Method and system for protecting the user of a pointing device from identification

Abstract

The invention relates to the field of computing and can be used for protecting the user of an information system from identification on the basis of their data input dynamics when using a pointing device. The technical result of the claimed invention is the possibility of concealing a user's pointing device dynamics by distorting the cursor trajectory set by the user, thus making it possible to effectively protect the user from identification on the basis of their data input dynamics when using a mouse or a pointing device having a similar function. The invention envisages a method for protecting the user of an information system from identification on the basis of their data input dynamics when using a pointing device, which consists in introducing random distortions with the aid of technical means into the coordinate signals from the pointing device before said signals are input into a computer, and further envisages a system for protecting a user from identification.

Description

[0001] METHOD AND SYSTEM FOR PROTECTION AGAINST IDENTIFICATION OF A MANIPULATOR USER

[0002] AREA OF TECHNOLOGY

[0003] The invention relates to the field of computer technology and can be used to protect the user of an information system from identification by handwriting of information entered from a mouse manipulator and manipulators with a similar function (trackball, touchpad, touch screen).

[0004] LEVEL OF TECHNOLOGY

[0005] A known solution, "Computer Mouse and Cursor Control Method" (patent CN101008874, published August 1, 2007), modifies the trajectory of a cursor on the screen, determined by a mouse, while accounting for the actual mouse speed and its impact on the actual mouse movement. A drawback of this known method is the ability to identify the user based on the cursor trajectory. Another drawback of this known method is the ability to identify the user based on the intervals between mouse button presses and releases, which, like the cursor trajectory, carry information about the user's identity and enable their identification.

[0006] A solution known as "Method and apparatus for controlling movement of cursor" (US Patent 5,477,236, published December 19, 1995) ensures the movement of the cursor of a mouse or trackball manipulator along a line close to the line displayed on the screen. The disadvantage of the known method is

[0007] SUBSTITUTE SHEET (RULE 26) the need to constantly display on the screen a line to which the cursor will approach, which creates inconvenience for the user.

[0008] Another known solution is "Method, device, and system for detecting mule accounts and accounts used for money laundering" (US 2022 / 0108319, published April 7, 2022). This solution proposes training legitimate users to react to intentionally distorted mouse cursor movements during system login. Upon login, the user's reaction to the distorted mouse cursor movement is checked, and if the user reacts incorrectly, login is denied. A drawback of this known method is that the distortions are introduced into the cursor position displayed on the user's screen, not into the cursor position transmitted to the information system. A drawback of this known solution is that it addresses the problem of user identification, but does not address the problem of protecting the information system user from identification based on the handwriting used to enter information from the mouse.

[0009] DISCLOSURE OF THE ESSENCE OF THE INVENTION

[0010] With the growing role of Internet technologies in society, the issue of protecting users' personal information transmitted online from unauthorized use by third parties is becoming increasingly pressing.

[0011] A user's interests are determined primarily based on their search queries and websites visited. The web browser records a history of visited websites and entered queries, which is accessible to the websites visited.

[0012] SUBSTITUTE SHEET (RULE 26) To keep statistics of visited websites and requests, if the web browser history is cleared when it is closed, websites try to uniquely identify the user by his actions in the current session.

[0013] In each website connection, the user is first recognized based on their actions in the current session, and then the user's history of previous sessions stored on a remote server is supplemented by the current session's history. As a result, information about the internet user's actions over an arbitrarily long period of time is stored on multiple remote servers. This also increases the potential for the use of internet user information for illegal purposes. For example, if a user enters their passport information on a website, each subsequent visit will identify them not simply as a unique anonymous user with a specific set of interests, but as a real person, which will pose a wide range of security threats.

[0014] One way to match the identity of an anonymous user to a unique characteristic is to analyze the "handwriting" of their computer inputs. One component of this "handwriting" is the statistical characteristics of the mouse cursor or similar device's movement trajectory. User identification based on the analysis of a user's "handwriting" is achieved with over 90% certainty. Therefore, protecting this information from misuse is a pressing issue.

