1

Wireless Sensor Network for Monitoring Temperature

Gloria F. Serrano Moya, José L. Zamorano Flores

Universidad Autónoma Metropolitana Unidad Azcapotzalco

gfsm@correo.azc.uam.mx Abstract We developed, implemented and tested a wireless sensor network based on technology XBee-ZNet 2.5 by Digi™, operating in transparent mode, aimed at measuring temperature on four locations in our work place, each node is equipped with one temperature sensor, one microcontroller and one XBee-ZNet 2.5 embedded module, the software stack is similar to the ZigBee 2006 standard, we use a precision centigrade temperature sensor of 0.5°C accuracy, the microcontroller carries out the data processing and transmits data via serial communications to each XBee module, the sensor data are collected in a coordinating node and forwarded via USB port to PC which displays data on X-CTU window. The nodes, four, are located at different points of our work place, transmit temperature data every minute, our system is rated for 0 to +50 C range. Our principal objective is to determine the temperature gradient. This parameter has several applications in atmospheric sciences, other places where noticeable temperature gradients can be experienced include the entrance (or exits) of air conditioned shops in the summer, or the entrance of caves and other protected or poorly ventilated areas. Rapid changes in temperature may cause discomfort and, in extreme cases, heat or cold stresses. These events are reported to the base station, which can take the appropriate action. The set of system design requirements are established according the specific application. The determination of temperature gradient is a work to be done, at the moment we have the temperature data sensed at four points of our work place obtaining satisfactory results. The results obtained are acceptable in a small area with the nodes separated a few meters. I Introduction Wired technologies have several drawbacks mainly the costs involved in wiring and his little flexibility when the network nodes arise. If wiring is damaged then the lines of communication are literally cut. On the other hand wireless technologies, like wireless sensors networks, WSN, has several advantages over traditional approach: The WSN nodes are easy to deploy and use, individual node is relatively cheap, allowing more data to be collected per unit cost; the nodes provide abundant measured data for processing to improve accuracy. Standardized, freely available and modular software components decrease cost of developing, modifying and maintaining the system; the deployed system allows near real-time response to events. This technology has demonstrated to be successful on various health monitoring machinery. The first application of sensor networks in field was on Great Duck Island [1] on habitat monitoring, to monitor the behavior of storm petrel. It was proposed too for wildfire monitoring, in [2], [3] and [4] it proposed a design for fire monitoring incorporating wireless sensors, and report results from field. However, few works was done to design a WSN system to monitor changes in temperature in indoors that may cause discomfort and, in extreme cases, heat or cold stresses. A WSN consists of a large number of wireless low power nodes with the ability to connect a large number of them into a single network, and able to take environmental measurements such as temperature, light, sound, humidity. The technology XBee-ZNet 2.5 uses the globally available, license-free 2.4GHz frequency band. It enables wireless applications using a standardized set of high level communication protocols based on the IEEE 802.15.4 standard for wireless personal area networks WPAN. The standard defines three kinds of devices that incorporate radios, with all

2

three found in a typical WSN network: a coordinator, which organizes the network and maintains routing tables; routers, which can talk to the coordinator, to other routers, and to reduced function end devices; and reduced function end devices, which can talk to routers and the coordinator, but not to each other. Each one has a profile, which defines the module application; some manufacturers have created private profiles that are integrated into their stacks and typically general-purpose serial UARTs. The sensors measurements are transmitted via wireless to a running application on a PC that makes decisions based on these sensor readings, and whose characteristics are determined by the particular application. One widely recognized issue is the limited power available to each wireless sensor node, but other challenges such as limited storage or processing capabilities play a significant role in constraining the application development. They operate on nonrenewable power sources and employ a short-range transceiver to send and receive messages. Sensor nodes are generally densely deployed in the area of interest. This dense deployment can be leveraged by the application. The most important technologies to implement sensor networks are classified into sensors, communications, processors, interfaces, and security. Many applications have been developed, since 2004, in consumer electronic device control, energy management and efficiency, health care, military, home and commercial building automation as well as industrial plant management. This case deals with monitoring the temperature of our work place environment, the nodes transmits the temperature sensor once a minute. In the future we expect to obtain the temperature gradient and a running application will make decisions based on these sensor readings when certain event occurs. II Methodology

A. The WSN Deploying the wireless sensor nodes can be done randomly or at a chosen spots in a given area. This is a continuous process, and usually is done to increase the coverage area and to acquire the physical parameters accurately. To build the WSN we use the type infrastructure based network, where the nodes communicate simultaneously with a single base station (coordinating node) which is connected to a PC via USB port, using a star topology; that is because our XBee radios have an original firmware that configure them as end terminals and one of them, with firmware as coordinating node, besides our future work includes a PC running application to generate results and graphs of temperature gradient and through PC we gain access to an infrastructure LAN and INTERNET, in the future, extend our application beyond the UAM. In figure 1 is shown the implemented system. The coordinating node is configured to operate in broadcasting mode and the End devices send an ID with the collected data, as we can see on figure 1 they are identified as: R1, R2, R3, R4; and configured in low power mode, we need only a few meters coverage, and from these modules we obtain the supply for the sensor and microcontroller board, henceforth sensor board.

3

Fig. 1. The system consists of four end devices and a coordinating node as base station.

The temperature monitoring WSN is based on XBee-ZNet 2.5[4] nodes by Digicom™ and a sensor board, each End device node send readings, from their sensor to the base station (coordinating node), periodically, every minute, the base station is connected to PC via USB port. In fact we have readings in real time.

