A data integrity authentication method for WSNs based on double watermark

Abstract

The present invention discloses a WSNs data integrity authentication method based on double watermark in the field of wireless sensor network technology, comprising the following steps: S1, randomly and dynamically grouping the collected data stream to make the length of each group of data irregular; S2, embedding the double watermark algorithm into the sensed data, wherein the embedding of the double watermark algorithm includes reversible watermark embedding and flag watermark embedding; S3, embedding the reversible watermark into the effective precision bit of the sensed data; S4, judging the embedding position of the flag watermark, and respectively embedding the flag group data and the inter-group data; S5, transmitting the embedded data to a base station for restoration. The present invention has the advantages of high data authentication security and the ability to effectively reduce the false detection rate.

Description

Technical Field

[0001] The invention relates to the technical field of wireless sensor networks, and in particular to a WSNs data integrity authentication method based on double watermarks. Background Art

[0002] Since wireless sensor networks are not restricted by space, can work at any time and place, and have low and simple deployment costs, and can transmit data in real time without affecting other nodes in the event of local damage, wireless sensor networks have been increasingly widely used in recent years. However, the communication lines of wireless sensor networks cannot be privately controlled, and the security of transmitted data cannot be guaranteed. If the receiver makes an incorrect prediction based on the attacked data, it may cause huge losses. Therefore, the data security of wireless sensor networks has gradually received widespread attention from researchers.

[0003] As a branch of information hiding technology, digital watermarking technology is simple to calculate, consumes less energy, and does not generate additional overhead. It is more suitable for the application of data security authentication in wireless sensor networks. Guo et al. [4] proposed a fragile watermarking scheme to detect whether the streaming data of wireless sensor networks has been tampered with. The scheme uses dynamic grouping, but the watermarking technology is irreversible. Kamel et al. [5] improved the performance of Guo's scheme and proposed a lightweight chain watermarking scheme, which is much faster than Guo's scheme. However, in some special fields, such as military, body area network, medical, etc., the data must be accurate and cannot have deviations, so the irreversible watermarking algorithm cannot be applied in some fields with high precision requirements.

[0004] Reversible digital watermarking technology, also known as lossless digital watermarking technology, inherits the advantages of traditional digital watermarking technology and can accurately restore the received watermarked data to the original data. Shi et al. [6] calculated the synchronization point through hash function for dynamic grouping, and embedded the reversible digital watermark into the sensor data by group. This scheme can accurately detect the tampered group. Liu et al. proposed a lightweight integrity authentication scheme based on reversible watermark in wireless body area network [7]. The data stream adopts static grouping, generates and embeds reversible digital watermark in a linked manner, and uses histogram shifting technology to prevent overflow. Sen et al.'s scheme [8] uses the unique ID of each node and its neighbor node information to generate a watermark, and then selects a random position to embed the watermark information. This work provides a security solution with relatively low computational and communication overhead. The scheme [9] uses data sampling time and pseudo-random sequence to generate reversible digital watermarks to resist cloning attacks. Zhang et al.

[10] proposed to use the data collection time of sensor nodes to dynamically calculate the embedding position of the watermark, saving energy and realizing real-time data authentication. Xiao et al. proposed an individual authentication scheme for sensor node data

[11] . They associated several data to generate watermark bits, which were then embedded into the original data. This scheme has better performance. Jiang et al.

[12] proposed a new recoverable data fusion protocol that combines reversible digital watermarking technology and homomorphic elliptic curve encryption technology. The original data read by the node is fragmented and embedded with watermarks at the sensing node. At the same time, the original data is ellipticly encrypted and sent to the cluster head node for fusion encryption. This scheme can ensure data integrity and confidentiality at the same time.

[0005] Among them, [1] JAAnastasov, AMCvetkovic, DMMilovic, DNMilic, GTDjordjevic, On physical layer security in WSN over GK fading channels during intercept events, Telecommunication Systems, 74 (2020) 95-102.

[0006] [2] A. Jabbari, JBMohasefi, Improvement of a User Authentication Scheme for Wireless Sensor Networks Based on Internet of Things Security, WirelessPersonal Communications, 116 (2021) 2565-2591.

