Browsing by Author "Walelgne, Ermias"
Now showing 1 - 8 of 8
- Results Per Page
- Sort Options
- Analyzing Throughput and Stability in Cellular Networks
A4 Artikkeli konferenssijulkaisussa(2018-07-09) Walelgne, Ermias; Manner, Jukka; Bajpai, Vaibhav; Ott, JörgThe throughput of a cellular network depends on a number of factors such as radio technology, limitations of device hardware (e.g., chipsets, antennae), physical layer effects (interference, fading, etc.), node density and demand, user mobility, and the infrastructure of Mobile Network Operators (MNO). Therefore, understanding and identifying the key factors of cellular network performance that affect end-users experience is a challenging task. We use a dataset collected using netradar, a platform that measures cellular network performance crowd- sourced from mobile user devices. Using this dataset we develop a methodology (a classifier using a machine learning approach) for understanding cellular network performance. We examine key characteristics of cellular networks related to throughput from the perspective of mobile user activity, MNO, smartphone models, link stability, location and time of day. We perform a network-wide correlation and statistical analysis to obtain a basic understanding of the influence of individual factors. We use a machine learning approach to identify the important features and to understand the relationship between different ones. These features are then used to build a model to classify the stability of cellular network based on the data reception characteristics of the user. We show that it is possible to classify reasons for network instability using minimal cellular network metrics with up to 90% of accuracy. - Clustering and predicting the data usage patterns of geographically diverse mobile users
A1 Alkuperäisartikkeli tieteellisessä aikakauslehdessä(2021-03-14) Walelgne, Ermias; Asrese, Alemnew; Manner, Jukka; Bajpai, Vaibhav; Ott, JörgMobile users demand more and more data traffic, yet network resources are limited. This creates a challenge for network resource management. One way of addressing this challenge is by understanding the data usage patterns of mobile users so that resources can be optimally allocated based on user traffic demand and data usage behavior. However, understanding and characterizing the data usage patterns of mobile users is a complex task. In this work, we investigate and characterize users’ data usage patterns and behavior in mobile networks. We leverage a dataset (∼113 million records) collected through a crowd-based mobile network measurement platform – Netradar – across five countries. Data usage behavior of users over a cellular network is primarily driven by user mobility, the type of subscription plan marketed by Mobile Network Operators (MNOs), network congestion, and network coverage. We apply an unsupervised machine learning approach to cluster mobile user types by considering different factors such as data consumption, network access type, the number of sessions created per user, throughput, and mobility. By defining data usage pattern of mobile users, we develop a user clustering model and identify three different mobile user groups (clusters). Our clustering model shows that the data usage patterns are unevenly distributed across the five countries studied, characterized by a small number of heavy users consuming the highest volume of data. We show how the types of applications installed by users correlate with data consumption patterns in some countries. Heavy users tend to install more traffic-demanding apps than users from the other two groups – regular and light users. Finally, we trained a classification model using the labeled dataset produced by our aforementioned user clustering method. The model helps classifying mobile users according to their usage patterns (i.e., heavy, regular, and light) with an accuracy of ∼80% in the test dataset. - Correlation-Based Feature Mapping of Crowdsourced LTE Data
A4 Artikkeli konferenssijulkaisussa(2018) Apajalahti, Kasper; Walelgne, Ermias; Manner, Jukka; Hyvönen, EeroThere have been efforts taken by different research projects to understand the complexity and the performance of a mobile broadband network. Various mobile network measurement platforms are proposed to collect performance metrics for analysis. Data integration would provide more thorough data analyses as well as prediction and decision models from one dataset to another. The crucial part of the data integration is to find out, whether two datasets have corresponding features (performance metrics). However, finding common features across datasets is a challenging task. For example, features might: 1) have similar names but be different metrics, 2) have different names but be similar metrics, or 3) be same metrics but have differences in the underlying methodology. We designed a feature mapping methodology between two crowdsourced LTE measurement-based datasets. Our method is based on correlations between the features and the mapping algorithm is solving a maximum constraint satisfaction problem (CSP). We define our constraints as inequality patterns between the correlation coefficients of the measured features. Our results show that the method maps measurement features based on their correlation coefficients with high confidence scores (between 0.78 to 1.0 depending on the amount of features). We observe that mapping score increases as a function of the amount of features. Altogether, our results show that this methodology can be used as an automated tool in the measurement data integration. - Factors Affecting Performance of Web Flows in Cellular Networks
