Table 2 Main features of the referenced approaches
From: Optimal reception of sub-sampled time-domain sparse signals in wired/wireless OFDM transceivers
Reference | NMSE/BER | Notes |
|---|---|---|
This work | In all cases, signal to channel noise ratio >30 dB | |
B E R<10−3 | AWGN, 16-QAM, N=256, R<N/8, S≤2 % | |
B E R<10−4 | AWGN, 4-QAM, N=1024, S≤2 % | |
B E R<10−3 | STBC-OFDM, 16-QAM, N=1024, R<3N/64, S≤2 % | |
N M S E<10−2 | Image of Fig. 7, R<N/4, S=10 % | |
This work | B E R<10−4 | AWGN, 16-QAM, N=256, S≤2 %, S N R>17d B |
CS input | B E R<10−2 | AWGN, 4-QAM, N=1024, S≤2 %, S N R>8d B |
[4] | N M S E<2.5 | SNR between -5dB and 4dB, |
50 measurements, 100 sensors | ||
N M S E>3 | SNR between -5dB and 4dB, | |
20 measurements, 50 sensors | ||
N M S E=0 | # m e a s u r e m e n t s># s e n s o r s | |
[5] | N M S E=0.18 | Measurements used: 25 % |
N M S E≈0 | Measurements used: 63 %, KL coefficients used | |
N M S E=0.2 | 31.25 % of measurements used with LARS or | |
11.72 % used with KL | ||
[7] | MSE=0 to 0.8 | e.g., sparseness 4 % or |
10 % (fraction of measurements used) | ||
[8] | 0.02<N M S E<0.1 | The proposed approach in [8] |
0.1<N M S E<1 | Compared approaches | |
[9] | Error Rate: | Sparseness between |
from 0.05 to 0.075 | 5/30,000 and 30/30,000 | |
[10] | N M S E<10−5 | S N R=50d B |
[11] | B E R<10−4 | Subcarriers used: |
≤86 (S=86/112≈77 %), S N R>40d B | ||
[14] | N M S E=0.0085 | S=4.4 %, S N R=16d B |
[15] | 4×4 MIMO, S N R=15d B | |
B E R<10−5 | Active: 4 of 256 subcarriers (S=1.56 %) | |
B E R>10−2 | Active: 32 of 256 subcarriers (S=12.5 %) | |
[25] | STBC 2×2-OFDM, 8K carriers, | |
no compression/sub-sampling: | ||
B E R<10−4 | Depending on: Alamouti or Golden encoding, | |
16-QAM or 64-QAM modulation | ||
13d B<S N R<22d B | equal/unequal transmit power (by 6dB) | |
[26] | S E R=10−4 | 3D MIMO |
@ S N R=15d B | STBC- 4×2, 4-QAM | |
S E R=10−4 | 3D MIMO | |
@ S N R=25d B | STBC- 4×2, 16-QAM |