[0015] The technical result of the claimed invention is the possibility of hiding the user's handwriting by distorting it

[0016] SUBSTITUTE SHEET (RULE 26) of the cursor trajectories specified by it, which ensures effective protection of the user from identification of his identity by the handwriting of the information entered using a “mouse” or a manipulator with a similar function.

[0017] The claimed technical result is achieved by intercepting signals from the manipulator before they are sent to the computer, introducing controlled random changes to the manipulator's cursor displacement, and transmitting them to the computer in a distorted form. The probability distributions of the random process corresponding to the distorted cursor trajectory are specified during system configuration. The distorted trajectory, however, remains "human-like," making it difficult for the information system to detect the use of a system to protect against user identification based on input handwriting.

[0018] The claimed technical result is achieved by using a method for protecting against user identification when entering data from a manipulator, which consists in the fact that data representing the values ​​of the dispersion and correlation function are entered into the processor 3 via the interface unit 2; the manipulator 1 periodically transmits messages to the processor 3 via the interface unit 2, containing the horizontal and vertical movements of the manipulator 1; the processor 3 calculates random distortions of the horizontal and vertical movement values ​​of the manipulator 1; the processor 3, via the interface unit 2, enters data representing the distorted signals of the manipulator coordinates into the computer 4.

[0019] In one embodiment of the method, the processor 3 calculates random distortions of the values ​​of the manipulator's movement using the convolution of a white Gaussian noise vector with the impulse response of a linear

[0020] SUBSTITUTE SHEET (RULE 26) of a filter that transforms white noise into noise with a given correlation function.

[0021] In one embodiment of the method, the processor 3 calculates the transmitted signals of the manipulator coordinates using a neural network trained on the trajectories of the manipulator movements of real users (see, for example, https: / / qudata.com / ru / news / how-do-neural-networks-learn-motion-interpretation-of-motion-modeling / ).

[0022] In one embodiment of the method, processor 3 checks, using Student's criterion, the statistical hypothesis that the mean vector of the manipulator's movement is equal to zero.

[0023] In one embodiment of the method, the variance of added distortions is reduced by accepting the statistical hypothesis that the average vector of the manipulator's movement is equal to zero.

[0024] In one embodiment of the method, the variance of added distortions when rejecting the statistical hypothesis that the average vector of the manipulator's movement is equal to zero is restored to the value previously entered into processor 3.

[0025] In one embodiment of the method, the processor 3 additionally transmits to the computer 4 data representing the actual coordinate signals of the manipulator.

[0026] In one embodiment of the method, processor 3 transmits data representing distorted signals of the manipulator coordinates to computer 4 immediately after calculation.

[0027] In one embodiment of the method, processor 3 generates a time delay for a certain time interval, receives the initial position of the manipulator at the moment the delay begins and the final position at the moment the delay ends, and at the moment the delay ends, calculates the points of the trajectory of the manipulator's movement from the initial position to the final position.

[0028] SUBSTITUTE SHEET (RULE 26) position, without using the manipulator movements transmitted during the delay period, and transmits the calculated distorted signals of the manipulator coordinates to computer 4.

[0029] In one embodiment of the method, delays are generated for random time intervals distributed in accordance with a given distribution of random numbers.

[0030] In one embodiment of the method, delays at random time intervals are uniformly distributed in the range of 0 - 100 milliseconds.

[0031] In one embodiment of the method, delays at random time intervals are distributed according to Rayleigh's law.

[0032] In one embodiment of the method, delays at random time intervals are distributed according to Gauss's law, and if the random number is less than zero, it is set to zero.

[0033] In one embodiment of the method, delays at random time intervals are distributed exponentially.

[0034] The claimed technical result is also achieved by using a system for protecting against user identification when entering data from a manipulator, implementing the method according to paragraph 1, comprising a manipulator 1, an interface unit 2, a processor 3, a computer 4, a display 5, a random number generator 6, wherein the manipulator 1, the processor 3, the computer 4, the random number generator 6 are connected to the interface unit 2 and exchange data through it, the display 5 is connected to the computer 4, wherein the processor 3 is configured to receive the movements of the manipulator 1 and random numbers from the generator 6 and to distort the movements of the manipulator 1 and transmit the real and distorted movements of the manipulator 1 to the computer 4, the computer 4 is configured to receive the movements of the manipulator 1 from the processor 3 and display the real movements of the manipulator 1 on the display 5 using

[0035] SUBSTITUTE SHEET (RULE 26) of the additional cursor and distorted movements of the manipulator 1 using the main cursor.

