Title: Priority-based Parallel Processing Scheme of Mass Data for Real-time DDS Monitoring System
Congratulations! - your paper #1570537448("Priority-based Parallel Processing Scheme of Mass Data for Real-time DDS Monitoring System") for ICUFN 2019 has been ACCEPTED and will be presented at the ICUFN 2019 conference in Zagreb, Croatia, 2-5 July, 2019
Reviewer(s)' Comments to Author:
Accept if room (3)
This paper suggests priority based monitoring system for Data Distribution System. The authors propose priority scheduling algorithm and evaluated the scheme using simulation.
This paper addresses an important problem in DDS. To guarantee the integrity of messages in DDS, it is necessary to accommodate monitoring system. The authors considered data-mining based monitoring system, which can provide complex monitoring across multiple messages in cost of expensive computation. To provide computation resources to message readers with high priority over semi-batch monitoring processes, the authors developed a simple scheduling algorithm.
The proposed system is very simple and the evaluation is not realistic. While, each tasks has dependency on other tasks in many cases, the authors assume that every task is orthogonal. Also, though the paper implies real-time environment, the scheduling algorithm does not consider real-time deadlines.
Also, there is no evaluation results proposed method in the paper.
Accept if room (3)
The paper proposes a system to monitor the DDS messages and detect and cope with errors.
The proposed system is compared to a basic fault checking system.
- It said that "parallel environment can be implemented by GPU or OpenCL " --> You meand CUDA or OpencL ?
- It is not clear how the simulation was conducted.
- line 6, Alg II: For i<-0 to 2 do --> you mean For i = 0 to 2 (as in Algorithm I) ?
Likely reject (2)
The authors proposed a priority-based monitoring system in Data Distribution Service environment.
The authors conduct some simulation experiments and confirm the effectiveness of the proposed system in terms of processing time and execution time.
The explain of the proposed system is too abstracted, and the reviewer can not understand the application of the proposed system in the actual network systems. The authors should present the concrete application scenario in the paper to show the importance of the proposed system.