DATA INTEGRATION IN BIOLOGICAL MODELS

P. Mariño, F. P. Fontán, M. A. Domínguez and S. Otero

Electronic Technology Department, University of Vigo, Spain

Keywords:

Database applications, wireless networks, information systems.

Abstract:

Biological research in agriculture needs a lot of specialized electronic sensors in order to fulﬁll different goals,

like as: climate monitoring, soil and fruit assessment, crop tracking, and so on. That research must be sup-

ported by consistent biological models able to simulate diverse environmental conditions, in order to predict

the right human actions before risky biological damage could be irreversible. In this paper is described an

experimental distributed network based on climatic an biological wireless sensors, for providing real measure-

ments in order to validate different biological models used for viticulture applications.

1 INTRODUCTION

The experimental wireless network is deployed in a

peninsula surrounded by two large sea arms called

“rias” in Spanish language.

Currently differences in productivity and quality

of grapes are broadly related with relative heights

and sea proximity from each of four zones but nev-

ertheless more rigorous biological and climatic re-

search (Gail, 2007) must be done, in order to pro-

vide accurate biological models for ecological sim-

ulations applied to viticulture. For that reason multi-

disciplinary work must be done among electronic en-

gineers, biologists and ecologists.

Each zone has an electronic zonal station (EZS),

in order to bring differences (microclimates), in mea-

surements like: temperature, relative humidity, leave

humidity, soil temperature, solar radiation, rain gauge

(tipping bucket), and other biological sensors. A data

logger and a radio modem is included in each EZS

in order to sense, process and transmit the data, en-

abling the development of an automatic wireless sen-

sor network (WSN), which nodes (the EZSs) are ac-

cessible from a wide area. These wireless communi-

cation capabilities allow that data could be remotely

monitored. The implementation of a warehousing ap-

proach, allows the data to be stored in a centralized

database that is responsible for query processing. The

stored data will be used for biological and ecological

models.

Firstly the paper describes the different elements

employed in the experimental network. These include

(a) the wireless nodes (b) the base station (c) the re-

peaters and (d) the data management. Finally some

measurements from EZSs are depicted.

2 DATA ACQUISITION SYSTEM

The electronic zonal stations (EZSs) are connected

with the base station (BS) by the UHF band (not li-

censed) between 869.4MHz to 869.65MHz, and the

BS is also connected through Internet to the Data Base

(DB), the biological and ecological models (BEMs),

and to the Web access (Fig. 1). Each zonal station

comprises an UHF radio modem that transmits the

sensors information to the BS through a data call. A

powering solar panel (PSP) is located near each EZS

for feeding its circuits. In order to reduce costs, the

BS makes a call to all the EZSs every 24 hours by

means of a polling procedure (Mariño et al., 2006).

During these calls the EZSs send all the informa-

tion that has been stored on that period. Therefore

the BS periodically executes the reading data process

and later database storage of the received information,

through an Ethernet local area network.

2.1 Data from Sensors

An electronic zonal station (EZS) is the basic acqui-

sition equipment of the distributed system, that car-

ries out the data registration (measurements and pro-

cessing), and the communication with the base station

(BS). In this way, each EZS comprises an automatic

measurement unit with data transfer capability. The

389

Mariño P., P. Fontán F., A. Domínguez M. and Otero S. (2008).

DATA INTEGRATION IN BIOLOGICAL MODELS.

In Proceedings of the Tenth International Conference on Enterprise Information Systems - DISI, pages 389-392

DOI: 10.5220/0001667603890392

 SciTePress

Figure 1: Network architecture and database interfaces.

data acquisition process is made inside the EZS by

the sensors and the data logger. Each EZS comprises

the following sensors: temperature, relative humidity,

leave humidity, soil temperature, solar radiation, rain

gauge (tipping bucket), and other biological and eco-

logical features depending on running models (Poza

et al., 2006).

All that sensors are integrated in the data logger.

