21-05-2012, 01:23 PM
Multi-Objective Optimization Control in Monitoring System for Facility Agriculture Based on Wireless Sensor Network
[attachment=22564]
Abstract
In order to reduce the impact of inherent network
time-varying transmission delay, packet loss and other
phenomena on performance of the remote monitoring system
based on wireless sensor network, a kind of multi-objective
optimization control strategy based on network quality of
service (QoS) management is proposed from the perspective of
improving the overall system performance. For the data
transmission between sensor nodes and sink node, the
optimization control strategy takes deadline miss ratio as the
QoS evaluation criteria, and utilizes the PID controller to
calculate the network bandwidth needs of all sensor nodes. On
this basis, by solving the optimization problem which contains
multiple objectives, the packet sending period of all sensor
nodes is adjusted so that the network bandwidth allocation can
adapt to the QoS changes, the QoS is maintained at a predetermined
desired level and other performance criteria are
achieved. Preliminary experimentations indicate that the
optimization control strategy is reasonable, effective and
practical, which can be widely used in greenhouses, croplands,
nurseries, etc.
Keywords- quality of service; wireless sensor network; facility
agriculture; optimization
I. INTRODUCTION
Wireless sensor network can effectively meet the
requirements of intellectualization, accuracy and distribution
needs of facility agriculture [1] remote monitoring system.
Therefore, it catches researchers continues attention [2].
However, the inherent network time-varying delay,
packet loss and other phenomena have a strong influence on
the performance of monitoring system for facility agriculture
based on wireless sensor network, even lead the system to be
instable [3]. Therefore some measures must be taken to
reduce the impact of the overall system performance. That is,
improve the network quality of service (QoS) [4].
In order to improve the QoS of network, researchers have
proposed a lot of solutions from the perspective of designing
system controllers, modifying network protocol, etc [5-6].
However, due to the difficulties of obtaining the accurate
mathematical model of wireless sensor network applications
system, and the difficulty to modify the standardized
network protocols, practicability and using effect of these
solutions still need further study. On this basis, self-adaptive
sampling method based on dynamic adjustment sensor
sampling period in accordance with changes of the QoS was
proposed by some researchers from the perspective of
network scheduling [7]. However, the sampling period is
generally based on signal characteristics and design
requirements, and after the completion of system design,
only a small sampling period fluctuation around the set value
is allowed, otherwise it would lead to serious degradation of
system performance, which will seriously affect the
practicability of the self-adaptive sampling method.
In this paper, a multi-objective optimization control
strategy based on the QoS management is presented. The
strategy which is based on self-adaptive sampling method
aims to improve the QoS of network, thus improve the
overall system performance by analyzing the relationship
between the sensor nodes data packet sending period and the
sampling period.
II. CONTROLLER DESIGN
A. System Architecture
Figure 1. System architecture
Fig. 1 demonstrates a facility agriculture remote
monitoring system based on star topology wireless sensor
network. The system is composed of the sensor nodes that
are distributed within the facility environment, the sink node
that collects and processes data, and the base station that
coordinates, optimizes, and manages the system operation.
Sensor nodes are time-triggered, sink node is event-triggered.
B. Basic Design Idea
In the system shown in Fig. 1, sensor nodes sample the
detected signal periodically. Sampled data is calculated,
coded and packed, then sent out to the sink node through the
[attachment=22564]
Abstract
In order to reduce the impact of inherent network
time-varying transmission delay, packet loss and other
phenomena on performance of the remote monitoring system
based on wireless sensor network, a kind of multi-objective
optimization control strategy based on network quality of
service (QoS) management is proposed from the perspective of
improving the overall system performance. For the data
transmission between sensor nodes and sink node, the
optimization control strategy takes deadline miss ratio as the
QoS evaluation criteria, and utilizes the PID controller to
calculate the network bandwidth needs of all sensor nodes. On
this basis, by solving the optimization problem which contains
multiple objectives, the packet sending period of all sensor
nodes is adjusted so that the network bandwidth allocation can
adapt to the QoS changes, the QoS is maintained at a predetermined
desired level and other performance criteria are
achieved. Preliminary experimentations indicate that the
optimization control strategy is reasonable, effective and
practical, which can be widely used in greenhouses, croplands,
nurseries, etc.
Keywords- quality of service; wireless sensor network; facility
agriculture; optimization
I. INTRODUCTION
Wireless sensor network can effectively meet the
requirements of intellectualization, accuracy and distribution
needs of facility agriculture [1] remote monitoring system.
Therefore, it catches researchers continues attention [2].
However, the inherent network time-varying delay,
packet loss and other phenomena have a strong influence on
the performance of monitoring system for facility agriculture
based on wireless sensor network, even lead the system to be
instable [3]. Therefore some measures must be taken to
reduce the impact of the overall system performance. That is,
improve the network quality of service (QoS) [4].
In order to improve the QoS of network, researchers have
proposed a lot of solutions from the perspective of designing
system controllers, modifying network protocol, etc [5-6].
However, due to the difficulties of obtaining the accurate
mathematical model of wireless sensor network applications
system, and the difficulty to modify the standardized
network protocols, practicability and using effect of these
solutions still need further study. On this basis, self-adaptive
sampling method based on dynamic adjustment sensor
sampling period in accordance with changes of the QoS was
proposed by some researchers from the perspective of
network scheduling [7]. However, the sampling period is
generally based on signal characteristics and design
requirements, and after the completion of system design,
only a small sampling period fluctuation around the set value
is allowed, otherwise it would lead to serious degradation of
system performance, which will seriously affect the
practicability of the self-adaptive sampling method.
In this paper, a multi-objective optimization control
strategy based on the QoS management is presented. The
strategy which is based on self-adaptive sampling method
aims to improve the QoS of network, thus improve the
overall system performance by analyzing the relationship
between the sensor nodes data packet sending period and the
sampling period.
II. CONTROLLER DESIGN
A. System Architecture
Figure 1. System architecture
Fig. 1 demonstrates a facility agriculture remote
monitoring system based on star topology wireless sensor
network. The system is composed of the sensor nodes that
are distributed within the facility environment, the sink node
that collects and processes data, and the base station that
coordinates, optimizes, and manages the system operation.
Sensor nodes are time-triggered, sink node is event-triggered.
B. Basic Design Idea
In the system shown in Fig. 1, sensor nodes sample the
detected signal periodically. Sampled data is calculated,
coded and packed, then sent out to the sink node through the