高，但帶給旅人的極大的方便，對於國家生產力 有重大的影響。撿了芝麻掉了西
瓜。本人常常出國，在機場各地，只要有插座的地方， 都有旅客使用，從未見機場
人員干涉。新建的機場，不但有電源插座，還有免費的無線區域網路。 本人數度因
錯過末班客運車而滯留印第安那波里機場，就直接在機場窩個數小時，等待早班 客
運車到普渡大學。在等待時間，有免費的電源及無線區域網路可用，在機場正好可
以 利用時間將旅途中積壓的工作做一個清理，(多半是信箱中的垃圾信件。) 在現代
的資訊社會中，台灣還有很大的改進空間。其中，交通系統中的電源供應，是 提高
旅人生產力的第一步，跨不出這一步，奢談其他。
日期: Fri, 23 Dec 2011 12:30:26 -0500 寄件者: MP2P 2012 <[email protected]>
收件者: Yao-Nan Lien <[email protected]>
副本: Yi-Shiuan Lin <[email protected]>, [email protected]
主旨: [MP2P 2012] Your paper #1569537241 ('Mobile Agent Tracking Technology over Opportunistic Network') has been accepted
Dear Prof. Yao-Nan Lien:
We are pleased to inform you that your paper “Mobile Agent Tracking Technology over Opportunistic Network”
(#1569537241) has been accepted in the final program of The 8th International Workshop on Mobile Peer-to-Peer Computing (MP2P 2012).
The reviews of your paper are enclosed below. When preparing the final version, please take the reviewers’
comments under consideration. Should your paper be approved, you will need to have the camera ready version uploaded by January 27, 2012. You will find instructions regarding the formatting requirements and final submission of your paper at http://www.percom.org/.
Once again, many thanks for contributing your work to MP2P 2012. If you have any additional questions, please feel free to contact us.
Sincerely,
Chih-Lin Hu, Ying Cai, and Hong Va Leong TPC Chair co-chairs
IEEE PerCom 2012 Workshop on Mobile Peer-to-Peer Computing
=========================================================================================
The reviews below are also posted on EDAS at http://edas.info/showPaper.php?m=1569537241.
=========================================================================================
======= Review 1 =======
> *** Summary: Please enter a short summary of the paper, including its main contributions and its relevance to PerCom.
The authors investigate the application of mobile agent on the opportunistic network characterized by “CenWits Search and Rescue System” applied in YuShan National Park as well as propose a control network technology to support rapid search of mobile agent. Under different objectives and constraints, they propose several placement models for control point network deployment. The authors design some efficient heuristic algorithms to solve the placement problem. In addition, they design a simple agent search algorithm based on the control network to support rapid agent search.
> *** Reasons to accept: Please give the main reasons for accepting the paper.
Please see detailed comments.
> *** Reasons to reject: Please give the main reasons for rejecting the paper.
Please see detailed comments.
> *** Detailed Comments: Please provide detailed comments to authors and TPC including strong points, weaknesses and any errors.
1. I don't think the control network idea can be easily integrated with opportunistic networks in reality. If we are able to deploy the infrastructure of a control network, then the communication in the target area is not totally opportunistic anymore.
2. In Section V, the authors should provide more details about the simulation parameters and data sets. It is very hard for readers to understand and judge the correctness of the simulation results without important information such as simulation environment (hardware and software), data set details (real or synthetic), etc.
3. The proceeding name of the second reference is wrong. Please correct it.
4. There are some grammatical errors and typos in this paper. For example, page 1, left column, “batter filed networking” => “battle field networking”. The authors should proofread the paper before publication.
> *** Originality: Not known or experienced before. A technique or a method not used before. Has this or similar work been previously reported? Are the problems and/or approaches in the paper completely new?
Marginal (3)
> *** Novelty: It may be not necessary for the paper to develop new techniques, or to generate new knowledge, but it should, at least, apply, or combine, them in a fresh and novel way or shed some new light on their applicability in a certain domain.
Marginal (3)
> *** Presentation: Adequate organization of the article and the language used in it, as to make its content clear, easily readable and understandable. Clarity in what has been achieved by the author of the article. Even technical papers on a narrow topic should be written such that non-experts can comprehend the main contribution of the paper and the methods employed. The paper shouldn't just be a litany of deep but obscure theorems. The information of the paper should be available to the reader with a minimum of effort.
Marginal (3)
> *** Overall Evaluation: Please rate the paper.
Accept if room (3)
> *** Reviewer Expertise: Please rate your expertise.
Familiar (2)
======= Review 2 =======
> *** Summary: Please enter a short summary of the paper, including its main contributions and its relevance to PerCom.
Presented is a mechanism for internet access point (referred to as control point in the paper) selection. The problem is formulated as max-flow assuming the network is known a prior.
