Review Reports - RAP-MAC: A Robust and Adaptive Pipeline MAC Protocol for Underwater Acoustic String Networks

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper introduces a novel MAC protocol with adaptive and fault recovery capabilities for underwater acoustic networks. The authors present some specific designs for marine environments, which have been verified through real sea experiments. The proposed algorithms comprehensively consider system throughput, adaptability, robustness, as well as time and energy efficiency, contributing meaningful test results. While the paper is well-written, minor revisions are suggested.

l The analysis of challenges facing underwater MAC protocols could be clearer. The authors should offer more explicit generalizations of existing issues, including factors leading to link interruptions, network standardization, hardware limitations, and special requirements of underwater network protocols. This would enhance readers' understanding of the problems addressed and innovations proposed.

l Simulation results demonstrate that the proposed RAP-MAC outperforms the PMAC protocol in both throughput and delivery ratio. Why was PMAC the only protocol chosen for comparison? It's recommended to explain the reasons for excluding other MAC protocols.

Comments on the Quality of English Language

Some cited references display formatting discrepancies.

Author Response

Point 1: The analysis of challenges facing underwater MAC protocols could be clearer. The authors should offer more explicit generalizations of existing issues, including factors leading to link interruptions, network standardization, hardware limitations, and special requirements of underwater network protocols. This would enhance readers' understanding of the problems addressed and innovations proposed.

Response 1: Thanks for the suggestion. We revised the related parts based on your comments. We have reorganized the challenges, commencing with physical channels, followed by underwater acoustic communication, and culminating in network considerations.

The modifications (Lines 50-62) are as follows: “However, MAC protocol design faces many challenges. First, underwater acoustic channels are considered to be highly challenging, characterized by low propagation speed, limited bandwidth, multipath, large Doppler effect, high noise and spatio-temporal variability. These characteristics result in long propagation delay, low data rate, high bit error rate and unstable links in underwater acoustic communication. Second, these aforementioned features bring high end-to-end delay, low delivery ratio, and topology changes to underwater networks. What’s worse, underwater devices are susceptible to damage because of the harsh marine environment. If any hop link in the underwater acoustic string network is disrupted due to device failure, the entire network can become paralyzed. Given these challenging issues and considering the hardware constraints such as half-duplex, limited available bandwidth and low computing power, the existing MAC protocols are insufficient in handling them effectively. “

Point 2: Simulation results demonstrate that the proposed RAP-MAC outperforms the PMAC protocol in both throughput and delivery ratio. Why was PMAC the only protocol chosen for comparison? It's recommended to explain the reasons for excluding other MAC protocols.

Response 2: We are grateful for the suggestion. Through a survey on existing work, we found that there is a very limited number of research on MAC protocols for underwater acoustic string networks. They focus on improving throughput and do not take into account the spatiotemporal variability, vulnerability, and unreliability of the underwater acoustic link, as well as the susceptibility of underwater nodes to failures. When the channel deteriorates, leading to a high packet loss rate, the network performance can be significantly impacted if the protocol algorithm is unable to effectively handle it. In addition, when the physical communication link is interrupted due to node failures, the network will be completely disabled. To address these issues, the proposed RAP-MAC protocol is specifically designed to enhance the robustness and reliability of the network, with adaptability to the changing underwater acoustic channel and fault recovery capability. This has been demonstrated through simulation experiments compared with PMAC, a representative MAC for the string topology.

Point 3: Some cited references display formatting discrepancies.

Response 3: We feel sorry for the problem. We have checked all the references using Google Scholar.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This paper proposed a systematic robust and adaptive pipeline MAC (RAP-MAC) protocol for underwater acoustic string networks (UA-SNs). The characteristics of the proposed protocol was illustrated. Simulations and ocean experiments were also performed to verify the feasibility of the proposed scheme. However, some concerns should be carefully addressed before its publication.

