CN117675990A - Smart phone continuous identity authentication method based on biological behaviors - Google Patents
Smart phone continuous identity authentication method based on biological behaviors Download PDFInfo
- Publication number
- CN117675990A CN117675990A CN202311844503.2A CN202311844503A CN117675990A CN 117675990 A CN117675990 A CN 117675990A CN 202311844503 A CN202311844503 A CN 202311844503A CN 117675990 A CN117675990 A CN 117675990A
- Authority
- CN
- China
- Prior art keywords
- user
- touch
- behaviors
- smart phone
- touch behavior
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 230000004791 biological behavior Effects 0.000 title claims abstract description 11
- 230000006399 behavior Effects 0.000 claims description 74
- 239000013598 vector Substances 0.000 claims description 18
- 238000012549 training Methods 0.000 claims description 12
- 238000012795 verification Methods 0.000 claims description 11
- 238000013145 classification model Methods 0.000 claims description 10
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 2
- 238000010187 selection method Methods 0.000 claims description 2
- 238000012360 testing method Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims 2
- 238000013480 data collection Methods 0.000 abstract 1
- 230000004927 fusion Effects 0.000 abstract 1
- 238000011160 research Methods 0.000 description 6
- 238000011161 development Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 241000700605 Viruses Species 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
Abstract
A smart phone continuous identity authentication method based on biological behaviors can authenticate the identity of a holder of a smart phone in real time. At present, the continuous authentication of the smart phone is carried out by adopting single touch data or a motion sensor, the single touch data is difficult to accurately identify the identity of the user, and the continuous data collection is required by using the motion sensor, so that the power consumption of the smart phone is increased. The invention has the innovation point that the fusion authentication of multiple touch data is realized by utilizing the combination of the touch data types of the multiple different gestures intelligently collected at present and the time sequence, and the accuracy of continuous identity authentication of the smart phone is ensured on the premise of improving user experience.
Description
Technical Field
The invention relates to the field of biological identification and identity authentication.
Background
With the development of mobile and intelligent industries, the use of smart phones has shown explosive growth [1] . Smartphones have become an indispensable part of our daily lives, occupy an important role in modern society, and currently billions of smartphones are in use. With the development of the latest technology, their capabilities and demands are increasing. However, the availability of the latest technology makes mobile phones vulnerable to various security threats, which means that users must pay attention to the security and privacy of mobile data at all times, where user identification has become a key component to ensure security and privacy. The main method of smart phone access is limited to the method of guiding identity verification through an Entry Point (Point-of-Entry) [2] For example, entering a PIN or password, locating or scanning a fingerprint using a stroke and pattern, facial recognition [3-6] . Smart phone user authentication based on passwords, pin codes and touch modes causes some security problems [7] One of their most major limitations is that users are authenticated only once when they log into the smartphone, and thus in the lost or unlocked state of the smartphone, the data stored therein is at risk of being hacked. In addition, some applications may require additional security layers and methods to enhance the user authentication process, such as two factorsAuthentication, however, such multi-factor approaches tend to create a poor user experience due to invasiveness.
In order to overcome the security problem of the conventional user identity recognition method, in recent years, one trend in the authentication field is to use an active recognition mechanism to recognize the behavior and context of a user, and allow an attacker to be accessed or discovered, thereby reducing illegal intrusion [8] . These active mechanisms require that the user's session data be stored, encoded, and matched with historical data to achieve identification. In the method of research, continuous authentication (continuous authentication, CA) is defined as a method and technique in which an authentication system effectively and reliably authenticates, verifies and identifies an individual by collecting detailed information of the physical properties and/or behavior patterns of the individual throughout a session [9] . CA is a strategy that can greatly improve security while also ensuring good user experience [10] 。
The current continuous authentication research method mainly adopts biological behavior characteristics to realize identity authentication, wherein the method comprises mobile equipment touch screen interaction, multi-mode sensor data and the like. The current mainstream research has the following problems:
1. most of the current touch behavior research is based on single touch behaviors, the behavior characteristics are sliding or clicking, the method ignores the help of different touch behaviors to the continuous authentication process, and in addition, the current research does not explain weights in the continuous authentication process for various characteristics of the single touch behaviors. The method of using partial features or zooming-in and zooming-out features can improve the continuous authentication performance to a certain extent.
2. Current research on mobile phone sensor characteristics easily ignores the influence of real-time performance in the continuous authentication process of the smart phone. In the continuous authentication process, the security of the smart phone may be reduced due to a longer authentication period, and the possibility of occurrence of abnormal data may be higher due to an excessively long authentication time, thereby causing the degradation of authentication quality. Meanwhile, the smart phone equipment which is sensitive to power consumption is greatly affected by acquiring the data of the mobile phone sensor in real time, and the long-time acquisition of the data can possibly lead to shortening of the service time of the mobile phone and influence on user experience.
