Short Word-Length Entering Compressive Sensing Domain: Improved Energy Efficiency in Wireless Sensor Networks
Round 1
Reviewer 1 Report
Accept the paper with minor English revision.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The authors show their trial on combining CS and short word-length techniques for energy-efficient communications, considering locolization and tracking between the network anchors and the fusion center. The results sound good agreement with their statements. Before publication, I suggest the following optional revision.
- Please show how the step size was optimized.
- If possible, please discuss on the range of T_max. Or, how large can it be, to the most?
- More detailed caption for the figures will be helpful for readers.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors present a high resolution TOA Estimation technique based on compressed sensing (CS) combined with (sigma) delta modulation (SDM/DM). However, the approach has severe flaws.
1) What is the benefit of adding SDM to CS within the context of TOA estimation? On top of that, SDM/DM adds additional complexity.
2) The proposed approach wans only studied under ideal conditions i.e., without communication channel. Clearly, both schemes seem to work excellently. The communication channel is a non-neglibile part of the system that cannot be skipped and hence, must be added to the simulation setup. On top of that, the work must be compared with state-of-the-art CS stand-alone in noisy environment (see the very relevant work by [Xiong-2015], not mentioned in the list of references.). Generally speaking, the proposed scheme is not necessarily outperforming CS stand-alone. All depends on the channel, that has been neglected in the work.
The question arises "Under what conditions the proposed scheme outperforms state of the art (CS stand-alone)?", "Is the price of additional complexity worth the performance benefit; under what conditions?"
3) What is the SNR of the proposed method (CS-DM,CS-SDM compared with CS stand-alone) ? There must be quite a large difference.
Minor comments.
The introduction on DM/SDM in Section 3 is taught at undergraduate level and hence, can be readiliy skipped.
Line 131. Why has the term TDOA been introduced? It is not used in the body of the script.
Reference:
[Xiong-2015] Xiong, W., Liu, C., Hu, S. et al. High Resolution TOA Estimation Based on Compressed Sensing. Wireless Pers Commun 84, 2709–2722 (2015). https://doi.org/10.1007/s11277-015-2762-y
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
Quality of the revised version has improved.
1) The newlly inserted Figure 15 tells the reader about the performance in AWGN. The plot shows the SNR ranging from 5 to 50 dB. Practical communication systems, however, operate in the range between 0 dB and 5..8 dB, which are not shown in the plot. The focus must be given to the relevant SNR range instead.
Rather than just adding some simulation results, it would be great to add the theoretical expression of NEM(SNR) for CS-DM/CS-SDM in AWGN.
2) It should be mentioned that additional complexity caused by adding DM/SDM has the benefit to outperform pure CS transmission (Fig. 15).
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
