Wireless Senor Network Performance In High Voltage Computer Science Essay

Published:

In this thesis we observe the wireless senor network performance in high voltage and harsh industrial environments. To observe the wireless sensor network performance in industrial environments we use ABB different testing power laboratories.

Wireless sensor network contain wireless field devices (motes) and central gate way. Gate way centralize information gathered and processed by motes, motes can communicate with each other and with gate way via radio.

In this thesis we observe the cause of data packet loss in wireless sensor network communication and the time delay in communication in industrial environments and focus on both the devices and the radio communication.

From harsh industrial environments and high voltage environments we were expecting electromagnetic interference in different frequencies band. These EMI can interfere in transmission band which can create a bit error or radio communication loss or these EMI can cause the change in the linearity/sensitivity of the motes.

Lady using a tablet
Lady using a tablet

Professional

Essay Writers

Lady Using Tablet

Get your grade
or your money back

using our Essay Writing Service!

Essay Writing Service

So in this thesis we monitor the different frequencies band for EMI in these industrial environments, also focused on WSN performance parameters and on the individual mote performance in these environments.

For WSN performance and motes performance a large data is collected, analyzed and simulate in Matlab.

TABLE OF CONTENTS

Introduction

A wireless sensor network consists of wireless field devices called modes. These sensor modes communicate wirelessly with each other with in its radio communication range. A sensor mode has micro processor small amount of memory for signal processing and task scheduling and one or more sensor devices. Wireless sensor net work has wide range of application in commercial industrial and defense application. For example environment monitoring, industrial sensing and diagnostic, powered grid station and multi target tracking.

In industrial sensing and diagnosing like power grids, heavy machine operating environments, affect the wireless sensor networks performance. These industrial environments contain high voltage heavy and changing current, high electric and magnetic field which cause electromagnetic interference in different frequency band.

Background

A lot of work is done in the field of wireless sensor network to improve the network performance. Wireless sensor net work operate in 2.4 GHz ISM band. Bluetooth, WLAN, RFID, s and lot of other medical and industrial devices operate in 2.4 GHz which make very crowded this band [12],So most of the work done in the field of wireless sensor network is to arrange interference free parallel operation with these devices. The work done and ongoing work in the field of wireless sensor network for interference free parallel operation with other technologies operating in 2.4 GHz band using different algorithm proto type and topologies and frequency hopping schemes, Which improves the wireless communication and response time in the field of wireless sensor net work too much but the previous studies in the field of WSN in industrial environment done in ABB shows that the wireless sensor devices performance in high voltage sparking environment in heavy current environment is much different from the non industrial environment[].

Purpose

The main purpose of this thesis is to investigate the performance of WSN in industrial environment in presence of high voltage and harsh environments and monitor these environments for electromagnetic interference.

To investigate what causes in these harsh industrial environment in packet loss and increase in response time in WSN communication.

From these harsh industrial environments electromagnetic interference affects the radio communication which can causes the packet loss or these EMI can change the linearity/sensitivity of field devices which causes the communication loss.

Scope

This theses describe the affect of electromagnetic interference cause by unintentionally by harsh industrial environments( high voltage, heavy machine operating )on WSN communication .the thesis report depend completely on practical experimental results which is done by deploying wireless network in realistic industrial environment.

The collected data is compared by non industrial with no other wireless devices communicating area called zero state data.

The observed data includes the network performance log and level of electromagnetic interference in different frequency band.

This thesis does not describe all possible industrial environments this report based some possible industrial environments.

Abbreviations

ACK

Acknowledgment

CSMA

Carries Sense Multiple Access

Lady using a tablet
Lady using a tablet

Comprehensive

Writing Services

Lady Using Tablet

Plagiarism-free
Always on Time

Marked to Standard

Order Now

DSSS

Direct Sequence Spread Spectrum

EMI

Electromagnetic Interference

FHSS

Frequency Hoping Spread Spectrum

ISM

Industrial, Scientific and Medical

MIC

Message Integrity Code

RSSI

Received Signal Strength Indicator

TDMA

Time division multiple access

WSN

Wireless Sensor Network

Structure

This report has the following structure.

Section 1 Introduction

This section describes the purpose and scope for this report as well as terms, abbreviations and acronyms used.

Section 2 Problem Description

This section describes problem statement.

Section 3 Relative Theory

This section describes the relative theory, uses and applications of wireless sensor network. This section also describes the used technology and definitions. .