Disclosure of Invention
The smart phone continuous authentication is to ensure the security and reliability of the smart phone. The continued authentication of the smart phone may prevent unauthorized persons from accessing the phone, thereby protecting personal information and privacy. In addition, the continuous authentication can also ensure the integrity of mobile phone software and hardware and prevent the invasion of malicious software and viruses. Thus, continuous authentication is an important step in ensuring the security and reliability of a mobile phone. In order to solve the defects of the continuous authentication of the smart phone, improve the authentication quality and strengthen the user identity security, the invention provides a method for continuous authentication of the smart phone based on biological behaviors, which introduces various user touch behaviors and realizes more efficient continuous authentication of the smart phone. In order to verify the proposed solution, the implementation steps are as follows:
step 1, constructing a user touch behavior sequence according to a user touch behavior log of the intelligent mobile phone;
step 2, performing Robust Scale and PCA operation on the user touch behaviors, and converting the user touch behaviors into user touch behavior feature vectors;
and step 3, inputting the user touch behavior feature vector into the single classification model to enable the model to learn the user touch behavior feature.
And 4, continuously authenticating the identity of the user touch behavior sequence by using the trained single classification model of the plurality of touch types.
Drawings
The drawings are for describing in detail the experimental steps of the method of the present invention. In the drawings:
FIG. 1 is a technical roadmap of a continuous identity authentication method of a smart phone based on biological behaviors;
fig. 2 is a diagram of a continuous identity authentication model of a smart phone.
Detailed Description
For a more complete description of the objects, technical routes and advantages of the present invention, reference should be made to the following detailed description of the present invention taken in conjunction with the accompanying drawings.
The specific embodiment of the invention is shown in fig. 1, and the specific steps are as follows:
step 1, constructing a user touch behavior sequence according to a user touch behavior log of the intelligent mobile phone;
step 2, performing Robust Scale and PCA operation on the user touch behaviors, and converting the user touch behaviors into user touch behavior feature vectors;
and step 3, inputting the user touch behavior feature vector into the single classification model to enable the model to learn the user touch behavior feature.
And 4, continuously authenticating the identity of the touch behavior of the user by using the trained single classification model of the plurality of touch types.
The detailed description of the individual steps is as follows:
step 1, constructing a user touch behavior sequence according to a user touch behavior log of a smart phone;
1.1, dividing the HMOG data into touch behavior data in all data according to different users and different touch behavior types.
1.2 forming a touch behavior sequence X by sorting the divided user touch behaviors according to the time sequence of the user touch behaviors, wherein X is a sequence formed by n touch behaviors and is expressed as
X＝[touch 1,touch 2,…,touchn]
Step 2, performing Scale and PCA operation on the user touch behaviors, and converting the user touch behaviors into user touch behavior feature vectors;
2.1 split each type of touch behavior per user in units of each user.
2.2, respectively extracting the characteristics of each type of touch behavior of each user by using a characteristic selection method, wherein the characteristics comprise Scale and PCA, and the influence of abnormal data is eliminated by using a Robust Scale method. PCA is used for extracting the characteristic vector of the touch behavior of the user, so that the characteristic quantity is reduced.
Step 3, inputting the user touch behavior feature vector into a single classification model to enable the model to learn the user touch behavior feature;
and 3.1, taking a single touch behavior feature vector of a single user as a training object, wherein the training set is only from the single user and only contains positive samples, a verification set is formed by the single user and other users, the single touch behavior feature vector of the single user is taken as legal sample data, the other users are taken as illegal sample data, and the test set is designed as the verification set.
3.2 inputting the training set and the verification set in 3.1 into an One Class SVM model, enabling the model to learn the touch behavior characteristics of the user through cross verification, enabling the model learning target to obtain a hyperplane capable of distinguishing whether the user is a legal user or an illegal user, and enabling the One Class SVM model optimization target to be expressed as follows:
where ζ is the relaxation variable, v is the fraction of outliers that sets an upper bound (within the training dataset considered outliers), also is the lower bound for the number of samples within the training dataset as support vectors, ρ is the offset value, i is the sample index, x is the training samples, and n is the total number of samples.
Step 4, continuously authenticating the touch behavior of the user by using the trained single classification model with a plurality of touch types:
and 4.1, respectively using different One Class SVM models for verification according to different touch behavior types of users.
Claims (5)
1. A smart phone continuous identity authentication method based on biological behaviors is characterized by comprising the following steps of
Step 1, constructing a user touch behavior sequence according to a user touch behavior log of the intelligent mobile phone;
step 2, performing Robust Scale and PCA operation on the user touch behaviors, and converting the user touch behaviors into user touch behavior feature vectors;
and step 3, inputting the user touch behavior feature vector into the single classification model to enable the model to learn the user touch behavior feature.
And 4, continuously authenticating the identity of the touch behavior of the user by using the trained single classification model of the plurality of touch types.
2. The continuous identity authentication method of a smart phone based on biological behaviors as claimed in claim 1, wherein the step 1 of constructing a user touch behavior sequence according to a smart phone user touch behavior log comprises the following steps:
2.1 dividing the touch behavior data in all the data according to the HMOG data set and different users and different touch behavior types.