The data logger is the EZS nucleus, it captures the

data from each sensor, automates the measurements,

synchronises the data and manages the communica-

tions. The data transmission is carried out by means

of the data logger and the UHF radio modem con-

nected to it. Next, the data captured by the EZS is sent

to the database (DB), through the base station (BS),

where they are saved. The communication process

setting, through the UHF radio modem connected to

the data logger, allows the control and programming

of several tasks as well as the acquisition of stored

data.

The data captured by the data logger are organised

in registers. The registers comprise the sensor out-

puts as well as the time and date. These registers are

then sent to the storage system where they are saved

for a future access. The data logger is programmed

for capturing and storing the sensors information each

minute. Due to the limited capacity of the storage sys-

tem integrated in the data logger, the data can only be

stored during a day (24 hours). Figure 2 illustrates

the data logger, the storage system, the UHF radio

modem and connections with the sensors and elec-

trical supply. Al these elements are placed inside a

box which protects them from the weather conditions.

This box and all the sensors are ﬁxed to a metallic

base located at the site.

Figure 2: EZS data acquisition and communications sys-

tem.

3 GLOBAL DATA MANAGEMENT

The information obtained from the EZSs are collected

by the BS and stored in the DB for later process, anal-

ysis and query. The BS requests and compiles the

data from the different EZSs to store them in the DB.

Also the BS is provided with an UHF radio modem

to make the polling query of each EZS in the wire-

less network. Therefore the BS is a PC connected to

a wireless network and Internet that executes the de-

veloped program to perform its operations ﬂowchart.

Since all the measured data must have the same

time reference for its later process, the BS obtains

the system reference clock from a real time network

server by the NTP synchronization protocol (Network

Time Protocol).

The data from the EZSs are centralized in a rela-

tional database. This DB presents one interface with

the BS through which all the system information is

introduced, and three interfaces to access this infor-

mation: general data access, access to interesting data

to analyse viticulture features, and query of data for

providing models. The interface between BS-DB and

ICEIS 2008 - International Conference on Enterprise Information Systems

390

queries-DB are executed directly by means of ODBC

(Open Database Connectivity).

The general data access will directly take place

through an Internet accessible Web page. Whereas

for queries related to the analysis of viticulture fea-

tures and models, the access is made through speciﬁc

views for each type of study (Mariño et al., 2006).

Figure 3 shows an example of the EZS data manage-

ment. This picture illustrates the structure of commu-

nications among equipments, interfaces and layers.

4 THE STATE-OF-THE-ART AND

THE IMPLEMENTED

NETWORK

Past decade has been very fruitful in the development

and application of several standards for mobile, no-

madic and ﬁxed wireless networks related with sen-

sors (Demirkol et al., 2006). Some speciﬁc problems

about this kind of networks have been well studied,

like: energy efﬁciency due to collisions, overemitting-

receiving, control of packets and idle listening; scal-

ability and changes adaptation in network size, node

density and topology; communication paradigms like

node-centric, data-centric and position-centric; and

many others.

Nevertheless this great researching effort over

wireless networks for sensors, there is no any ac-

cepted MAC for them, because this kind of sensor

networks has a very big dependence of the applica-

tion. Recent surveys about the most advanced wire-

less networks like MANETs (Conti and Giordano,

2007) show poor real results in front of expected

ones, because the great complexity involved in simu-

lated MAC protocols, on big programming tools, was

not after validated with implementation, integration

and experimentation over real equipment (chips, mi-

crocircuits, modems, antennas, and others). In this

way, a particular ﬁeld of application, called “wire-

less sensor networks” (WSN) is proposed for envi-

ronmental monitoring, industry and precision agricul-

ture, among other sectors of activity. The WSNs are

featured by a stronger interdisciplinary collaboration

for creative projects, and a change in the communica-

tion paradigm from node-centric to data-centric one,

because the main point is the transfer of data from the

application ﬁeld, and not the communication between

all the network nodes.

4.1 The Wireless Sensor Network

(WSN)

Several comparatives among general wireless stan-

dards like ZigBee, Ultra Wide Band (UWB) and WiFi

have been made in order to evaluate some examples of

application included industrial wireless sensors. Also,

more speciﬁc WSN applications could be found about

environmental research like: hydrology (Moore et al.,

2000), ﬁre monitoring (Ruiz et al., 2005), and others.