Section 4 Test and analysis

This section describes the monitoring and measurement tools and used methodology for monitoring and measuring

This section also describes the Planning of tests, test results, Analysis of tests.

Section 5 Conclusion

This section describes the thesis report Conclusion

Section 6 Future work

This section describe future work

Section 7 Appendix

This section contains the used prototyped data sheets, used product data sheets, and detailed numerical results.

Section 8 References

This section specifies source material and further reading.

In section <> Enclosures are supplementing information.

Problem Description

In high voltage and harsh industrial environments where the unintentional electromagnetic interference present , finding out what factor involve in data packet loss and delay in response time of WSN communication in these environments.

In harsh industrial environments and high voltage environments there are two main expected possibilities in time delay or packet loss in radio communication due to electromagnetic interference. The first possibility can be the electromagnetic interference from industrial environment interferes in transmission band which can cause data packet loss due to bit error or delay in response time.

Other possibility can be the device sensitivity or linearity change due to electromagnetic interference which can also cause data packet loss or delay in response time in WSN radio communication in above mention industrial environments.

Both possibilities can cause individually or mutually error in WSN communication, so estimate from the WSN performance the percentage loss of WSN communication due to bit error and percentage loss of WSN due to change in the linearity of the devices.

Relative Theory

WSN and Transmission band

Wireless Sensor network are currently in the international arena, involving a high degree of cross multi disciplinary, cutting edge knowledge and highly integrated hot area of research. [1]

Wireless Sensor Network has a very broad application prospectus are continuously used in the field of military, defence, agriculture and industrial control.

It also has wide application in the field of biomedical, environmental monitoring, disaster zone and in many more fields.

Due to usage demand of this Wireless Sensor Network technology in every field, there is lot of work done in the field of Wireless Sensor Network technology to improve the performance and make technology faster accurate and reliable.

As Wireless Sensor Network operate in 2.4 G Hz frequency band.

2.4 G Hz band is licensed free and available worldwide and has high band width, due to this a trend to using 2.4 G Hz band increasing rapidly which make this band very crowded.

Bluetooth, RFID, Wi-Fi, microwave, WIRELESS SENSOR NETWORK, and lot of medical and home appliances use 2.4 G Hz frequency band. These all devises and application are in use at same time, and in many places at same time and at same places, like in hospital and in industrial control system, so for reliable communication in this crowded 2.4 G Hz ISM band, the primary requirements of any technology used in this 2.4 G Hz band is a communication with minimum interference with other devices and same devices using this band.[]

Work done in the Field of WSN

There is lot of work done in Wireless Sensor Network and other technologies to overcome this inter band intentional interference. Different intelligent algorithm and protocol developed in WIRELESS SENSOR NETWORK field which use different network topologies, frequencies hoping scheme, time synchronization and combination of different scheme, which make wireless network communication other devices operate in2.4 G Hz band with reliable communication results. But still a lot of research work is ongoing and needed in the improvement of these technologies.

Lady using a tablet
Lady using a tablet

This Essay is

a Student's Work

Lady Using Tablet

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Examples of our work

But when we talk about the real environments the results are different from expected theoretical results in some specific environments like the use of Wireless Sensor Network in industrial environments, in grid station and the environment with high voltage, high current and strong electric and magnetic field, in all these environments some time wiles transmission is not 100% reliable, as it is in normal environments.

Radio frequency interference or electromagnetic interference can be the outcome of these environments which can interfere with communication and can create bit error or can change the device sensitivity, in the result loss or delay in radio transmission.

So we can say there are two types of RFI/EMI which can cause interference in transmission band or affect the signal or device.

The one type of EMI is known and intentional , that is the interference from the same or other devices operating in 2.4 G Hz band , like blue tooth ,RFIDs ,Wi-Fi ,WIRELESS SENSOR NETWORK ,and almost all work is done in Wireless Sensor Network and other technology operating in 2.4 G Hz is to overcome this inter band operating interference problem with same devices and other technologies and devices [8].

But the 2nd type of the EMI/RFI is unintentional mean that the RFI or EMI from the industrial environments ,from the grid station ,sparking or due to strong electric and magnetic field changes or from other sources , in this thesis our focus is this 2nd kind of EMI and these harsh environments .[10]

Related Work

There is some work done in the field of wireless sensor network performance in these harsh industrial and other environments by using different technologies and devices in Wireless Sensor Network.

But mostly they concluded results are not about the reasoning, just about the transmission performance loss.