B. Wireless sensor network hardware design The nodes of WSN are composed of the XBee-ZNet 2.5 platform mounted on interface board RS232 with an independent sensor board. This separation allows hardware and software development of WSN to proceed independently. A node block diagram is shown in figure 2. The sensor board, as shown in Fig.3, is a separate component from XBee-ZNet 2.5 and connects to the platform through a 10 pin connector. Fig. 2. Node Block Diagram. Fig. 3. The sensor board, as a separate component from XBee-ZNet 2.5. Connection through a 10pin connector.

4

The sensor board, consists of: a precision integrated-circuit temperature sensor, the CI LM35[6]. The output voltage is linearly proportional to the temperature, in Celsius Degrees. Its principal characteristics are:

Output voltage CatV 1505.1

Scale factor linear CmV 0.10

Current drain Less than A60

The other component of the sensor is the microcontroller ATMega8L [7] PDIP package, which has attractive characteristics like: Operating Voltages

V5.57.2

Power Consumption at CVMhz 25,3,4

Active mA6.3 Idle Mode mA1

Peripheral Features

6-channel ADC Four Channels 10-bit Accuracy Two Channels 8-bit Accuracy

Programmable Serial USART

The total node power supply comes from the Interface board XBee RS-232 of XBee-ZNet 2.5

which supply voltage is V3.3

The maximum current is wasted at transmission )3.3(295 VamA the rest of time we have to

minimize energy waste in order to save energy. The readings do not show decimals so we do not need more than 8 bit accuracy The pins of XBee-ZNet 2.5 RF module can connect directly to devices that have a UART

interface. The USART of ATMega8L is fully compatible with the UART. Microcontroller tasks:

1. Take sample of sensor output voltage (analog), every minute. 2. Converts sample to digital binary. 3. Converts binary to BCD. 4. Obtain ASCII code and compose information to be displayed at X-CTU Terminal. 5. Transmits information, in ASCII code, of collected data via USART to RF module.

We determine ADC VREF according equation 1:

5

REF

in

V

VADC

256 (1)

Where:

VADC

VV inREF 56.2

256

With: resolutionADC256

VVin 5.1

CADCfromvaluebinaryADC 150 (We use C150 with V50.1 sensor output

voltage since REFV minimum of ADC is V2 )

In figure 4 is shown the electrical diagram of sensor board. In order not to have problems with message lost because of poor radio connectivity, the power for the RF transceiver, micro-controller and the current sensing circuitry are obtained from the AC mains.

Fig. 4. Electrical Diagram of Sensor Board.

6

III Results Since the coverage area is about a few meters we do not have problems of packet losses, so the information transmitted by the nodes was displayed on X-CTU terminal every minute without errors. Figure 5 shows two locations of the nodes in our Communications Laboratory. In figure 6 we can see the base station (a) and the received data displayed on X-CTU terminal (b) This is our first deployment of these nodes platform; it was very interesting to see how the wireless sensor network performed. Although, at the moment, the readings collected by the wireless sensor network are unusable, yet, to make issues to take scientific conclusions, the fidelity of the acquired sensor readings gave insight into overall network behavior. Fig. 5. Two nodes locations in Communications Laboratory of UAM-Azcapotzalco (a) (b) Fig. 6. (a) Base Station coordinating point connected to PC via USB port. (b) Received Data displayed on X-CTU terminal

7

IV Discussion The use of a design-in module, like XBee-ZNet 2.5, will generally provide the shortest time to develop applications. It could represent the most economical choice of all, especially in research areas where quantities are modest or uncertain. We think that if you can locate a module with a profile that meets your needs use it, in this case, we avoid not only much of the engineering costs, but the investment in test equipment and test fees as well. This work is about to be completed by means of intelligent user interfaces, decision assistance, and alarm functions. There is software ad-hoc to take out this task among the most popular; National Instruments LabVIEW software unfortunate it is very expensive. That is why we are looking for alternatives like Visual Basic V. Conclusion We designed and implemented a real-time temperature monitoring system using wireless sensor networks. We obtained satisfactory results, the best part; we think, is the fidelity of the acquired sensor readings observed in the data displayed at the terminal X-CTU, giving insight into overall network behavior. It needs some improvements in order to apply our system to more different fields and especially, when sensor networks need many sensor nodes to cover a larger area. References [1] J.Anderson, “Wireless sensor networks for habitat monitoring,” in Proceedings of the 1st ACM international workshop on wireless sensor networks and applications, ACM Press, New York,2002, pp.88-97. [2] Z. Chaczko, and F. Ahmad, “Wireless sensor network based system for fire endangered area,” in Proceedings of the 3rd international conference on information technology and applications, 2005, Wuhan, China, Vol. 2, pp. 203 -207. [3] D. M. Doolin and N. Sitar. "Wireless sensor for wild fire monitoring Proceedings of SPIE Symposium on Smart Structures & Materials" NDE 2005 San Diego, California, 2005.pp.6-10 [4] Thierry Antoine-Santoni , "Using Wireless Sensor Network for Wildfire detection. A discrete event approach of environmental monitoring tool, ” Environment Identities and Mediterranean Area, ISEIMA '06. First international Symposium on , 2006.pp.115-120 [5] http://ftp1.digi.com/support/documentation/90000866_C.pdf , XBee™ ZNet 2.5/XBee-PRO™ ZNet 2.5 OEM RF Modules [6] www.national.com/ds/LM/LM35.pdf LM35 technical specifications [7] www.atmel.com/atmel/acrobat/doc2486.pdf ATMega8L