RFID RSS Fingerprinting System for Wearable Human Activity Recognition
Round 1
Reviewer 1 Report
The manuscript presents a cost-effective target tracking system based o processing data received from RFID tags. It uses RFID fingerprinting for inferring the patient's position in real-time.
Next, I send some comments that must be solved to improve the paper:
1) The title of the paper addresses the reader to the patient tracking problem, but the paper shows daily activity recognition instead of a real-time tracking application, the title could include the daily activity recognition problem instead. You must clearly define the main objective of the paper, is it tracking the position of the patient or is it about daily activity recognition?
2) The abstract must show some results obtained.
3) In the Introduction Section, there isn't a description of the structure of the paper.
4) Can you explain with more detail why the Euclidean distance is the best choice in RF-based applications?
5) Can you present the system's response when an atypical activity comes on?
6) Is this system useful to elderly care? It uses many sensors, and they are placed in locations of difficult access.
7) In the Introduction Section, the authors should mention the weak points of former works. A comparative overview table showing the key differences between other approaches ant the proposed method is welcome.
8) Can you present an analysis of the precision and the recall metrics of your proposal?
9) The document needs a new revision to correct typos and other mistakes, for example:
- et.al. --> et al. In the introduction, the paragraphs about references 14 and 16 are very similar.
- Section 2.2:"Equation 6.1consider the effect..."
- Please define the acronyms before using them: RFID, ADL, RSSI, LOS, NLOS
- In Eq. (1), the definition of "a" and "b" aren't in italics.
- Figure 3 is miss referenced in Section 2.3, it referenced Figure 2
- Section 3: "...for both correct both positions..."
- The last image in Figure 9 is incorrect.
Author Response
Please see our response as attached.
Author Response File: Author Response.docx
Reviewer 2 Report
Major Comments: The usefulness of the results cannot be evaluated because no information is provided on the room in which the experiment was performed, the position of the reader antenna, the performance of the antenna such as directivity and so on. Without knowing the details of the experimental situation, the validity of RSSI of each tag cannot be evaluated. In conclusion, you said "The proposed RSS fingerprinting approach achieved an accuracy of 98% for most sampled activities.", how did you find the results? Minor comments: The meaning of the abbreviation ADL is not defined. Equation 6.1 -> Equation (1) The paragraph structure has not been sufficiently proofread.
Author Response
Please see our response as attached.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
The problem of multipath propagation is greatly influenced by surrounding media such as walls and objects.If the structure of the room used for the experiment changes, the experimental results will differ greatly.Please add information such as the size of the room where the experiment was conducted, the material of the wall, the size and location of objects placed in the room that affect radio wave propagation. I will ask you again. Equation 6.1 cannot be found in your paper. Where is Equation 6.1 shown in the paper?
Author Response
Please see the response as attached.
Author Response File: Author Response.docx