I n ABB the already work done in the field of Wireless Sensor Network in high voltage and harsh industrial environments by using ABB field devices and using wireless HART technology which shows that the lost of communication and retransmission and delay in harsh and high voltage environments but it's not concluding and reasoning why its happen[p] . Whether it's due to bit error, interference in transmission or signal level interference, or affect on wireless devices which can change device sensitivity or linearity.

In some cases we repeat the some tests and experiments which already done at ABB to verify the results and try to know the original reasoning by monitoring the device log and environmental noise and level of interference and observing the device behavior .

Used Technology

In wireless sensor network different technologies are used on the different demands the most popular technologies in WIRELESS SENSOR NETWORK filed are ZigBee and wireless HART, but for our experiments we use wireless HART technology the reason is that, the ABB research scientist in their research paper "A comparison of wireless HART and ZigBee for industrial application" concluded that for industrial application wireless HART is more suitable then ZigBee [11].

Wireless HART is part of HART7 Standard.

In this thesis for experiments we use DUST Network technology products, Dust Network Company and ABB are member of HARTS 7 standardization committee.

Wireless HART

Wireless HART is secure, robust, low-cost, low rate compatible with existing HART devices wireless solutions, mainly used in the process of industrial monitoring.

Wireless HART protocol work in 2.4 G Hz ISM frequency band, protocol based on IEEE 802.15.4 standard.

Technical features of wireless HART

Highly reliable self-organizing network

TDMA to avoid message conflicts

Increased frequency hopping communication point to point communication

Automatic request retransmission ensures the success rate of packet transmission

Mesh routing to improve the reliability of end to end communication

OSI Layer Model

Wireless HART Stack

Application Layer

Command oriented, predefined data types and applications

Presentation Layer

Session Layer

Transport layert layer

Reliable stream transport, negotiated segment sizes, transfer of large data sets

Network Layer

Power optimized, redundant path, self healing, wireless mesh network

Data link Layer

Secure and reliable, Time Synched TDMA/CSMA, Frequency Agile with ARQ

Physical Layer

2.4 GHz wireless, 802.15.4 based radios, 10 dBm Tx power

Figure 1: OSI layer Model and Wireless HART Stack

The main technical features include the combination of direct sequence spread spectrum (DSSS) and frequency hopping spread spectrum (FHSS) use the adaptive frequency hopping mechanism that can effectively suppressed the sudden interference [4].

In wireless HART another technical feature include the combination of Carrier sense multiple access (CSMA) and Dynamic time division multiple access (TDMA)give full play of advantages of TDMA&CSMA [3].

In wireless HART the network layer using intelligent mesh network technology. Due to interference when path is interrupted the device switches on other communication path of good quality [3].

In wireless HART protocol the transport layer used connection oriented data transmission technology by end to end retransmission mechanism to ensure high reliability of data transmission [2].

The wireless hart protocol use intelligent network management.

Wireless HART network elements

Network Manager

Security Manager

Access Point

Field Device

The network manager is responsible for the management, scheduling and optimization of overall network. It is also responsible for collecting, diagnostics and maintaining the overall network and responsible for reporting to host application about diagnostics [3]

Used products

In this thesis for WSN performance in high voltage and industrial environments, the wireless sensor network products used of DUST SmartMesh network technology.

SmartMesh network has one manager and have up to 250 mode module [7]. SmartMesh networks are reliable, ultra low power and wireless mesh networks used for different monitoring applications [7].

The used DUST SmartMesh products are

The SmartMesh IA-510 D2510 Network Manager[7]

The M2510 Evaluation Mote Module[7]

P1060763

The M2510 Evaluation Mote Module The IA-510 D2510 Network Manager

Figure 2: Dust WSN Network Mote Module and Network Manager[7]

The SmartMesh network manager is responsible for network configuration, management, and gateway functionality for field devices or motes [6].

SmartMesh network managers allow programmatic access to network control commands, by using host interface application via XML API and Serial API [6].

The SmartMesh M2510 motes are ultra low-power wireless transceivers and onboard radio to send packets [6]. (For complete specification of products see Appendix at en

WSN Performance

Wireless Sensor network performance depends on different performance parameters.

Network performance parameters are

Data Reliability

Path stability

Latency

Data Reliability

The data Reliability of the network is percentage of expected data packets that the manager actually received [6].

Data Reliability= (Number of Packet Received) ÷ (number Packet received +number of packets lost) x 100%

Path Stability

The Path Stability of network is percentage of data packets transmit successfully [6].

Packet transmits successfully when transmitter mode get acknowledgement for that packet.