[0007] [3]G.Yang,L.Dai,Z.Q.Wei,Challenges,Threats,Security Issues and NewTrends of Underwater Wireless Sensor Networks,Sensors,18(2018).

[0008] [4]H.P.Guo,Y.J.Li,and S.Jajodia,“Chaining Watermarks for DetectingMalicious Modifications to Streaming Data,”Information Sciences,vol.177,no.1,pp.281-298,Jan 1,2007.

[0009] [5]I.Kamel,and H.Juma,“Simplified Watermarking Scheme for SensorNetworks,”International Journal of Internet Protocol Technology,vol.5,no.1-2,pp.101-111,2010.

[0010] [6]X.Shi,and D.Xiao,“A Reversible Watermarking Authentication Schemefor Wireless Sensor Networks,”Information Sciences,vol.240,pp.173-183,2013.

[0011] [7]X.Y.Liu,Y.Ge,Y.S.Zhu et al.,“A Lightweight IntegrityAuthentication Scheme based on Reversible Watermark for Wireless Body AreaNetworks,”Ksii Transactions on Internet and Information Systems,vol.8,no.12,pp.4643-4660,Dec 31,2014.

[0012] [8] A.Sen, T.Chatterjee, and S.DasBit, "LoWaNA: Low Overhead Watermarkbased Node Authentication in WSN," Wireless Networks, vol.22, no.7, pp.2453-2467, Oct, 2016.

[0013] [9] T.Guan and Y.Chen, "A Node Clone Attack Detection Scheme based on Digital Watermark in WSNs," First IEEE International Conference on ComputerCommunication and the Internet (ICCCI), pp.257-260, 2016.

[0014]

[10] GYZhang, L.Kou, LGZhang et al., "A New Digital WatermarkingMethod for Data Integrity Protection in the Perception Layer of IoT," Securityand Communication Networks, 2017.

[0015]

[11] Y. Xiao, and G. Gao, "Digital Watermark-Based Independent Individual Certification Scheme in WSNs," IEEE Access, vol.7, pp.145516-145523, 2019.

[0016]

[12] W.Jiang, and Z.Zhang, "Reversible Digital Watermarking-basedProtocol for Data Integrity in Wireless Sensor Network," Journal onCommunications, vol.39, no.3, pp.118, 2018.

[0017] However, the inventors of this application found that the above technology has at least the following technical problems in the process of implementing the technical solution of the invention in the embodiment of this application:

[0018] In some special fields, such as military, body area network, medical, etc., data must be accurate and cannot be biased, so reversible digital watermarks are needed to protect data privacy, and attention must also be paid to issues such as node resource energy consumption and false detection rate.

[0019] Based on this, the present invention designs a WSNs data integrity authentication method based on double watermark to solve the above problems. Summary of the invention

[0020] In order to solve the technical problems mentioned in the current background technology, the purpose of the present invention is to provide a WSNs data integrity authentication method based on double watermarks.

[0021] In order to achieve the above object, the present invention adopts the following technical solution:

[0022] A WSNs data integrity authentication method based on double watermarks includes the following steps:

[0023] S1, randomly and dynamically group the collected data stream so that the length of each group of data is irregular;

[0024] S2, embedding a double watermark algorithm into the perception data, wherein the embedding of the double watermark algorithm includes reversible watermark embedding and flag bit watermark embedding;

[0025] S3, embedding the reversible watermark into the effective precision bits of the sensed data;

[0026] S4, determining the embedding position of the flag watermark, and embedding the flag group data and the inter-group data respectively;

[0027] S5. Transmit the embedded data to the base station for restoration.

[0028] Preferably, the dynamic grouping includes:

[0029] The sink node and the relay node are both set with an array of specified length, and a group leader with irregular data length related to the array is randomly generated.

[0030] Preferably, the embedding of the flag group data includes:

[0031] The group determines whether to embed the watermark into the group head and group tail and the nearby flag group data. If the conditions are met, the flag group data is used to construct a flag check matrix, and the watermark is generated through the matrix and embedded into the flag group data.

