Design and Analysis of a Wireless Telemetry System Utilizing IEEE 802.16 by Terry Moltzan Advisor: Dr.
TABLE OF CONTENTS 1. Introduction 1.1 Motivation 2. System Selection 3 3 3 2.1 IEEE 802.16 Standard 5 2.2 Redline Equipment 6 2.3 Link Budget 9 3. Test Bed 10 3.1 Test Bed Objectives 10 3.2 Test Bed Architecure 11 3.3 Hardware Selection 13 3.4 Software Selection 14 4.Test Procedure 19 5. Test Results 23 5.1 Packet Size Effects 23 5.2 Received Signal Strength Effects 25 5.3 Additive Noise Effects 25 5.4 Interference Effects 25 5.5 Long Range Tests 28 6.
1. Introduction 1.1 Motivation: Advanced Acoustic Concepts (AAC) has several projects that require a communications link between remote sites. The distances and data rates vary from project to project, but they desire a solution that can handle many of their needs by using standards based off the shelf equipment that can easily be deployed. To be quickly and cost affordable to deploy a wireless system was determined to be necessary.
discussed. It became clear that the emerging IEEE 802.16 standard was the standard that best met the criteria of the task. The top contending systems on the basis of being 1) standards based 2) long range capable and 3) capable of throughput in excess of 40Mbps were the Redline Communications and Airspan offerings. Technically the two systems were very similar in every aspect. The table in Figure 1.1 shows how similar the two systems are in the basic system specifications.
2.1 IEEE 802.16 Standard: The IEEE 802.16 standard is designed as the air interface for wireless metropolitan area networks (MANs). A wireless MAN provides a radio connection between an exterior mounted antenna on a building and a central base station. The MAN is used as a replacement for a wired access network such as cable, DSL, or fiber optic cable. The ability to connect sites with out an expensive infrastructure between them can provide economic advantages as well as advantages in ease of deployment.
sacrificing performance. 99.999 percent link availability is planned into the standard. By making these adjustments the throughput on the link is always maximized for the given conditions. This “self tuning” of the link will provide for the ability to quickly establish links with out having to manually adjust the equipment. In the 2-11 GHz frequencies the MAC layer also issues automatic repeat requests (ARQ) to insure higher data reliability.
Figure 2.1 Throughput for various modulation schemes. The over the air throughput is much higher than the Ethernet rate since it also includes added bits for error detection and link layer retransmission, as well as link management. The AN-50 radios operate in the 100MHz available in the 5.8GHz unlicensed band. Each radio link occupies one 20 MHz wide channel. The spectrum is broken into 9 partially overlapping channels.
The AN-50 functions as a wireless Ethernet bridge. For all of the testing the IP addresses of the two radios were set to 192.168.25.2 and 192.168.25.4. The subnet mask was set to 255.255.255.0. The computers hooked to the terminals were given IP addresses of 192.168.25.100 and 192.168.25.200. The subnet mask was the same as for the terminals themselves. The Redline radios are capable of issuing flow control frames. This feature is very nice for use with UDP traffic.
2.3 Link Budget Consideration: A link budget for the Redline radio was considered for the distance of 20 miles. The system was considered with 28dBi panel antennas and a transmit power of 20dBm. The free space path loss of a 5.8GHz signal at a distance of 20 miles is (λ ) 2 calculated by the equation PL = (4π )2 (d )2 =138 db. The receive power is then 20+28+28-138 = -62dbm. The receive sensitivity of the AN-50 is specified as being –86dbm, therefore the link margin is 24db.
3.Test Bed 3.1 Test Bed Objectives: A test bed for evaluating a high-speed radio link is required. This test bed should be able to simulate a real world link and as many of the parameters that a real world system has to face as possible. The test bed also has to be able to generate and monitor the traffic on the link. The link that is to be tested is a high speed RF link between remote terminals.
quality of shielding was required so that there would be minimum leakage of RF power. This insures that the energy is flowing through the test bed as designed, rather then having radiation leakage that may interfere with another portion of the test bed circuit. The IF interconnects have to be shielded as well, but the low frequency makes it much less critical. The loss on the IF cable is not important since the gain of the transceiver is adjusted to compensate for the cable loss.