Path stability = (Packets Successfully Transmitted) ÷ (Packets Successfully Transmitted +Packets Failed) x 100%

Latency

Latency is average time required for a data packet to travel from originating mote to the manager [6].latency is measured in millisecond.

Data Latency = (Packet Received) - (Packet Timestamp)

Ideal Latency = [---->10ms (packet time to reach destination)] + [<------10ms (ACK to transmitter)] +offset

Test and Analysis

To observe the wireless sensor network performance in high voltage and industrial environments we did some experiments in high voltage labs and harsh industrial environments.

But for reference we did some experiments in nonindustrial environments.

Non-Industrial Environment tests

It does not mean that the environment is completely interference free because some intentional interference like WLAN ,Wireless devices ,RFID,s are working around the environments, so we can't ignore these intentional interference but to minimizing the effect s the test run for whole week and collected results are mostly of night times and on weekend.

Testing Strategy

The test took place at ABB AB Corporate Research, in office environment room. The test runs for whole week to collect the best average results in this environment.

I used five wireless field devices for this experiment. In which "mote 1" is working as gate way.

The distance of field devices is not more than 7 meters from AP.

For Monitoring and measurement DUST's SmartMesh Console Software and DUST's SmartMesh API & Command Line Interface was used.

For measurement of interference FSH view software and FSH 6 Remote Spectrum [9] analyzer used.

For calculation and simulation MATLAB is used.

Results

Network Statistics

In "Network Statistic" the each performance parameter is for overall wireless sensor network in given time interval. Each time interval is of 15 minutes for this test.

Figure 3: Network Statistics of nonindustrial environment

Here we can observe that the "Network Stability" is above 99.00% and "Latency" is below 0.6 second, which is expected result in non industrial environment. The "Reliability" of the network is 100%, which means that the manager receive all expected data.

If we compare with ideal results, the Latency has very different the reason is that in practical situation results are,

Theoretical expected Latency + delay=Measured Latency

The best Latency I measured is "0.5 second" in 5 meters rang using DUST Network technology products.

Figure4: Packet Transmitted Vs Fail in nonindustrial environments

Here we can see that in each time interval the number of packets for which no acknowledgement was received is less than 10 packets in 15 minutes interval, with average around 5 packets.

The packet can fail to receive acknowledgement by any reason, field device buffer showing the message of full, field device terminate packet, duplicate packet received, packet dropped by the field device or bit error.

Path Statistics

In "Path Statistic" the each performance parameter is for individual field device or single path.

Figure 5: Average RSSI Value for Transmission in nonindustrial environments

Here we can see that in non industrial environment the average RSSI values of transmission is "-57.33 dBm".

Figure 6: Path stability for mote to AP in nonindustrial environment

Here we can see the average path stability of motes to AP. Here path stability is not for whole network just for single path.

The path stability varies from 100% to 98.5% in non industrial environments for mote to AP path.

Result analysis and test summary

In non -industrial environment test we observe that the wireless sensor network performance parameter has 100% Reliability, above 99% Network Stability with average RSSI value of transmission is around -57.33dBm. The packets fail to get acknowledgement is less than 10 packets in each 15 minutes interval and average latency is 0.57 second.

No network layer message integrity error occurs, no single packet drop by the mote in this test.

The performance of the wireless sensor network in non-industrial environment test is as good as we are expecting. So we can use these results for comparative study of wireless sensor network performance in industrial environments and high voltage.

In non industrial environments test we observe that the different network performance parameter affected by increasing the distance mainly the average RSSI value of transmission, and average latency.

Industrial Machines Lab environment Tests

To observe the performance of wireless sensor network in industrial environment, we choose the ABB AB Lab as shown in figure below.

Machine1.jpgMachine2.jpg

550 Kw-523 A--- 690 V 450 Kw-230 A--- 930 V

Figure 7: ABB Lab with two Motor Machines

These kinds of environments are often seen in industries. So to observe the WSN performance we placed four field devices in both sides of the machine, very near almost 20 cm rang.

Expected is that when these motor run, some non intentional electromagnetic interference produced which can disturb the WSN performance.

Testing Strategy

The test took place at ABB AB Corporate Research, in Power Lab.

I used five wireless field devices for this experiment. In which "mote 1" is working as gate way, remaining four filed devices placed on the both sides of the both motor in rang of 20 cm.

The distance between field devices and AP is not more than 10 meters.

The test is divided in to three steps.

Test in the lab when machines are not running.

Test in the lab when machines are running.

