Comprehensive Analysis Of Modern Satellite Vulnerabilities And Emerging Attack Surfaces

The talk outlines how a satellite service is divided into space, ground, user and link segments, and where each one can be attacked. Because independent researchers rarely get access to flight hardware, the speaker studies open-source cubesat projects and a ground station system instead.

The first case covers the internal bus protocol of a small satellite, which runs over a CAN bus. The node identifier is not range-checked and is not required to be unique, so a compromised component can impersonate another and send spoofed telemetry. A housekeeping mode lets nodes transmit without being polled, and the frame reassembly logic does not check frame indexes, so crafted fragments can overwrite parts of a reassembled packet. That makes false sensor values, such as voltage readings, possible. The second case examines a widely used open-source satellite networking library, where the checksums do not cover packet headers and rely on hardcoded tables that anyone can reproduce. Its management functions also allow remote memory reads and writes.

The ground station case involves a Python web application in a supplier's test environment. An API key field accepted template syntax, and the length limit existed only in the browser. Bypassing a template sandbox led to a shell, then stored credentials, then admin functions for firmware upload and remote memory access. The takeaway is that open space protocols often lack security design, and ground stations that command satellites remain a critical target.