Stefan Forsström

Distributed context support for ubiquitous mobile awareness services

Context-aware applications and services require ubiquitous access to context information about the users or sensors such as preferences, spatial & environmental data, available connectivity, and device capabilities. Systems for the brokering or the provisioning of context data via wireless networks do so with centralized servers or by employing protocols that do not scale well with real-time distribution capabilities. In other cases, such as the extending of presence systems, the data models are limited in expressive capabilities and consequently incur unnecessary signaling overhead. This paper presents a distributed protocol, the Distributed Context eXchange Protocol (DCXP), and an architecture for the real-time distribution of context information to ubiquitous mobile services: We present the architecture and its principle operation in a sample ubiquitous mobile awareness service. Preliminary results indicate that our approach scales well for the ubiquitous provision of context data in real-time to clients on the Internet via 3G wireless systems. Performed measurements show that DCXP can reduce the time to process context data with a factor of 20 compared to similar approaches.

The MediaSense Framework

Mobile telecommunication is evolving rapidly. People no longer only communicate with each other regardless of time and place, but also share other information that is important for tasks with which they are involved. In response to this growing trend the MediaSense framework addresses the intelligent delivery of any information to any host, anywhere, based on context-aware information regarding personal preferences, presence information, and sensor values. This includes another challenge to achieve seamless delivery, especially of multimedia content, and multimodal services via heterogeneous connections.

Location-Based Ubiquitous Context Exchange in Mobile Environments

Context-aware applications and services require ubiquitous access to context information of users. The limited scalability of centralized servers used in the provisioning of context information mandates the search for scalable peer-to-peer protocols. Furthermore, unnecessary signaling must be avoided in large-scale context networks, when location-based services only require nodes in a certain area with which to communicate context. To this end, we propose a lightweight model for composing and maintaining unstructured location-scoped networks of peer-to-peer nodes, which gossips in order to ensure quality of service for each user. The model is implemented in a prototype application running in a mobile environment, which is evaluated with respect to real-time properties. This model can also be extended to include more context dimensions, other than location.

Home care with NFC sensors and a smart phone

The worlds aging population and limited hospital resources together with advancements in information technology provides a thorough basis for further development of tele-healthcare and assisted living. For ordinary people, and especially elderly people, to be diagnosed and treated in their homes there is a definite need for technical equipment that is extremely easy to handle. One way to facilitate tele-healthcare is to take advantage of modern cell phone technology and the infrastructure surrounding these units. Many elderly people are today comfortable with using cell phones and even advanced smart phones can be purchased to a cost that is low compared to corresponding specialized medical equipment. The concept presented in this work is based upon Near Field Communication (NFC), a wireless feature that is currently being widely launched in cell phones. Amongst other capabilities, NFC technology allows a phone to act as a 13.56 MHz RFID reader. This work shows how NFC-equipped phones can be used to read passive NFC tags that in turn are equipped with general resistive sensors and how sensor data are read and communicated to a web server without the need to push a single button.

Enabling ubiquitous sensor-assisted applications on the internet-of-things

The increase in sensor-assisted applications such as sensing campaigns and interactions in large-scale populations put new demands on seamless sharing of sensor information on a global scale. Existing support is limited due to simplified information models reflecting narrow application scenarios and poorly scaling architectures. In response to this, we present an application architecture for pervasive internet-of-things applications which circumvents these limitations. Our architecture enables applications to utilize information from sensors and wireless sensor networks via a peer-to-peer overlay. The overlay shares sensor information in an generic information model, which is extensible and thus enables intelligent application behavior. Our evaluation of the architecture indicates that it supports the sharing of sensor information on a global scale in sensor-assisted applications, with low response times. Finally, we evaluate the feasibility of running such applications in end devices with limited resources in a sensor-assisted application prototype.

Real-Time Distributed Sensor-Assisted mHealth Applications on the Internet-of-Things

Existing sensor-assisted mHealth applications would benefit from large-scale sharing of sensor information in real-time. Existing communication solutions are however limited in this respect, because of centralized application-level communication. In response to this, we presents a distributed communication solution for mHealth applications which circumvents these limitations. Our Internet-of-Things architecture enables mHealth applications to utilize information from sensors and wireless sensor networks via a peer-to-peer overlay, where sensor information is organized in an information model which is stored in the overlay itself. We present a proof-of-concept application and evaluation results regarding the architectures real-time capabilities. The results indicate that a fully distributed architecture can support real-time sensing in mHealth applications and the support is available as an open source platform, MediaSense. Current work is focused on evaluating scalability in very large scale scenarios using field trials.

The Updated MediaSense Framework

Mobile telecommunication is evolving rapidly. People no longer only communicate with each other regardless of time and place, but also share other information that is important for the tasks with which they are involved. In response to this growing trend, the Media Sense framework addresses the intelligent delivery of any information to any host, anywhere, based on context-aware information regarding personal preferences, presence information, and sensor values. The focus of this paper is the overall framework, the handling of user profiles, management of context information, real-time distribution and proof of concepts.

Combining Fog Computing with Sensor Mote Machine Learning for Industrial IoT

Digitalization is a global trend becoming ever more important to our connected and sustainable society. This trend also affects industry where the Industrial Internet of Things is an important part, and there is a need to conserve spectrum as well as energy when communicating data to a fog or cloud back-end system. In this paper we investigate the benefits of fog computing by proposing a novel distributed learning model on the sensor device and simulating the data stream in the fog, instead of transmitting all raw sensor values to the cloud back-end. To save energy and to communicate as few packets as possible, the updated parameters of the learned model at the sensor device are communicated in longer time intervals to a fog computing system. The proposed framework is implemented and tested in a real world testbed in order to make quantitative measurements and evaluate the system. Our results show that the proposed model can achieve a 98% decrease in the number of packets sent over the wireless link, and the fog node can still simulate the data stream with an acceptable accuracy of 97%. We also observe an end-to-end delay of 180 ms in our proposed three-layer framework. Hence, the framework shows that a combination of fog and cloud computing with a distributed data modeling at the sensor device for wireless sensor networks can be beneficial for Industrial Internet of Things applications.

Evaluating Ubiquitous Sensor Information Sharing on the Internet-of-Things

Next generation context-aware mobile applications will require a continuous update of relevant information about a users surroundings, in order to create low latency notifications and high quality of experience. Existing mobile devices already contain a large number of built in sensors which are capable of producing huge amounts of sensor data, exceeding both the capacity of the local storage and the Internet connection. Therefore, we will in this paper study the limits when sharing contextual information from mobile devices, as well as finding the impact of this information overload for the Internet-of-Things. Furthermore, we present an evaluation model for assessing the effort required to present applications with relevant context information. In conclusion, the model shows that one feasible solution for the future Internet-of-Things is a peer-to-peer based solution which can control the flow of information without any centralized authority, to circumvent earlier limitations.

Scenarios, Research Issues, and Architecture for Ubiquitous Sensing

This paper describes research issues and work-in-progress concerning ubiquitous sensing. We present scenarios where the current approaches are deficient in addressing the needs for ubiquitous sensing in services and applications on the Future Internet, involving the massive sharing of information from sensors via heterogeneous networks. We propose an information-centric architecture for real-time ubiquitous sensing which capitalizes on the proposed locator/identifier split, thus extending the Network of Information (NetInf) approach. From this we identify the challenges for which we present work-in-progress within the framework of the EU-funded MediaSense project. Firstly, we integrate sensors as addressable objects, exposed by means of sensor gateways and relocatable abstract interfaces. Sensor information is thus made available to applications solely based on identity. Secondly, sensor information is made available in a distributed data model towards searching and browsing. Finally, we evaluate the effectiveness of the architecture in proof-of-concept applications for intelligent commuting, environmental monitoring and seamless media transfer, utilizing two different sensor platforms.