[0032] Preferably, the embedding of the inter-group data includes:

[0033] It is determined whether to embed the watermark into the remaining inter-group data of the group. If the condition is met, the non-repeating decimal sequence specified within the base station and the node is transformed and embedded into the inter-group data as a watermark.

[0034] Preferably, the grouping determination includes:

[0035] The queue data of the specified length in the data stream is allowed to enter in sequence, and it is determined whether the data meets the group head and group tail flag calculation scheme. If it does, the data in the queue is used to form a flag check matrix. If the specified number of data flags in the matrix are verified to be correct, the data that meets the requirements of the group head and group tail flag calculation scheme will be used as the head node data and tail node data respectively.

[0036] Preferably, embedding of the effective precision bit watermark includes:

[0037] A reversible watermark is generated using the effective precision bits of the group header data to obtain a binary watermark sequence. Each time, a watermark of a specified size is taken from the watermark sequence and embedded into the effective bits of two adjacent data, and it is determined whether the length of the watermark sequence meets the group length requirement. If the number of bits in the watermark sequence is less than the data in the embedded group, it is embedded cyclically.

[0038] Preferably, the embedding of the flag watermark includes:

[0039] Embed the watermark into a special flag in a special data set; and

[0040] The watermark is embedded into the normal flag bit in the remaining data set.

[0041] Preferably, transmitting the embedded data to the base station for restoration includes:

[0042] Determine whether the data to be restored meets both the reversible digital watermark authentication conditions and the flag watermark authentication conditions. If not, the verification fails and double authentication of individual data is performed.

[0043] Preferably, the dual authentication of the individual data includes:

[0044] The data in the remaining data set in the group after embedding the double watermark with the effective accuracy of the watermark and the flag watermark are divided into small groups, and the attack situation of each two data is collected and judged. If attacked, the malicious attack is detected, and the first data of the current group corresponding to it is marked as unverified, and the other data and the next data are combined into a new group for re-judgment.

[0045] One or more technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages:

[0046] 1. The present invention uses double-watermark WSNs data integrity authentication to reduce the false detection rate of data integrity authentication while reducing resource consumption, and can also resist multiple attacks to ensure data security;

[0047] 2. The present invention uses double verification at the individual level, which not only retains the advantages of group verification, but also skips the attacked data and restores other unattacked data in the group to the maximum extent;

[0048] 3. The present invention ensures data security by embedding double watermarks and the randomness of watermarks and groups. The generation, embedding and data recovery of watermarks do not involve complex operations, and neither hash functions nor encryption algorithms are used, which greatly reduces computing resources.

[0049] 4. The present invention further ensures the robustness of grouping by randomly selecting the group length and using the mark check matrix to determine the group head and group tail data;

[0050] 5. The present invention can effectively reduce the false detection rate and false alarm rate of data integrity authentication through the double watermark algorithm.

[0051] In summary, the present invention has the advantages of high data authentication security and the ability to effectively reduce the false detection rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0053] Figure 1 is a flow chart of the authentication algorithm of the present invention;

[0054] Figure 2 This is a data transmission layout diagram of the wireless sensor network of the present invention;

[0055] Figure 3 An example diagram of a mark check matrix of the present invention;

[0056] Figure 4 Two packet data distribution diagrams cached by the relay node of the present invention;

[0057] Figure 5 It is a layout diagram of the generation process of the common flag watermark of the present invention;

[0058] Figure 6 A flowchart of dual authentication at the individual level of the present invention;

[0059] Figure 7 A line graph showing the effect of packet length on false detection rate when data is attacked using different tampering methods of the present invention;

[0060] Figure 8A line graph showing the influence of the tampering rate on the false detection rate when the data is attacked by different tampering methods of the present invention;

[0061] Fig. 9 It is a curve diagram of the recovery rate of data within the algorithm group of the present invention;

[0062] Fig.10 A comprehensive comparison curve diagram of the algorithm of the present invention and algorithms [6],

[11] and

[12] ;

[0063] Fig.11 : is a comparison curve of the false detection rate of the algorithm of the present invention and algorithms [6],

[11] and

[12] ;

[0064] Fig.12 The false alarm rate comparison curve of the algorithm of the present invention and algorithms [6],

[11] and

[12] is shown in FIG. DETAILED DESCRIPTION

[0065] The following is a description of the implementation of the present invention by means of specific embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification.