1. In the abstract, the authors should address the characteristics and novelty of the proposed protocol, and statements for simulations and experiments should be shortened. And numerical results should be addressed in the abstract.

2. The language of the article needs improvement, as in Line 16, 17.

3. In the introduction, the authors should address the characteristics of the proposed protocol as in the contribution, instead of addressing the simulations and experiments in the contribution, especially given that “the US Navy’s Seaweb and PLUSNet have over 20 nodes and a coverage range of over 100km”.

4. In the section of experiments and results, more typical protocols besides PMAC should adopted as the compared schemes in the simulations.

5.The section of experiments and results is suggested to be divided into two subsections, one for simulation results and another for ocean experiments.

6. In the simulation results, as in Fig. 7 and Fig.8, more numerical results should be analyzed.

Comments on the Quality of English Language

The language of the article needs improvement, as in Line 16, 17.

Author Response

Point 1: In the abstract, the authors should address the characteristics and novelty of the proposed protocol, and statements for simulations and experiments should be shortened. And numerical results should be addressed in the abstract.

Response 1: We are grateful for the suggestion. We have already rewritten the abstract according to your comments. The modifications are as follows: “The development of underwater acoustic networks is a significant expansion of Internet-of-Things technology to underwater environments. These networks are essential for a variety of marine applications. For many practical uses, it is more efficient to collect marine data from a remote location over multiple hops, rather than direct point-to-point communications. In this article, we will focus on the underwater acoustic string network (UA-SN) designed for this type of application. We propose a Robust and Adaptive Pipeline Medium Access Control (RAP-MAC) protocol to enhance the network's transmission efficiency, adaptability, and robustness. The protocol includes a scheduling-based concurrent algorithm, online real-time configuration adjustment function, a rate mode adaptive algorithm, and a fault recovery algorithm. We conducted simulations to compare the new protocol with another representative protocol, validating the RAP-MAC protocol's adaptability and fault recovery ability. Additionally, we carried out two large-scale sea trials. The results of these experiments indicate that the RAP-MAC protocol ensures effectiveness and reliability in large-scale multi-hop UA-SNs. In the South China Sea, we were able to achieve a communication distance of 87km with a throughput of 601.6bps, exceeding the recognized upper bound of underwater acoustic communication experiment performance by 40km·kbps.”

Point 2: The language of the article needs improvement, as in Line 16, 17.

Response 2: Thanks for the suggestion. We have improved the expression of this sentence. The modification (Lines 13-15) is: “In the South China Sea, we were able to achieve a communication distance of 87km with a throughput of 601.6bps, exceeding the recognized upper bound of underwater acoustic communication experiment performance by 40km·kbps.”

Point 3: In the introduction, the authors should address the characteristics of the proposed protocol as in the contribution, instead of addressing the simulations and experiments in the contribution, especially given that “the US Navy’s Seaweb and PLUSNet have over 20 nodes and a coverage range of over 100km”.

Response 3: We sincerely appreciate the comment. We revised the related parts based on your comments. We still included our sea trial efforts in the contribution part. Even though the US Navy’s Seaweb project deployed over 20 nodes, and had a coverage of over 100km, they did not achieve the same network throughput as we did. Further, in the area of underwater wireless networks, many of the research still remain in the simulation stage. Not many existing studies have conducted sea trials, especially large-scale sea trials as we did. It takes a lot of effort to develop the hardware and software needed for the sea trials. Moreover, the cost of sea trials is considerably high. It also takes a lot of time and careful planning to conduct tests in the sea. Therefore, the sea experiment data are very valuable. Hence we consider them as an important contribution to the research community.

The modifications (Lines 77-97) are as follows: “(1) In order to establish a practical UA-SN in real-world scenarios, taking into consideration the network features, channel characteristics and physical layer parameters, a systematic cross-layer optimized RAP-MAC is proposed. The RAP-MAC protocol design is characterized by four key highlights: parallel transmission method to prevent conflicts and improve network throughput, network regulation strategy to enhance the controllability of the network, cross-layer optimized rate adaptation algorithm to improve the adaptability to ever-changing ocean environment, and fault recovery algorithms to increase the network robustness. The comparative analysis with PMAC through simulations confirmed the adaptive ability to environment changes and the fault recovery ability of RAP-MAC.

