kumoh national institute of technology
Networked Systems Lab.

Review Comment

NSL > Education> Review Comment
Syamsul - WFCS 2017
By :
Date : 2017-04-20
Views : 225

Dear author,

We regret to inform you that your paper Dynamic Rate Adaptation Scheme for Industrial Wireless Networks using RSSI has not been selected for presentation at WFCS 2017, to be held in Trondheim, Norway, from the 31st of May to the 2nd of June 2017. Find attached the reviews for your submission, which we believe can help you improving your work even further.

We would like to thank you for submitting to WFCS 2017. The conference program for WFCS 2017 will appear on the conference homepage soon. We are looking forward to meeting you in Trondheim nevertheless.

Best Regards

----------------------- REVIEW 1 ---------------------
TITLE: Dynamic Rate Adaptation Scheme for Industrial Wireless Networks using RSSI
AUTHORS: Syamsul Rizal, Jin-Yong Im, Jae-Min Lee and Dong-Seong Kim

Overall evaluation: 1 (weak accept)

----------- Overall evaluation ----------- Paper describes a rate adaptation scheme for multicast in IEEE 802.11 wireless networks using OFDM. The main concept is that stations use individual OFDM subcarriers to send their CTS and ACKs concurrently, thus allowing fast acknowledged multicast transmission. This is a useful method in some industrial scenarios. Unfortunately, the detailed description of the Algorithms is partly unclear, thus I can only recommend weakly to accept this submission. If the submission is accepted, the descriptions must be improved for final submission, addressing the listed comments below, most importantly comments no.7, 9, 10, 12, 14, 16.

1. Eq(1): replace Tb (= bit duration) by Rb (= bit rate). j and l are unclear. Anyway, best to remove (1), it is not used further.
2. II.B: delete In real-time networks
3. II.B: delete last sentence For instance, in 64QAM, the receiver could incorrectly decode the symbol, even when a symbol shifts slightly with respect to the axes. (bad wording not worth fixing)
4. II.D: ... without any deviation from the standard frame format. (delete, unclear wording not worth fixing)
5. II.D: replace tactical by real-time
6. III.A: How are C_ID and S_ID assigned to each station?
7. III.A: First, the transmitter calculates the average RSSI..: Average over all CTSs received earlier from this station on its subcarrier frequency, or average over all stations in this sector? What is R(i), does it refer to user i, or sector i?
8. III.A: .. to transmit the data to the users whose SID are recorded..: using multicast or unicast?
9. III.B: The orthogonality of each subcarrier ensures the safety of transmission..: With OFDMA, this requires precisely synchronized transmissions from stations. How to ensure this? What if the unknown propagation times (distances) differ?. Replace "safety" -> "reliability"
10. III.C, eqn(1): Optimization with respect to the rates R(i)s? Trivially, just take the maximum allowed R(i) ? Explain what prevents this solution? If the selected D is too small, there may be no feasible solution to the optimization problem. Algorithm 2 doesnt seem to optimize, but searches for feasible values, and stops transmision if the required rate would be > Mmax. Please clarify!
11. Eqn(1), delete superfluous =.
12. below eqn(3): R(i) refers to sector i, while in Algorithm 1 it was R(i) for station i?
13. users: Here, "user" = "station" in earlier paragraphs. Use consistent terminology!
14. Algorithm 2: What is N? How is D(i) calculated: what values to select for payload size P and processing times in (2)? What is ra(i,R[i])? a real number or a set? What is script R?
15. Algorithm 2: Start retransmission: retransmission per unicast? or multicast?
16. Eqn(4): How is epsilon Rx determined, what values does it take? What are alpha i and alpha i prime, and where do they enter the algorithm?

----------------------- REVIEW 2 ---------------------
TITLE: Dynamic Rate Adaptation Scheme for Industrial Wireless Networks using RSSI
AUTHORS: Syamsul Rizal, Jin-Yong Im, Jae-Min Lee and Dong-Seong Kim

Overall evaluation: -3 (strong reject)

----------- Overall evaluation ----------- In this paper authors propose a rate adaptation algorithm for multicast communication in wireless industrial networks. In particular, the presented solution checks the network condition using the received signal strength indicator (RSSI) within the clear-to-send (CTS) frames. Moreover, the proposed algorithm has been evaluated using simulation.

The paper has to be reject because its contents are practically the same presented in Rate adaptation algorithm for multicast communication in tactical networks published in: Military Communications Conference, MILCOM 2015 - 2015 IEEE.
Only some new introductory sections have been added, but no new research results. Most of the sections have been written using cut&paste. Then, same algorithms and same figures to show algorithm evaluation.

----------------------- REVIEW 3 ---------------------
TITLE: Dynamic Rate Adaptation Scheme for Industrial Wireless Networks using RSSI
AUTHORS: Syamsul Rizal, Jin-Yong Im, Jae-Min Lee and Dong-Seong Kim

Overall evaluation: -2 (reject)

----------- Overall evaluation ----------- The paper targets wireless sensor networks that transmit data to multiple users, targeting industrial sensor networks.
The method adjusts the modulation scheme based on feedback collected from the receiver nodes. That way the transmission quality is controlled.
Simulations are performed to evaluate the proposed approach.

The addressed problem is valuable in order to enable wireless technologies in timing critical applications.
The paper provides a lot of background information which seems not always needed.
This makes it hard to filter the relevant content.
Also many places lack citations in Section 2.

