ITR: Architecture for Surviving Denial-of-Service Attacks on Battery-powered Mobile Computers

点击翻译按钮获取中文摘要

The ongoing proliferation of battery-powered computing devices has created a new type of "denial of service" attack: If an attacker can drain a device's battery, for example, by having it repeatedly execute a energy-hungry program, the device will be rendered inoperable. Unlike other denial-of-service attacks where the attacker must keep up the attack in order to continue to deny the service, the attacker can quit attacking a battery-powered device once she has fully discharged thebattery, and move on to attack another device. Just as the advent of computer networks enabled an increase in the number of computer viruses, Trojan horses, and other computer security breaches, the rising availability of and increasing dependence on mobile computing devices will lead to the creation and spread of "power-related security attacks." The battery in a mobile computing device is thus a point of vulnerability and must be protected. The purpose of this research is to defend against attacks on the battery.In a typical mobile computer, the battery is expected to give a certain battery life under a set of usage conditions where the user is actively using the device for a small fractionof the time, and the device is idle the rest of the time. When the device is idle, power management software puts thedevice into low power standby and sleep modes, extending the device's battery life. If an attacker can prevent the device from entering low power modes by keeping it active, the battery life can be drastically shortened. There are three main methods for an attacker to drain the battery: (1)Service request attacks, where repeated requests are made to the victim for services, typically over a network--even if the service is not provided, the victim must expend energy deciding whether or not to honor the request; (2) benign power viruses, where the victim is made to execute a valid but energy-hungry task repeatedly, and (3) malignant power viruses, where an attacker modifies a program to make it consume more energy than it would otherwise.The ongoing research defends against these attacks by defining (1) a power-secure architecture for mobile computing devices that guarantees a minimum battery life, and (2) a design flow for identifying power-related security vulnerabilities. This work guarantees a minimum battery life by guarding against attacks on the device's battery, including service request attacks, benign power viruses, and malignant power viruses. The overall power-secure architecture employs twofundamental security features in the system, multi-layer authentication and energy signature monitor: The multi-layer authentication ensures that all untrusted services rendered consume less than a certain amount of energy. Additional resources are committed only to those requesters who have obtained further levels of trust. The energy signature monitor catches those intrusions that have entered the system to execute an energy-hungry application or service. The researchconsists of the following tasks:* Classification of services to guarantee minimum mission time* Generation, capture, and validation of energy signaturesfor trusted service requests* Validation of the architecture by implementing power-relatedsecurity attacksThis project has a number of direct benefits to the stateof the art in information technology. First, it protects battery-powered mobile computing systems, an increasingly important portion of the information technology infrastructure, from a potentially devastating form of security attack. Second, it explores a new problem and attracts attention to an areathat should be studied by a large community of researchers.The broader impact of the this research is to ensure that mobile computing continues to be attractive to a growing community. By protecting against battery-based security attacks, this work ensures the ongoing adoption of mobile computing by a wider segment of society.