Tracking via SSD: a new method for identifying devices on the web

New Privacy Threat in the Network via Drive Performance Analysis

User tracking methods in the global network are constantly evolving, adapting to increased security measures in modern browsers. Traditional tools, such as advertising and tracking cookies, blocking third-party trackers, and anti-fingerprinting technologies based on screen parameters or available fonts, force tracking system developers to look for new vectors to identify devices. Recent research by cybersecurity experts has revealed an unusual and stealthy method for monitoring the activity of website visitors. This method is based on a detailed analysis of micro-delays during input-output operations of solid-state drives (SSDs) through a standard browser.

The technology uses the basic capabilities of the JavaScript programming language, which runs client-side in any modern web browser. Since JavaScript has access to measuring the execution time of certain operations, malicious or analytical scripts can create temporary loads on the device’s data storage subsystem. The obtained results of reading and writing speeds allow the formation of a unique digital fingerprint of the user’s drive.

Architectural Features of SSDs as a Basis for Identification

Each solid-state drive, regardless of brand and model, has unique physical and software characteristics. The production process of NAND flash memory chips is accompanied by microscopic defects and non-uniformities of silicon wafers. As a result, individual memory blocks have slightly different access, read, and program times for cells. In addition to the physical differences of memory chips, the built-in firmware of the drive controller plays a huge role.

The SSD controller performs complex tasks of load balancing, garbage collection, and wear leveling. The algorithms of these processes are proprietary intellectual technologies of each individual manufacturer. During drive operation, specific micro-delays occur, which depend on many factors. Below are the main parameters that shape the unique behavior of the device:

  • Current wear state of memory cells and the total number of rewrite cycles.
  • Size and architecture of the drive’s internal cache and its clearing algorithms.
  • Specifics of controller operation during parallel processing of data streams.
  • Current level of file structure fragmentation at a low level.

Mechanism of Tracking via JavaScript

To access drive characteristics, a script running in a browser does not require elevated administrator privileges or the installation of additional software. Browser APIs provide standard tools for working with local data storage, such as IndexedDB or the Origin Private File System. The script initiates a series of rapid write and read operations of small blocks of information, while simultaneously measuring the execution time of each request with microsecond accuracy.

Since operating systems and browsers apply their own caching mechanisms to speed up performance, tracking system developers use special methods to bypass these limitations. They create a volume and structure of requests that force the system to access the physical storage device directly. By measuring the time intervals between sending a request and receiving confirmation, the script captures delay patterns inherent in specific hardware.

Comparison of user tracking methods efficiency on the internet
Tracking Method User Blocking Complexity Resistance to Browser Cache Clearing Dependence on Operating System Settings
Standard cookies Low, deleted via settings Absent, completely erased Does not depend on OS parameters
Canvas and audio fingerprinting Medium, blocked by protective extensions High, configuration remains stable Partially depends on video and audio drivers
SSD activity analysis High, requires browser timer limitations Maximum, based on hardware properties Depends on hardware state and internal disk algorithms

Impact on Privacy and Bypassing Incognito Modes

The main danger of this approach is that it completely invalidates the effectiveness of standard privacy tools. Using incognito mode, changing virtual private networks (VPNs), spoofing network IP addresses, and clearing local browser databases do not change the physical properties of the storage drive installed in the computer. When a user revisits a site that uses storage delay analysis, the system recognizes the unique hardware profile.

This method allows linking different web browsing sessions of the same user, even if they use different browsers on the same device. Since the time characteristics of the drive remain stable over a long period, the digital fingerprint of the disk becomes a long-lasting identifier that is practically impossible to reset without physically replacing the computer hardware or reflashing the drive controller.

Technical Challenges of Implementation and Limitations for Attackers

Despite the high level of danger, creating a stable tracking system based on drive speed analysis presents certain engineering difficulties. The speed of input-output operations is affected by the current level of general load on the user’s computer. If other heavy processes are running in the operating system during testing, such as software updates or antivirus scanning, the results of delay measurements can be significantly distorted.

To minimize the impact of external factors, tracking systems use complex statistical models and multiple re-tests. They filter out random spikes in activity and isolate permanent hardware constants. However, this requires the user to stay on the page longer and consumes a significant amount of CPU computing resources, which can be noticed by attentive users due to increased resource usage in the task manager.

Prospects for Protection and Countermeasures Against the New Threat Vector

Modern web browser developers and security researchers are already working on creating effective protection mechanisms against this type of data collection. The main direction of countermeasure is the intentional reduction of precision of high-resolution timers in JavaScript. If a script cannot measure microsecond time intervals, it will not be able to record microscopic delays in the operation of the solid-state disk hardware controller.

Another promising method of protection is adding random artificial noise to read and write operations performed by web applications. This will make measurement results unstable and unsuitable for building an accurate digital profile. Users seeking the maximum level of privacy are advised to restrict the operation of heavy scripts on untrusted resources using specialized extensions and monitor security updates of their web browsers.

Pavlo Zaslonov
About The Author

Pavlo Zaslonov

Cybersecurity expert, knows everything about IP hiding and modern chatbot vulnerabilities.

0 Comments

Leave a Reply

2500
Please enter a comment
Please enter your name