> *** Reasons to accept: Please give the main reasons for accepting the paper.
o. relevance to mp2p high
> *** Reasons to reject: Please give the main reasons for rejecting the paper.
o. assumptions not realistic
o. proposed CN mechanism not plausible for the motivating app o. lack of tech novelty
> *** Detailed Comments: Please provide detailed comments to authors and TPC including strong points, weaknesses and any errors.
o. The proposed mechanism is motivated by opportunistic networks such as Cenwits designed for hiker tracking, where the hikers carry mobile nodes that log and pass the trajectory info to each other. It is also known as a delay-tolerant network.
o. The problem formulated is however a little hypothetical. Assuming there are plenty of control points (CP) around, the problem to solve is how a set of mobile nodes can find /contact their targets in a cost-efficient way. The CPs have different cost. It's essentially to max the # of flows under the resource constraint - a max-flow problem.
o. Having formulated the problem as max-flow, the authors then went off to formulate it as a max-coverage problem, which obscures the original formulation. Applying yet another hypothetical utility function, the max-coverage is extended to max-coverage-utility.
o. WIth all the above said, heuristics are used for the actual computation. The three heuristics do not differ significantly and the results show that really the computation does not scale to the size of the network.
o. The main concern is the fact that global info is required for the computation, which is exactly how the proposed solution contradicts the motivating app. The solution does not scale to the network.
o. The 2nd major concern is the various unrealistic assumptions. These assumptions diverge from the characteristics of the motivating application.
o. The 3rd concern is lack of technical novelty. The application is not new. The proposed heuristics are not
o. The 3rd concern is lack of technical novelty. The application is not new. The proposed heuristics are not significantly better and have appeared in the authors' prior work [4]
o. There are two major differences communicating mobile agents vs. mobile nodes in a DTN. (1) The mobile agents are software components residing in a connected network. THe mobile nodes, in particular in the cenwits scenario, rarely connect to each other. (2) The mobile agents reside in computationally-resourceful nodes. The mobile nodes are low-end devices. Such an application of idea working for mobile agent paradigm is IMHO not suitable for highly dynamic DTN.
> *** Originality: Not known or experienced before. A technique or a method not used before. Has this or similar work been previously reported? Are the problems and/or approaches in the paper completely new?
Low (2)
> *** Novelty: It may be not necessary for the paper to develop new techniques, or to generate new knowledge, but it should, at least, apply, or combine, them in a fresh and novel way or shed some new light on their applicability in a certain domain.
Marginal (3)
> *** Presentation: Adequate organization of the article and the language used in it, as to make its content clear, easily readable and understandable. Clarity in what has been achieved by the author of the article. Even technical papers on a narrow topic should be written such that non-experts can comprehend the main contribution of the paper and the methods employed. The paper shouldn't just be a litany of deep but obscure theorems. The information of the paper should be available to the reader with a minimum of effort.
Marginal (3)
> *** Overall Evaluation: Please rate the paper.
Accept if room (3)
> *** Reviewer Expertise: Please rate your expertise.
Expert (3)
======= Review 3 =======
> *** Summary: Please enter a short summary of the paper, including its main contributions and its relevance to PerCom.
In this paper, the authors proposed to use mobile agent to enhance the message transfer efficiency on an opportunistic network. In particular, they designed the efficient heuristic to solve the placement problem of control points and investigated the corresponding application in YuShanNet.
> *** Reasons to accept: Please give the main reasons for accepting the paper.
Please see detailed comments.
> *** Reasons to reject: Please give the main reasons for rejecting the paper.
> *** Detailed Comments: Please provide detailed comments to authors and TPC including strong points, weaknesses and any errors.
This paper proposed to use mobile agent to enhance the message transfer efficiency on an opportunistic network. In particular, the authors designed the efficient heuristic to solve the placement problem of control points and
investigated the corresponding application in YuShanNet. Experiments were conducted by simulations and
experimental results showed their heuristics are reasonable. The paper is well-writen and well-organlized. This paper will be interesed by most of the intended readers.
> *** Originality: Not known or experienced before. A technique or a method not used before. Has this or similar work been previously reported? Are the problems and/or approaches in the paper completely new?
High (4)
> *** Novelty: It may be not necessary for the paper to develop new techniques, or to generate new knowledge, but it should, at least, apply, or combine, them in a fresh and novel way or shed some new light on their applicability in a certain domain.
High (4)
> *** Presentation: Adequate organization of the article and the language used in it, as to make its content clear, easily readable and understandable. Clarity in what has been achieved by the author of the article. Even technical papers on a narrow topic should be written such that non-experts can comprehend the main contribution of the paper and the methods employed. The paper shouldn't just be a litany of deep but obscure theorems. The information of the paper should be available to the reader with a minimum of effort.
High (4)
> *** Overall Evaluation: Please rate the paper.
Accept (4)
> *** Reviewer Expertise: Please rate your expertise.
Familiar (2) Regards,
Chih-Lin Hu, Ying Cai, and Hong Va Leong TPC Chair co-chairs
IEEE PerCom 2012 Workshop on Mobile Peer-to-Peer Computing