Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917620
Eri Nakagawa, K. Moriya, H. Suwa, Manato Fujimoto, Yutaka Arakawa, K. Yasumoto
Automatic recognition of activities of daily living (ADL) can be applied to realize services to support user life such as elderly monitoring, energy-saving home appliance control, and health support. In particular, “real-time” ADL recognition is essential to realize such a service that the system needs to know the user's current activity. There are many studies on ADL recognition. However, none of these studies address all of the following problems: (1) privacy intrusion due to the utilization of high privacy-invasive devices such as cameras and microphones; (2) limited number of recognizable activities; (3) low recognition accuracy; (4) high deployment and maintenance costs due to many sensors used; and (5) long recognition time. In our prior work, we proposed a system which solves the problems (1)– (4) to some extent by using user's position data and power consumption data of home electric appliances. In this paper, aiming to solve all the above problems including (5), we propose a new system by extending our prior work. To realize “real-time” ADL recognition while keeping good recognition accuracy, we developed new power meters with higher sensing frequency and introduced new techniques such as adding new features, selecting the best subset of the features, and selecting the best training dataset used for machine learning. We collected the sensor data in our smart home facility for 11 days, and applied the proposed method to these sensor data. As a result, the proposed method achieved accuracy of 79.393% in recognizing 10 types of ADLs.
{"title":"Toward real-time in-home activity recognition using indoor positioning sensor and power meters","authors":"Eri Nakagawa, K. Moriya, H. Suwa, Manato Fujimoto, Yutaka Arakawa, K. Yasumoto","doi":"10.1109/PERCOMW.2017.7917620","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917620","url":null,"abstract":"Automatic recognition of activities of daily living (ADL) can be applied to realize services to support user life such as elderly monitoring, energy-saving home appliance control, and health support. In particular, “real-time” ADL recognition is essential to realize such a service that the system needs to know the user's current activity. There are many studies on ADL recognition. However, none of these studies address all of the following problems: (1) privacy intrusion due to the utilization of high privacy-invasive devices such as cameras and microphones; (2) limited number of recognizable activities; (3) low recognition accuracy; (4) high deployment and maintenance costs due to many sensors used; and (5) long recognition time. In our prior work, we proposed a system which solves the problems (1)– (4) to some extent by using user's position data and power consumption data of home electric appliances. In this paper, aiming to solve all the above problems including (5), we propose a new system by extending our prior work. To realize “real-time” ADL recognition while keeping good recognition accuracy, we developed new power meters with higher sensing frequency and introduced new techniques such as adding new features, selecting the best subset of the features, and selecting the best training dataset used for machine learning. We collected the sensor data in our smart home facility for 11 days, and applied the proposed method to these sensor data. As a result, the proposed method achieved accuracy of 79.393% in recognizing 10 types of ADLs.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114186433","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917619
Akane Ishida, Kazuya Murao, T. Terada, M. Tsukamoto
Refrigerators are commonly used by multiple users in the home and office. However, expired food is sometimes left in the refrigerator, and users may eat food belonging to others since food is often arranged in the refrigerator in a disorderly manner. This happens because food is not organized by owner. If food can be linked with its owner, users will not eat food belonging to others, and food in the refrigerator will be consumed prior to expiration by informing the owner of the expiration date. The simplest way to link food with its owner is to input the name of the owner manually every time he or she puts food in the refrigerator, which, however, is tedious and impractical. We propose a method that identifies who put food in the refrigerator by using pressure sensors, an accelerometer, and a gyroscope attached to the refrigerator door. The method analyzes the motions of opening/closing a refrigerator door and the pressure distribution of gripping the door-handle. In the future, we aim to link food with its owner. From the experiment, we confirmed that the method achieves 90.3% accuracy in user identification for a group of four.