P C P C G e n e ra to r R e c e iv e r E th e rn e t C o n n e c tio n A N -5 0 A N -5 0 IF C o n n e c tio n T - 58 T - 58 T -5 0 T -5 0 S im u la te d R F L in k Fig. 3.1 General Test bed Architecture. The physical portion of the test bed, showing the terminals and the basic interconnects. A) Attenuation B) Attenuation Additive Noise C) A tte n u a tio n A N -5 0 T T- 5- 058 A t t e n u a t io n Fig. 3.2 RF link simulation environments.
3.3 Hardware Selection: The signal attenuation block shown in the above section was implemented by using fixed value attenuators, as well as a step attenuator. The set of fixed attenuators were procured from the Electrical and Computer Engineering Department RF lab. The attenuators are fixed N-type attenuators from Weinschel Engineering. Not all of the attenuators were rated for use at this frequency (5.
AN-50 terminal and the T-58 transceiver. The radios, transceivers and some of the attenuators are shown in Fig. 3.3 Advanced Acoustic Concepts supplied the PCs that are used to generate and receive traffic. One terminal was a Compaq Evo N620c laptop with a Pentium M 1.6Ghz processor and 512 MB RAM and a gigabit Ethernet port. In early testing the other terminal was an IBM ThinkPad transnote. This machine did not have the resources to support the throughput we needed.
A Visual Basic (VB) script was written to do some analysis on several different data files at once as well. This allows for complete test set data to be calculated once and imported in one easy step. The Omnicor software has been installed and tested on the Compaq laptop and the IBM Laptop. Figure 3.4 is a diagram of the IP Traffic measurement environment. As is evident in the figure there are many parameters that can be set for testing purposes.
From the IP Generator – Parameters tab the protocol as well as the destination IP address and port can be specified. These values need to match with the values that are set under the IP Answering – Parameters + Statistics tab. The parameters #x button access the parameters screen shown in Figure 3.5. Fig 3.5 IP Generator Parameter Tab. This tab sets the protocol to use as well as the destination IP address and port. Fig 3.6 IP Answering Parameters Tab. This tab sets the source IP address and port.
An important part of the test bed is the web interface for the AN-50. The system status screen is shown below in Figure 3.7. This screen shows the current received power (RSSI), signal to noise and distortion ratio (SINADR), as well as statistics on packet retransmissions on the wireless link. Redline has released RF Monitor, a tool that automatically polls this web interface and plots the results.
sweep, the higher the received power level is. It was observed that small aiming differences did not result in a significant change in signal quality since the beam pattern of the antennas is reasonable flat near zero degrees. Fig. 3.7 System Status Page. The system status page of the Redline web interface displays information such as Received Signal Strength (RSSI), Signal to Noise and Distortion Ratio (SINADR), and packet retransmission data. Fig. 3.8 A typical output graph from RF Monitor.
-----------------------------------------------------------Server listening on UDP port 5001 Receiving 1470 byte datagrams UDP buffer size: 128 MByte ------------------------------------------------------------ [1932] local 192.168.25.100 port 5001 connected with 192.168.25.200 port 1039 [ ID] Interval Transfer Bandwidth Jitter Lost/Total Datagrams [1932] 0.0- 1.0 sec 5.04 MBytes 42.3 Mbits/sec 0.126 ms -842149805/ 4244 (-2 e+007%) [1932] 0.0- 1.0 sec 842150451 datagrams received out-of-order [1932] 1.
generation was set to produce random packet contents for a duration of 1,000,000 packets. The average result over the test interval was extracted by using a VB script. The script can be found in Appendix 1. The traffic generation and data logging functions are not integrated in IP Traffic, so the data logging had to be started manually before the traffic generation, and then stopped after the traffic generation section had completed.
Cobleigh Hall, on the MSU campus and a remote site located twenty miles north on Rocky Mountain Road. The Cobleigh Hall antenna was located on the roof, approximately 80 feet above ground level. The antenna at the Rock Mountain Road site was mounted on an 8-foot mast. The path between the locations was line of site with no blockage within the Fresnel zone. Figure 4.1 shows a terrain map of the area with the two antenna locations labeled.
Fig. 4.2 Fresnal Zone Clearance Fig. 4.3 The Remote Site for the Long Range Test. In this photo the generator is visible in front of the SUV, as well as the antenna and mast at the rear of the vehicle. The laptop and radio were set up on a table at the rear of the vehicle.