Test for EMI in lab when filed devices are off.

For Monitoring and measurement DUST's SmartMesh Console Software and DUST's SmartMesh API & Command Line Interface was used.

For measurement of interference FSH view software and FSH-6 Remote Spectrum analyzer used with different receiver sensitivity of antenna.

For calculation and simulation MATLAB is used

Results

Network Statistics

Figure 8: Network Statistics in Machine lab test

In Network stability graph the blue line shows the network stability of the network when machines were not running, and red line shows the network stability when machines were running. The graph shows that the stability is around 99.5% when machines were not running and in machine running test the network stability drop up to 96.5% with an average of 97.8%. More than 2% loss in same positions and same distance of the field devices shows that the machines running environment affect the WSN performance.

In Latency graph the green line is the average latency of the WSN before the machines were running which is average 0.57 seconds where as red line is latency when machines were running which shows the average latency 0.68 seconds. An increase of 100 ms in latency shows that the WSN performance decreases in machines running environments.

Figure 9: Packets transmitted verses packets fail in Machine lab test

Here we can see that in blue line graph less than 10 packets fail in the test when machines were not running but in machine running test the number of packets fails to get acknowledgement is up to 46 with an average of 30 packets in 15 minutes interval. As the number of fail increase the number of transmitted packets also increased.

Path Statistics

Figure 10: Path statistics in machine lab test

Here we can see that the path stability of the WSN field devices before the machines were running is more than 99% where as when the machines were running the path stability decreases up 89%. The decreased in path stability up to 10% shows that some factor in environments create disturbance in WSN.

Mote statistics

Mote statistics are the performance of the mote individually. In mote statistics the collected results are averaged for individual mote for long duration tests.

Figure 11: Mote performance in Machine lab test

In graph the green line shows the data packets terminated by the motes in machine lab test when machines were not running is less than 5 packets in 5hours test, where as the red line shows in that graph the number of packets terminated by the motes in machine lab test when machine were running and the number of terminated packet is up to 43, with an average of the 30 packets. When the packet is terminated by the mote, motes show the message that buffer is full, which means that the running machine environment affect the motes.

In graph we can also see that the number of duplicate packets received by mote is less than 10 in machine lab test when machines were not running but when machine were running the number of duplicate packets increased up to 550 packets in 5 hours long test which shows that the machines running environment affect the motes that's why it's slow down the mote processing, and packet retransmitted again and again

Figure 12: Mote network layer MIC error and mote temperature

Here we can see that in machine lab test when machines are not running there is no network layer message integrity code error but when machines were running there are few errors in 5 hours test. But the error probability is very, very low in machine lab test.

In graph we can see that the temperature of the mote also increased when machines are running.

Measurements of EMI in 2.4 GHz

For measurements of electromagnetic interference in industrial environments, remote spectrum analyzer is used and reports are created by comparing the results.

Results are collected in the presence of field devices when they are operating, and when they are off

Centre Frequency

: 2.44175 GHz

Frequency Offset

: 0 Hz

Span

: 83.5 MHz

Reference Level

: -20 dBm

Reference Offset

: 0.0 dB

Start Frequency

Stop Frequency

Number of Chanel

Chanel Spacing

: 2.400 GHz

:2.4835 GHz

:15

: 5 MHz

C:\Users\minhas\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.Word\flab7bmp.bmp

Figure 13: 2.4 GHz Spectrum in Machine Lab when no machines were operating

The figure show the interference in 2.4 GHz band in machine lab test when no machine is running. But the field devices are operating. Here we can see that some interference in 2.4 GHz band because of the WSN field devices are operating.

In any environment when the wireless sensor field devices are operating, the FSH-remote spectrum analyzer shows the interference in 2.4 GHz band and that interference in same environment is at different rate and with different probabilities.

Centre Frequency

: 2.44175 GHz

Frequency Offset

: 0 Hz

Span

: 83.5 MHz

Reference Level

: -20 dBm

Reference Offset

: 0.0 dB

Start Frequency

Stop Frequency

Number of Chanel

Chanel Spacing

: 2.400 GHz

:2.4835 GHz

:15

: 5 MHz

C:\Users\minhas\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.Word\flab32bmp.bmpFigure 14: 2.4 GHz Spectrum when machines are operating

Here we can see that there is no interference in 2.4 GHz band, in this test machines are running and WSN field devices are off. The spectrum is monitor with help of remote spectrum analyzer and antenna is placed very near of running machines and monitor all around the machines. But didn't observe any interference in 2.4 GHz band in machine lab when machines are running and WSN field devices are off.