{"title":"A user identification method based on features of opening/closing a refrigerator door","authors":"Akane Ishida, Kazuya Murao, T. Terada, M. Tsukamoto","doi":"10.1109/PERCOMW.2017.7917619","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917619","url":null,"abstract":"Refrigerators are commonly used by multiple users in the home and office. However, expired food is sometimes left in the refrigerator, and users may eat food belonging to others since food is often arranged in the refrigerator in a disorderly manner. This happens because food is not organized by owner. If food can be linked with its owner, users will not eat food belonging to others, and food in the refrigerator will be consumed prior to expiration by informing the owner of the expiration date. The simplest way to link food with its owner is to input the name of the owner manually every time he or she puts food in the refrigerator, which, however, is tedious and impractical. We propose a method that identifies who put food in the refrigerator by using pressure sensors, an accelerometer, and a gyroscope attached to the refrigerator door. The method analyzes the motions of opening/closing a refrigerator door and the pressure distribution of gripping the door-handle. In the future, we aim to link food with its owner. From the experiment, we confirmed that the method achieves 90.3% accuracy in user identification for a group of four.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126267502","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917579
K. Jaffrès-Runser, G. Jakllari, Tao Peng, Vlad Nitu
This paper discusses the design and development efforts made to collect data using an opportunistic crowdsensing mobile application. Relevant issues are underlined, and solutions proposed within the CHIST-ERA Macaco project for the specifics of collecting fine-grained content and context data are highlighted. Global statistics on the data gathered for over a year of collection show its quality: Macaco data provides a long-term and fine-grained sampling of the user behavior and network usage that is relevant to model and analyse for future content and context-aware networking developments.
{"title":"Crowdsensing mobile content and context data: Lessons learned in the wild","authors":"K. Jaffrès-Runser, G. Jakllari, Tao Peng, Vlad Nitu","doi":"10.1109/PERCOMW.2017.7917579","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917579","url":null,"abstract":"This paper discusses the design and development efforts made to collect data using an opportunistic crowdsensing mobile application. Relevant issues are underlined, and solutions proposed within the CHIST-ERA Macaco project for the specifics of collecting fine-grained content and context data are highlighted. Global statistics on the data gathered for over a year of collection show its quality: Macaco data provides a long-term and fine-grained sampling of the user behavior and network usage that is relevant to model and analyse for future content and context-aware networking developments.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"230 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133448861","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917530
P. Giridhar, T. Abdelzaher
In this demo we present a tool that allows us to visualize the real world events on a map interface using the contents shared by users on Twitter and Instagram. Social networks have become popular in recent times for sharing contents about observations made by users. Our tool incorporates a novel algorithm that analyzes the data from both Twitter and Instagram for fusing the contents corresponding to the same event thereby enhancing the corroboration of the event detection techniques for the individual networks. In addition to providing a much cleaner information our tool leverages the various data available from both the social networks (text, images, geo-data) to improve the overall experience of the user visualizing the events.
{"title":"Visualization of events using Twitter and Instagram","authors":"P. Giridhar, T. Abdelzaher","doi":"10.1109/PERCOMW.2017.7917530","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917530","url":null,"abstract":"In this demo we present a tool that allows us to visualize the real world events on a map interface using the contents shared by users on Twitter and Instagram. Social networks have become popular in recent times for sharing contents about observations made by users. Our tool incorporates a novel algorithm that analyzes the data from both Twitter and Instagram for fusing the contents corresponding to the same event thereby enhancing the corroboration of the event detection techniques for the individual networks. In addition to providing a much cleaner information our tool leverages the various data available from both the social networks (text, images, geo-data) to improve the overall experience of the user visualizing the events.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131957157","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917611
Christian Krupitzer, Martin Breitbach, Johannes Saal, C. Becker, Michele Segata, R. Cigno
Mobile IoT devices enable new classes of systems, such as cyber-physical systems. These systems pose challenges as they should seamlessly interact with users and other systems. In this paper, we address the problem of interaction between mobile pervasive IoT devices. Our contributions are threefold. First, we present a concept for a framework for coordination of mobile IoT devices. Second, we implement a reusable robot platform using the Mindstorms toolkit and a customizable adaptation logic for their coordination based on our framework. Third, we show its usability with two applications: an intelligent vehicle highway system as well as a smart vacuum cleaner.