[0036] BRIEF DESCRIPTION OF DRAWINGS

[0037] Fig. 1 shows the structural diagram of the system for protecting against user identification based on the handwriting of the information entered from the manipulator.

[0038] Fig. 2a shows examples of the trajectories of the manipulator cursor specified by the user (user trajectory) and the corresponding trajectories distorted according to the claimed method when configured to obtain trajectories with high smoothness and a large average deviation from the user trajectory.

[0039] Fig. 26 shows examples of the trajectories of the manipulator cursor specified by the user (user trajectory) and the corresponding trajectories distorted according to the claimed method when configured to obtain trajectories with less smoothness and a smaller average deviation from the user trajectory.

[0040] Fig. 2c shows an example of a manipulator cursor trajectory specified by the user (user trajectory), a trajectory distorted according to the claimed method, and a trajectory distorted by adding white noise.

[0041] Fig. 3 shows possible time diagrams of the change in the speed V of the manipulator movement and the variance D of the noise introduced by the processor and the region of acceptance of the hypothesis H1 about the cursor approaching the stopping point and the opposite hypothesis H1.

[0042] Fig. 4 shows the trajectories of the cursors as the manipulator moves from point A to point B.

[0043] SUBSTITUTE SHEET (RULE 26) The system for protecting against user identification based on the handwriting of the information entered contains a manipulator 1, an interface unit 2, a processor 3, a computer 4, a display 5, and a random number generator 6.

[0044] The following designations are used: main cursor 7, additional cursor 8, trajectory 9, distorted according to the claimed method (distorted trajectory), trajectory 10, specified by the user (user trajectory), trajectory 11, distorted by adding white noise.

[0045] Computer 4 can be any type of computing device capable of connecting an external manipulator. The manipulator is connected to processor 3 and computer 4, and processor 3 is connected to random number generator 6 via interface unit 2.

[0046] IMPLEMENTATION OF THE INVENTION

[0047] During normal interaction between a computer and mouse-type manipulators and similar devices, the manipulator, at equal time intervals At (usually I = 10 ms), sends a message to one input of the input-output port, which contains the movements and states of the manipulator buttons, and to the other input - a synchronization signal, which allows the computer to determine the need to receive the message.

[0048] The main characteristics of a user's handwriting that allow for identification are:

[0049] - The distance traveled from one stopping point of the manipulator to another;

[0050] - Manipulator movement speed;

[0051] - Angular velocity of the manipulator;

[0052] SUBSTITUTE SHEET (RULE 26) - Curvature of the manipulator's trajectory.

[0053] The method of protection against user identification based on the handwriting of the information input allows for the effective concealment of all the listed characteristics of the user's handwriting by introducing distortions into the cursor trajectory as a function of time, and, consequently, into all the derived characteristics of the trajectory.

[0054] When the user begins moving the manipulator, to protect against identification, white Gaussian noise can be added to the manipulator movements received in the message. The resulting movement di (i=l,2) can then be represented as where c? / is the movement corresponding to the user's intention, bi is the movement corresponding to the user's noise, c, is the noise generated by the random number generator b with the variance Dr, here and below i=1 corresponds to the horizontal movement, / =2 corresponds to the vertical movement.

[0055] Distorted coordinate signals from the manipulator can be transmitted to the computer immediately after calculation or with a delay. The proposed method implements these two operating modes (instantaneous point transmission mode and delayed point transmission mode).