(2) Many existing researches on underwater networks are simulation-based, with few network systems conducting experiments in real ocean environments. In order to further validate the practicability of RAP-MAC, we established the UA-SN with 21 nodes in the selected areas of the North Sea and the South Sea of China for experiments respectively. The sea trial verification took nearly a month, during which we mobilized more than 40 people and 25 boats. Through a great deal of effort, the experimental results from two sea areas fully demonstrated good performance of the UA-SN using RAP-MAC from the abovementioned four aspects.

(3) The experiment result of a communication distance of 87km and a throughput of 601.6bps for UA-SNs is unparalleled. In our best effort, no other network systems in current published research have achieved such a communication distance with such a comparable throughput. Furthermore, this achievement exceeds the recognized upper limit of underwater acoustic communication test performance by 40km·kbps. “

Point 4: In the section of experiments and results, more typical protocols besides PMAC should adopted as the compared schemes in the simulations.

Response 4: Thank you for the comment. Through a survey on existing work, we found that there is a very limited number of research on MAC protocols for underwater acoustic string networks. They focus on improving throughput and do not take into account the spatiotemporal variability, vulnerability, and unreliability of the underwater acoustic link, as well as the susceptibility of underwater nodes to failures. When the channel deteriorates, leading to a high packet loss rate, the network performance can be significantly impacted if the protocol algorithm is unable to effectively handle it. In addition, when the physical communication link is interrupted due to node failures, the network will be completely disabled. To address these issues, the proposed RAP-MAC protocol is specifically designed to enhance the robustness and reliability of the network, with adaptability to the changing underwater acoustic channel and fault recovery capability. This has been demonstrated through simulation experiments compared with PMAC, a representative MAC for the string topology.

Point 5: The section of experiments and results is suggested to be divided into two subsections, one for simulation results and another for ocean experiments.

Response 5: Thanks for the suggestion. Section 5 “Experiments and Results” is divided into four subsections. Among them, subsection 1 lists the metrics for evaluating the performance of the proposed MAC protocol in simulations and ocean experiments. Subsection 2 focuses on simulations, subsection 3 focuses on ocean experiments, and subsection 4 provides a summary of the simulations and ocean experiments, along with a discussion on the applicability of the RAP-MAC protocol.

Point 6: In the simulation results, as in Fig. 7 and Fig.8, more numerical results should be analyzed.

Response 6: We sincerely appreciate the comment. More analysis for Figures 7 and 8 is added (Lines 474-475, 480-486 and 493-496, Page 14). In response to the variations in throughput and end-to-end delivery ratio depicted in the figures, we have provided a more detailed analysis of the underlying causes. The added content is as follows:

“Lines 474-475: When a fault occurs, the PMAC protocol cannot recover on its own and the network is directly interrupted, resulting in a throughput of 0 after the network interruption. Thus, this leads to a lower overall average throughput.

Lines 480-486: Assuming that node 5 fails, according to the RAP-MAC protocol algorithm rules, node 4 needs to skip node 5 and directly connect to node 6. Since nodes 4 and 7 can send data simultaneously, in order to avoid conflicts at node 6, the data transmission delay cannot exceed the single-hop propagation delay. Therefore, after the fault recovery, not only the rate decreases, but also the maximum supported packet length is reduced. This results in a substantial decline in network throughput after fault recovery, leading to a decrease in overall average throughput.

Lines 493-496: In accordance with the proposed algorithm rules, following the recovery from a fault, the rate mode decreases. The lower rate mode has a higher encoding redundancy, which leads to a higher communication success probability, bringing a higher delivery ratio.”