Given the hilly nature of the vineyard zones (Sec-

tion 1), the coverage challenges for linking the EZSs

with the BS (Section 3) were founded in power, data

speed and acceptable error ratio. For example, in

the Meis zone the coverage area was over 5km, with

difference in heights about 200m, very prone to in-

terpose obstacles in the line-of-sight (LOS) among

EZS-BS. In order to achieve a wireless network with

very low cost and reduced power consumption, be-

cause static nodes are transmitting infrequently (low

duty cycle) only two-way small data packets, the Eu-

ropean ISM band (868-870MHz) was selected, where

one channel with a data rate of 20kbps is available.

In this ISM band the used radio modems for linking

EZS-BS, have the following features: 10-500mW of

transmitting power, 25kHz of channel spacing, half-

duplex communication, 10% duty cycle and 36 sec-

onds of maximum emission time (must be controlled

by the data logger, Section 2.1). To avoid an obstacle

in the LOS between the BS and the EZS, a repeater

station (RS) is inserted with other ISM radio modem

and a directive antenna, linking the EZS (2,1km) with

the BS (5,17km).

5 FUTURE DEVELOPMENTS

Experimental work over the implemented WSN is be-

ing made in the following lines:

• Addition of new climatic and biological sensors to

the EZSs.

• Deployment of new EZS over the four different

vineyard zones for providing more spatial resolu-

tion to biological and ecological models.

• Design of a wireless broadband (20Mbps) net-

work in order to provide crop tracking by real time

images, and infrared cameras (zonal isotherm

maps), by WiMax (IEEE 802.16) equipment over

the 5GHz ISM band (Fig. 1).

• Integrate those images in the global data manage-

ment system (Section 3), for giving to the biologi-

cal and ecological researchers new knowledge for

future enhancement of models.

DATA INTEGRATION IN BIOLOGICAL MODELS

391

Figure 3: Communication example between elements, interfaces and layers.

6 CONCLUSIONS

The authors have developed an experimental dis-

tributed network based on the WSN paradigm for

wireless sensors. This WSN is based on the European

ISM band for providing a low cost and low power

consumption network, bringing real measurements to

validate different biological and ecological models

used for viticulture applications. Also a global data

management system is designed to integrate consis-

tently the measured data in the models. New develop-

ments in the experimental wireless network are being

tested to add real time images and infrared cameras

information, by means of broadband network stan-

dards.

REFERENCES

Conti, M. and Giordano, S. (2007). Multihop ad hoc net-

working: The theory. In Communications Maga-

zine, IEEE, volume 45, pages 78–86, Toronto, Ont.,

Canada.

Demirkol, I., Ersoy, C., and Alagoz, F. (2006). MAC pro-

tocols for wireless sensor networks: a survey. IEEE

Communications Magazine, 44(4):115–121.

Gail, W. B. (2007). Climate control. IEEE Spectrum,

44(5):20–25.

Mariño, P., Fontán, F. P., Machado, F., and Otero, S. (2006).

Distributed sensors network applied to the rain im-

pairment study on radiocommunication systems. In

Industrial Informatics, 2006 IEEE International Con-

ference on, pages 1036–1041, Singapore.

Moore, R. J., Jones, D. A., Cox, D. R., and Isham, V. S.

(2000). Design of the hyrex raingauge network. Hy-

drology and Earth System Sciences, 4:521–530.

Poza, F., Mariño, P., Otero, S., and Machado, F. (2006).

Programmable electronic instrument for condition

monitoring of in-service power transformers. IEEE

Transactions on Instrumentation and Measurement,

55(2):625–634.

Ruiz, L. B., Braga, T. R. M., Silva, F. A., Assuncao, H. P.,

Nogueira, J. M. S., and Loureiro, A. A. F. (2005). On

the design of a self-managed wireless sensor network.

IEEE Communications Magazine, 43(8):95–102.

ICEIS 2008 - International Conference on Enterprise Information Systems

392