[0056] The second option introduces a time delay between the actual movement of the manipulator and the moment the signal about the movement is transmitted to the computer, which further complicates the identification of the information system user based on the manipulator's input patterns. In this case, the initial position of the manipulator is determined at the beginning of the delay and the final position of the manipulator at the end of the delay. After the delay ends,

[0057] SUBSTITUTE SHEET (RULE 26) calculates the trajectory of the manipulator's movement from the initial position to the final position, independent of the user's actual movements between these positions. The points of the calculated trajectory are transmitted to the computer. For example, the initial and final points can be connected by a straight line, along which a set of transmitted points is selected and noise is added to them before transmission. Thus, during the delay period, there is no leakage of information about the user's handwriting, even in the form of a noisy trajectory.

[0058] When working according to the second option in the algorithm described below, in formula (1), when calculating the trajectory transmitted to the computer, the movement of the manipulator cn+b received in the message is taken to be the movement along the calculated straight line or other deterministic line connecting the points of the beginning of the delay and the end of the delay.

[0059] Furthermore, the trajectory between the start and end points of the delay can be calculated using a neural network. In this case, the neural network, programmed into processor 3, is pre-trained to transform the white noise vector into a manipulator trajectory similar to the trajectories of real users from the training set (see, for example, https: / / qudata.com / ru / news / how-do-neural-networks-learn-motion-interpretation-of-motion-modeling / ). Such a neural network, in its calculated trajectory, takes into account the general properties of a person's operating style with a manipulator, rather than replicating the style of a specific user from the training set.

[0060] Transferring movements (1) to the computer already provides protection from user identification, but it will be obvious to the receiving party that the manipulator's trajectory was generated automatically (see Fig. 2c). This is undesirable, as it can lead to failure of the information

[0061] SUBSTITUTE SHEET (RULE 26) for the system from working with a given user. To give the trajectory a human-like shape, new functions must be implemented.

[0062] Using the method of mathematical modeling and comparison with real user trajectories, the author found that human mouse movement is well described by a Gaussian random process with a normalized correlation function of the form where t=0, At, 2At, ... , (sl)At, At is the time interval between messages transmitting the movements of the manipulator, ₽i>0, yi>l, ₽2>0, Y2>1, s=l,2,3,....

[0063] Using the convolution spectrum theorem and the Wiener-Khinchin theorem, we can establish that a random process with a given correlation function Ri is obtained by convolving a white Gaussian noise vector with an impulse response hi, calculated using the formula where F is the Fourier transform.

[0064] Thus, the process of displacement of the manipulator Si with the correlation function / ?, and unit variance is calculated by the formula where hi(j) is the j-th sample of the impulse response (3), c (w) is the i-th element of the white noise vector of length n, complemented by a vector of zeros of the same length.

[0065] A process with dispersion Di is obtained by multiplying Si by , / 5^.

[0066] Using formula (4), it is possible to calculate random movements of the manipulator when the user is idle, if necessary,

[0067] SUBSTITUTE SHEET (RULE 26) substituting random numbers obtained from generator 6 as di. When the manipulator moves, the displacement can be calculated using the same formula (4), substituting di (1) into it.

[0068] According to formulas (1) and (4), when the manipulator moves, the distorted trajectory Si(m) consists of a smoothed trajectory cii(m), corresponding to the user's intention, and a noise component bi(m)+d(m) with a normalized correlation function / ?, and a dispersion somewhat increased due to bi(m) compared to the dispersion Di of the component Ci(m), obtained from the random number generator 6. It should be noted that the correlation function / ?, is unknown to the receiving party (information system) and can change in each session, therefore, the user's trajectory cannot be reconstructed by the receiving party even with a small amount of added noise c.

[0069] The type of the resulting trajectory depends on the parameters of the correlation functions Pi, yi, 02, Y2 and the variances of the introduced noise Di, D2. The larger yi, the stronger the simulated "hand tremor" effect along the i-th direction. The smaller 0i, the less the distorted trajectory fluctuates relative to the one specified by the user in a given direction. The greater the variance Di, the greater the average amplitude of the distortion in a given direction. By adjusting these parameters, it is possible to specify different user styles in terms of the characteristics listed above: the distance traveled between stopping points, speed, angular velocity, and curvature (see Fig. 2a and Fig. 2b). In particular, with an increase in 0, the characteristic distance traveled by the manipulator and the speed of the manipulator increase. With an increase in 0i, the average curvature of the trajectory increases. A simultaneous increase in Di and 0i allows for an increase in the average angular velocity.

