Transcript
Mobile WiMAX™ Throughput Measurements Application Note Products: |
R&S®CMW500
|
R&S®CMW270
Steffen Heuel Heinz Mellein 09.2011-1SP10
Application Note
This application note explains the basics of data throughput measurements for the mobile WiMAX™ air interface. In addition this application note provides the throughput reference measurement setup based on the R&S®CMW270 or R&S®CMW500 communication tester, including representative results.
02
Rohde & Schwarz Application Note 1SP10 3
Table of Contents 1 Introduction .............................................................................................................. 6 2 OSI layers and data throughput ............................................................................... 7 2.1 Mobile WiMAX™ PHY layer ................................................................................... 7 2.1.1 Mobile WiMAX™ TDD radio frame structure..................................................... 9 2.1.2 Mobile WiMAX™ PHY channel encoding ........................................................ 10 2.1.3 Mobile WiMAX™ MIMO operation .................................................................... 11 2.2 Mobile WiMAX™ MAC layer................................................................................. 12 2.3 IP layer
............................................................................................................ 13
2.4 Transport layer ..................................................................................................... 13 2.5 Upper layer applications...................................................................................... 14 3 Throughput Measurements .................................................................................... 15 3.1 TCP-Throughput................................................................................................... 17 3.1.1 Bandwidth – Delay Product.............................................................................. 17 3.1.2 Packet Loss ....................................................................................................... 18 3.1.3 Upstream Bandwidth ........................................................................................ 19 4 Simulation and Prototype Results ......................................................................... 20 4.1 UDP/ICMP Throughput Measurement Results................................................... 20 4.2 TCP/FTP Throughput Measurement Results ..................................................... 21 5 Test Setup ............................................................................................................ 23 5.1 Test Setup A ......................................................................................................... 24 5.2 Test Setup B ......................................................................................................... 27 5.3 Additional Information ......................................................................................... 28 6 Application Setup.................................................................................................... 29 6.1 R&S®CMW270 or R&S®CMW500 Setup ............................................................ 29
02
Rohde & Schwarz Application Note 1SP10 4
6.2 UDP-Throughput Test Setup ............................................................................... 31 6.3 TCP-Throughput Test Setup................................................................................ 32 6.4 FTP-Throughput Test Setup ................................................................................ 34 6.5 Video-Stream Setup ............................................................................................. 37 6.6 Additional Information ......................................................................................... 42 7 Conclusion ............................................................................................................ 44 8 Abbreviations .......................................................................................................... 45 9 Literature
............................................................................................................ 46
10 Additional Information .......................................................................................... 47 11 Ordering Information ............................................................................................ 48
02
Rohde & Schwarz Application Note 1SP10 5
Introduction Mobile WiMAX™ PHY layer
1 Introduction Data rates in communication systems are basically determined by the capacity of the interfaces involved. Therefore, in wireless communication systems the air interface may appear as a bottleneck of the entire system and should be analyzed with due care. The capacity of the mobile WiMAX™ air interface depends on the OFDMA signal structure and the duplexing scheme, in particular. Thus, the possible data rate across the air interface is determined by the physical layer (PHY) implementation. However, on top of those PHY data rates, the e2e (end-to-end) data throughput from an applications viewpoint is of great interest and depends on upper layer implementations as well. It is obvious that both, PHY data rates and upper layer data throughput measurements are an important task in all stages of a mobile WiMAX™ product development. This includes in particular the R&D phase and the production cycle, in order to ensure the quality and conformity of the reference design and the final product. This application note will evaluate the data rates on the mobile WiMAX™ physical layer according to [1], and the throughput rates for various applications at upper layers (chapter 3). Chapter 4 will present reference measurement results achieved. Detailed explanations of easy to use test setups including the R&S®CMW270 or R&S®CMW500 communication tester in order to measure data and throughput rates across the mobile WiMAX™ air interface are provided in chapter 5 and 6. The main focus there is to evaluate the maximum possible rates for various applications and to compare them against the maximum rates that can be achieved in theory. Chapter 2 will give a brief introduction to the ISO/OSI layer model of communications, in order to provide the very basics of the following data rate and throughput measurement considerations. Mobile WiMAX™ according to [1] is basically designed to provide the lower layers of an IP based network, i.e. it shall provide mobile, broadband wireless access to the internet. Thus, all applications considered in this document will be IP based applications. By the way, the terms throughput and bandwidth are often used synonymous for the amount of data transferred across a communication network per time unit. Existing bandwidth measurement tools show more or less accurate results without sufficiently specifying what bandwidth or throughput exactly they measure. Thus, such kind of tools should be used carefully.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
6
OSI layers and data throughput Mobile WiMAX™ PHY layer
2 OSI layers and data throughput This chapter provides the very basics of the ISO/OSI layer model according to Figure 1, as they are relevant for the throughput measurements across the mobile WiMAX™ air interface.
Figure 1: ISO/OSI layer model Mobile WiMAX™ layer 1 (PHY) and layer 2 (MAC) are specified in detail in [1], however, the following sub clauses shall summarize all air interface capacity relevant information.