But in machine lab when machines are operating the electromagnetic interference are observed in KHz band.

Result analysis and test summary

In Machine lab test the WSN performance dropped when machines were running compare to machine lab test when machines were not running.

The drop in Network stability is around 2% in average test and latency increased up to 100ms.

The path stability drop up to 10% and the number of packets fail to get acknowledgement increased up to average 30 packets in 15 minutes interval.

In mote statics we observe some network layer MIC error but with very, very few in 5 hours test. But in fail packets over 5 hours, average number of packets terminated by mote is also increased up to average 30 packets compare to the result when machines are not running.

But when interference observe in 2.4 GHz band in machine lab test with machines running and WSN field devices off , no interference observed in 2.4 GHz band. But interference observed in band 0-100 KHz.

So from results we conclude that these kind of industrial environments affect the WSN performance, but these kinds of industrial environments produced no Interference in 2.4 GHz transmission band.

But produced EMI level and this kind of industrial environment affect the WSN performance, specifically performance of the motes.

DC High Voltage environment tests

To observe the performance of wireless sensor network in DC high voltage industrial environment, we choose the ABB power Lab. the test setup used is given below.

The set up consist a transformer to control the voltage, a secondary transformer to increase the voltage and rectifier circuit to convert AC into DC current.

The voltage level increased step by step in (In KV). This kind of high voltage environments can affect the WSN performance and WSN field devices. The high voltage environments can produce Corona; electric Corona is sparking or lightning due to ionization in air in high voltage environments.

P604607.jpg

Figure 15: DC high voltage setup

These kinds of environments are can be seen in industries. So to observe the WSN performance we placed two field devices on the top of DC voltage setup and in middle on a silver ring.

Testing Strategy

The test took place at ABB AB Corporate Research, in Power Lab.

I used three wireless field devices for this experiment. In which "mote 1" is working as gate way, remaining two filed devices placed on Dc high voltage experimental equipment.

The distance between field devices and AP is not more than 10 meters.

The test is divided in to two steps.

WSN performance test in the lab when no DC high voltage in experimental setup.

WSN performance test in the lab when DC high voltage is changing step by step in experimental setup.

For Monitoring and measurement DUST's SmartMesh Console Software and DUST's SmartMesh API & Command Line Interface was used.

For measurement of interference FSH view software and FSH-6 Remote Spectrum analyzer used with different receiver sensitivity of antenna.

For calculation and simulation MATLAB is used.

Results

Network Statistics

Figure 16: Network Statistics in DC High voltage test

In Network stability graph the blue line shows the network stability of the network when Voltage level from experimental setup is zero or no voltage and red line shows the network stability when DC high voltage is present in setup and changing step by step .The graph shows that the stability is around 99% when voltage level is zero but when DC high voltage is present in experimental setup and changing the network stability drop up to 95.5% with an average of 96.7%. More than 2.5% loss in same positions and same distance of the field devices shows that the DC high voltage environment affects the WSN performance.

In Latency graph the green line is the average latency of the WSN when voltage level zero, which is average 0.6 seconds where as red line shows the latency when DC high voltage is present, which is average 0.77 seconds. An increase of 170 ms in latency shows that the WSN performance decreases

Figure 17: Packets transmitted verses packets fail in DC high voltage test

Here we can see that in light blue line graph less than 10 packets fail in the test when voltage level is zero but when DC high voltage (In KV) is present and increasing step by step the number of packets fails to get acknowledgement is up to 32 with an average of ~25 packets in 15 minutes interval. As the number of fail increase the number of transmitted packets also increased.

Path Statistics

Figure 18: Path Statistics in DC High voltage test

Here we can see that the path stability of the WSN field devices around 99% when voltage level is zero where as when the Dc high voltage (In KV) is present and increasing step by step the path stability decreased up 89%. The decreased in path stability up to 10% shows that some factor in environments create disturbance in WSN.

Mote statistics

Figure 19: Mote performance in DC high voltage

In above graph the blue line shows the data packets terminated by the motes in DC high voltage test when voltage level is zero the number of terminated packet is zero. Where as red line shows in that graph the number of packets terminated by the motes in DC high voltage test when voltage level is increasing step by step is up to 14 packets. The results are averaged for one hour test results.

If we compare the results of packets terminated by mote in DC high voltage test and machine lab test then we find that the number of average packet terminated by mote per hour is more in DC high voltage lab test.