The main contribution is in the System Model section which is confusing, as the system model is usually given.
Restructuring the paper into system model (what is mostly in the section 2) and a dedicated section on the proposed method, would present the contribution better.

While evaluations are performed, the setup is not clear to me. How many simulation runs were performed, and what is shown in the graph? (if average values, then the STD would be good to add.)

Other comments:
- Abstract: "This paper proposed..." -> "This paper proposes..."
- Abstract: "...modify the frames in MAC layer..." -> "...in the MAC..."
- Abstract: "...in terms of reduced the delay..." -> "...reduced delay..."
- Section 1: "For sensor networks application..." -> "For sensor network applications..."
- Section 1: "...as the most common topology in sensor networks in the mesh or ad-hoc network." -> "...is the..." and a citation would be good to back the claim.
- Section 1: "These limitations result in the inability of the transmitter to send high-quality data in a short time." What is high quality data? Metric?
- Section 1: Previous work is mentioned but no citation is provided!
- Equation 1: T_b is not defined
- Section 2B: "In real-time networks, these limitations are aggravated by the movement of people and machinery, harsh environmental conditions, and the possible presence of stationary obstacles that hamper signal propagation." Real-time networks only means that there are temporal constraints but makes no assumption on the environment!
- What is the difference between wireless mesh and wireless backbone network?
- Section 2D: The section is about rate adaption but starts discussing routing?!
- Section 2D: Abbreviation RA is not defined.
- Section 2D: "We adopted the direct method in this study because this method is more efficient in terms of choosing the transmission rate." -> citation
- Section 2D: "...The MAC layer method use s frame/packet..." uses
- Tactical network is mentioned in Section 1 and 2 but not defined.
- Section 2: "however, it does not provide any time constraint in the current state, which is crucial in military networks." -> The assumptions for the work are not clear, what properties/requirements does a military network have? And whats its relation to industrial sensor networks?
- Section 3A: "$R[i]toM_+$" its all in math environment (including "to")
- Section 4: "Although the difference between the delays in the two schemes is only in seconds, it has a big impact on data transmission, as the transmission rate..." It looks like it is only a fraction of a second on the figure.
- Section 4: "This means that more packets can transmitted for a smaller delay." -> "can be transmitted"
- Section 4: "This is because the probability is higher in the proposed..." -> Which probability?
- Section 4: "The users that has its data transmission stopped" -> "have"

----------------------- REVIEW 4 ---------------------
TITLE: Dynamic Rate Adaptation Scheme for Industrial Wireless Networks using RSSI
AUTHORS: Syamsul Rizal, Jin-Yong Im, Jae-Min Lee and Dong-Seong Kim

Overall evaluation: -2 (reject)

----------- Overall evaluation ----------- In this paper the authors propose a dynamic rate adaptation algorithm using OFDM to overcome packet errors in a wireless network. The authors validate their findings using MATLAB simulations.

The paper is farly written with less care towards the technical method and the grammar.

Consistency in the terminology is not maintained in the algorithms by the authors, which creates confusion to the reviewer.

The paper would benefit from simulations with more insights to support the claims.

Overal it is felt that the techical depth and quality of writing is on the weak side.


Page 1:
This paper proposed a rate (mentioned in past) Why do you say that the common the topology used is mesh (justify and spell check)?
High quality data in a short time – what type of quality?
Please cite your previous work for reference.
Signal-to-Noise Ratio (SNR) capitalize acronyms and its ratio not ration.

Page 2:
Wireless channel is dependent on factors like loss . which result in packet loss (wrong sentencing) A wireless mesh network is a fully wireless network (wireless is mentioned already)

Page 3:
Please provide the info Rate Adaptation is RA to avoid confusion.
In direct method, transmitter can change its transmission rate to any data rate, in this same paragraph you mention this method uses lower transmission rate, which is contradictory.
Use s -> uses
Bit error rate (capitalize)
[5], [11] references based on PHY manipulation was not elaborated, they can be removed from the context if it is not relevant to the work.
Round trip time (capitalize)

Page 4:
Available transmission rate: 2, 5.5, 11 Mbps DCF, PCF (capitalize) Under sectorization, the coverage area was given as a terminology, but it is not explained through the Fig 3.
Parts for sectors must be marked in the Fig 3 for easy understanding and relatability with the work.
CID and CID are two different terms when you explain your system model, it is confusing when you use mixed up terms.
What is M? should be elaborated as a transmission constant for instance, so that it will increase the readability of the algorithm.

Page 5:
Like M, R is also not properly explained in both the following algorithms.
Line 20 of algorithm, shouldnt it be R(i) In Fig 4, Short Interframe Spacing (SIFS) and its impact is unexplained.
In Fig 4
b, the transmission is not illustrated as transmissions from several transmitters, because it is shown it happens under the same time interval from a single transmitter. Is it a broadcast?
Equation for time constraint D was interesting and well written.

Page 6:
Receiving ability of the receiver: do you mean the channel quality or the power level or some QoS aspect of the network.
is introduced in the paper as a weighting value, but it is not used anywhere in the algorithm.
What is ra? Is it RA or R ?

Pages 7 and 8:
In the start of the paper, they mentioned that there is a constant loss because of real time environment, is this loss functionality added in the simulation or is it an ideal case?
The diagrams in the pages use same legend dots which can be confusing if the print is Black and white Justification for the sentence, however, it increases the reliability w.r.t the remaining users communicating is needed as Fig 9 gives a crucial decline in the proposed system.
The paper lacks future research directions and more insights into the overall proposed method.