2.1 Mobile WiMAX™ PHY layer The mobile WiMAX™ PHY layer according to [1] corresponds to the ISO/OSI physical layer 1 (Figure 1) and determines the maximum data capacity of the related air interface. The following parameters are responsible: • • • •
02
CP-OFDM parameters o Bandwidth and corresponding FFT size o Cyclic Prefix length Duplex scheme o TDD or FDD o Volume of downlink and uplink sub frame in case of TDD Modulation scheme o QPSK, 16QAM, 64QAM Channel encoding scheme o Repetition rate o FEC coding rate
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
7
OSI layers and data throughput Mobile WiMAX™ PHY layer
Figure 2 summarizes the data capacity determining parameters on mobile WiMAX™ PHY layer according to [1]. BW Nused Ndata n G NFFT Fs Jƒ Tb Tg Ts M rFEC rRep
Nominal channel bandwidth { 10 MHz, 8.75 MHz, 7 MHz, 5 MHz, 3.5 MHz } Number of used subcarriers (data and pilot subcarriers) Number of data subcarriers (Over-)Sampling factor { 8/7, 28/25 } Ratio of cyclic prefix (CP) time to useful time ( default G = 1 / 8 ) Fast Fourier Transform size.Smallest power of 2 greater than Nused {512,1024 } Sampling frequency Fs = floor (n·BW/8000) x 8000 Subcarrier spacing Jƒ = Fs / NFFT OFDM symbol time Tb = 1 / Jƒ Cyclic Prefix (CP) time Tg = G · Tb CP-OFDM symbol time Ts = Tb + Tg = ( 1 + G ) · Tb QAM modulation order { 2 (QPSK), 4 (16QAM), 6 (64QAM) } FEC coding rate { 1/2, 2/3, 3/4, 5/6 } Repetion coder rate { 0, 2, 4, 6 } Figure 2: Mobile WiMAX™ OFDM parameters according to [1]
The instantaneous data rate R that can be achieved across the mobile WiMAX™ air interface is determined by the number of bits per CP-OFDM symbol of duration Ts. According to [1] a CP-OFDM symbol is defined by its pilots and the number Ndata of data sub carriers in the frequency domain. There, each occupied data sub carrier is M modulated by the 2 -QAM modulation scheme. Thus, the instantaneous data rate R is given by
R=
M N data Ts
Table 1 summarizes the maximum instantaneous gross data rates depending on the nominal bandwidth and modulation scheme. Obviously, the maximum instantaneous PHY data rate of 42 Mbps can be achieved with the maximum nominal bandwidth BW and 64QAM in the default PUSC [1] sub carrier permutation mode. BW [MHz] 10 8.75 7 5 3.5
n 28/25 8/7 8/7 28/25 8/7
FS [MHz] 11.2 10 8 5.6 4
NFFT 1024 1024 1024 512 512
Jƒ [kHz] 10.9 9.8 7.8 10.9 7.8
Ndata 720 720 720 360 360
Tb [Ns] 91.4 102.4 128 91.4 128
Ts [Ns] 102.9 115.2 144 102.9 144
RQPSK [Mbps] 14 12.5 10 7 5
R16QAM [Mbps] 28 25 20 14 10
R64QAM [Mbps] 42 37.5 30 21 15
Table 1: Maximum instantaneous gross data rates, PUSC, G = 1/8
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
8
OSI layers and data throughput Mobile WiMAX™ PHY layer
However, digital communication across any interface requires a certain data processing (e.g. channel encoding) and signaling (e.g. broadcast information) overhead. Thus, the available PHY capacity is not entirely available for upper layer applications, i.e. a portion of it has to be excluded for mandatory signaling purposes. Furthermore, the duplex scheme has a significant impact on the possible data rates towards one direction. In case of time division duplex, the most common scheme for mobile WiMAX™, the downlink and the uplink direction have to share the available resources over time. All those aspects shall be discussed in the subsequent sub clauses.
2.1.1 Mobile WiMAX™ TDD radio frame structure In case of a TDD mobile WiMAX™ air interface, the available PHY resources are organized in terms of radio frames according to Figure 3. Due to the time division duplexing, every single radio frame of a certain length (default 5 ms) is divided into a downlink sub frame and an uplink sub frame. Those two parts do not necessarily have the same size. Both sub frames are separated by well specified transition gaps. Each and every downlink sub frame starts with the preamble first, followed by some common broadcast signaling overhead. The uplink sub frame data resources are shortened by some mandatory uplink signaling, e.g. the ranging zone.
Figure 3: Mobile WiMAX™ TDD radio frame Table 2 depicts the number of available CP-OFDM symbols per 5 ms TDD radio frame depending on the nominal bandwidth and FFT size (column 1 and 2). The total number of symbols (column 3) is distributed asymmetrically across the downlink sub frame and the uplink sub frame. Column 6 depicts the distribution for maximum downlink sub frame, column 7 depicts the distribution for maximum uplink sub frame size. However, there are always a number of symbols to be excluded, which are allocated to common signaling purposes. The number of signaling overhead symbols for the downlink and uplink are depicted by column 4 and 5.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
9
OSI layers and data throughput Mobile WiMAX™ PHY layer
BW [MHz]
FFT size
Total number of symbols
DL signaling overhead symbols
UL signaling overhead symbols
10 8.75 7 5 3.5
1024 1024 1024 512 512
47 42 33 47 33
13 13 13 17 17
3 3 3 3 3
Max DL sub frame symbol distribution DL UL 35 12 30 12 24 9 35 12 24 9
Max UL sub frame symbol distribution DL UL 26 21 24 18 18 15 26 21 18 15
Table 2: Mobile WiMAX™ TDD frame OFDM symbol distribution
1
Furthermore, the available OFDMA resources are divided into slots, which by default contain 48 modulation symbols in total each. Thus, there is always an integer multiple of 48 modulation symbols on air. In case of QPSK modulation, this means an integer multiple of 96 bits, in case of 16QAM an integer multiple of 192 bits and in case of 64 QAM an integer multiple of 288 bits. Taking into account the TDD radio frame structure according to Figure 3, the asymmetrical OFDM symbol distribution and the required signalling overhead symbols according to Table 2, the available PHY capacity for payload data will be as listed Table 3. Now the payload gross data rate is calculated by the number of available slots times 48 modulation symbols divided by the total frame length of 5 ms. BW [MHz] 10 8.75 7 5 3.5
Max DL slots 330 240 150 135 45
Max DL payload data rate [Mbps] QPSK 16QAM 64QAM 6.336 12.672 19.008 4.608 9.216 13.824 2.880 5.760 8.640 2.592 5.184 7.776 0.864 1.728 2.592
Max UL slots 210 175 140 102 68
Max UL payload data rate [Mbps] QPSK 16QAM 64QAM 4.032 8.064 12.096 3.360 6.720 10.080 2.688 5.376 8.064 1.958 3.917 5.875 1.306 2.611 3.917
Table 3: Maximum payload gross data rates, PUSC, G = 1/8
2.1.2 Mobile WiMAX™ PHY channel encoding In addition to the already discussed modulation and OFDMA symbol allocation, the mobile WiMAX™ PHY signal processing is responsible for the channel encoding too. The channel encoder basically includes two steps. A first step performs a forward error correction (FEC) encoding of various types and coding rates. The coding rate rFEC is the ratio of the number of FEC block input bits and FEC block output bits. Every FEC type used by mobile WiMAX™ adds redundancy to the input bits. Thus, the coding rate rFEC is ever less than 1. Possible values for rFEC are 1/2, 2/3, 3/4 and 5/6. 1
Important note:
The calculation of the DL signaling overhead size is based on the assumption that the default broadcast messages (e.g. DL-MAP) are channel encoded with maximum repetition factor 6. Reducing that repetition factor would consequently reduce the size of the DL signaling overhead. Thus, the number of DL data symbols per frame would be increased. When doing so, the DL reference values of table 2, 3, 4 and 5 have to be re-calculated!