In graph we can see that in DC high voltage when voltage level is zero there is no network layer message integrity code error but when voltage level change step by step in DC high voltage (In KV) test there are few errors in 1 hours test.

Measurements of EMI in 2.4 GHz

For measurements of electromagnetic interference in industrial environments, remote spectrum analyzer is used and reports are created by FSH view software by comparing the results.

The result is collected during DC high voltage test but the distance of receiver antenna is more from DC voltage test equipment as compared to the distance of mote around 4 meters more.

flab25bmp.bmp

Figure 20: 2.4 GHz Spectrum DC high voltages Test

There is no interfernce obsevered in 2.4 GHz transmission band in DC high voltage test.

Result analysis and test summary

In DC high voltage test the WSN performance dropped when DC high voltage is present and increasing step by step compare to DC high voltage test when voltage level is zero.

The drop in Network stability is around 2.5% in average test and latency increased up to 170ms.

The path stability drop up to 10% and the number of packets fail to get acknowledgement increased up to average 25 packets in 15 minutes interval.

In mote statics we observe some network layer MIC error but with very, very few in 1 hour test. But in fail packets over 1 hour, average number of packets terminated by mote is also increased up to average 14 packets compare to the result when voltage level is zero.

But when interference observed in 2.4 GHz band in DC high voltage test, no interference observed in 2.4 GHz when motes are not transmitting.

From results we conclude that these DC high voltages affect the WSN performance, but these kinds of industrial environments produced no Interference in 2.4 GHz transmission band.

The performances of motes or filed devices are specifically more affected instead of transmission error.

AC High Voltage environment tests

To observe the performance of wireless sensor network in AC high voltage industrial environment, we choose the ABB power Lab. The test setup used is given below.

The voltage level increased step by step. This kind of high voltage environments can affect the WSN performance and WSN field devices.

P6046075.JPG

Figure 21: AC high voltage setup

Testing Strategy

The test took place at ABB AB Corporate Research, in Power Lab.

I used three wireless field devices for this experiment. In which "mote 1" is working as gate way, remaining two filed devices placed on Ac high voltage experimental equipment. Field devices are placed on 2nd and 5th red box as shown in figure.

The distance between field devices and AP is up to 15 meters.

The test is divided in to two steps.

WSN performance test in the lab when no AC high voltage in experimental setup.

WSN performance test in the lab when AC high voltage is changing step by step in experimental setup.

For Monitoring and measurement DUST's SmartMesh Console Software and DUST's SmartMesh API & Command Line Interface was used.

For calculation and simulation MATLAB is used.

Results

Network Statistics

Figure 22: Network Statistics in AC High voltage test

In AC high voltage test when AC voltage is increased step by step during the test the both motes stop transmission and SmartMesh software show that the motes are lost, even get no response of the ping. But when voltage test stopped after few minutes both motes reconnect the WSN and start transmitting. I did not use any reboot command to reboot motes or manager and wait until the motes itself reconnect the WSN.

In Network stability graph the green line shows the network stability of the network when Voltage level from experimental setup is zero or no voltage and it's showing the graph of before the test and after the AC voltage test finished. Red line shows the network stability when AC high voltage(In KV) is present in setup and changing step by step the interval where the line is broken is showing that motes stop transmitting due to high voltage affect on device. The time when motes stop the high voltage is in its peak during test and it produced some flash light for micro seconds in other equipment .The graph shows that the stability is around 99% when voltage level is zero but when AC high voltage is present in experimental setup and changing, the network stability drop up to 95.1% and then suddenly mote lost from network.

In Latency graph the green line is the average latency of the WSN when voltage level zero, which is average 0.7 seconds where as red line, shows the latency when AC high voltage is present, which increased up to 1.4 seconds before the motes lost from network.

Figure 23: Packets transmitted verses packets fail in AC high voltage test

Here we can see that in blue line less than 10 packets fail in the test when voltage level is zero but when AC high voltage (in KV) is present and increasing step by step the number of packets fails to get acknowledgement is up to 35 with an average of ~25 packets in 15 minutes interval. As the number of fail increase the number of transmitted packets also increased. When motes reconnect, again number of fail start decreasing.

Path Statistics

Figure 24: Path statistics in AC high Voltage Test

Here we can see that the path stability of the WSN field devices during AC high voltage test is dropped up to 86% with an average of 90%, broken portion of line shows that the mote lost and network performance is zero but when the AC high voltage test stop the mote again reconnect the network and then stability increased gradually and maintain around 99%.