A4 Artikkeli konferenssijulkaisussa(2018) Walelgne, Ermias; Kim, Setälä; Bajpai, Vaibhav; Neumeier, Stefan; Manner, Jukka; Ott, JörgStudies show that more than 95% of the traffic generated by smartphones typically consists of short-lived TCP flows towards websites. The content of such websites often is distributed across multiple servers which requires clients to resolve multiple DNS names and establish multiple TCP connections to fetch the webpage in its entirety. Studies have shown that network latency in a mobile network (attributed to DNS lookup and TCP connect times) contributes heavily to poor experience when browsing such websites. However, there is little understanding of the factors that contribute to high DNS lookup and TCP connect times. In this paper, we take this further by measuring the Domain Name System (DNS) lookup time and the TCP connect time to popular websites from ∼25K subscribers of a cellular network operator in Finland. Using a month-long dataset (Oct 2016) of these measurements, we show that LTE offers considerably faster DNS lookup time compared to legacy cellular networks (such as HSPA+ and UMTS). We also show that the model of the device and the proximity of the DNS server to the subscribers impacts the DNS lookup time. Furthermore, the TCP connect time is also affected by the radio technology. We observe that LTE offers a significantly low latency profile such that the TCP connect time to popular websites is reduced by ∼80% compared to legacy cellular networks. The presence of ISP caches also considerably improves TCP connect times. Using a ping test, we also observe that legacy radio technologies (such as HSPA+ and UMTS) suffer from higher packet loss than LTE - Measuring the Feasibility of Teleoperated Driving in Mobile Networks
A4 Artikkeli konferenssijulkaisussa(2019) Neumeier, Stefan; Walelgne, Ermias; Bajpai, Vaibhav; Ott, Jörg; Facchi, ChristianTeleoperated Driving is the remote control driving of a vehicle by a human driver. The concept of Teleoperated Driving requires the use of mobile networks, which typically experience variable throughput, variable latency and uneven network coverage. To investigate whether Teleoperated Driving can be possible with contemporary mobile networks, we have conducted measurements while driving with vehicles in the real world. We used complementary measurement setups to obtain results that can be compared. The dataset consists of about 5200 km (4660 minutes) driving measurements. Results show that Teleoperated Driving could be possible, but the high variance of network parameters makes it difficult to use the system at all times. It appears that the speed of the vehicle and the distance to the base station may not influence Teleoperated Driving, while handover with changed radio technology, signal strength and distance to the teleoperation station may have an impact. Possible mitigations to overcome these problems along with a basic whitelisting approach is discussed. - Measuring web latency in cellular networks
Poster(2018) Asrese, Alemnew; Walelgne, Ermias; Lutu, Andra; Alay, Ozgu; Ott, JörgThis work presents a new methodology to measure the performance of web browsing over operational Mobile Broadband (MBB) networks. We designed a web performance measurement tool that collects both the web QoS metrics and the web rendering time in a browser window. We used MONROE [1], European wide measurement platform, to deploy our tool and to conduct a web browsing measurement over operational MBB networks. The results from the initial deployment show that different operators across countries and within the same country have a significant difference in web browsing performance (e.g. in the median case TIM is 75 ms faster than I WIND regarding time to first byte (TTFB)). - Measuring Web Latency in Cellular Networks
Abstract(2018) Asrese, Alemnew; Walelgne, Ermias; Lutu, Andra; Alay, Ozgu; Ott, JörgThis work presents a new methodology to measure the performance of web browsing over operational Mobile Broadband (MBB) networks. We designed a web performance measurement tool that collects both the web QoS metrics and the web rendering time in a browser window. We used MONROE [1], European wide measurement platform, to deploy our tool and to conduct a web browsing measurement over operational MBB networks. The results from the initial deployment show that different operators across countries and within the same country have a significant difference in web browsing performance (e.g. in the median case TIM is 75 ms faster than I WIND regarding time to first byte (TTFB)). - Poster: Using Crowdsourcing Data for Adaptive Video Streaming in Cellular Network
Poster(2018-05-04) Walelgne, Ermias; Asrese, Alemnew; Bajpai, Vaibhav; Ott, Jörg; Manner, Jukka