{"title":"RoCoSys: A framework for coordination of mobile IoT devices","authors":"Christian Krupitzer, Martin Breitbach, Johannes Saal, C. Becker, Michele Segata, R. Cigno","doi":"10.1109/PERCOMW.2017.7917611","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917611","url":null,"abstract":"Mobile IoT devices enable new classes of systems, such as cyber-physical systems. These systems pose challenges as they should seamlessly interact with users and other systems. In this paper, we address the problem of interaction between mobile pervasive IoT devices. Our contributions are threefold. First, we present a concept for a framework for coordination of mobile IoT devices. Second, we implement a reusable robot platform using the Mindstorms toolkit and a customizable adaptation logic for their coordination based on our framework. Third, we show its usability with two applications: an intelligent vehicle highway system as well as a smart vacuum cleaner.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131322576","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917609
Yi Ren, Ren-Jie Wu, Y. Tseng
IoT (Internet of Things) has attracted a lot of attention recently. IoT devices need to report their data or status to base stations at various frequencies. The IoT communications observed by a base station normally exhibit the following characteristics: (1) massively connected, (2) lightly loaded per packet, and (3) periodical or at least mostly predictable. The current design principals of communication networks, when applied to IoT scenarios, however, do not fit well to these requirements. For example, an IPv6 address is 128 bits, which is much longer than a 16-bit temperature report. Also, contending to send a small packet is not cost-effective. In this work, we propose a novel framework, which is slot-based, schedule-oriented, and identity-free for uploading IoT devices' data. We show that it fits very well for IoT applications. We propose two schemes, from an ideal one to a more practical one. The main idea is to bundle time slots with certain hashing functions of device IDs, thus significantly reducing transmission overheads, including device IDs and contention overheads.
物联网(Internet of Things)最近引起了人们的广泛关注。物联网设备需要以不同的频率向基站报告其数据或状态。基站观察到的物联网通信通常具有以下特征:(1)大规模连接，(2)每个数据包的负载较轻，以及(3)周期性或至少大部分可预测。然而，当应用于物联网场景时，当前通信网络的设计原则并不能很好地满足这些要求。例如，IPv6地址是128位，比16位的温度报告长得多。此外，争着发送一个小数据包是不划算的。在这项工作中，我们提出了一个新的框架，该框架基于插槽，面向时间表，并且无需身份来上传物联网设备的数据。我们证明它非常适合物联网应用。我们提出了两种方案，从理想方案到比较实际的方案。其主要思想是将时隙与设备id的某些散列函数捆绑在一起，从而显著降低传输开销，包括设备id和争用开销。
{"title":"The hint protocol: Using a broadcast method to enable ID-free data transmission for dense IoT devices","authors":"Yi Ren, Ren-Jie Wu, Y. Tseng","doi":"10.1109/PERCOMW.2017.7917609","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917609","url":null,"abstract":"IoT (Internet of Things) has attracted a lot of attention recently. IoT devices need to report their data or status to base stations at various frequencies. The IoT communications observed by a base station normally exhibit the following characteristics: (1) massively connected, (2) lightly loaded per packet, and (3) periodical or at least mostly predictable. The current design principals of communication networks, when applied to IoT scenarios, however, do not fit well to these requirements. For example, an IPv6 address is 128 bits, which is much longer than a 16-bit temperature report. Also, contending to send a small packet is not cost-effective. In this work, we propose a novel framework, which is slot-based, schedule-oriented, and identity-free for uploading IoT devices' data. We show that it fits very well for IoT applications. We propose two schemes, from an ideal one to a more practical one. The main idea is to bundle time slots with certain hashing functions of device IDs, thus significantly reducing transmission overheads, including device IDs and contention overheads.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129903320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917625
Adwait Dongare, Craig Hesling, Khushboo Bhatia, Artur Balanuta, R. Pereira, Bob Iannucci, Anthony G. Rowe
Infrastructure monitoring applications currently lack a cost-effective and reliable solution for supporting the last communication hop for low-power devices. The use of cellular infrastructure requires contracts and complex radios that are often too power hungry and cost prohibitive for sensing applications that require just a few bits of data each day. New low-power, sub-GHz, long-range radios are an ideal technology to help fill this communication void by providing access points that are able to cover multiple kilometers of urban space with thousands of end-point devices. These new Low-Power Wide-Area Networking (LPWAN) platforms provide a cost-effective and highly deployable option that could piggyback off of existing public and private wireless networks (WiFi, Cellular, etc). In this paper, we present OpenChirp, a prototype end-to-end LPWAN architecture built using LoRa Wide-Area Network (LoRaWAN) with the goal of simplifying the design and deployment of Internet-of-Things (IoT) devices across wide areas like campuses and cities. We present a software architecture that exposes an application layer allowing users to register devices, describe transducer properties, transfer data and retrieve historical values. We define a service model on top of LoRaWAN that acts as a session layer to provide basic encoding and syntax to raw data streams. At the device-level, we introduce and benchmark an open-source hardware platform that uses Bluetooth Low-Energy (BLE) to help provision LoRa clients that can be extended with custom transducers. We evaluate the system in terms of end-node energy consumption, radio penetration into buildings as well as coverage provided by a network currently deployed at Carnegie Mellon University.