[0066] Embodiment 1

[0067] The present invention provides a technical solution: a WSNs data integrity authentication method based on double watermark, comprising the following steps:

[0068] S1, randomly and dynamically group the collected data stream so that the length of each group of data is irregular;

[0069] S2, embedding a double watermark into the perception data, wherein the embedding of the double watermark algorithm includes reversible watermark embedding and flag bit watermark embedding;

[0070] S3, embedding the reversible watermark into the effective precision bits of the sensed data;

[0071] S4, determining the embedding position of the flag watermark, and embedding the flag group data and the inter-group data respectively;

[0072] S5. Transmit the embedded data to the base station for restoration.

[0073] Through the above steps, it is not difficult to find that in the process of double-watermark WSNs data integrity authentication, through the double-watermark-based WSNs data integrity authentication scheme, by embedding watermarks into grouping methods at different positions, the consumption of computing resources can be effectively reduced, and the other data in the group where the tampered data is located can be lostless restored, effectively reducing the false detection rate of the scheme.

[0074] In the present invention, the data stream is first randomly grouped, and then a double watermark is embedded into the sensed data, so that a reversible watermark is embedded into the effective precision bit of the collected data; the flag bit watermark is irreversible, and is divided into two cases according to the embedding position to be embedded into the negligible precision bit of the data: the first case is to embed it into the group head and group tail and the nearby data (flag group data), and use the flag group data to construct a flag check matrix, and generate a watermark through the matrix and embed it into the flag group data, so as to ensure the robustness of the grouping; the other case is to embed it into the remaining data of the group, that is, the inter-group data, and transform the non-repeating decimal sequence specified in the base station and the node as a watermark and embed it into the inter-group data; experimental results show that the present invention has the ability to completely restore the effective precision bit of the data, resist malicious attacks, and can restore the untampered data in the group while detecting the tampering position and type. In addition, the present invention has high detection efficiency, extremely low resource consumption, and good comprehensive performance.

[0075] In order to better complete the random grouping of data streams, the dynamic grouping includes:

[0076] The sink node and the relay node are both set with an array of specified length, and a group leader with irregular data length related to the array is randomly generated.

[0077] In this embodiment, both the aggregation node and the relay node specify an array J with a length of preferably 90. According to formula (1), a group length l is randomly generated to ensure that the data length of each group is irregular. The generation of l is related to the length of J, that is, l = randi([15,90], 1)(1), so the value range of l is [15,90].

[0078] In order to embed a type of situation in the flag bit watermark, the embedding of the flag group data includes:

[0079] The group determines whether to embed the watermark into the group head and group tail and the nearby flag group data. If the conditions are met, the flag group data is used to construct a flag check matrix, and the watermark is generated through the matrix and embedded into the flag group data.

[0080] In order to embed another type of situation in the flag bit watermark, the embedding of the inter-group data includes:

[0081] It is determined whether to embed the watermark into the remaining inter-group data of the group. If the condition is met, the non-repeating decimal sequence specified within the base station and the node is transformed and embedded into the inter-group data as a watermark.

[0082] Furthermore, the grouping determination includes:

[0083] The queue data of the specified length in the data stream are allowed to enter in sequence, and it is determined whether there is data that meets the group head and group tail flag calculation scheme. If it does, the data in the queue is used to form a flag check matrix. If the specified number of data flags in the matrix are verified to be correct, the data that meets the requirements of the group head and group tail flag calculation scheme will be used as the head node data and tail node data respectively.