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
10
OSI layers and data throughput Mobile WiMAX™ PHY layer
For example, in case of coding rate 5/6 each 5 bits FEC block input result in 6 bits FEC block output. Or in other words, the FEC block output gross data rate is 20 % higher than the input net data rate. Obviously, the coding rate 1/2 adds more redundancy than coding rate 5/6. Thus, in order to achieve the maximum net data rate coding rate 5/6 in conjunction with modulation scheme 64QAM shall be used. A second step of the channel encoder is a simple repetition coder, to increase data protection. Repetition coding simply means to transmit multiple copies of the same FEC output block, subsequently. [1] specifies valid repetition factors as 0, 2, 4 and 6. Since repetition coding is only applied in conjunction with QPSK and the focus is on the maximum throughput rates, the repetition factor is set to 0 for all considerations in this application note, i.e. no impact of the repetition coder is assumed. FEC coding rate 1/2 2/3 3/4 5/6
Max DL payload data rate [Mbps] QPSK 16QAM 64QAM 3.168 6.336 9.504 n/a n/a 12.672 4.752 9.504 14.256 n/a n/a 15.840
Max UL payload data rate [Mbps] QPSK 16QAM 64QAM 2.016 4.032 6.048 n/a n/a 8.064 3.024 6.048 9.072 n/a n/a 10.080
Table 4: Maximum payload net data rates, BW 10 MHz, PUSC, G = 1/8 Exemplarily, Table 4 depicts the mobile WiMAX™ PHY net payload data rate for the maximum nominal bandwidth of 10 MHz. The numbers given by Table 3 (10 MHz line) are simply adopted for the various FEC coding rates. Please note, 64QAM is not mandatory on the uplink, and might not be supported by the device under test. All maximum PHY rates for all different bandwidths are given in Table 5.
2.1.3 Mobile WiMAX™ MIMO operation Multiple antenna (MIMO) implementations have two basic goals: Increasing the performance, i.e. reduce the bit error rate and increasing the air interface capacity. Throughput measurements are of interest for both types of MIMO implementations. Mobile WiMAX™ according to [1] offers therefore two basic MIMO types, known as matrix A and matrix B. Matrix A uses two downlink transmit antennas and requires one receive antenna at the mobile station only. Its goal is to improve the performance. Matrix B requires 2 downlink transmit antennas and 2 receive antennas at the mobile station. Therefore, it can double the air interface capacity with respect to a single antenna implementation (SISO). Thus, measuring the maximum throughput across a matrix A implementation would not change the reference values. However, considering matrix B implementations, all reference values on PHY level should be doubled! Please note, MIMO of type matrix A and matrix B according to [1] affect the downlink direction only.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
11
OSI layers and data throughput Mobile WiMAX™ MAC layer
2.2 Mobile WiMAX™ MAC layer The mobile WiMAX™ MAC layer according to [1] corresponds to the ISO/OSI data link layer (Figure 1) and determines the maximum capacity of the related air interface. It delivers the data towards the PHY layer in terms of MAC PDUs (Protocol Data Unit). Such a MAC PDU is composed according to Figure 4 of a 6 bytes header, a payload of variable length and a 4 bytes CRC check sum. MTU – Maximum Transmission Unit The maximum transmission unit refers to the size of the largest PDU that a given ISO/OSI layer 2 implementation, such as the mobile WiMAX™ MAC layer or e.g. Ethernet according to IEEE 802.3, can handle without fragmentation. The MTU size for mobile WiMAX™ is determined by the maximum length of a MAC PDU, which is 2 KB (2047 Bytes) according to Figure 4. This includes 10 Bytes of header and CRC check sum. As mentioned in the previous section, the mobile WiMAX™ PHY divides all incoming MAC PDUs into FEC input blocks which match to the slot structure. If the payload size delivered by higher layers exceeds the maximum MTU size, it will be fragmented by the mobile WiMAX™ MAC into 2 KB portions. Header 6 bytes
Payload Maximum 2037 bytes
CRC 4 bytes
Figure 4: Mobile WiMAX™ MAC PDU format Due to the signaling overhead caused by the 6 byte MAC header and the 4 byte CRC, the upper layer payload rate is slightly reduced with respect to the PHY payload rate according to Table 4, for instance. Assuming full MTU size deliveries, the payload rate is reduced approximately by 0.5 %. By the way, the default Ethernet MTU size according to IEEE 802.3 is 1500 bytes, and this is how standard IP networks handle data. Therefore, common internet upper layer implementations often adapt their MTU sizes with respect to the default Ethernet MTU size of 1500 Bytes, which not fully covers the maximum mobile WiMAX™ MAC MTU size of 2 KB. Thus, it is very likely that a single MAC PDU includes a single IP datagram and a fraction of the subsequent one.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
12
OSI layers and data throughput IP layer
2.3 IP layer The Internet Protocol (IP) covers basically the ISO/OSI network layer (Figure 1) function. It is a connection-less protocol designed for packet-switched (PS) networks. Thus, IP does not really care about the physical link – it simply assumes there is one. Indeed, the MAC and PHY layers provided by mobile WiMAX™ establish and maintain the mobile radio link, in order to serve the IP layer. Today, the internet uses mainly IP version 4 implementations. IPv4 handles data by means of datagrams, including a 20 Bytes IP header. The header includes in particular the 32 Bit source IP address and the 32 Bit destination IP address. The maximum total length of an IPv4 datagram is 64 KB (65535 Bytes).