{"title":"OpenChirp: A Low-Power Wide-Area Networking architecture","authors":"Adwait Dongare, Craig Hesling, Khushboo Bhatia, Artur Balanuta, R. Pereira, Bob Iannucci, Anthony G. Rowe","doi":"10.1109/PERCOMW.2017.7917625","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917625","url":null,"abstract":"Infrastructure monitoring applications currently lack a cost-effective and reliable solution for supporting the last communication hop for low-power devices. The use of cellular infrastructure requires contracts and complex radios that are often too power hungry and cost prohibitive for sensing applications that require just a few bits of data each day. New low-power, sub-GHz, long-range radios are an ideal technology to help fill this communication void by providing access points that are able to cover multiple kilometers of urban space with thousands of end-point devices. These new Low-Power Wide-Area Networking (LPWAN) platforms provide a cost-effective and highly deployable option that could piggyback off of existing public and private wireless networks (WiFi, Cellular, etc). In this paper, we present OpenChirp, a prototype end-to-end LPWAN architecture built using LoRa Wide-Area Network (LoRaWAN) with the goal of simplifying the design and deployment of Internet-of-Things (IoT) devices across wide areas like campuses and cities. We present a software architecture that exposes an application layer allowing users to register devices, describe transducer properties, transfer data and retrieve historical values. We define a service model on top of LoRaWAN that acts as a session layer to provide basic encoding and syntax to raw data streams. At the device-level, we introduce and benchmark an open-source hardware platform that uses Bluetooth Low-Energy (BLE) to help provision LoRa clients that can be extended with custom transducers. We evaluate the system in terms of end-node energy consumption, radio penetration into buildings as well as coverage provided by a network currently deployed at Carnegie Mellon University.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131715853","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917583
Chelsea Dobbins, S. Fairclough
The experience of negative emotions in everyday life, such as anger and anxiety, can have adverse effects on long-term cardiovascular health. However, objective measurements provided by mobile technology can promote insight into this psychobiological process and promote self-awareness and adaptive coping. It is postulated that the creation of a mobile lifelogging platform can support this approach by continuously recording personal data via mobile/wearable devices and processing this information to measure physiological correlates of negative emotions. This paper describes the development of a mobile lifelogging system that measures anxiety and anger during real-life driving. A number of data streams have been incorporated in the platform, including cardiovascular data, speed of the vehicle and first-person photographs of the environment. In addition, thirteen participants completed five days of data collection during daily commuter journeys to test the system. The design of the system hardware and associated data streams are described in the current paper, along with the results of preliminary data analysis.
{"title":"A mobile lifelogging platform to measure anxiety and anger during real-life driving","authors":"Chelsea Dobbins, S. Fairclough","doi":"10.1109/PERCOMW.2017.7917583","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917583","url":null,"abstract":"The experience of negative emotions in everyday life, such as anger and anxiety, can have adverse effects on long-term cardiovascular health. However, objective measurements provided by mobile technology can promote insight into this psychobiological process and promote self-awareness and adaptive coping. It is postulated that the creation of a mobile lifelogging platform can support this approach by continuously recording personal data via mobile/wearable devices and processing this information to measure physiological correlates of negative emotions. This paper describes the development of a mobile lifelogging system that measures anxiety and anger during real-life driving. A number of data streams have been incorporated in the platform, including cardiovascular data, speed of the vehicle and first-person photographs of the environment. In addition, thirteen participants completed five days of data collection during daily commuter journeys to test the system. The design of the system hardware and associated data streams are described in the current paper, along with the results of preliminary data analysis.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"13 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129436456","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917628
V. Chamola, C. Tham, G. Chalapathi
With the various technological advances, mobile devices are not just being used as a means to make voice calls; but are being used to accomplish a variety of tasks. Mobile devices are being envisioned to practically accomplish any task which could be done on a computer. This is hurdled by the limited computational resources available with the mobile devices due to their portable size. With the mobile devices being connected to the Internet, leveraging cloud services is being seen as a promising solution to overcome this hurdle. Computationally intensive tasks can be offloaded to the Cloud servers. However, owing to the latency and cost associated with using cloud services, edge devices (termed cloudlets) stationed near the mobile devices are being seen as a prospective alternative to replace/assist the Cloud services. The mobile devices have an easier access to the cloudlets being situated in their vicinity and can offload their task requests to them to be served at a lower cost. This paper considers a network of such connected cloudlets which provide service to the mobile devices in a given area. We address the issue of task assignment in such a scenario (i.e. which cloudlet serves which mobile device) aimed towards improving the quality of service experienced by the mobile devices in terms of minimizing the latency. Through numerical simulations we demonstrate the performance gains of the proposed task assignment scheme showing lower latency as compared to the traditional scheme for task assignment.