Fig. 4.4 View From the Cobleigh Hall Site Towards the Remote Site. 5. Test Results A series of tests have been done on the Redline AN-50 system to evaluate its performance. Bench tests covering the effects of packet size, received signal strength, SNR and various userdefined parameters within the radio were conducted. A long-range test was also conducted by using two sector antennas to establish a link at a range of about 20 miles.
overhead becomes less important as the payload size increases. The packet size was increased to a maximum of 1480 bytes, the size of the Ethernet frame payload. Packets that are larger than this are dropped by the AN-50. In order to use larger packets, such as MPLS packets, on a network with the AN-50 they need to be fragmented before being sent to the AN-50. Throughput Vs.
the testing software. This is where it first becomes apparent that Iperf is better suited to run tests at high speeds. As the signal level is lowered it gets closer to the thermal noise floor. This in effect decreases the SNR without the use of a noise source. The results are shown in Figure 5.2. The same tests were run using Iperf as the software to both generate and measure the traffic levels. The results shown in Figure 5.3 show a similar trend.
5.3 Additive Noise Effects: The effect of injecting various amounts of noise into the signal was investigated. This was done at several different levels of signal attenuation. For a fixed level of signal attenuation the noise power was adjusted by varying the level of attenuation on the front of the noise source. The results are shown in Figure 5.4. This only changes the amount of noise added to the signal, and not the power level of the information signal itself. Throughput Vs.
probably due to the use of OFDM, which allows for sub-channels to be utilized, some of which are not sharing any bandwidth with the interferers. Throughput Vs Interfering Power (Signal at 5.765 GHz) RSSI = -64dbm 40 Throughput (Mbps) 39 38 37 36 35 34 33 32 -40 -35 -30 -25 -20 -15 -10 -5 0 5 Recieved Interfering Power at 5.815GHz (dbm) Fig 5.5 Throughput Vs Received Out of Band Interference.
Throughput Vs Interfereing Power (Signal at 5.765 GHz) RSSI = -64dbm 45 Throughput (Mbps) 40 35 30 25 20 15 10 5 0 -70 -60 -50 -40 -30 -20 -10 0 Recieved Interfering Power (dbm) at 5.775GHz Fig. 5.6 The Effect of Interference In an Overlapping Channel. In this series of tests the occupied bandwidth of the signal is from 5.765GHz to 5.785GHz. The signal bandwidth is from 5.755GHz to 5.775GHz. Throughput Vs Interfering Power (Signal at 5.
environment. The alignment of the antennas was done by first using a visual alignment and then fine-tuning by using the built in buzzer in the transceivers. Aiming proved not to be critical, as the small changes in angle did not change the received power level much since the beam pattern is relatively flat. The range as given by the user interface of the Redline terminal was 17.03 miles. The severe difference in throughput in different channels shown in Figure 5.
The effect of using Redline’s proprietary encryption was investigated as well. The results show little difference if it is on or off. For all of these results the modulation format was set to be adaptive with a maximum bit rate of 54Mbps with a step size of two levels. This proved to give better results than any of the attempts to have a fixed level of modulation. Encryption Throughput (Mbps) On 19.48 Off 19.94 Fig 5.10 The Effect of Encryption on Throughput. 6.
Glossary 31
Appendix 1 This macro takes the raw IP Traffic data files that are saved as text files and stored in the C:\SampleData file. Sub Macro2() Dim objFSO As New FileSystemObject Dim DataDir As Folder Dim DataDirFiles As Files Dim DataFile As File Dim stream As TextStream Set DataDir = objFSO.GetFolder("C:\SampleData") j=2 For Each DataFile In DataDir.Files Set stream = DataFile.OpenAsTextStream stream.ReadLine stream.ReadLine stream.ReadLine stream.
Data = CDbl(temp3(1)) Else Data = 0 End If If Data = 0 And Ender = 0 Then Start = I Else If Data > 0 Then RunningTotal = Data + RunningTotal Ender = I End If End If I=I+1 Wend Average = RunningTotal / (Ender - Start) ActiveSheet.Cells(j, 1) = DataFile.Name ActiveSheet.Cells(j, 2) = Average ActiveSheet.Cells(j, 3) = Start ActiveSheet.Cells(j, 4) = Ender ActiveSheet.
