EURASIP Journal on Advances in Signal Processing

Table 5 Confusion matrix of the testing samples (SNR_tr= − 10 dB)

From: A robust modulation classification method using convolutional neural networks

Actual type	Classification result
	M-ASK			M-FSK			M-PSK			M-QAM			OFDM	LFM	MSK
	2	4	8	2	4	8	2	4	8	4	16	64
2ASK	950	33	16	0	0	0	0	0	0	0	0	1	0	0	0
4ASK	65	553	382	0	0	0	0	0	0	0	0	0	0	0	0
8ASK	21	329	650	0	0	0	0	0	0	0	0	0	0	0	0
2FSK	0	0	0	1000	0	0	0	0	0	0	0	0	0	0	0
4FSK	0	0	0	0	1000	0	0	0	0	0	0	0	0	0	0
8FSK	0	0	0	0	1	997	0	0	0	0	0	0	0	0	2
2PSK	0	0	0	0	0	0	1000	0	0	0	0	0	0	0	0
4PSK	0	0	0	0	0	0	0	956	44	0	0	0	0	0	0
8PSK	0	0	0	0	0	0	0	93	907	0	0	0	0	0	0
4QAM	0	0	0	0	0	0	0	0	0	999	0	1	0	0	0
16QAM	0	0	0	0	0	0	0	0	1	8	726	265	0	0	0
64QAM	0	0	0	0	0	0	0	0	0	6	192	802	0	0	0
OFDM	0	0	0	0	0	0	0	0	0	0	0	0	985	15	0
LFM	0	0	0	0	0	0	0	0	0	0	0	0	0	1000	0
MSK	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1000

Back to article page