Design, Characterization and Management of a Wireless Sensor Network for Smart Gas Monitoring
Vana Jeliˇci´cy, Michele Magno_, Giacomo Paci_, Davide Brunelliz and Luca Benini_
_ Department of Electronics, Computer Sciences and Systems, University of Bologna, Italy E-mail: fmichele.magno, giacomo.paci, luca.beninig@unibo.it
yFaculty of Electrical Engineering and Computing, University of Zagreb, Croatia E-mail: vana.jelicic@fer.hr
z Department of Computer and Information Science, University of Trento, Povo (TN), Italy E-mail: davide.brunelli@disi.unitn.it
Abstract
Air quality monitoring in indoor environments is of great significance for comfort and health, especially nowadays that people spend more than 80% of the day indoor. We propose flexible wireless system able to detect polluted air and dangerous situations in a complex and large environment. It is important for ambient intelligent systems to be unobtrusive and to optimize the power consumption of the platforms in order to be able to live on batteries for several years. We present a system with aggressive energy management that involves three levels: sensor level, node level and network level. The sensor board we designed is a wireless sensor network (WSN) node, with very low sleep current consumption (only 8 _A). It contains two modalities — a gas sensor and a Pyro electric Infrared (PIR)sensor. The network is multimodal: it uses information from the PIR sensor and neighbor nodes to detect the presence of people and to modulate the duty cycle of the node and the Metal Oxide Semiconductor (MOX) gas sensor. In this way we reduce the nodes’ activity and energy requirements, providing reliable service at the same time. We simulate the benefits of the context-aware adaptive duty-cycling of the gas sensor activity and we demonstrate a significant lifetime extension compared to the continuously driven gas sensor (several years vs. several days).
Keywords—Wireless sensor network, gas sensor, metal oxide semiconductor technology, energy management, people detection.
I. Introduction
Monitoring Indoor Air Quality (IAQ) is very important for people’s comfort, health and safety. Nowadays we spend most of our time indoor, thus the lack of ventilation causes the Sick Building Syndrome (SBS), with symptoms like headaches, nausea, dizziness, eye and throat irritation [1]. Earlier, ventilation has been controlled only by the information from the CO2 sensors. In the recent several years Volatile Organic Compounds (VOCs)have been proven to show better results about persons’ comfort(being able also to sense the smell in the room). Important sources of VOCs in a building are people (bio effluents), furniture, building materials, paints, etc. [2]. Besides comfort, detection of dangerous situations, like gas leakage (e.g. CH4 or CO) is very important. CH4 (methane) is a principal constituent of the natural gas, used in almost every household for cooking or heating. When it reaches a certain concentration in air (5–15%),it is flammable and explosive [3]. CO (carbon monoxide) sources are tobacco smoke, gas heaters and stoves, leaking chimneys, etc. It is colorless, odorless and tasteless, hence impossible to notice without a sensing device. In smaller quantities (e.g. 100 ppm) it causes a headache and dizziness after a couple of hours of exposure. Higher concentrations (e.g. 3200 ppm) cause headaches and dizziness after 5–10 min, and death within 30 min. Very high concentrations (e.g. 12800 ppm)
cause unconsciousness after a couple of breaths, followed by death in less than 3 min [4].
Implementation of Wireless Sensor Networks (WSNs) in IA monitoring avoids installation costs due to wire depositions, introducing at the same time power efficiency as a main challenge. In fact, wireless sensor nodes are mainly battery-powered. Even if they are equipped with power harvesting units [5],energy is a limited resource and should be managed wisely. A WSN should be autonomous and self-sustainable, able to function for several years with battery power supply. To fulfill this goal, power consumption should be minimized. In general, wireless transceivers consume a major amount of energy and many power management policies have been investigated and applied to reduce their activity. However, even sensors can be power-hungry, and gas sensors do consume significant power. Their power consumption is in the same order of magnitude of transceivers’ power consumption (typically 60–70 mow) and they should be active most of the time to sense the gas concentration and ensure a good quality of service. Thus, we need power management techniques that schedule both energy-consuming sensors and the transceiver [6].