Appendix 2 Iperf commands and options Command line option Environment variable option Description Client and Server options -f, --format [bkmaBKMA] $IPERF_FORMAT A letter specifying the format to print bandwidth numbers in. Supported formats are 'b' = bits/sec 'B' = Bytes/sec 'k' = Kbits/sec 'K' = KBytes/sec 'm' = Mbits/sec 'M' = MBytes/sec 'g' = Gbits/sec 'G' = GBytes/sec 'a' = adaptive bits/sec 'A' = adaptive Bytes/sec The adaptive formats choose between kilo- and mega- as appropriate.
Default is 5001, the same as ttcp. -u, --udp $IPERF_UDP Use UDP rather than TCP. See also the -b option. -w, --window #[KM] $TCP_WINDOW_SIZE Sets the socket buffer sizes to the specified value. For TCP, this sets the TCP window size. For UDP it is just the buffer which datagrams are received in, and so limits the largest receivable datagram size. -B, --bind host $IPERF_BIND Bind to host, one of this machine's addresses. For the client this sets the outbound interface.
-o (only for Windows, from v1.2 or higher) . Redirect output to given file. -c, --client host $IPERF_CLIENT If Iperf is in server mode, then specifying a host with -c will limit the connections that Iperf will accept to the host specified. Does not work well for UDP. -P, --parallel # $IPERF_PARALLEL The number of connections to handle by the server before closing. Default is 0 (which means to accept connections forever).
throughput 0x08 IPTOS_RELIABILITY reliability 0x04 IPTOS_LOWCOST cost 0x02 maximize minimize -T, --ttl # $IPERF_TTL -F (from v1.2 or higher) . Use a representative stream to measure bandwidth, e.g. :$ iperf -c -F -I (from v1.2 or higher) . Same as -F, input from stdin. The time-to-live for outgoing multicast packets. This is essentially the number of router hops to go through, and is also used for scoping. Default is 1, link-local.
Appendix 3 40
Appendix 4 42
Appendix 5 Radio System Test Plan AAC Wireless Telemetry Project Task Order 1 1. Introduction and background This test plan document outlines the procedures for acceptance and performance testing of the Redline Communications AN50 radio system being used provide point-to-multipoint wireless telemetry under AAC Task Order #1 by Montana State University (MSU).
MSU has constructed a lab test bed to support bench acceptance and performance testing of the AN50 radio equipment. The lab test bed is located in the AAC Tech Ranch facility consists of the following components major components. • IP Traffic generators. MSU has procured and installed Omnicor IP traffic generator measurement software capable of generating and measuring up to sixteen simultaneous traffic streams consisting of either UDP or TCP packets.
o Multiple P2P connections with cross talk and noise injection o PMP connection, one BS and three SSs, without or with noise injection. These configurations are shown in Figure 3. A) SS 1 BS 1 N o is e S o u rc e B) SS 1 B S 1 N o is e S o u rc e SS 2 B S 2 C) SS 1 SS 2 BS 1 SS 3 N o is e S o u rc e Figure 3. Testbed configurations for P2P and PMP RF performance measurements a) P2P with noise; b) P2P with cross talk and noise; c) PMP with noise • Data collection and analysis.
Figure 4, in conjunction with the IP layer traffic data to assess system performance. Figure 4. AN50 Radio System web interface diagnostics All data will be collected and archived under computer control for post-test analysis. We will use standard tools such as Matlab to analyze the data and generate reports. 3. Field test bed We will test the radio system in the field after satisfactory completion of the lab bench tests.
in Figure 5. These tests will include power consumption, transmit output power, RF frequencies and the various control functions. We will not test for environmental conformance. RF conformance tests (e.g., throughput versus range) will be carried out through emulation using the lab test bed and calibrated attenuators.
We will also test the web-based administrative interface to assure that the test points and control functions and system configuration data are all functioning as specified. The GUIs for these features are shown in Figure 6. These features include system configuration, SNMP configuration, subscriber ID and configuration, link configuration and wireless link statistics, RF error codes and software update status.
Figure 6. AN50 Web-based administrative and monitoring features 5. Performance tests A. Throughput measurements.
• PMP) Same as above, but with one channel being kept at a constant power as a reference and varying the loss on the second channel. C. Throughput versus SNR ratio measurements. We will test the sensitivity of the radio system throughput to added white noise using the RF noise injection system described above. The P2P and P2MP configurations will be tested with procedures as follows. • P2P) Set-up a P2P link and inject noise from a broadband noise source.