[0070] SUBSTITUTE SHEET (RULE 26) When the user reaches the target of the manipulator's movement, the cursor movement must stop, and, consequently, any distortion of the trajectory must cease. This is achieved by reducing the variance of the added noise c / when the user reaches the target – see formula (1) – based on the following user behavior characteristics.

[0071] Upon seeing the proximity of the target, the user reduces the speed of the manipulator until it stops when the cursor hits the target, and the speed vector of the manipulator's movement will be, on average, close to zero.

[0072] To determine the need to reduce the variance of the introduced distortions at each n consecutive movements of the manipulator, the statistical hypothesis HO about the equality of the average vector of the manipulator velocity at the level of a% is tested using the Student's criterion. If the hypothesis is accepted, then the variance of the added noise decreases exponentially with the exponent -k and is equal to zero upon reaching the minimum value of AЬ. When the hypothesis HO is rejected, the variance increases according to the same law, increasing at most to the initial value of Di, starting from the minimum value of AЬ.

[0073] If the HO hypothesis is accepted, the variance will reach its minimum value Ai and be equal to zero, thus halting the introduction of distortions. In a% of cases, the hypothesis that the mean manipulator velocity vector is equal to zero will be incorrectly rejected. However, if the manipulator's further movement shows its velocity tending toward zero, the HO hypothesis will become accepted, and the noise variance will begin to decrease. When the manipulator begins moving toward a new target, the HO hypothesis will be continually rejected, and the variance will quickly restore its original value Di. Possible graphs of velocity change and the corresponding noise variance are shown in Fig. 3.

[0074] SUBSTITUTE SHEET (RULE 26) The method of protection against user identification by the handwriting of the information entered in the case of connection to the computer via a cable works as follows.

[0075] As a first, one-time step in implementing the claimed method, data representing the values ​​of the dispersion parameters and the correlation function (a>0, ₽i>0, yi>l, ₽2>0, Y2>1, Di>0, D2>0, 0) are input into the processor 3 through the interface unit 2. <Ai<Di, 0<^2<D2). Процессор 3 вычисляет и сохраняет в своей памяти п отсчётов корреляционных функций перемещений манипулятора по горизонтали Ri, перемещений манипулятора по вертикали R2 по формуле (2).

[0076] Processor 3 calculates and stores in its memory n readings of the impulse characteristics hi, 112 of the filter, which transforms white noise into a random process with a correlation function Ri, R2 according to formula (3).

[0077] In another implementation of the method, the first step involves loading a neural network, trained to transform a white noise vector into a manipulator trajectory similar to the trajectories of real users from a training set, into the processor 3 program. The neural network can take into account the introduced variances and correlation functions.

[0078] The following process then occurs.

[0079] The manipulator 1, via the interface unit 2, transmits to the processor 3 with a frequency of At messages containing the horizontal and vertical movements of the manipulator, the movements of the scroll wheel and the button codes.

[0080] Processor 3, using the input displacements and impulse response readings hi, calculates, using formula (4), the reading Si of the convolution of the vector of trajectory points di, calculated using formula (1), and the impulse response hi at / =1.2. In another embodiment, the readings Si

[0081] SUBSTITUTE SHEET (RULE 26) of the distorted trajectory is calculated using the neural network introduced into the processor program 3 at the first step.

[0082] (Si, S2) is the point of the new distorted trajectory of the manipulator, which the processor 3 transmits to the computer 4 through the interface unit 2.

[0083] During the implementation of the method, a check is made for the need to begin stopping the cursor movement by gradually reducing the variance of the added noise in the event of a slowdown in the movement of the manipulator, for example, when approaching a target, which may be a button in the graphical interface of a program or a web link.