Figure 25: Average RSSI value for Transmission

Here in graph we can notice that there is not considerable change in RSSI value of transmission in high voltage test after first 15 minutes interval in over all tests before motes lost and after it reconnect. In 1st interval its value is around -63dBm,-65 dBm ,after 1st time interval motes maintain almost average same level before motes lost and after voltage test stop when it reconnects again.

Average RSSI value of transmission from mote to Gate way more affected by increasing distance then by any other factors.

Result analysis and test summary

During AC high voltage test we observe that the motes lost from the network and reconnect again after few minutes when voltage stop.

If its interference in transmission band then it should start as soon as test stop but mote take more than 15 minutes to reconnect the network which shows that during high voltage test the mote devices affects.

So there is very low probability of interference in transmission band.

In AC high voltage test the WSN performance dropped when AC high voltage is present and increasing step by step compare to AC high voltage test when voltage level is zero or no voltage and also after test when motes reconnect the network.

The drop in Network stability is around 4% in average test and latency increased up to 800ms which is really noticeable delay.

The path stability drop up to 12% and the number of packets fail to get acknowledgement increased up to 36 packets in 15 minutes interval.

From results we conclude that these AC high voltages environments affect the WSN performance, even transmission shut down completely, if we ignore this brake down of transmission even than AC high voltage affect the WSN performance.

.

Conclusions

In this thesis we observed the WSN performance in high voltage, industrial and non industrial environments.

During tests we observed that the WSN performance affected in high voltage and harsh industrial environments as compared to non industrial environments.

As we can see in section (4) in machine lab test, AC high voltage test and Dc high voltage test, the transmission loss or time delay is more as compared to nonindustrial environments. We can also see that in high voltage and machine lab test the transmission loss or time delay is more due to filed device (motes) errors, and delay in processing.

In section (4) we can also see that in high voltage and machine lab tests no electromagnetic interference observed in 2.4 GHz transmission band.

The observed level of interference spectrum is very low.

WSN performance loss in industrial and high voltage environments is not due to electromagnetic interference in transmission band.

So loss in WSN performance in industrial and high voltage environments is due to the affects of these environments on field devices.

We also observe that in WSN the RSSI value for transmission is more affected by increasing the distance than any other conditions or environment impacts.

In section (4.4) we can see that in AC high voltage test the motes lost from network. This kind of unwanted situation or performance loss is because the motes devices are not protected from outside interference or radiation, or the electromagnetic compatibility problem for specific level of interference.

As WSN filed devices working at high frequencies and processor working at high frequencies are very sensitive so some outside frequencies noise or radiation can interfere and can generate some unwanted current or voltage level which can cause the transmission loss and time delay.

Future Work

In this thesis we focused more on overall performance of WSN in high voltage and industrial environments and transmission band interference. There is need to be worked on specifically on motes performance in WSN by measuring the change in current ,processing delay device temperature ,and other parameters to find out the exact causes of the transmission loss and time delay by modes or field devices. On the basis more specific mote performance results, need to be work on motes protection from radiation, interference and other harsh impacts from industrial environments.

Also need to work more sensitively for field devices on electromagnetic compatibility problems.

Appendix

D2510 Datasheet [6]

Parameter

Min

type

Max

Units

Comments

NORMAL OPERATING CONDITIONS

Operational supply Voltage range (between Vcc and GND)

4.0

5.0

5.0

V

Including noise and load regulation

Peak current

210

mA

, 3V3 out = 0 mA

Average current

100

140

mA

+5V_IN at 5.0 V, 25 °C,

+3V3 out = 0 mA

Operating temperature range

-40

85

°C

Antenna Specifications

Frequency range

2.4

2.4835

GHz

Impedance

50

Ω

Gain

+2 dBi

Pattern

Omni-directional

DETAILED RADIO SPECIFICATIONS

Operating frequency

2.4000

2.4835

GHz

Number of channels

15

Channel separation

5

MHz

Occupied channel

2.7

MHz

At -20 dBc

bandwidth

Modulation

IEEE 802.15.4 direct sequence spread spectrum (DSSS)

Raw data rate

250

kbps

Receiver sensitivity

-90

dBm

At 1% PER, , 25° C

Output power, EIRP

-2

dBm

Vcc = 3 V, 25° C +2 dBi antenn

Range* Indoor-outdoor

25

200

m

m

25° C, 50% RH, 1 meter above ground, +2 dBi Omni-directional antenna

* values when power amplifier disable

SMARTMESH IA-510 M2510 [5]

.