods,the proposed method avoids the direct transmission of secret information,resists steganalysis by random-generated test paper and solves the puzzles of small capacity,large data sharing,high search cost in traditional coverless information hiding methods without embedding and only depends on the correct key to obtain transmitted secret information.2)A full key dependent coverless test paper disguise method by interval extension based double authentications was proposed.Firstly,secret information was transformed into coded sequence by random codebook sequence which is generated by user keys and initialized codebook sequence to hinder the direct transmission of secret information.Secondly coded sequence was transformed into 16-decimal sequence and then used to create test database.Finally,candidate answers’ order and question set index were employed to express the combination of elements in 16-decimal sequence and random offset,where secret information is hidden by multiple-choice orders and authentication information is hidden in blank-fillings.In restoration process,the extracted information from multiple-choices can be authenticated by the information extracted from blank-fillings through the interval extension based double authentication strategy.The experimental results and analysis show that the proposed method can’t only avoid low hidden capacity,expensive search cost and the massive carrier transmission in SCIH,but also has good authentication ability to restore information and the restored information is fully dependent on user keys.The proposed method can be easily extended to other types of test questions with good universality.3)A double authentication and backup test paper disguise method combined with codebook extension and stem hash was proposed.Secret byte sequence was backed up as backup byte sequence by Lagrange(2,4)interpolation polynomials based on GF(28)and then the backup byte sequence was further transformed into backup index sequence.A test database divided into 8 sets was created and then the stem MD5 values were mapped as stem hash values by stem set indexes where backup index sequence were expressed by employing answers’ orders and stem hash values.In restoration,legal interval and parameters of interpolation polynomials were applied to authenticate extracted information and then used to generate candidate recovery values,and the most reliable candidate recovery value was selected to reconstruct secret information.The experimental results and analysis show that compared with SCIH and former proposed test paper disguise strategies,the proposed method introduce double authentications and reliability calculation to mark the reliability of restored information and ensure the quality of reconstructed information.The proposed method has good self-repairing ability with secret information backup strategy.Moreover,the proposed method improves the hidden capacity of single question by using stem hash and candidate answers’ order.Finally,the proposed method can provide enough security with full key dependence,and the difference of questions hidden secret and questions without secret can’t be distinguished without correct key.