[0084] In this embodiment, the packet length is randomly generated using the rand() function, so the packet judgment when extracting the watermark is mainly to find the group head and the corresponding group tail. A queue Q with a length of 12 is designed, and the data in the data stream is allowed to enter in sequence. If the data of Q(6) and Q(7) meet the group head and group tail flag calculation scheme, the data in the queue is used to form a flag check matrix. If the flags of more than 3 data in the matrix are verified correctly, Q(6) is considered to be the tail node or Q(7) is the head node. Compared with ordinary nodes, sink nodes have more energy and stronger computing power, so sink nodes can meet the need to cache up to 3 groups of data and at least 1 group of data. Define arrays Head and Tail to store the head node and tail node read from the data stream respectively. Each time, the first group head Head(1) in the array Head is read, and the corresponding group tail is found in Tail.

[0085] In order to better complete the embedding of the reversible watermark, the embedding of the effective precision watermark includes:

[0086] A reversible watermark is generated using the effective precision bits of the group header data to obtain a binary watermark sequence. Each time, a watermark of a specified size is taken from the watermark sequence and embedded into the effective bits of two adjacent data, and it is determined whether the length of the watermark sequence meets the group length requirement. If so (that is, when the watermark sequence is short), it is embedded cyclically.

[0087] In this embodiment, a reversible digital watermark is embedded into the effective precision bits of the perceived data using a difference expansion technique, and the reversible watermark w is generated by the effective precision bits G1 of the group header data, to obtain a binary watermark sequence W. Each time, 1 bit of the watermark w is taken from W and embedded into the effective bits of two adjacent data. If the length of W is less than 1 / 2 of the group length, the embedding is repeated cyclically. The data obtained after the original data Si is embedded with the reversible watermark is called the watermark data Si'. To ensure that the watermark is secret and imperceptible, the watermark sequence W is calculated as follows: Where n is the number of valid precision bits of the data, X[j] is the number of valid bits of the G1 data, and satisfies the value X[j]∈[1,9], j∈[1,n], l is the group length, and as the parameter of the chaotic sequence, r ranges from [3.5699456,4], where r is 4. The last 6 data of the previous group and the first 6 data of the current group are regarded as special data sets, and the other data in the group are called remaining data sets. For the convenience of calculation, special data sets do not need to be embedded with reversible watermarks.

[0088] More specifically, when embedding the flag watermark, the embedding of the flag watermark includes:

[0089] Embed the watermark into a special flag in a special data set; and

[0090] The watermark is embedded into the normal flag bit in the remaining data set.

[0091] In this embodiment, the generation and embedding of the watermark occurs in the relay node, and the relay node needs to cache two groups of data, namely the previous group of data and the current group of data; the flag bit is composed of two decimal places, which are called b1 and b2 respectively; due to the different embedding positions, the watermark generation scheme is also different; embedding the watermark into a special data set is called special flag bit embedding, and embedding the watermark into the remaining data sets is called ordinary flag bit embedding; the main function of the special flag bit embedding is to determine the group header and group tail data during recovery to ensure the correctness of the grouping; the flag bit watermark embedding adopts the LSB embedding algorithm.

[0092] It should be added that when the special flag watermark is generated and embedded, the first flag of the group header is the unit digit of the sum of the digits X[j] of the data G1, and the second flag is the unit digit of the product of the digits X[j] of the watermark data G1; then the flag watermarks Hb1 and Hb2 of the group header satisfy the formula: and The first flag bit Tb1 of the group tail data Gl is always 0, so in order to prevent collision when verifying the head and tail, the flag bit watermark of the head must satisfy Hb1∈[1,9], Hb2∈[0,9], and the value of the second flag bit Tb2 of the tail is related to the second flag bit Hb2 of the group head of the current group: Hb2+Tb2=9(5); This scheme proposes a flag check matrix to ensure accurate verification of the group head and the group tail and avoid group confusion; take the previous group tail and its first five bits of data and the current group head and its last five bits of data, a total of 12 bits of data as the flag group data, to form a 3*4 matrix A, and put the current group head and the previous group The tail data are placed at the positions of matrix A(2,2) and A(2,3) respectively. When calculating the flag of data m in matrix A, a 3*3 matrix B is formed with data m as the center and the 8 data around it. If there is no data, fill it with 0. Compare the data m with the remaining data in matrix B from the upper left. If m is greater than the compared data, the corresponding position is 1, otherwise it is 0. When m encounters data A(2,2) and A(2,3), it directly takes 1. After comparing data m with the remaining data in matrix B, an 8-bit binary sequence will be generated. The binary sequence is converted to decimal, and the last two digits of the decimal data are taken as the flag.

