Software-Based Memory Erasure with relaxed isolation requirements: Extended Version

Sergiu Bursuc,Reynaldo Gil-Pons,Sjouke Mauw,Rolando Trujillo-Rasua

2024-01-12

Abstract:A Proof of Secure Erasure (PoSE) is a communication protocol where a verifier seeks evidence that a prover has erased its memory within the time frame of the protocol execution. Designers of PoSE protocols have long been aware that, if a prover can outsource the computation of the memory erasure proof to another device, then their protocols are trivially defeated. As a result, most software-based PoSE protocols in the literature assume that provers are isolated during the protocol execution, that is, provers cannot receive help from a network adversary. Our main contribution is to show that this assumption is not necessary. We introduce formal models for PoSE protocols playing against provers aided by external conspirators and develop three PoSE protocols that we prove secure in this context. We reduce the requirement of isolation to the more realistic requirement that the communication with the external conspirator is relatively slow. Software-based protocols with such relaxed isolation assumptions are especially pertinent for low-end devices, where it is too costly to deploy sophisticated protection methods.

Cryptography and Security

What problem does this paper attempt to address?

The core problem that this paper attempts to solve is: **How to design a secure Proof - of - Secure - Erasure (PoSE) protocol without relying on the device isolation assumption to prevent malicious provers from outsourcing the proof of erasure to external attackers?**

### Detailed Explanation

#### Research Background
1. **Security Issues of Internet of Things (IoT) Devices**
- Due to their ubiquity, connectivity, and limited computing resources, IoT devices are vulnerable to malware infections.
- Once infected, these devices not only become victims but also are used as tools to attack more advanced infrastructures and services.
- Detecting whether IoT devices are infected with malware is crucial for maintaining network security.

2. **Existing Solutions and Their Limitations**
- **Hardware Approaches**: Memory erasure can be achieved through direct hardware operations, but it is costly and not suitable for all devices.
- **Software Approaches**: Existing software - based PoSE protocols usually assume that the prover is isolated during the protocol execution, that is, it cannot obtain help from network adversaries. This assumption is difficult to guarantee, especially on low - power and low - cost devices.

#### Main Contributions of the Paper
1. **Proposing New Models and Protocols**
- A new class of PoSE protocols, called PoSE - DB (Proof of Secure Erasure with Distance Bounding), is introduced. It ensures security through a distance - bounding mechanism without strict device isolation.
- Three specific PoSE - DB protocols are designed, and formal security proofs are provided.

2. **Relaxing Isolation Requirements**
- The isolation requirement is relaxed to a more realistic one, that is, the communication between external attackers and verifiers is relatively slow.
- This relaxation makes the protocol particularly suitable for low - power devices because the cost of deploying complex protection methods on these devices is too high.

3. **Specific Protocol Designs**
- **The First Protocol**: Based on the storage and query of random bit sequences, it is suitable for bandwidth - limited scenarios.
- **The Second Protocol**: Using graph - labeling techniques, it verifies the storage of node labels by rapid polling.
- **The Third Protocol**: Also based on the graph structure, but it makes a trade - off between performance and security.

4. **Theoretical Analysis and Proof**
- A formal security definition is proposed to ensure that in the presence of remote attackers, the prover must erase most of its memory to pass the verification successfully.
- The distance - bounding mechanism is used to ensure that the response time meets expectations, thereby preventing outsourcing attacks.

#### Formula Representation
The distance - bounding mechanism involved in the paper can be represented by the following formulas:

- **Time Threshold**: $\Delta$, used to measure the time difference of each round of challenge - response.
- **Block Size**: $w$, the size of each memory block.
- **Memory Size**: $m\cdot w$, where $m$ is the number of blocks in the memory.
- **Round**: $r$, the number of interaction rounds.

For example, the selection of the time threshold $\Delta$ should ensure that remote attackers cannot complete the response within the specified time:

\[
\Delta=\text{Select a suitable value so that remote attackers cannot complete the response within }\Delta\text{ time}
\]

In this way, the paper provides a new method to ensure the secure erasure of memory without relying on strict isolation assumptions, which is of great significance for enhancing the security of IoT devices.