ISOT Mil-STD-1553 Dataset

The dataset is a combination of datasets collected in a simulated environment consisting of normal activities and a series of attacks against the MIL-STD-1553 databus.

The simulation was conducted using Abaco R15-USB-2M USB interface box and Abaco BUSTOOLS/1553 GUI Software for MIL-STD-1553 bus analysis, simulation, and data logging.

Figure 1 depicts the simulated bus components. The bus architecture used in the simulation consists of a bus monitor (BM) and five different avionic systems, including a flight control computer (FCC) as RT3, a mission computer (MC) as BC, an inertial reference unit (IRU) as RT1, a mission control keyboard (MCK) as RT4, and a multi-function display unit (MFD) as RT2. The MC serves as bus controller (BC), while the remaining components are remote terminals.

Figure 1. Simulated MIL-STD-1553 bus components

 

The components were the only ones used to execute the normal activities. To simulate the attack, we added a rogue terminal as RT0.

A baseline dataset consisting of normal activities was generated by running the simulation for 10 min. Subsequently, 6 different attack scenarios were run separately against the baseline architecture, by introducing RT0 as rogue terminal. Table 1 provides a summary of the different datasets and Table 2 shows the different fields involved in the raw data. The dataset is provided as separate CSV files.

 

 

Table 1. Outline of the collected MIL-STD-1553 datasets

 

Dataset # Type of data Attack type Attack description

Number of

messages

Simulation length
1 Normal N/A N/A 23,000 10 min
2 Mix normal/malicious

Targeted/Basic DOS

(Attack 1)

rogue terminal (RT0) targets the FCC (RT3) by sending random

words in a loop

148 30 sec
3 Mix Normal/malicious Multi-target DOS (Attack 2) Rogue terminal (RT0) sends broadcast messages in a loop 373 30 sec
4 Mix Normal/malicious

Subtle Fake data injection

(Attack 3)

rogue terminal (RT0) replicates one of the messages sent by one of the legitimate RTs by slightly

altering one of the data items

970 30 sec
5 Mix Normal/malicious

Noisy fake data injection

(Attack 4)

similar to the above attack, except that in this case fake data

is randomly generated.

970 30 sec
6 Mix Normal/malicious Logic attack (Attack 5)

a rogue RT broadcast mode code value 4 (0x04), which corresponds to the Transmitter

Shutdown, which is unusual

981 30 sec
7 Mix Normal/malicious Hybrid Logic/fake data injection (Attack 6) Combination of the attacks in datasets 5 and 6 1004 30 sec

 

Table 2. Raw dataset format

 

Fields Description
msgID Sequence number associated with the message by the simulator
timestamp Message timestamp in seconds
error Indicate whether the error bit of the status word related to the message is set; contains TRUE/FALSE accordingly.
modeCode Indicate whether the message is a mode command message; contains TRUE/FALSE accordingly.
channel Channel associated with the message
connType Communication type
sa Address of sending RT
ssa Sub-address of the sending subsystem from the sending RT
da Address of receiving RT
dsa Sub-address of the receiving subsystem at the receiving RT
wc Word count: number of data words included in the message
modeCode value Mode code value when applicable
txRsp Transmit command response time in ms
txSts Transmit status word
rxRsp Receive command response time in ms
rxSts Receive status word
dw0 … dw31 Values of the data word included in the message ranging from dw0 to dw31; N/A is used when there is no data words for a field.

 

Malicious Indication of whether the message is malicious: contains TRUE/FALSE accordingly.
injected

Indication of whether part of the data included in the message is injected:

contains TRUE/FALSE accordingly.

gap Inter-message gap time in ms
msgTime Message time in ms

 

 

 

 

To cite this dataset use:

Hadeer Saad, Issa Traore, Paulo Quinan, Karim Ganame, Oussama Boudar, “A Collection of Datasets for Intrusion Detection in MIL-STD-1553 Platforms”, in Artificial Intelligencefor Cyber- Physical Systems Hardening, I. Traoré, I. Woungang, and S. Saad, Eds. Springer, 2022, Chapter 4.