{"title":"An empirical characterization of cellular network performance","authors":"A. Seetharam, Peter Walker","doi":"10.1109/ICCNC.2016.7440682","DOIUrl":null,"url":null,"abstract":"With the proliferation of smartphones and the growth of cellular technology (e.g., 4G LTE), individuals are constantly accessing the Internet on the go. But, there is limited understanding of the variation in cellular network performance (i.e., throughput, latency, packet loss) for various wireless service providers. In this paper, we perform an empirical characterization of cellular network performance based on a large scale measurement study of the major wireless service providers in California. We describe the system and software application (CalSPEED) used to measure various network parameters such as TCP download/upload throughput, latency and packet loss for the different wireless service providers. Six rounds of measurement are conducted for a period of approximately three years in California, (referred to as field tests) starting from spring 2012. During each field test, personnel drove with mobile clients to different parts of the state and collected data by executing the software. Data is collected over approximately 1500 locations in the state during each field test1. Our data analysis demonstrates quantitatively that i) cellular network performance varies considerably with location and is poor overall - a large number of locations in our study report low network throughput, high latency and high loss rates ii) all cellular providers are not the same - network performance varies considerably among the different providers iii) on average, cellular network performance is improving over time iv) cellular network performance depends on a number of factors such as download/upload, server location and type of device used.","PeriodicalId":308458,"journal":{"name":"2016 International Conference on Computing, Networking and Communications (ICNC)","volume":"292 1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2016-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2016 International Conference on Computing, Networking and Communications (ICNC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCNC.2016.7440682","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
With the proliferation of smartphones and the growth of cellular technology (e.g., 4G LTE), individuals are constantly accessing the Internet on the go. But, there is limited understanding of the variation in cellular network performance (i.e., throughput, latency, packet loss) for various wireless service providers. In this paper, we perform an empirical characterization of cellular network performance based on a large scale measurement study of the major wireless service providers in California. We describe the system and software application (CalSPEED) used to measure various network parameters such as TCP download/upload throughput, latency and packet loss for the different wireless service providers. Six rounds of measurement are conducted for a period of approximately three years in California, (referred to as field tests) starting from spring 2012. During each field test, personnel drove with mobile clients to different parts of the state and collected data by executing the software. Data is collected over approximately 1500 locations in the state during each field test1. Our data analysis demonstrates quantitatively that i) cellular network performance varies considerably with location and is poor overall - a large number of locations in our study report low network throughput, high latency and high loss rates ii) all cellular providers are not the same - network performance varies considerably among the different providers iii) on average, cellular network performance is improving over time iv) cellular network performance depends on a number of factors such as download/upload, server location and type of device used.