2.4 Transport layer There are two common transport layer protocols representing ISO/OSI layer 4 (Figure 1) functions. With UDP, there is a layer 4 connection-less protocol, i.e. with no additional transmission protection. However, the most important one is the connection-oriented TCP protocol. Both protocols are of great interest with respect to throughput measurements in a mobile WiMAX™ environment. The TCP protocol handles data by means of segments of a certain length including a 20 byte header. For instance, the TCP header includes a 16 bit source and destination port. Furthermore, a checksum and sequence numbering scheme enable acknowledgement procedures for secure communications. The length of the TCP segment ideally fits onto the lower layer MTU sizes, typically the Ethernet MTU size of 1500 Bytes. The entire TCP data processing flow towards the mobile WiMAX™ PHY layer is illustrated by Figure 5. It is obvious, that every upper layer contributes some overhead, typically a layer specific header of a certain size, which reduces slightly the payload throughput. Table 5 depict approximate maximum payload data rates for various layers, including UDP and TCP.
TCP
HEADER
IP
HEADER
Payload (typical 1500 Bytes)
Payload (max. 64 KB)
MAC
CRC
Payload (max. 2KB) etc.
PHY
Slot
Slot
Slot
Slot
Slot
Slot
Slot
Figure 5: TCP data processing towards the mobile WiMAX™ PHY
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
13
OSI layers and data throughput Upper layer applications
2.5 Upper layer applications Now on top of the layers 1 – 4 an application can be implemented (Figure 1). In fact, there is a vast variety of applications out there with different Quality of Service requirements. However, for the measurements presented in this document the most common applications are considered only, which includes FTP, HTTP and streaming applications.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
14
Throughput Measurements Upper layer applications
3 Throughput Measurements The most common method of performing throughput measurements is sending a large file from one peer to another. By dividing the file size over the transfer time duration the throughput rate in bits per second is achieved. This method measures the applicationthroughput of the established link, describing the throughput on the application layer of the OSI-model excluding protocol overhead such as transport layer or network layer, retransmitted (ARQ, HARQ) packets due to loss, corruption or error messages for example. Table 5 depict the maximum capacity of the different ISO/OSI layer implementations across the mobile WiMAX™ air interface. The PHY capacity for every valid nominal bandwidth, modulation scheme and FEC coding rate is calculated according the considerations of the previous sub clause. The figures for the MAC, IPv4, UDP and TCP layer are approximations assuming full size MTU transmission. Thus, those figures are upper bounds for the expected rates, and might be used as benchmarks. As the following sub clauses will show, R&S®CMW based throughput measurements reach those upper bounds. However, with respect to connection-oriented protocols, such as the most common TCP protocol, further aspects have an impact to the maximum throughput. Since the TCP protocol applies a transmission protection scheme based on regular peer entity reception acknowledgments, the throughput depends on the peer-to-peer round trip time and the packet loss rate. Both aspects will be discussed in more detail in the following section.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
15
Throughput Measurements Upper layer applications
Max downlink reference rates PHY 10 MHz QPSK 16 QAM 64 QAM
1/2 3/4 1/2 3/4 1/2 2/3 3/4 5/6
MAC
3,17 4,75 6,34 9,50 9,50 12,67 14,26 15,84
3,15 4,72 6,30 9,44 9,44 12,59 14,17 15,74
2,30 3,46 4,61 6,91 6,91 9,22 10,37 11,52
2,29 3,43 4,58 6,87 6,87 9,16 10,30 11,44
8.75 MHz QPSK 16 QAM 64 QAM
1/2 3/4 1/2 3/4 1/2 2/3 3/4 5/6
16 QAM 64 QAM
1/2 3/4 1/2 3/4 1/2 2/3 3/4 5/6
16 QAM 64 QAM
1/2 3/4 1/2 3/4 1/2 2/3 3/4 5/6
1,44 2,16 2,88 4,32 4,32 5,76 6,48 7,20
1,43 2,15 2,86 4,29 4,29 5,72 6,44 7,15
16 QAM 64 QAM
1/2 3/4 1/2 3/4 1/2 2/3 3/4 5/6
3,13 4,69 6,26 9,37 9,37 12,50 14,07 15,63
PHY
MAC
2,27 3,41 4,55 6,82 6,82 9,09 10,23 11,37 1,42 2,13 2,84 4,26 4,26 5,68 6,39 7,10
1,30 1,94 2,59 3,89 3,89 5,18 5,83 6,48
1,29 1,93 2,57 3,86 3,86 5,15 5,79 6,44
1,28 1,92 2,56 3,84 3,84 5,12 5,76 6,39
3,07 4,60 6,14 9,20 9,20 12,27 13,81 15,34
3,01 4,51 6,01 9,01 9,01 12,02 13,53 15,03
2,02 3,02 4,03 6,05 6,05 8,06 9,07 10,08
2,00 3,00 4,01 6,01 6,01 8,01 9,01 10,01
2,23 3,35 4,46 6,69 6,69 8,92 10,04 11,16
2,19 3,28 4,37 6,56 6,56 8,74 9,84 10,93
1,68 2,52 3,36 5,04 5,04 6,72 7,56 8,40
1,67 2,50 3,34 5,01 5,01 6,68 7,51 8,34
0,43 0,64 0,86 1,29 1,29 1,72 1,93 2,15
0,43 0,64 0,85 1,28 1,28 1,71 1,92 2,13
UDP
TCP
1,99 2,98 3,98 5,97 5,97 7,96 8,95 9,95
1,95 2,93 3,90 5,86 5,86 7,81 8,79 9,76
1,91 2,87 3,82 5,74 5,74 7,65 8,61 9,56
1,63 2,44 3,25 4,88 4,88 6,51 7,32 8,13
1,59 2,39 3,19 4,78 4,78 6,37 7,17 7,97
1,30 1,95 2,79 3,90 3,90 5,21 5,86 6,51
1,27 1,91 2,73 3,82 3,82 5,10 5,74 6,37
0,95 1,42 1,90 2,84 2,84 3,79 4,27 4,74
0,93 1,39 1,86 2,79 2,79 3,72 4,18 4,64
0,63 0,95 1,26 1,90 1,90 2,53 2,84 3,16
0,62 0,93 1,24 1,86 1,86 2,48 2,79 3,10
24:18 1,66 2,49 3,32 4,97 4,97 6,63 7,46 8,29 18:15 1,39 2,09 2,79 4,18 4,18 5,58 6,28 6,97
1,37 2,05 2,73 4,10 4,10 5,46 6,15 6,83
