Transmitter Packet Corruption Error Model

SWITCH TO API

Description

The Transmitter Packet Corruption Error Model introduces configurable packet corruption to a Transmitter component. This model simulates bit errors that occur during data transmission due to channel noise, interference, or hardware imperfections. By attaching this model to a transmitter, users can evaluate system robustness to data corruption and test error detection and correction mechanisms at higher protocol layers.

---

Example Use Cases

• Protocol Validation: Test checksums, CRCs, and error correction codes under realistic corruption conditions.
• Link Margin Analysis: Evaluate communication system performance with degraded signal quality.
• Fault Injection Testing: Systematically introduce errors to verify system resilience.
• Monte Carlo Studies: Characterize mission reliability under varying corruption probabilities.

---

Module Implementation

The packet corruption error model is attached to a Transmitter component and controls the probability that individual packets are corrupted before transmission.

Corruption Probability

The model exposes a single parameter, the packet corruption fraction $P_c\in [0,1]$, which defines the probability that any given packet will be corrupted:

$$P(\text{corrupt})=P_c$$

When $P_c=0$, no packets are corrupted. When $P_c=1$, every packet is corrupted.

Corruption Mechanism

When the model is active, the parent transmitter uses $P_c$ to determine whether each packet should be corrupted. For each packet, a uniform random value $r\in [0,1)$ is generated:

$$\text{corrupt}=\{\begin{array}{ll}\text{true}& \text{if }r<P_c\\ \text{false}& \text{otherwise}\end{array}$$

If corruption occurs, the transmitter flips a single random bit within the packet data. The bit flip is performed by selecting a random byte index $i$ and bit position $b\in [0,7]$, then applying an XOR operation:

$$d_i\leftarrow d_i\oplus 2^b$$

This single-bit error is sufficient to invalidate most checksums and CRCs, causing the packet to fail validation at the receiver.

---

Assumptions/Limitations

• Only a single bit is flipped per corrupted packet; multi-bit errors are not modelled.
• The corruption probability is applied per-packet, not per-bit.
• Bit errors are uniformly distributed across the packet; burst errors are not modelled.
• The model does not distinguish between different error sources (noise, interference, hardware faults).
• Retransmission and error recovery must be handled at higher protocol layers.
• Detaching the model from a transmitter resets the corruption probability to zero.