Implicit User Re-authentication for Mobile Devices

Question

1. What have the authors contributed in "Implicit user re-authentication for mobile devices" ?

2. how is the proposed system able to distinguish between normal use and attack?

3. Why do the authors need to map some fields to integer values?

4. How much energy is consumed by each run of the anomaly detection process?

Accepted Answer

The authors propose an implicit user re-authentication system for portable computers that requires no application changes or hardware modifications.. The authors describe these automated model generation and user detection techniques, and explain how to efficiently implement them in a wireless distributed system composed of servers and battery-powered portable devices.

Accepted Answer

The proposed system is able to distinguish between normal use and attack with an accuracy of approximately 90% with a latency of 5 minutes.

Accepted Answer

Due to the limitation of the software that the authors use for model construction [19], the authors need to map some of the fields to integer values such as location, source IP, and destination IP fields.

Accepted Answer

The energy consumption by each run of the anomaly detection process isEanomaly = Panomaly × Tanomaly = 22.45 W × 20 s = 449 J. (11)Given 5-minute detection latency, the total daily energy consumption of anomaly detection is 129.3 kJ, i.e., 42.6% of Ebat .

Accepted Answer

The proposed implicit re-authentication system provides a way to protect sensitive information on these computers from attack without inconveniencing the user.

Accepted Answer

Attacker #3 did not produce enough network data for identification, resulting in a FAR of 94.1% when the single-metric network-based technique was used.

Accepted Answer

Pusara [6] built a signature-based intrusion detection system based on a combination of keystroke dynamics, mouse movements, and graphical user interface events, in which the attacker’s behavior was known a priori.

Accepted Answer

The proposed system is able to distinguish between normal use and attack with an accuracy of approximately 90% with a detection latency of 5 minutes for network activity and 10 minutes for file access.

Accepted Answer

Experimental results indicate that the authors can achieve an FAR of 13.7% and an FRR of 11% with a detection latency of 5 minutes by combining filesystem and network based re-authentication techniques.

Accepted Answer

a low threshold value would result in low false acceptance rate (FAR) at the cost of high FRR, while a high threshold value corresponds to a low FRR and a high FAR.

Accepted Answer

Since the log file has to be sent to the server every 5 minutes (see Section 6), or 288 times per day, the total daily energy consumption in the feature extraction process is 1.324 kJ.

Accepted Answer

For each record, it computes the Euclidean distance between the record and the K cluster centers and assigns the record to the closest cluster.

Accepted Answer

With three months of training data, the accuracy of the neural network based approach did not exceed 84%, while the accuracy of the K-means clustering based approach increased to 98% within 90 days.

Accepted Answer

In the absence of a wireless 1 The detection latency T should be small enough to detect an attack before theattacker is capable of harming the system.

Accepted Answer

since there are sufficient data for the filesystem model to accurately detect the attack during these periods, the multiple-metric approach achieves a FAR of 11.8%.

Implicit User Re-authentication for Mobile Devices

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What have the authors contributed in "Implicit user re-authentication for mobile devices" ?

2. how is the proposed system able to distinguish between normal use and attack?

3. Why do the authors need to map some fields to integer values?

4. How much energy is consumed by each run of the anomaly detection process?

5. What is the proposed implicit user re-authentication system?

6. What is the FAR of the attacker #3?

7. What is the recent example of a signature-based intrusion detection system?

8. how accurate is the detection of attacks on portable computers?

9. How did the researchers achieve an FAR of 13.7%?

10. What is the threshold value for a low false acceptance rate?

11. How many times per day does the feature extraction process consume energy?

12. What is the algorithm's method for assigning the record to the closest cluster?

13. How many years of training did the accuracy of the neural network based approach exceed 98%?

14. How is the detection latency of a portable computer system determined?

15. How does the multiple-metric approach achieve a FAR?

Figures

Citations

Touch me once and i know it's you!: implicit authentication based on touch screen patterns

It's a Hard Lock Life: A Field Study of Smartphone (Un)Locking Behavior and Risk Perception

Continuous and transparent multimodal authentication: reviewing the state of the art

Security, Privacy, and Usability in Continuous Authentication: A Survey.

WearIA: Wearable device implicit authentication based on activity information

References

A K-Means Clustering Algorithm

An Intrusion-Detection Model

Data Compression Using Adaptive Coding and Partial String Matching

Intrusion detection using neural networks and support vector machines

Intrusion detection via static analysis

Related Papers (5)

Implicit authentication through learning user behavior

Touch me once and i know it's you!: implicit authentication based on touch screen patterns

Modeling user search behavior for masquerade detection

Authenticating mobile phone users using keystroke analysis

Protecting private data on mobile systems based on spatio–temporal analysis