1,34 2,02 2,69 4,03 4,03 5,38 6,05 6,72
1,34 2,00 2,86 4,01 4,01 5,34 6,01 6,68
1,33 1,99 2,84 3,98 3,98 5,31 5,97 6,63 26:21
1,26 1,88 2,51 3,77 3,77 5,02 5,65 6,28
1,23 1,84 2,46 3,69 3,69 4,92 5,53 6,15
0,98 1,47 1,96 2,94 2,94 3,92 4,41 4,90
0,97 1,46 1,95 2,92 2,92 3,89 4,38 4,86
24:9 0,43 0,65 0,86 1,30 1,30 1,73 1,94 2,16
IPv4
DL:UL = 26:21 symbols
35:12
3.5 MHz QPSK
TCP
24:9
5 MHz QPSK
UDP
30:12
7 MHz QPSK
IPv4
DL:UL = 35 : 12 symbols
Max uplink reference rates
0,97 1,45 1,93 2,90 2,90 3,87 4,35 4,83 18:15
0,42 0,63 0,84 1,26 1,26 1,67 1,88 2,09
0,41 0,61 0,82 1,23 1,23 1,64 1,84 2,05
0,65 0,98 1,31 1,96 1,96 2,61 2,94 3,26
0,65 0,97 1,30 1,95 1,95 2,59 2,92 3,24
0,64 0,97 1,29 1,93 1,93 2,58 2,90 3,22
Table 5: Maximum reference rates
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
16
Throughput Measurements TCP-Throughput
3.1 TCP-Throughput In addition to the overhead caused by TCP header, further aspects have a significant impact on the TCP throughput. Those are • • •
window size packet loss uplink capacity
Unfortunately, there is no simple formula considering all kinds of limitations. So the throughput will be limited by the minimum of one calculation – to forestall, this is the window size of 64 KB. To understand those additional dependencies, it is necessary to look at the TCP threeway-handshake algorithm as illustrated in Figure 6.
Figure 6: TCP three-way-handshake Since TCP is a connection-oriented protocol, it has to make sure that there is a connection towards the peer port. Thus, prior to every data transmission of a certain size (known as the window size), an acknowledgement by the server port is awaited upon a sync sequence originated by the client port. It is obvious, that the Round Trip Time (RTT) as well as the window size have an impact on the TCP throughput. This dependency can be easily determined by the bandwidth-delay product.
3.1.1 Bandwidth – Delay Product The upper bound TCP throughput can be calculated by the bandwidth – delay product, which is the product of the Window Size (representing the bandwidth) and the Round Trip Time (representing the delay). The Window Size is usually 64 KB and the Round Trip Time across a mobile WiMAX™ air interface is due to the 5 ms TDD radio frame structure greater than 20 ms (typical values are 30 – 40 ms). Figure 7 depicts the TCP throughput vs. the RTT for various window sizes (WS).
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
17
Throughput Measurements TCP-Throughput
15.84 Mbps WiMAXTM PHY Limit
20 ms RTT
Figure 7: TCP Throughput vs. RTT Table 6 depicts representative values of the TCP throughput vs. RTT for various window sizes. According to the previous considerations, the important region for mobile WiMAX™ according to [1] is obviously beyond 20 ms RTT and below 15.84 Mbps throughput. WS [KB] RTT [ms] 20 30 35 40 45 50 55 60 65 70 75 80
128 [Mbps] 52,43 34,95 29,96 26,21 23,30 20,97 19,07 17,48 16,13 14,98 13,98 13,11
64 [Mbps] 26,21 17,48 14,98 13,11 11,65 10,49 9,53 8,74 8,07 7,49 6,99 6,55
32 [Mbps] 13,11 8,74 7,49 6,55 5,83 5,24 4,77 4,37 4,03 3,74 3,50 3,28
16 [Mbps] 6,55 4,37 3,74 3,28 2,91 2,62 2,38 2,18 2,02 1,87 1,75 1,64
8 [Mbps] 3,28 2,18 1,87 1,64 1,46 1,31 1,19 1,09 1,01 0,94 0,87 0,82
Table 6: Upper bound TCP throughput considering WS and RTT
3.1.2 Packet Loss Another important factor is the Packet Loss, e.g. due to channel outage or error The TCP-Throughput is reduced and the upper bound can be calculated ([4] and [5]). Figure 8 depicts the TCP-Throughput depending on packet error rate PER for various RTTs. The reference sensitivity level PER value for mobile WiMAX™ terminals in a -4 stationary AWGN channel is specified as 4.3 % (4.3·10 ). The reference input level at the terminal for this case is specified as approx. -72 dBm [1] for the most sensitive modulation scheme 64QAM. Thus, it is strongly recommended to set the test RF downlink level with the R&S®CMW to minimum -70 dBm, in order to avoid TCP throughput loss due to packet loss and to achieve maximum rates.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
18
TCP-Throughput
Thus, Figure 8 shows in particular, that under weak RF reception conditions with significant packet error rates, e.g. under fading conditions, the TCP throughput will be strongly reduced!
15.84 Mbps WiMAXTM PHY Limit
Figure 8: TCP Throughput limited by MSS and PER, RTT TM
It is obvious that the TCP throughput not only depends on the mobile WiMAX MAC and PHY parameters, but on upper layer parameters as well. Reference measurements with the R&S®CMW - simulating upper layer parameter variations confirm those dependencies.
3.1.3 Upstream Bandwidth Asymmetric communication systems, like the mobile WiMAX™ TDD air interface gain more DL capacity by saving UL capacity. The TCP protocol sends ACKs which require some UL bandwidth, hence, the throughput can be narrowed if the UL capacity is too small. However, using the default downlink and uplink symbol distributions according to Table 2, there is always sufficient uplink capacity.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
19
Simulation and Prototype Results UDP/ICMP Throughput Measurement Results
4 Simulation and Prototype Results The following two chapters provide UDP/ICMP and TCP/FTP-Throughput TM measurement results in a network simulation setup and prototype WiMAX setup working with the R&S®CMW270 or R&S®CMW500 communication tester.
