Network Intrusion Detection on MQTT enterprise network traffic FPA attack scenario benchmark leaderboard · SOTA2 Research

Network Intrusion Detection on MQTT enterprise network traffic FPA attack scenario

100ASRCNN-LSTM-GRU

Evaluation Results

Method	Links
CNN-LSTM-GRU2026.01		100	1	0.237	0.237	1.673	1.673	0	0	1	0
CNN-LSTM-GRU2026.01		100	45,737	0.684	0.665	0.711	0.602	5,391	8,655	0	5,391
CNN-LSTM-GRU2026.01		100	34,562	0.616	0.601	0.823	0.776	1,626	3,227	19	1,626
CNN-LSTM-GRU2026.01		100	2,716	0.518	0.509	1.01	0.983	0	0	141	0
CNN-LSTM-GRU2026.01		100	—	0.65	0.633	0.767	0.687	—	—	—	—
SF-CNN-LSTM2026.01		100	5,762	0.767	0.746	0.544	0.453	2,068	2,577	0	2,068
SF-CNN-LSTM2026.01		100	69,920	0.994	0.993	0.021	0.003	68,464	69,374	0	68,464
SF-CNN-LSTM2026.01		100	7,334	0.886	0.866	0.268	0.107	4,984	5,613	0	4,984
SF-CNN-LSTM2026.01		100	—	0.969	0.965	0.079	0.043	—	—	—	—
Hybrid CNN-LSTM2026.01		100	6	0.694	0.668	0.728	0.672	0	2	0	0
Hybrid CNN-LSTM2026.01		100	140	0.897	0.878	0.327	0.01	90	96	0	90
Hybrid CNN-LSTM2026.01		100	329	0.68	0.67	0.783	0.77	0	0	1	0
Hybrid CNN-LSTM2026.01		100	82,541	0.955	0.952	0.155	0.019	77,538	78,982	9	77,538
Hybrid CNN-LSTM2026.01		100	—	0.954	0.951	0.158	0.022	—	—	—	—
DCNNBiLSTM2026.01		80.19	66,570	0.979	0.977	0.077	0.031	63,403	64,436	0	63,403