Which statements are most accurate regarding presentation attacks? Select all that apply.

An adversary tries to alter the data sensed by the phasor measurement units installed on the power grid network so that the state estimation process estimates incorrect state vectors, which can cause some of the nodes to go down. Given this information, which type of attack is being introduced by the adversary?

Which attacks are forms of causative attacks? Select all that apply.

Which statements about the validity of biometrics are true? Select all that apply.

Imagine a scenario where Boeing designed a pilot monitoring system that deploys in response to a failing MCAS system. If an MCAS system is engaged and the pilot is detected to be stressed, MCAS will automatically disengage. In this system, consider a brain mobile interface application where the pilot wears a Neurosky headset that senses brain signals (EEG) at 400 Hz. Each brain data point is a 32-bit floating point number. The brain signal is collected by a central controller in the plane and sent to a server, where complex machine learning algorithms are employed to determine the stress level of the pilot. Additionally, the aircraft is equipped with sensors, such as the AoA, pitch monitoring, and other relevant sensors. The data rate from the AoA is 5 kbps, and the data rate from the other relevant sensors is 300 kbps. Using the data from these sensors, the MCAS disable system attempts to predict MCAS failures. If the system detects that the pilot is stressed and an MCAS failure is predicted, the auto-disable facility should disable MCAS. The auto-disable feature only has 5 seconds to make a decision after collecting 5 seconds worth of data. There are two options for performing all of the related computation: (a) use a GPU server at the control center, or (b) use a fog server that is onboard the aircraft. The GPU server upload speed is 1 Mbps, whereas the fog server upload speed is 5 Mbps. However, the computation speed of the GPU server is 1500 kbps (in other words, it can finish the computation on 1500 kb of data in 1 second), whereas the fog server has a computational speed of 200 kbps. What is the computation time for the fog server in seconds?

Imagine a scenario where Boeing designed a pilot monitoring system that deploys in response to a failing MCAS system. If an MCAS system is engaged and the pilot is detected to be stressed, MCAS will automatically disengage. In this system, consider a brain mobile interface application where the pilot wears a Neurosky headset that senses brain signals (EEG) at 400 Hz. Each brain data point is a 32-bit floating point number. The brain signal is collected by a central controller in the plane and sent to a server, where complex machine learning algorithms are employed to determine the stress level of the pilot. Additionally, the aircraft is equipped with sensors, such as the AoA, pitch monitoring, and other relevant sensors. The data rate from the AoA is 5 kbps, and the data rate from the other relevant sensors is 300 kbps. Using the data from these sensors, the MCAS disable system attempts to predict MCAS failures. If the system detects that the pilot is stressed and an MCAS failure is predicted, the auto-disable facility should disable MCAS. The auto-disable feature only has 5 seconds to make a decision after collecting 5 seconds worth of data. There are two options for performing all of the related computation: (a) use a GPU server at the control center, or (b) use a fog server that is onboard the aircraft. The GPU server upload speed is 1 Mbps, whereas the fog server upload speed is 5 Mbps. However, the computation speed of the GPU server is 1500 kbps (in other words, it can finish the computation on 1500 kb of data in 1 second), whereas the fog server has a computational speed of 200 kbps. What is the computation time for the GPU server in seconds?

Suppose that an adversary gained access to the previously sensed data. During which phase of the mobile security threat model will the adversary be able to introduce an attack?

Imagine a scenario where Boeing designed a pilot monitoring system that deploys in response to a failing MCAS system. If an MCAS system is engaged and the pilot is detected to be stressed, MCAS will automatically disengage. In this system, consider a brain mobile interface application where the pilot wears a Neurosky headset that senses brain signals (EEG) at 400 Hz. Each brain data point is a 32-bit floating point number. The brain signal is collected by a central controller in the plane and sent to a server, where complex machine learning algorithms are employed to determine the stress level of the pilot. Additionally, the aircraft is equipped with sensors, such as the AoA, pitch monitoring, and other relevant sensors. The data rate from the AoA is 5 kbps, and the data rate from the other relevant sensors is 300 kbps. Using the data from these sensors, the MCAS disable system attempts to predict MCAS failures. If the system detects that the pilot is stressed and an MCAS failure is predicted, the auto-disable facility should disable MCAS. The auto-disable feature only has 5 seconds to make a decision after collecting 5 seconds worth of data. There are two options for performing all of the related computation: (a) use a GPU server at the control center, or (b) use a fog server that is onboard the aircraft. The GPU server upload speed is 1 Mbps, whereas the fog server upload speed is 5 Mbps. However, the computation speed of the GPU server is 1500 kbps (in other words, it can finish the computation on 1500 kb of data in 1 second), whereas the fog server has a computational speed of 200 kbps. What is the communication time for the fog server in seconds?

Consider a restaurant recommendation application (Live2Eat), which shows you nearby restaurants for a given location. Assume that Live2Eat automatically updates the location as the user moves, and that it also updates the nearby restaurants. It has **two options** for obtaining the location information. The first option: by using GPS, which is more accurate. The second option: by using the mobile tower-based cellular network, which is far less accurate. Suppose that there is a Live2Eat user who is driving down a street, and the GPS signal is lost at time t = 0. Also suppose that the average speed of traffic is 10 km/h (6.21 mph). The error in GPS localization is 25 m (0.016 miles), while the error in mobile tower-based localization is 300 m (0.19 miles). Consider that location information is requested by Live2Eat every minute. When should you switch from GPS to cellular?

What are examples of inferred context? Select all that apply.