4.1 UDP/ICMP Throughput Measurement Results As already explained, the UDP (and ICMP) protocol is independent of the RTT due to its connection-less behavior, i.e. due to abdication of reverse link acknowledgement signaling. Hence, the the reference measurements are done using all available 2 modulation types and FEC rates with iPerf [6] and a R&S®CMW270 or R&S®CMW500. The results are plotted in Figure 9 resp. Table 7 and match the previous calculations. As discussed before, with increasing order of QAM modulation, and decreasing FEC coding rate, there is a linear increase of throughput. The almost negligible losses of the UDP and ICMP rates with respect to the reference rates given by Table 5 are caused by the MAC and IP layer signalling overhead only.
Figure 9: UDP/ICMP Throughput vs. Modulation All UDP throughput measurements show a similar result: the UDP-Throughput is approximately 3.2% less than the PHY-Throughput due to the overhead, as evaluated earlier. Additionally the ICMP-Throughput is plotted. ICMP is another connection-less protocol for network maintenance purposes. For instance the ICMP ECHO function (commonly known as the "ping" Test) echoes every Ethernet packet.
2
02
Common software tool to measure IP network peformance [6]
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
20
Simulation and Prototype Results TCP/FTP Throughput Measurement Results
Modulation 64QAM 5/6 64QAM 3/4 64QAM 2/3 64QAM 1/2 16QAM 3/4 16QAM 1/2 QPSK 3/4 QPSK 1/2
UDP Throughput
ICMP Throughput
[Mbps] 15.30 14.10 12.30 9.18 9.21 6.14 4.61 2.98
[Mbps] 15.68 14.11 12.67 9.50 9.50 6.34 4.74 3.17
Table 7: Maximum UDP/ICMP throughput measurement results
4.2 TCP/FTP Throughput Measurement Results TCP Throughput Simulation measurement results are shown in Figure 10 using a 64 KB and 128 KB window size. As the limit depends mainly on the Window Size and TM Round Trip Time, these results can be used as an upper bound in a WiMAX SiSO 10 MHz channel.
“grey“ Theory
Figure 10: Simulated TCP Throughput of a 15.84 MBit channel using different WS Prototype measurements are depicted in Figure 11. These tests match the simulated curves and can be used as reference as well. Exemplarily, maximum TCP throughput measurements (Figure 11, Table 8) have been achieved using prototype mobile WiMAX™ device, which confirm the 64 KB window size simulation results. For the measurements the maximum mobile WiMAX™ bandwidth and 64QAM modulation along with 5/6 FEC coding rate has been used.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
21
Simulation and Prototype Results TCP/FTP Throughput Measurement Results
Measurement range
Figure 11: TCP throughput vs. RTT, WS 64kB RTTset [ms] 38 45 50 55 60 65 70 75 80
TCP-calculatedThroughout [Mbps] 13.797 11.651 10.486 9.533 8.738 8.066 7.490 6.991 6.554
TCP-measuredThroughput [Mbps] 10.680 9.850 9.230 8.280 7.840 7.150 6.610 6.280 6.120
Table 8: TCP throughput vs. RTT, WS 64 KB
Note: All throughput measurements were done with limited bandwidth inputs, since some prototypes dropped the connection if the input was higher than the available channel capacity. If this is the case, limit the input bandwidth to the PHY bandwidth.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
22
Test Setup TCP/FTP Throughput Measurement Results
5 Test Setup This chapter describes two different throughput measurement test setups. The most common setup A, using two PCs and one R&S®CMW270 or R&S®CMW500 communication tester as a mobile WiMAX™ base station emulator (BSE) is explained in detail first. Then, it is shown how to come along with one additional PC (setup B) only. Table 9 depicts required HW components for all discussed setups. Setup A 1 1 1 1 1 1
R&S®CMW Server PC Client PC Ethernet Cable RF Cable DUT
Setup B 1 1 1 1 1
Table 9: Hardware Requirements However, before the setups will be discussed, there is a need to explain the HW composition of the R&S®CMW270 or R&S®CMW500. The instrument, according to Figure 12, is composed by a Windows™ PC (acting as instrument controller) and a Power PC (PPC). The latter one performs the mobile WiMAX™ protocol stack.
2 x Rear LAN switch R&S®CMW
Ethernet Switch IPPPC WiMAX air I/F
WiMAX stack (PowerPC) B
Instrument Controller (Windows PC)
IPFMR
LAN Front
Figure 12: R&S®CMW HW composition
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
23
Test Setup Test Setup A
5.1 Test Setup A The Server PC and the R&S®CMW are connected by an Ethernet cable. The Client PC holds the mobile WIMAX™ device under test (e.g. USB stick or PCMCIA card). The DUT is attached to the CMW RF1 COM front panel connector.
Figure 13: Setup A Logic This setup includes: • • • • •
Server and client PC R&S®CMW270 or CMW500 including options CMW-B660A and CMW-B661A DUT RF cable Ethernet cable
Setup Hardware as shown in Figure 14: 1. Connect the Server Ethernet Interface to the instruments rear panel LAN switch 2. Connect DUT to RF1 COM 3. Connect DUT (PCMCIA or USB) to Client
Figure 14: Setup A
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
24
Test Setup Test Setup A
Table 10 shows the network configuration for this test. Please note that this IP configuration is only an example! It is highly recommended to verify your local network configuration for difference in order to avoid network disturbances.
IP Subnet Mask
Server PC 100.100.100.91 255.255.255.0
CMW PPC 100.100.100.60 255.255.255.0
Client PC 100.100.100.11 255.255.255.0
Table 10: Setup A IP settings To configure the network IP address of the Server and the DUT use the TCP/IP properties in the Windows Network Connections Setup menu, depicted in Figure 15.
Figure 15: Server PC TCP/IP Properties Set the R&S®CMW power PC IP and the DUT IP in the Configuration Parameters as shown in Figure 16. Make sure that the client PC hosting the mobile WiMAX™ device corresponds to the IP address destination.
Figure 16: R&S®CMW IPv4 Interface settings
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
25
Test Setup Test Setup A
Since the server shall communicate with the client (IP 100.100.100.11) via the R&S®CMW PPC (IP 100.100.100.60) it requires an appropriate router table entry. Use the server PC DOS command shell to create this router entry: route add –p 100.100.100.11 100.100.100.60 Note: The DOS command shell command "route print" allows the verification of the router entry added above.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
26
Test Setup Test Setup B
5.2 Test Setup B Setup B according to Figure 17 reduces hardware requirements but still permits all throughput test varieties. However, it is required to install the server on the R&S®CMW270 or R&S®CMW500 communication tester windows PC.