2.2 forming a touch behavior sequence X by sorting the divided user touch behaviors according to the time sequence of the user touch behaviors, wherein X is a sequence formed by n touch behaviors and is expressed as
X＝[touch 1 ,touch 2 ,…,touch n ]
3. The smart phone continuous identity authentication method based on biological behaviors according to claim 1, wherein the step 2 of performing Scale and PCA operation on the user touch behaviors, converting the user touch behaviors into user touch behavior feature vectors includes the steps of:
3.1 splitting each type of touch behavior per user in units of each user.
And 3.2, respectively carrying out feature extraction on each type of touch behaviors of each user by using a feature selection method, wherein the feature extraction comprises Scale and PCA, and the influence of abnormal data is eliminated by a Robust Scale method. PCA is used for extracting the characteristic vector of the touch behavior of the user, so that the characteristic quantity is reduced.
4. The continuous identity authentication method of smart phone based on biological behavior according to claim 1, wherein the step 3 of inputting the user touch behavior feature vector into a single classification model to enable the model to learn the user touch behavior feature comprises the following steps:
4.1 training objects are single touch behavior feature vectors of a single user. The training set is only from a single user, only contains positive samples, a verification set is formed by the single user and other users, the touch behavior feature vector of the single user is used as legal sample data, the other users are used as illegal sample data, and the test set is designed to be the same as the verification set.
4.2, inputting the training set and the verification set in 4.1 into the One Class SVM model, and after cross verification, enabling the model to learn the touch behavior characteristics of the user. The model aims at obtaining a hyperplane capable of distinguishing whether the user is a legal user or an illegal user, and the One Class SVM model optimization targets are expressed as follows:
s.t.(ω T φ(x i ))＞ρ-ζ i ,i＝1,…,n
ζ i ＞0
where ζ is the relaxation variable, v is the fraction of outliers that sets an upper bound (within the training dataset considered outliers), also is the lower bound for the number of samples within the training dataset as support vectors, ρ is the offset value, i is the sample index, x is the training samples, and n is the total number of samples.
5. The continuous identity authentication method of a smart phone based on biological behaviors according to claim 1, wherein the step 4 of using a trained single classification model of a plurality of touch types to continuously identify the touch behaviors of the user comprises the following steps:
and 5.1, respectively using different One Class SVM models for verification according to different touch behavior types of users.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311844503.2A CN117675990A (en) | 2023-12-29 | 2023-12-29 | Smart phone continuous identity authentication method based on biological behaviors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311844503.2A CN117675990A (en) | 2023-12-29 | 2023-12-29 | Smart phone continuous identity authentication method based on biological behaviors |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117675990A true CN117675990A (en) | 2024-03-08 |
Family
ID=90080955
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311844503.2A Pending CN117675990A (en) | 2023-12-29 | 2023-12-29 | Smart phone continuous identity authentication method based on biological behaviors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117675990A (en) |
-
2023
- 2023-12-29 CN CN202311844503.2A patent/CN117675990A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yi et al. | Glassgesture: Exploring head gesture interface of smart glasses | |
| Li et al. | Unobservable re-authentication for smartphones. | |
| Biggio et al. | Adversarial biometric recognition: A review on biometric system security from the adversarial machine-learning perspective | |
| Giuffrida et al. | I sensed it was you: authenticating mobile users with sensor-enhanced keystroke dynamics | |
| Li et al. | Using feature fusion strategies in continuous authentication on smartphones | |
| US10108792B2 (en) | Biometric-based authentication method, apparatus and system | |
| US20170227995A1 (en) | Method and system for implicit authentication | |
| Shahzad et al. | Behavior based human authentication on touch screen devices using gestures and signatures | |
| Lu et al. | Safeguard: User reauthentication on smartphones via behavioral biometrics | |
| CN105068743B (en) | Based on the mobile terminal user identity authentication method for more referring to touch-control behavioural characteristic | |
| CN106572097B (en) | A kind of mobile device-based hybrid identity identifying method | |
| Wang et al. | Improving reliability: User authentication on smartphones using keystroke biometrics | |
| US9202035B1 (en) | User authentication based on biometric handwriting aspects of a handwritten code | |
| Sun et al. | A 3‐D hand gesture signature based biometric authentication system for smartphones | |
| Smith-Creasey et al. | A continuous user authentication scheme for mobile devices | |
| Yang et al. | PersonaIA: a lightweight implicit authentication system based on customized user behavior selection | |
| Giot et al. | Keystroke dynamics authentication | |
| CN111625789A (en) | Multi-core learning fusion mouse and keyboard behavior feature-based user identification method | |
| US11611881B2 (en) | Integrated systems and methods for passive authentication | |
| Stockinger | Implicit authentication on mobile devices | |
| CN116437355A (en) | Radio frequency fingerprint-based wireless equipment identity authentication method and device | |
| Buriro et al. | SWIPEGAN: swiping data augmentation using generative adversarial networks for smartphone user authentication | |
| Yang et al. | Bubblemap: Privilege mapping for behavior-based implicit authentication systems | |
| Zhang et al. | Multimodal continuous user authentication on mobile devices via interaction patterns | |
| Saifan et al. | A Survey of behavioral authentication using keystroke dynamics: Touch screens and mobile devices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |