Rossitza Goleva

Improving Activity Recognition Accuracy in Ambient-Assisted Living Systems by Automated Feature Engineering

Ambient-assisted living (AAL) is promising to become a supplement of the current care models, providing enhanced living experience to people within context-aware homes and smart environments. Activity recognition based on sensory data in AAL systems is an important task because 1) it can be used for estimation of levels of physical activity, 2) it can lead to detecting changes of daily patterns that may indicate an emerging medical condition, or 3) it can be used for detection of accidents and emergencies. To be accepted, AAL systems must be affordable while providing reliable performance. These two factors hugely depend on optimizing the number of utilized sensors and extracting robust features from them. This paper proposes a generic feature engineering method for selecting robust features from a variety of sensors, which can be used for generating reliable classification models. From the originally recorded time series and some newly generated time series [i.e., magnitudes, first derivatives, delta series, and fast Fourier transformation (FFT)-based series], a variety of time and frequency domain features are extracted. Then, using two-phase feature selection, the number of generated features is greatly reduced. Finally, different classification models are trained and evaluated on an independent test set. The proposed method was evaluated on five publicly available data sets, and on all of them, it yielded better accuracy than when using hand-tailored features. The benefits of the proposed systematic feature engineering method are quickly discovering good feature sets for any given task than manually finding ones suitable for a particular task, selecting a small feature set that outperforms manually determined features in both execution time and accuracy, and identification of relevant sensor types and body locations automatically. Ultimately, the proposed method could reduce the cost of AAL systems by facilitating execution of algorithms on devices with limited resources and by using as few sensors as possible.

Reliable Platform for Enhanced Living Environment

The aim of this paper is to present an idea of the platform for enhanced living environment that will allow flexible and reliable use of cloud computing and sensor networks for highly customized services and applications. The architecture is based on sensors using IEEE 802.15.4 and zigbee protocols. Furthermore, the access to the cloud could be done by any available wired or wireless technology. We propose an application layer service using peer port for reliable and scalable data transmission. The presented solution is dynamic, flexible and conforms to the health and home automation standards.

Data and Traffic Models in 5G Network

This chapter presents data and traffic analyses in 5G networks. We setup experiments with Zigbee sensors and measure different traffic patterns by changing the environmental conditions and number of channels. Due to the differences in read, write operations, message fragmentations and backoff of the Carrier Sense Multiple Access/Collision Avoidance algorithm we demonstrated that the traffic flows are changing dynamically. This leads to different behaviour of the network domain and requires special attention to network design. Statistical analyses are performed using Easyfit tool. It allows to find best fitting probability density function of traffic flows, approximation toward selected distributions as Pareto and Gamma and random number generation with selected distribution. Our chapter concludes with future plan for distribution parameters mapping to different traffic patterns, network topologies, different protocols and experimental environment.

Traffic sources measurement and analysis in UMTS

In this paper, we show statistical analyses of several types of traffic sources in a UMTS network, namely, voice, video, and data sources. For each traffic source type measurements were collected in order to, on the one hand, gain better understanding of the statistical characteristics of the sources and, on the other hand, enable forecasting traffic behavior in the network. The latter can be used to estimate service times and Quality of Service. Based on evaluation of packet inter-arrival times, we propose the Polya distribution to be used in network simulations or in estimations of available capacity in opportunistic sessions. Our results can be applied for dynamic network configuration in order to avoid potential network congestions or overflows.

Automated Ambient Open Platform for Enhanced Living Environment

The aim of this paper is to present an idea of open platform for Enhanced Living Environment that will allow flexible and reliable use of the cloud-computing, sensor, mobile, local and body area networks for highly customized services and applications. The platform architecture consists of sensors working on IEEE 802.15.4, ZigBee, EnOcean standards, Ethernet, GSM, 3G, 4G, 5G gateways to the cloud, and peer port application at application layer. The personal enhanced living environment consists of body area and home automation network. They are key factors for patient’s comfortable living outside hospitals. Home, car, working place, park automation support the mobility, social integration, possibility of the patients to manage everyday tasks. Experiments carried out using sensors and holter prove the vitality of the solution partially. The platform is capable of working independently from the network connectivity, uses multi-homing, is open to migration to new technologies

Generic IP Network Traffic Management from Measurement through Analyses to Simulation