Figure 17: Setup B Logic This setup includes: • • • • •
Client PC R&S®CMW270 or CMW500 including CMW-B660A and CMW-B661A options, which provide a switch board to connect an external PC, DUT RF cable Ethernet cable
Setup Hardware as shown in Figure 18 1. Connect front LAN (R&S®CMW, Windows) to rear LAN switch 2. Connect DUT to RF1 COM 3. Connect DUT (PCMCIA or USB) to Client
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
27
Test Setup Additional Information
Figure 18: Setup B Configure the Network as shown in Table 11.
IP Subnet Mask
Server (CMW PC) 100.100.100.91 255.255.255.0
CMW PPC 100.100.100.60 255.255.255.0
Client PC 100.100.100.11 255.255.255.0
Table 11: Setup B IP settings The server shall communicate with the client (IP 100.100.100.11) via the R&S®CMW PPC (IP 100.100.100.60) it requires an appropriate router table entry. Use the server PC (R&S®CMW Windows PC) DOS command shell to create this router entry: route add –p 100.100.100.11 100.100.100.60 Note: The DOS command shell command "route print" allows the verification of the router entry added above.
5.3 Additional Information Check your data output by using a bandwidth meter and your CPU workload while running the throughput tests if you determine problems of extreme low throughput. Latter can be done using the Windows Task Manager (Ctrl + Alt + Del).
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
28
Application Setup R&S®CMW270 or R&S®CMW500 Setup
6 Application Setup The following chapters describe the R&S®CMW setup for maximum DL/ULThroughput tests and different application setups for UDP, TCP, FTP tests, measuring the maximum throughput. Additionally a video-stream setup is provided.
6.1 R&S®CMW270 or R&S®CMW500 Setup 1. Start WiMAX™ Signaling 2. Set Frequency and Bandwidth • Frequency: DUT Frequency • Bandwidth: 10 MHz
Figure 19: WiMAX™ Signaling 3. Open Configuration 4. Maximum DL Throughput Test Settings • DL-Symbols: 35 • IP-Address-Destination: 100.100.100.11 • Modulation Coding Rate: 64QAM(CTC) 5/6
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
29
Application Setup R&S®CMW270 or R&S®CMW500 Setup
Figure 20: WiMAX™ DL Configuration 5. Maximum UL Throughput Test Settings • DL-Symbols: 26 • Slots: 210 • Modulation Coding Rate. 64 QAM 5/6 (note: depends on DUT)
Figure 21: WiMAX™ UL Configuration
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
30
Application Setup UDP-Throughput Test Setup
6.2 UDP-Throughput Test Setup The UDP-Throughput measurement is done with iPerf. To avoid path problems it is recommended to copy iperf.exe to C:\WINNT\system32. In this example the maximum DL measurement in described, please remember to exchange iPerf Server and iPerf Client to measure the maximum UL throughput. 1. Start the CMW and set the described parameters 2. Connect your DUT and check the established connection, for example sending a ping (ping 100.100.100.11) from the Server. Client (iPerf Server operates), Figure 22 1. Start the command shell: Start - Run - cmd.exe 2. Run iPerf in Server mode using a WS of 64kB: iperf –s –u –w 64K
Figure 22: iPerf UDP, Server on Client Server, Figure 23 Start the command shell: Start - Run - cmd.exe 3. Run iPerf in Client mode using a WS of 64kB: iperf –c 100.100.100.11 –b 15.84M –w 64K Note that –b 15.84M indicates the UDP bandwidth. 15.84M refers to maximum DL settings of your WiMAX™ DUT, you might adapt this value.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
31
Application Setup TCP-Throughput Test Setup
Figure 23: iPerf UDP, Client The iPerf Server reports the Client the transferred amount of data. The Client itself reports the transferred bandwidth, which is in this case 15.6Mbps while 15.8Mbps should be send. Anyway, the UDP-Throughput is in this case not 15.6Mbps it is 15.4Mbps, which results in 15.8Mbps adding 3.2% UDP overhead.
Figure 24: iPerf UDP-Throughput Results
6.3 TCP-Throughput Test Setup The TCP-Throughput measurement is done with iPerf. To avoid path problems it is recommended to copy iperf.exe to C:\WINNT\system32 – if this is not possible, copy it to any folder and adapt the path in the command shell. In this example the maximum DL throughput measurement is described, please remember to exchange iPerf Server and iPerf Client to measure the maximum UL throughput. 3. Start the CMW and set the described parameters 4. Connect your DUT and check the established connection, for example sending a ping to the DUT (ping 100.100.100.11) from the Server. Client (iPerf Server operates!), Figure 25 5. Start the command shell: Start - Run - cmd.exe 6. Run iPerf in Server mode using a WS of 64kB: iperf –s –w 64K
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
32
Application Setup TCP-Throughput Test Setup
Figure 25: iPerf TCP Server running on Client Server, Figure 26 7. Start the command shell: Start - Run - cmd.exe 8. Run iPerf in Client mode using a WS of 64kB: iperf –c 100.100.100.11 –w 64K
Figure 26: iPerf TCP Client running on Server The iPerf measurement results are plotted within the command line on the Server and Client. Figure 27 depicts the iPerf Client running on the Server. iPerf measured a TCP throughput of 11.7Mbps for the maximum DL and 3.2Mbps for the maximum UL with a RTT of 31ms and a WS of 64kB. Setup Max DL Settings Max UL Settings
02
FTP DL Throughput [Mbps] 11,7
FTP UL Throughput [Mbps] 3,2
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
33
Application Setup FTP-Throughput Test Setup
Figure 27: iPerf TCP DL Results on Server Changing iPerf Server and Client for the UL throughput measurement and adapting the R&S®CMW270 or R&S®CMW500 settings to “maximum UL” (16QAM ½) results in a UL throughput of 3.2 Mbps. iPerf in TCP mode prints the same throughput results in the Server and Client interface, since TCP is connection oriented.