There are several examples in literature of sensor systems for monitoring IAQ. In [7], an automated decentralized indoor climate control system is presented, including stationary wiredmulti-gas sensor modules and wearable wireless devices. Energy consumption of the system is not mentioned. Postulate teal. [8] present a Wi-Fi network for indoor and outdoor air quality monitoring with MOX sensor arrays from Figaro [9]. They are focused on advanced onboard processing and data publishing on the Web. Power consumption of the nodes is quite high(8 W). Choi et al. [10] present design and implementation of sensor board for air pollutant monitoring applications, based on an Mote, with IEEE 802.15.4/Sigsbee communication protocol. They developed an automated sensor-specific power management system and use pulse mode of the gas sensors, but the current consumption of their solution is still quite high (about 100 mA).
We focus on the power consumption reduction of the gas monitoring WSN through design of an ultra-low-power node. The power consumption of our node in sleep state is only24 _W. Information about presence of people and the messages received from the other nodes in the network enable contextawareadaptation of the gas sensor duty cycle. That features essential for achieving average power consumption of the network low enough to survive at least for one year on a battery power supply, without losing important information from the environment at the same time.
Most convenient gas detectors for WSN applications are those fabricated in Metal Oxide Semiconductor (MOX) technology, due to the small form factor and power efficiency. In [11], it is shown how the sensitivity, selectivity and response time of Moggs sensors strongly depend on the sensing layer temperature. Substantial study of the pulse mode for three different fabricated MOX gas sensors is presented in [12]. From these two papers it is important to emphasize power savings of an order of a magnitude compared with typical commercial off-the-shelf (COTS) Moggs sensors (the fabricated sensors consume only about 9 mow).In [13], authors study the dynamic behavior of low-power COMOX sensors (COTS sensors from Figaro) operated with pulsed temperature profiles and conclude that sensor thermal dynamics changes as a function of the CO concentration. They propose two parameters describing the sensor response shape, to provide an indication of the gas concentration, regardless of any calibration. These papers show the intensive effort to reduce the power consumption of the MOX gas sensors with pulse operational mode. Upon the state-of-the-art and expectations of further COTS gas sensors power consumption reduction, we decided to develop our system with MOX technology gas sensors. We base our simulations of node duty cycle on the research results of the gas sensor pulse mode.
From the literature we notice the lack of cooperation between the research field of MOX gas
sensors fabrication and the network cooperation research field. Our main contribution is to merge knowledge from both of these fields, which would enable more energy-efficient and flexible system design. In this paper we present a preliminary study of a flexible, context-aware Westford smart gas monitoring. The energy consumption reduction, that enables several years long battery lifetime, is performed on three levels:
· sensor level: pulse mode operation of the MOX gas sensor;
· node level: ultra-low sleep power consumption, dutycycling,activity modification based on people presence detection;
· network level: activity modification based on the messages received from the neighbor nodes.
To enhance the ambient intelligence of our system, we merge the information from gas sensors and Pyro electric Infrared sensors(PIRs) that detect people presence. From our experience [14, 15],PIR sensors, although very low-resolution, are able to give useful information when densely deployed and cooperating. Another advantage is that they have almost no impact on people’s privacy.
This paper is organized as follows. Section II gives an overview of the network we designed and Section III the characterization of the network in terms of energy consumption reduction. Section IV presents several case studies with simulations that show the benefits of the proposed system. The conclusions, together with future work plans, are brought in Section V.
II. Network architecture
The network we propose consists of several sensor nodes organized in an IEEE 802.15.4/Sigsbee network. For this preliminary work intended for gas monitoring in smaller apartments/buildings, the nodes are organized in a star configuration. One of the nodes is the Sigsbee coordinator, and the others are end devices. The coordinator is always on and it is mains powered, thus its energy consumption is not an issue. On the other hand, end devices are battery-powered and reducing their energy consumption is a crucial task for the longevity of the network. The hardware architecture of the nodes is the same, they differ only in the software application. In order to reduce the power consumption of the hardware as much as possible, we decided to design our own wireless sensor node, with the possibility of controlling activities of its components.
