Yohan Baga, Fakhreddine Ghaffari, D. Declercq, E. Zante, Michael Nahmiyace
{"title":"Reduction of frames storage size in AFDX reception end-system using a lossless compression algorithm","authors":"Yohan Baga, Fakhreddine Ghaffari, D. Declercq, E. Zante, Michael Nahmiyace","doi":"10.1109/DASC.2017.8102086","DOIUrl":null,"url":null,"abstract":"The growth of bandwidth needs and reliability requirements has determined Avionics Full-Duplex Switched Ethernet (AFDX) networks as the new generation of on-board communication mediums. AFDX belongs to the deterministic, real-time and Ethernet-based network family. The AFDX terminals are called End-Systems (ES). The frames arriving at an ES have to be stored in a reception buffer to avoid frames losses or corruptions due to slowdowns in the ES layers. Little attention is carried to the issue of buffer dimensioning which is generally set to a very large size. However, a too large buffer size leads to costs in terms of memory resources and energy. In this paper, we propose to reduce the reception buffer size by using an LZW-based compression algorithm implemented in hardware. To do that, we interpret frames as sequences of hexadecimal source symbols, and we use a set of 4 parallel dictionaries to encode sequences of source symbols in fix-length words. We realize compression gain measured on sets of frames comprising several millions of symbols, and we obtain until 22% of memory gain when the dictionaries sizes are optimally dimensioned.","PeriodicalId":130890,"journal":{"name":"2017 IEEE/AIAA 36th Digital Avionics Systems Conference (DASC)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 IEEE/AIAA 36th Digital Avionics Systems Conference (DASC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/DASC.2017.8102086","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
The growth of bandwidth needs and reliability requirements has determined Avionics Full-Duplex Switched Ethernet (AFDX) networks as the new generation of on-board communication mediums. AFDX belongs to the deterministic, real-time and Ethernet-based network family. The AFDX terminals are called End-Systems (ES). The frames arriving at an ES have to be stored in a reception buffer to avoid frames losses or corruptions due to slowdowns in the ES layers. Little attention is carried to the issue of buffer dimensioning which is generally set to a very large size. However, a too large buffer size leads to costs in terms of memory resources and energy. In this paper, we propose to reduce the reception buffer size by using an LZW-based compression algorithm implemented in hardware. To do that, we interpret frames as sequences of hexadecimal source symbols, and we use a set of 4 parallel dictionaries to encode sequences of source symbols in fix-length words. We realize compression gain measured on sets of frames comprising several millions of symbols, and we obtain until 22% of memory gain when the dictionaries sizes are optimally dimensioned.