Figure 28: iPerf TCP UL Results on Server
6.4 FTP-Throughput Test Setup The FTP throughput tests use a FTP Server and FTP Client. The Client establishes a connection to the Server and requests a file. Note that the FTP Server operates on the Server and the connection is established from the client, since the Client requests the data. This example shows how to setup the FTP Server, establish the connection from the client, transfer data and calculate the FTP throughput.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
34
Application Setup FTP-Throughput Test Setup
1. Start the CMW and set the described parameters 2. Connect your DUT and check the established connection, for example sending a ping from the Server to the Client (ping 100.100.100.11). Server In this example the freeware Quick ‘n Easy FTP Server is used, but any FTP Server can be used. Remember that some FTP Server have Upload limitations which results in lower throughput. 3. Start the FTP Server 4. Click Start, Figure 29 5. Share a folder
Figure 29: FTP Server running on Server Now the connection from the DUT to the FTP Server using the Windows build in FTP Client will be established. Client: DL throughput Start command shell on your Client (start – run – cmd.exe) 6. type in ftp -A 7. type in open 100.100.100.91 (Server IP) 8. ls will list your sharing (optional) 9. get test3.zip for example will download the file from the server
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
35
Application Setup FTP-Throughput Test Setup
Figure 30: FTP Client connecting to Server, max DL Client: UL throughput 5. Establish the connection (optional) 6. Type in send test3.zip 7. Set the FTP filename on Client PC, for example test3.zip (optional)
Figure 31: FTP Client, sending file, max UL The FTP Client plots the throughput results in KB/s, therefore this value has to be multiplied by 8. Consequently the FTP throughput results for the Maximum DL settings in kBit/s: Setup Max DL Settings Max UL Settings
FTP DL Throughput [Mbps] 11,28
FTP UL Throughput [Mbps] 0,82 3,12
Table 12: FTP Throughput Results for Maximum DL Settings
Figure 32: FTP Throughput Results for Maximum DL settings
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
36
Application Setup Video-Stream Setup
6.5 Video-Stream Setup VLC media player is used for video-streaming. The Server broadcasts a video file while the client connects to the stream. This chapter explains in detail how to setup the VLC media player (Version vlc-0.9.8a-win32) on the Server and Client to stream a video over a WiMAX™ Air Interface. Server 1. Install VLC media player 2. Open VLC media player and select Media
Streaming (Figure 33)
Figure 33: Server, VLC Stream File 3. Select a video file and Stream (Figure 34)
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
37
Application Setup Video-Stream Setup
Figure 34: Server, VLC Open File 4. Select RTP, insert the Client IP address 100.100.100.11 (Figure 35) 5. Select a profile, for example H264 (Figure 35) 6. Stream transmits the file to the Client
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
38
Application Setup Video-Stream Setup
Figure 35: Server, VLC Stream Output 7. Detailed information, for example about the codec, errors or streaming-rate can be gathered using the VLC console (Figure 36). Therefore, open Tools Add Interface Console (Figure 37).
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
39
Application Setup Video-Stream Setup
Figure 36: Server, VLC Console
Figure 37: Server, VLC Add Interface - Console The video will be streamed to the Client and the generated traffic can be monitored using a Bandwidth Meter or simply by the Windows Network Packet counter. Anyway, to view the video the Client has to be setup. Client 8. Install VLC media player 9. Open VLC media player and select Media
02
Open Network (Figure 38)
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
40
Application Setup Video-Stream Setup
Figure 38: Client, VLC (1) 10. Select the RTP protocol, insert the Client IP address 100.100.100.11 and select play (Figure 39)
Figure 39: Client, VLC (2) 11. The video-stream starts automatically after a short buffering time (Figure 40). Please consider the VLC console for occurring problems.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
41
Application Setup Additional Information
Figure 40: Client, VLC Stream
6.6 Additional Information It is highly recommended to install a tool like Bandwidth Meter (shareware) or BitMeter (freeware) to monitor the traffic of the specific network device (Figure 41).
Figure 41: running Bandwidth Meter 1. Install Bandwidth Meter 2. Right click on icon, go to Adapter and select the used Ethernet interface, Figure 42.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
42
Application Setup Additional Information
Figure 42: select Ethernet interface
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
43
Conclusion Additional Information
7 Conclusion Data throughput measurements are a versatile task and require a versatile test platform, which is truly provided by the R&S®CMW270 or R&S®CMW500 communication tester. This test platform offers all capabilities from physical level data rate measurements up to higher level application related throughput evaluation. It has been outlined that throughput rates are not only limited by radio parameters of the WiMAX™ air interface according to [1].
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
44
Abbreviations Additional Information
8 Abbreviations BS BSE CP CRC DL DUT FDD FEC FFT FTP HTTP ICMP IFFT IP ISO/OSI LAN MAC MIMO MTU OFDM OFDMA PDU PER PUSC RF QAM QPSK RTT SISO TCP TDD UDP UL WiMAX WS
02
Base Station BS Emulator Cyclic Prefix Cyclic Redundancy Check Downlink Device Under Test Frequency Division Duplex Forward Error Coding Fast Fourier Transform File Transfer Protocol Hypertext Transfer Protocol Internet Control Message Protocol Inverse FFT Internet Protocol International Standardisations Organisation/Open System Interconnecton Local Area Network Medium Access Control Multiple Input Multiple Output Maximum Transmission Unit Orthogonal Frequency Division Multiplex Orthogonal Frequency Division Multiple Access Protocol Data Unit Packet Error Rate Partical usage of sub channelisation Radio Frequency Quadrature Amplitude Modulation Quadrature Phase Shift Keying Round Trip Time Single Input Single Output Transport Control Protocol Time Division Duplex User Datagram Protocol Uplink Worldwide Interoperability for Microwave Access Window Size
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
45
Literature Additional Information
9 Literature [1] [2]
[3] [4]
[5]
[6] [7] [8] [9] [10]
02
IEEE 802.16™ Air Interface for Broadband Wireless Access Systems Source: www.wimaxforum.org WiMAX Forum™ Mobile System Profile Source: www.wimaxform.org Neues von Rohde & Schwarz, „IP-basierte Applikationstests an mobilen WiMAX™-Endgeräten“, Nr. 199 / 2009 Mathis, M., Semke, J., