Imagine you need to ship a precious, time-sensitive package across a large, complicated city. You have many routes to choose from, involving multiple delivery vehicles and checkpoints. You don’t just care about the fastest route; you care about the safest, most reliable path where every driver and warehouse is completely trustworthy.
This is the central challenge addressed by the CASTOR project.
The modern reality of the Computing Continuum drives the need for guaranteed trust. This complex environment includes the sprawling growth of Edge Computing and smart devices, everything from factory sensors and connected cars to city-wide surveillance drones. These are mobile, resource-constrained devices operating in dynamic spaces, all generating massive amounts of data. The network isn’t one monolithic system: it’s a heterogeneous mess of different technologies (Wi-Fi, 5G, Bluetooth) spanning multiple organizational domains, such as utility companies or different internet providers. Finding a trustworthy path in this constantly shifting, high-stakes domain (where low delay and high security are mandatory for drone systems and connected vehicles) is the project’s central objective.
The network is essentially a huge map (a graph) made of routers (nodes) connected by links (edges). When data flows from a source to a destination, it moves along a path composed of these nodes and links.
The Routing Challenge: Balancing Trust and Performance
Traditionally, networking focuses on speed, bandwidth, and minimizing cost. CASTOR introduces a complex layer of accountability: explicit trust. For sensitive data, a path must satisfy two distinct sets of requirements simultaneously:
- Trustworthiness Requirements: Every router along the path must possess verifiable qualities like high integrity (proven software/hardware state) and confidentiality.
- Network Requirements: The links must meet performance standards like low latency and sufficient bandwidth.
The problem is that trust is not a simple “on/off” switch. CASTOR treats trust as a probabilistic value that encompasses uncertainty and disbelief, searching for the optimal path in a complex, multi-objective optimization problem.
The Roadblocks to the Perfect Path
Finding this ideal path is difficult because of several complex roadblocks identified in the CASTOR architecture:
- Conflicting Goals: The requirements often pull in opposite directions. The absolute shortest or fastest path may traverse a router with a low Actual Trust Level (ATL), while the most secure path might be too slow for real-time drone commands.
- Systematically Measuring Trust: In line with the Zero Trust paradigm, CASTOR does not assume any inherent trust in the network topology. Instead, it deploys the relevant Trust Sources within each router, enabling the secure collection of evidence regarding the trustworthiness of critical network functions, both in terms of configuration and operational state. Given that evidence may be insufficient or potentially conflicting, CASTOR provides a comprehensive Trust Assessment Framework to measure and quantify trust across the network elements. This framework supports trust evaluation not only at the router level but also across the entire network topology.
- Defining Trust Numerically: Translating abstract concepts like “high integrity” into a mathematical language that an optimizer can understand is difficult. CASTOR overcomes this by using Subjective Logic to project trust opinions into scalar values that can be calculated alongside traditional network metrics.
- Shared Resource Competition: When multiple services request paths at once, they compete for the same highly trustworthy node. Managing these interdependencies without causing network congestion is a massive scaling challenge.
- Paths Go Stale: The network map is constantly shifting. A path deemed optimal today might become insecure tomorrow if a router’s trust level drops due to a new vulnerability. CASTOR requires re-optimization to happen in near real-time to maintain compliance with Secure Service Level Agreements (SSLAs).
A Foundation for Verifiable Trust
Navigating these roadblocks requires more than just faster processing; it requires a fundamental shift in how network paths are calculated. CASTOR moves beyond traditional “best-effort” routing by integrating continuous trust assessment directly into the optimization process. By bridging the gap between abstract security requirements and real-time network engineering, the project ensures that the path of least resistance is also the path of highest integrity. Ultimately, CASTOR’s goal is to ensure that when critical data is sent, the network doesn’t just hope for the best; it explicitly identifies and guarantees a route that is both fast and demonstrably trustworthy from start to finish.