[0093] It should also be added that when generating and embedding the common flag bit watermark, an array J with a length of 90 is specified inside the base station and relay node, and the elements inside the array are not repeated and are in random order. The generation of the common flag bit is related to the array J and the valid bit digits of the data to be embedded; read the current group G, obtain each digit X[j] in the valid precision bit of the corresponding data Gi, calculate the inverted position based on it, convert the data in J corresponding to Gi into a binary sequence M, invert the corresponding position in the binary sequence M to generate a watermark w, and embed it into the data Gi as a common flag bit watermark; the specific example is shown in the figure Figure 1 As shown; the selection of the corresponding data inversion position in J is determined by the data Gi. Since Gi is different, the selection of the inversion position is random, and the same position can be inverted multiple times; the calculation formula for the inversion position r is: r[j]=mod(X[j],5)+1(6).

[0094] Furthermore, when extracting and restoring the watermark, transmitting the embedded data to the base station for restoration includes:

[0095] Determine whether the data to be restored meets both the reversible digital watermark authentication conditions and the flag watermark authentication conditions. If not, the verification fails and double authentication of individual data is performed.

[0096] In this embodiment, data recovery authentication must meet both the reversible digital watermark authentication condition and the flag watermark authentication condition. If one of the conditions is not met, the authentication is considered unsuccessful. Since the effective precision bits of the header data are not changed when the watermark is embedded, the watermark sequence W' to be verified is generated according to formula (2) using the header data and compared with the reversible digital watermark extracted from the data Si". At the same time, the data flag bit must be authenticated. If the data authentication is unsuccessful, double authentication of the individual data is performed.

[0097] In order to realize the dual authentication of individual data within a group, the dual authentication of individual data includes:

[0098] The data in the remaining data set in the group after embedding the double watermark with the effective accuracy of the watermark and the flag watermark are divided into small groups, and the attack situation of each two data is collected and judged. If attacked, the malicious attack is detected, and the first data of the current group corresponding to it is marked as unverified, and the other data and the next data are combined into a new group for re-judgment.

[0099] In this embodiment, the data in the remaining data set in which both the data valid precision bit watermark and the flag bit watermark are embedded in the group G" after the double watermark is embedded are divided into small groups, and each two data are verified as a small group; if these data are attacked, a malicious attack will be detected, and the first data of the current group in which the malicious attack is detected will be marked as unverified, and the second data of the current group and the next data will be combined into a new group; and so on, until the erroneous data is skipped, and normal calculation can be restored; while ensuring integrity authentication, this scheme restores the valid precision bits of other data that have not been tampered with in the group as much as possible.

[0100] It should be noted that if Figure 3 As shown in , 19.01 is the tail data of the previous group, and 18.98 is the head data of the current group. According to the rules of the flag check matrix, a 3×4 matrix is ​​constructed, as shown in Matrix A. According to the calculation method of the head and tail flags above, when the head data G1=18.98, it can be calculated that the flag of G1 is 66, so G1 is 18.9866 after the flag is embedded. Assuming that the tail flag is 06, the tail data is 19.0106 after the flag is embedded. Taking m=19.11 as an example, the constructed 3×3 matrix is ​​shown in Matrix B. To calculate the flag of this data, it is necessary to compare it with other data in B in turn. When encountering the tail and head data, the binary takes 1. According to the above method, after comparison, it can be obtained that the binary sequence corresponding to m is (11101011)2=235, so 19.11 is 19.1135 after the flag is embedded.

