{"title":"The impact of content popularity distribution on energy efficient caching","authors":"N. Osman, T. El-Gorashi, J. Elmirghani","doi":"10.1109/ICTON.2013.6602957","DOIUrl":null,"url":null,"abstract":"This paper evaluates the effect of content popularity distribution on the power consumption of cache-based services in an IP over WDM network. We consider the Zipf, Bimodal, Pareto and Equal Popularity distributions for content popularity. The results reveal that using our MILP model, reductions in power consumption under the Zipf, Pareto, Bimodal and Equal Popularity distributions are up to 47%, 71%, 72% and 71%, respectively. We also study the impact of the sizes of media files and the impact of the power efficiency of caches on the network power consumption. The results confirm that under the Zipf and Equal Popularity distributions the power consumption rises as the average size of media files and cache power consumption increases. Under the Bimodal and Pareto distributions, the average cache size in the network is not affected by the increase in media file sizes.","PeriodicalId":376939,"journal":{"name":"2013 15th International Conference on Transparent Optical Networks (ICTON)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2013-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"19","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2013 15th International Conference on Transparent Optical Networks (ICTON)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICTON.2013.6602957","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 19
Abstract
This paper evaluates the effect of content popularity distribution on the power consumption of cache-based services in an IP over WDM network. We consider the Zipf, Bimodal, Pareto and Equal Popularity distributions for content popularity. The results reveal that using our MILP model, reductions in power consumption under the Zipf, Pareto, Bimodal and Equal Popularity distributions are up to 47%, 71%, 72% and 71%, respectively. We also study the impact of the sizes of media files and the impact of the power efficiency of caches on the network power consumption. The results confirm that under the Zipf and Equal Popularity distributions the power consumption rises as the average size of media files and cache power consumption increases. Under the Bimodal and Pareto distributions, the average cache size in the network is not affected by the increase in media file sizes.