[0084] This is implemented as follows. Processor 3 uses the Student's t-test to test the hypothesis HO about the equality of the mean cursor movement vector (u, 02) to zero in n cycles. If the hypothesis HO is accepted (the user stops the manipulator), then the variances Di and D2 in the memory of processor 3 are multiplied by e~ кIf the hypothesis HO is rejected (the user moves the manipulator quickly enough), then the variances Di and D2 in the memory of processor 3 are multiplied by e к If the result is greater than the initial value Di entered into processor 3, then the variance is set equal to the original value. If the result is less than Ai, then Di is set equal to zero. If the HO hypothesis is rejected, and Di = 0, then D / =Ai is set.

[0085] Thus, when using the method discussed above, if the user does not move the manipulator, processor 3 transmits zero movements to computer 4, and the cursor remains stationary. When the user moves the manipulator, processor 3 distorts the movements transmitted by the manipulator using the algorithm discussed above before transmitting them through interface unit 2.

[0086] SUBSTITUTE SHEET (RULE 26) computer 4. When the user reaches the target of the manipulator movement and, accordingly, reduces its speed, the dispersion of the introduced distortions decreases down to zero, and the distortion of the trajectory stops, allowing the user to hit the target with the cursor.

[0087] In one embodiment of the method, the processor 3 transmits to the computer 4 both the actual movements of the manipulator 1 previously received by it and the distorted movements of the manipulator 1 calculated by it. This is used by the computer to display the distorted movement of the manipulator 1 along the distorted trajectory 9 using the main cursor 7 and the actual movement of the manipulator 1 along the user's trajectory 10 using the additional cursor 8.

[0088] In one embodiment of the method, the processor 3 immediately after receiving the movement of the manipulator 1 calculates the distorted movement and transmits it to the computer 4.

[0089] In one embodiment of the method, processor 3 generates a time delay for a certain time interval, receives the initial position of manipulator 1 at the start of the delay and the final position at the end of the delay, calculates the points of the trajectory of movement of manipulator 1 from the initial position to the final position using formulas (1) and (4) or using a neural network at the end of the delay, and transmits the calculated distorted signals of the coordinates of manipulator 1 to computer 4. When calculating the trajectory points using formulas (1) and (4), the points of any line connecting the initial and final positions of the manipulator are used as the movements transmitted by the user. After transmitting the trajectory points to computer 4, a new initial position of manipulator 1 is received, a new delay is generated, and the process

[0090] SUBSTITUTE SHEET (RULE 26) is repeated. The duration of delays can be calculated according to a deterministic or random law.

[0091] Thus, the distortion of the cursor trajectory does not allow the identification of the user of the information system by the handwriting of the information entered from the manipulator.

[0092] The considered method can be used in a system for protecting against identification of an information system user based on the handwriting of the information entered from the manipulator. Manipulator 1, processor 3, computer 4, and random number generator 6 interact via interface unit 2. Computer 4 displays the information received from processor 3 on display 5. Processor 3 receives coordinate signals from manipulator 1, random numbers from generator 6 (an option is possible in which random numbers are generated by processor 3 without using generator 6), distorts the received coordinate signals using the proposed method, transmits the distorted coordinate signals to computer 4, which displays them on display 5 in the form of main cursor 7 and (if necessary) additional cursor 8. The computer can interact with arbitrary information systems, including the Internet, and transmit to them the distorted manipulator coordinate signals received from processor 3.Information about the user's handwriting is hidden from the information system by distorting the manipulator's coordinate signals.

[0093] Thus, the essential features of the claimed invention (method and system) ensure the achievement of the technical result.

[0094] Although the present invention has been described with respect to certain embodiments, it is not limited to a particular

[0095] SUBSTITUTE SHEET (RULE 26) in the form set forth in this document. On the contrary, the scope of the present invention is limited only by the appended claims. Furthermore, although a feature may apparently be described in relation to specific embodiments, those skilled in the art will understand that various features of the described embodiments may be combined in accordance with the invention. Furthermore, although listed separately, multiple means, elements, circuits, or method steps may be implemented using, for example, a separate circuit, unit, or processor. Furthermore, although individual features may be included in different claims, they may be effectively combined, and inclusion in different claims does not mean that the combination of features is not feasible and / or effective.Furthermore, the inclusion of a feature in one category of a claim does not imply a limitation to that category, but rather indicates that the feature is equally applicable to other categories of the claim, if appropriate. Furthermore, the order of features in the claims does not imply any specific order in which these features must be included. Reference signs in the claims are provided merely as illustrative examples and should not be construed as limiting the scope of the claims in any way.