[0101] It should also be noted that the simulation experiment was carried out using Matlab, and the data used were selected from the real data of the Berkeley Laboratory, including temperature, humidity, light, voltage, etc. The data needs to be preprocessed before the experiment, and the blank data items and data items with incorrect formats need to be deleted. The watermark embedding process of the algorithm of the present invention is carried out at the relay node, and the data is restored at the base station. The specific process is as follows Figure 1 As shown, it can be mainly divided into the following steps:

[0102] 1. Use formula (1) to group the data stream at the relay node;

[0103] 2. Generate a reversible digital watermark using the effective precision bits of the group header, and embed the watermark into the original data using interpolation and extension technology to obtain data S';

[0104] 3. Generate an irreversible digital watermark, and use LSB technology to embed the data S' twice to obtain data S". Embed the watermark generated by the mark check matrix into the special data set, and use the specified random array J and data S' to calculate to obtain the common mark watermark, and embed it into the remaining data sets;

[0105] 4. Transmit the data S" to the base station for restoration. Similarly, the restoration process in the base station is also divided into the verification of reversible watermarks and irreversible watermarks. If both watermarks are correct, the original data is restored;

[0106] Moreover, after a large number of experimental tests, this algorithm can normally process large amounts of data and achieve the desired experimental results.

[0107] The present invention provides a WSNs data integrity authentication method based on double watermarks; it reduces the false detection rate of data integrity authentication on the basis of reducing resource consumption, and can also resist multiple attacks. It is difficult for existing data security authentication schemes for wireless sensor networks to meet the above requirements at the same time, and most of the group authentication schemes will judge the entire data group as tampered when an abnormality occurs in a data item in the group, resulting in a waste of resources; based on the above scheme, the comprehensive performance is further improved compared with the prior art.

[0108] 1) Improve the data recovery rate: In most existing grouping schemes, after detecting abnormal data in the group, the entire group will be lost, resulting in low data recovery rate and waste of resources. The present invention optimizes this problem and proposes a double verification based on the individual level, which not only retains the advantages of group verification, but also skips the attacked data and maximizes the recovery of other unattacked data in the group;

[0109] 2) Reduce resource consumption: This algorithm ensures data security by embedding double watermarks and the randomness of watermarks and groups. The generation, embedding and data recovery of watermarks in this algorithm do not involve complex operations. Therefore, compared with algorithms [6],

[11] and

[12] , this algorithm does not use hash functions or encryption algorithms, and the computing resources are greatly reduced.

[0110] 3) Grouping robustness: The grouping length of this algorithm is randomly selected, and a sign check matrix is ​​designed to determine the group header and group tail data. According to theoretical analysis and experimental proof, the probability of a false header or false tail is close to 10-4. Compared with the previous algorithm that uses synchronization points for dynamic grouping, this algorithm further ensures the robustness of grouping;

[0111] 4) Compared with the other three algorithms, the dual watermark algorithm proposed in this invention can effectively reduce the false detection rate and false alarm rate of data integrity authentication. Fig. 9 (a) and (b) show the comparison of the false detection rate and false alarm rate of this algorithm with those of algorithms [6],

[11] and

[12] under modification attack. Figure 1 It can be seen that the false detection rate and false alarm rate of the present invention are lower.

