Oscar Gama

Quality of service support in wireless sensor networks for emergency healthcare services

The reliable and efficient operation of emergency healthcare (e-emergency) services poses quality demands to the systems and underlying communication infrastructures. In this context, most existing wireless body sensor networks fall short in meeting these demands as they only offer an unreliable service delivery. e-Emergency systems must provide quality of service (QoS) support so that a pervasive and trustable assistance is provided to patients under health risk. This paper discusses the need for QoS in wireless e-health and e-emergency services. To demonstrate this need, some current and relevant e-health projects with QoS requirements are presented. The study reveals the importance of providing QoS support in this emerging field of application and provides a summary characterizing the e-health proposals herein presented.

Autonomy Suitability of Wireless Modules for Ambient Assisted Living Applications: WiFi, Zigbee, and Proprietary Devices

Ambient Assisted Living is pushing the development of innovative wireless monitoring solutions. Most of the available solutions are based on well known wireless communication standards, such as Wi-Fi, Bluetooth or ZigBee. Moreover, there are also a few proprietary solutions, including those designed specifically for biomedical monitoring applications. The success of those solutions depends on many factors, and power autonomy is one key issue, since system recharging requires the user to temporarily remove the device. In this paper, we evaluate the autonomy of three different wireless modules, proposed for use on Ambient Assisted Living applications. We present the measured and expected autonomy for modules built from off-the-shelf components to assess its suitability for real scenario applications. From the results, it is evident that may be worthless to design extremely low power acquisition electronics, sacrificing the acquisition performance, when the main power consumption comes from the wireless subsystem that, which is one order of magnitude higher.

Quality of Service in Wireless e-Emergency: Main Issues and a Case-Study

Due to its critical nature, emergency healthcare (e-emergency) systems should be totally reliable, efficient and support real-time traffic. Therefore e-emergency networks must provide proper quality of service (QoS) levels. After assessing the relevance of QoS deployment in different e-health contexts, this paper presents a pragmatic case-study intended to be deployed in a hospital room containing patients with high risk abnormalities, whose vital signals are being monitored by personal wireless body sensor networks. After justifying the unsuitability of ZigBee standard in this e-emergency scenario, the use of Low-Power, Real-Time (LPRT) protocol for wireless sensor networks, is proposed as an adequate candidate for such task. For the present case-study, the protocol is able to fulfill quantitatively the required QoS levels.

Trade-off Analysis of a MAC Protocol for Wireless e-Emergency Systems

Wireless sensor networks are envisioned to be deployed in healthcare. Since emergency and intensive care applications need to assure reliable and timely data delivery, they have increased demands for quality of service, including at the MAC layer. Amongst MAC protocols available for WSNs, the Low Power Real Time (LPRT) presents suitable characteristics to be deployed in emergency platforms due to its rational bandwidth allocation, low energy consumption, and bounded latency. Yet, this protocol may present a significant packet loss ratio in a wireless channel with bit error ratio. In order to define a MAC protocol more robust to bit error conditions and able to fulfill the required quality of service, solutions based on short size beacons and multiple retransmissions are proposed and tested. The results showed that such strategies led to meaningful improvements regarding packet loss ratio, without compromising significantly the energy consumption.

A platform to emulate ambient assisted living environments

According to the opinion of clinicians, emerging medical conditions can be timely detected by observing changes in the activities of daily living and/or in the physiological signals of a person. To accomplish such purpose, it is necessary to properly monitor both the persons physiological signals as well as the home environment with sensing technology. Wireless sensor networks (WSNs) are a promising technology for this support. After receiving the data from the sensor nodes, a computer processes the data and extracts information to detect any abnormality. The computer runs algorithms that should have been previously developed and tested in real homes or in living-labs. However, these installations (and volunteers) may not be easily available. In order to get around that difficulty, this paper suggests the making of a physical model to emulate basic actions of a user at home, thus giving autonomy to researchers wanting to test the performance of their algorithms. This paper also studies some data communication issues in mobile WSNs, namely how the orientation of the sensor nodes in the body affects the received signal strength, as well as retransmission aspects of a TDMA-based MAC protocol in the data recovery process.

Design of a MAC protocol for e-emergency WSNs

The wide adoption of WSNs in healthcare is still conditioned by quality of service (QoS) issues, namely at the MAC level. Medium access protocols currently available for WSNs are incapable of providing the required QoS to healthcare applications in scenarios of emergency or intensive medical care. To fill this lacuna, this paper introduces a MAC protocol presenting novel concepts to assure the QoS of e-emergency WSNs. Preliminary validation tests showed that the proposed protocol presents a good performance regarding data transmission robustness without sacrificing the power consumption efficiency.

A model to improve the accuracy of WSN simulations

Simulation studies have been extensively adopted in the networking research community. Nevertheless, the performance of the software components running within the network devices is often not modeled by generic network simulators. This aspect is particularly important in wireless sensor networks (WSN). As motes present very limited computing resources, the overhead of the software components cannot be ignored. Consequently, WSN simulation results may diverge significantly from the reality. After showing experimentally the validity of this assumption, the paper proposes a set of generic equations to model the performance of WSN software components. Validation tests using contention and multiplexing-based MAC protocols show that the inclusion of the proposed model in a WSN simulator improves the confidence degree in the simulation results significantly.

A time-slot scheduling algorithm for e-health wireless sensor networks

For e-health wireless sensor networks presenting significant traffic loads, MAC protocols based on deterministic scheduling algorithms are consensually considered more adequate than protocols based on random access algorithms. Indeed, TDMA-based MAC protocols are able to control the delay bound and save power by eliminating collisions. However, these protocols always require some expedite scheme to assign the superframe time-slots to the network devices that need to transmit data. Knowing that patients of an e-health wireless network are normally monitored by the same number and types of motes, originating a regular traffic pattern, a simple collaborative time-slot allocation algorithm can be achieved, as introduced in this paper. In the proposed algorithm, the announcement of time-slot allocation by the network coordinator is avoided, which helps to improve the packet delivery ratio and reduce the energy consumption in the e-health wireless network.

A protocol extension for selective reprogramming of WSNs

Wireless sensor networks (WSNs) are expected to operate for long time periods, often in places of difficult access. Thus, the ability to reprogram remotely and selectively sensor nodes becomes crucial in the maintenance and management of these networks. From the operating systems mostly used in WSNs, only TinyOS supports natively a reprogramming protocol (Deluge). The epidemic nature of Deluge, without supporting selective reprogramming, compromises the potentially vast application of WSNs. This paper proposes a new protocol as an extension of Deluge, allowing selective reprogramming of sensor nodes based on the platform type or node identifier. Aspects such as robustness and efficiency are taken into account in the protocol design and implementation, while assuring backward compatibility with Deluge. This new reprogramming protocol provides the required flexibility for deploying versatile WSNs, comprising heterogeneous nodes and applications.

A Different Approach in an AAL Ecosystem: A Mobile Assistant for the Caregiver

Currently the Ambient Assisted Living and the Ambient Intelligence areas are very prolific. There is a demand of security and comfort that should be ensured at people’s homes. The AAL4ALL (ambient assisted living for all) project aims to develop a unified ecosystem and a certification process, allowing the development of fully compatible devices and services. The UserAccess emerges from the AAL4ALL project, being a demonstration of its validity. The UserAccess architecture, implementation, interfaces and test scenario are presented, along with the sensor platform specially developed for the AAL4ALL project.