[0096] SUBSTITUTE SHEET (RULE 26)

Claims

Invention formula 1. A method for protecting against user identification when entering data from a manipulator, which consists in the fact that data representing the values ​​of the dispersion and correlation function are entered into processor 3 via interface unit 2; manipulator 1 periodically transmits messages containing the horizontal and vertical movements of manipulator 1 to processor 3 via interface unit 2; processor 3 calculates random distortions of the horizontal and vertical movement values ​​of manipulator 1; processor 3 enters data representing distorted coordinate signals of manipulator 1 into computer 4 via interface unit 2.

2. The method according to paragraph 1, characterized in that the processor 3 calculates random distortions of the values ​​of the movement of the manipulator 1 using the convolution of the white Gaussian noise vector with the impulse response of a linear filter that converts white noise into noise with a given correlation function.

3. The method according to claim 1, characterized in that the processor 3 calculates the transmitted signals of the coordinates of the manipulator 1 using a neural network trained on the trajectories of movement of the manipulator 1 of real users.

4. The method according to paragraph 1, characterized in that the processor 3 checks, using the Student’s criterion, the statistical hypothesis that the average vector of movement of the manipulator 1 is equal to zero.

5. The method according to paragraph 4, characterized in that the variance of the added distortions is reduced when accepting the statistical hypothesis about SUBSTITUTE SHEET (RULE 26) equality of the average vector of movement of manipulator 1 to zero.

6. The method according to paragraph 4, characterized in that the variance of the added distortions, when rejecting the statistical hypothesis that the average vector of movement of manipulator 1 is equal to zero, is restored to the value previously entered into processor 3.

7. The method according to claim 1, characterized in that the processor 3 additionally transmits to the computer 4 data representing the actual coordinate signals of the manipulator 1.

8. The method according to claim 1, characterized in that the processor 3 transmits data representing the distorted signals of the coordinates of the manipulator 1 to the computer 4 immediately after the calculation.

9. The method according to claim 1, characterized in that the processor 3 generates a time delay for a certain time interval, receives the initial position of the manipulator 1 at the moment of the beginning of the delay and the final position at the moment of the end of the delay, at the moment of the end of the delay calculates the points of the trajectory of movement of the manipulator 1 from the initial position to the final position, without using the movements of the manipulator 1 transmitted during the delay period, and transmits the calculated distorted signals of the coordinates of the manipulator 1 to the computer 4.

10. The method according to paragraph 9, characterized in that delays are generated at random time intervals distributed in accordance with a given distribution of random numbers.

11. The method according to paragraph 10, characterized in that the delays at random time intervals are uniformly distributed in the range of 0 - 100 milliseconds. SUBSTITUTE SHEET (RULE 26) 12. The method according to paragraph 10, characterized in that the delays at random time intervals are distributed according to Rayleigh's law.

13. The method according to paragraph 10, characterized in that the delays at random time intervals are distributed according to the Gaussian law, and if the random number turns out to be less than zero, then it is set to zero.

14. The method according to paragraph 10, characterized in that the delays at random time intervals are distributed according to an exponential law.

15. A system for protecting against user identification when entering data from a manipulator, implementing the method according to paragraph 1, comprising a manipulator 1, an interface unit 2, a processor 3, a computer 4, a display 5, a random number generator 6, wherein the manipulator 1, the processor 3, the computer 4, the random number generator 6 are connected to the interface unit 2 and exchange data through it, the display 5 is connected to the computer 4, wherein the processor 3 is configured to receive the movements of the manipulator 1 and random numbers from the generator 6 and to distort the movements of the manipulator 1 and to transmit the real and distorted movements of the manipulator 1 to the computer 4, the computer 4 is configured to receive the movements of the manipulator 1 from the processor 3 and to display on the display 5 the real movements of the manipulator 1 using an additional cursor and the distorted movements of the manipulator 1 using the main cursor. SUBSTITUTE SHEET (RULE 26)

Images