The aim of this chapter is to present different approaches to network traffic management applicable to the IP, transport and application layers in IP, 3G, WiMAX and 4G technologies. The proposed technology for analysis is flexible enough to different types of traffic in opportunistic networks. We start with traffic measurements and obtain accurate data for detail network simulations and precise analysis. Then, we highlight the self-similar nature of the incoming traffic at network nodes. In our next analysis, we look at mapping the measured data with the Polya arrival process by Pareto and gamma distributed inter-arrival times. Polya, Pareto and gamma distributions have the capability to change shape and scale in a way to simulate different types of observed traffic. A proper analytical description of the end-recipient traffic flows and point process of self-similarity inputs are applied for a better user behavior specification. During an end-to-end simulation, more complex queuing models with priorities are proposed. The behavior of the system at its bounds is shown. We map the data from measurements and simulations with the application layer requirements, cross-layer Quality of Service (QoS) and Quality of Experience (QoE) parameters. This is done by traffic fractality analyses, codec-dependent resource reallocation and Fibonacci backward difference traffic moments analyses. All of them demonstrate special moments in the breakdown of the shaping effect. Finally, we express views on openresearch issues for offering optimization in the Internet traffic analyses.

New constant service time Polya/D/n traffic model with peaked input stream

Abstract In this paper, we study the buffer queueing behaviour in high-speed networks. Some limited analytical derivations of queue models have been proposed in literature but their solutions are often a great mathematical challenge. We propose to use the Polya distribution to overcome such limitations. The specific behaviour of an IP interface with bursty traffic and long-range dependence is investigated by a version of the “classical” M/D/ n queueing model called Polya/D/ n . This is queueing system with a Polya input stream (a negative binomial distributed number of arrivals in a fixed time interval), a constant service time, multiple servers, and infinite waiting rooms. The model is considered a renewal process because of its quasi-random input stream and constant service time. We develop balance equations for the state of the system and obtain results for the packet loss and delay. The finding that the Polya distribution is adequate to model bursty input streams in IP network interfaces has motivated the proposal to evaluate the Polya/D/ n system. It is shown that the variance in the input stream significantly changes the characteristics of the waiting system. The suggested model is new and allows defining different bursty traffic and evaluating losses and delays relatively easily.

AALaaS intelligent backhauls for P2P communication in 5G mobile networks

Ambient Assisted Living as a Service (AALaaS) implementation is an emerging area of research and many different solutions could be found worldwide. In this paper, a decentralized foggy solution that well be context-aware and data-oriented is proposed. The new intelligent Peer-to-Peer framework supporting the Radio Access Network (RAN) management, which analyzes dynamically the traffic from the mobile device to different backhaul networks, and suggests most effective routing is fully decentralized and allows data residing. Many access networks will be based on zigbee or similar technology and will allow distributed data collection and dynamic network topologies. Simulation and laboratory experiments of zigbee access network demonstrates type of data to be stored and transmitted in the fog.

Traffic shaping measurements and analyses in 3G network

Traffic shaping effect may have significant impact on end-to-end Quality of Service (QoS) provisioning. Therefore, it should be carefully studied in order to allow the creation of appropriate traffic models to be used for simulations. First, to demonstrate the traffic shaping effect, we present statistical analyses on real-time measurements of diverse traffic sources (voice and video over IP) in a 3G network. By comparing the statistical distributions of the packet inter-arrival times for both the forward and backward direction, we can demonstrate directly the end-to-end traffic shaping effect introduced by the IP core network. Hence, we argue that distributed QoS management approach is needed. Additionally, we give the mean, variance, mean standard deviation, skewness, and kurtosis of the inter-arrival times, which can be used as input for simulation models. The accurate validation of the probability distributions is ensured by the Wolfram Mathematica and Crystal Ball statistical tools. Second, for the same set of measurements, we propose and defend with evaluations the use of the gamma distribution as best fitting function to traffic dynamics. Our proposal is applicable for traffic environments found in delay-tolerant networks, opportunistic networks, Internet of Things, sensor networks etc.

On the comparison of location based software solutions used for tracking purposes in ambient assisted living applications

Wearable and handheld devices, such smart phones, have come a long way in a very short time as they are capable of providing an extremely vast number of services simply at the click of a button, or even automatically. The diversity of the different embedded sensors (accelerometer, light sensors, etc.) lay the required foundation for the emergence of specialized applications that are capable of serving all kinds of purposes. By taking advantage of this specialized hardware found in smart phones, not only it has become feasible to constantly sense and gather information from the devices environment, but it has also become an effortless task to transmit the collected data to any individual or organization in a plethora of forms. This paper serves as an evaluation method for a location based Android application, SeniorTracker, by presenting a comparison between two applications that are both used for tracking down a device. While both have quite a similar purpose, the way they go about it as well as the ways the application itself can be utilized, drastically differs. SeniorTrackers main purpose lies in the retrieval of the coordinates of a device in case it is located somewhere it is not expected to be, allowing monitoring of persons that potentially will be lost, otherwise. The comparison was mainly focused on the performance of both applications when it comes to location retrieval and energy consumption while keeping in mind the respective tasks that each of them were designed to fulfill.