{"title":"Latency aware mobile task assignment and load balancing for edge cloudlets","authors":"V. Chamola, C. Tham, G. Chalapathi","doi":"10.1109/PERCOMW.2017.7917628","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917628","url":null,"abstract":"With the various technological advances, mobile devices are not just being used as a means to make voice calls; but are being used to accomplish a variety of tasks. Mobile devices are being envisioned to practically accomplish any task which could be done on a computer. This is hurdled by the limited computational resources available with the mobile devices due to their portable size. With the mobile devices being connected to the Internet, leveraging cloud services is being seen as a promising solution to overcome this hurdle. Computationally intensive tasks can be offloaded to the Cloud servers. However, owing to the latency and cost associated with using cloud services, edge devices (termed cloudlets) stationed near the mobile devices are being seen as a prospective alternative to replace/assist the Cloud services. The mobile devices have an easier access to the cloudlets being situated in their vicinity and can offload their task requests to them to be served at a lower cost. This paper considers a network of such connected cloudlets which provide service to the mobile devices in a given area. We address the issue of task assignment in such a scenario (i.e. which cloudlet serves which mobile device) aimed towards improving the quality of service experienced by the mobile devices in terms of minimizing the latency. Through numerical simulations we demonstrate the performance gains of the proposed task assignment scheme showing lower latency as compared to the traditional scheme for task assignment.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128957221","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-03-13DOI: 10.1109/PERCOMW.2017.7917621
Md Abdullah Al Hafiz Khan, Nirmalya Roy
Activity recognition using smartphone has great potential in many applications like healthcare, obesity management, abnormal behavior detection, public safety and security etc. Typical activity detection systems are built on to recognize a limited set of activities that are present in the training and testing environments. However, these systems require similar data distributions, activity sets and sufficient labeled training data in both training and testing phases. Therefore, inferring new activities is challenging in practical scenarios where training and testing environments are volatile, data distributions are diverge and testing environment has new set of activities with limited training samples. The shortage of labeled training data samples also degrades the activity recognition performance. In this work, we address these challenges by augmenting the Instance based Transfer Boost algorithm with k-means clustering. We evaluated our TransAct model with three public datasets - HAR, MHealth and DailyAndSports and demonstrated that our TransAct model outperforms traditional activity recognition approaches. Our experimental results show that our TransAct model achieves ≈ 81% activity detection accuracy on average.
{"title":"TransAct: Transfer learning enabled activity recognition","authors":"Md Abdullah Al Hafiz Khan, Nirmalya Roy","doi":"10.1109/PERCOMW.2017.7917621","DOIUrl":"https://doi.org/10.1109/PERCOMW.2017.7917621","url":null,"abstract":"Activity recognition using smartphone has great potential in many applications like healthcare, obesity management, abnormal behavior detection, public safety and security etc. Typical activity detection systems are built on to recognize a limited set of activities that are present in the training and testing environments. However, these systems require similar data distributions, activity sets and sufficient labeled training data in both training and testing phases. Therefore, inferring new activities is challenging in practical scenarios where training and testing environments are volatile, data distributions are diverge and testing environment has new set of activities with limited training samples. The shortage of labeled training data samples also degrades the activity recognition performance. In this work, we address these challenges by augmenting the Instance based Transfer Boost algorithm with k-means clustering. We evaluated our TransAct model with three public datasets - HAR, MHealth and DailyAndSports and demonstrated that our TransAct model outperforms traditional activity recognition approaches. Our experimental results show that our TransAct model achieves ≈ 81% activity detection accuracy on average.","PeriodicalId":319638,"journal":{"name":"2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121062355","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：