[0112] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A WSNs data integrity authentication method based on double watermark, characterized in that: The following steps are involved: S1, randomly and dynamically group the collected data stream so that the length of each group of data is irregular; S2, embedding a double watermark algorithm into the perception data, wherein the embedding of the double watermark algorithm includes reversible watermark embedding and flag bit watermark embedding; S3, embedding the reversible watermark into the effective precision bits of the sensed data; The reversible watermark generated by the effective precision bit of the group header data is used to obtain a binary watermark sequence. Each time, a watermark of a specified size is taken from the watermark sequence and embedded into the effective bits of two adjacent data, and it is determined whether the length of the watermark sequence meets the group length requirement. If it does, it is embedded cyclically. The reversible digital watermark is embedded into the effective precision bit of the perceived data using the difference expansion technology, and the reversible watermark w is generated by the effective precision bit G1 of the group header data to obtain a binary watermark sequence W. Each time, 1 bit of watermark w is taken from W and embedded into the effective bits of two adjacent data. If the length of W is less than l / 2 of the group length, it is embedded cyclically. The data obtained after the original data Si is embedded with the reversible watermark is called the watermark data Si'; to ensure that the watermark is secret and imperceptible, the watermark sequence W is calculated as follows: Where n is the number of effective precision bits of the data, X[j] is the number of effective bits of the G1 data, and satisfies the value X[j]∈[1,9], j∈[1,n], l is the group length, and as the parameter of the chaotic sequence, r ranges from [3.5699456,4], where r is 4; the last 6 data of the previous group and the first 6 data of the current group are regarded as special data sets, and the other data in the group are called the remaining data sets; for the convenience of calculation, the special data sets do not need to be embedded with reversible watermarks; S4, determining the embedding position of the flag watermark, and embedding the flag group data and the inter-group data respectively; The grouping determines whether it is embedded in the group header and group tail and the flag group data nearby. If the conditions are met, the flag group data is used to construct a flag check matrix, and the watermark is generated through the matrix and embedded in the flag group data; the flag check matrix ensures the accuracy of the verification of the group header and group tail to avoid group confusion; take the previous group tail and its first five bits of data and the current group header and its last five bits of data, a total of 12 bits of data as the flag group data, to form a 3*4 matrix A, and put the current group header and the previous group tail data in the matrix A(2,2) and A(2,3) respectively; when calculating the flag bit of data m in matrix A, take data m as the center and form a 3*3 matrix B with the 8 data around it, and fill in 0 if there is no data; compare the size of data m with the remaining data in matrix B from the upper left in turn. If m is greater than the comparison data, the corresponding position is 1, otherwise it is 0. When m meets When the data is A(2,2) or A(2,3), directly take 1; after comparing the data m with the remaining data of the matrix B, an 8-bit binary sequence will be generated. The binary sequence is converted into decimal, and the last two digits of the decimal data are taken as the flag bit; determine whether to embed the watermark into the remaining inter-group data of the group. If the conditions are met, transform the non-repeating decimal sequence specified by the base station and the node as a watermark and embed it into the inter-group data; among them, the first case is to embed it into the group head and the group tail and the nearby data, namely the flag group data, and use the flag group data to construct a flag check matrix, and generate a watermark through the matrix and embed it into the flag group data, in order to ensure the robustness of the group; the other case is to embed it into the remaining data of the group, namely the inter-group data, and transform the non-repeating decimal sequence specified by the base station and the node as a watermark and embed it into the inter-group data; S5, transmitting the embedded data to the base station for restoration; Data recovery authentication must meet both the reversible digital watermark authentication conditions and the flag watermark authentication conditions. If one of the conditions is not met, the authentication is considered unsuccessful. Since the effective precision bits of the group header data are not changed when the watermark is embedded, the watermark sequence W' to be verified is generated using the group header data according to the calculation formula of the watermark sequence W, and compared with the reversible digital watermark extracted from the data Si". At the same time, the data flag bit must be authenticated. If the data verification is unsuccessful, double authentication of the individual data is performed. The double authentication of the individual data includes: dividing the data in the remaining data set with the effective precision of the watermark and the flag watermark embedded in the group after the double watermark is embedded into small groups, collecting and judging the attack situation of each two data. If attacked, a malicious attack is detected, and the first data of the corresponding current group is marked as unverified, and the second data of the current group and the next data are combined into a new group for re-judgment.

2. According to claim 1, a WSNs data integrity authentication method based on double watermark is characterized in that: The dynamic grouping includes: The sink node and the relay node are both set with an array of specified length, and a group leader with irregular data length related to the array is randomly generated.

3. A WSNs data integrity authentication method based on double watermark according to claim 1, characterized in that: The grouping determination includes: The queue data of the specified length in the data stream are allowed to enter in sequence, and it is determined whether there is data that meets the group head and group tail flag calculation scheme. If it does, the data in the queue is used to form a flag check matrix. If the specified number of data flags in the matrix are verified to be correct, the data that meets the requirements of the group head and group tail flag calculation scheme will be used as the head node data and tail node data respectively.

4. A WSNs data integrity authentication method based on double watermark according to claim 1, characterized in that: The embedding of the flag watermark includes: Embed the watermark into a special flag in a special data set; and The watermark is embedded into the normal flag bit in the remaining data set.

Images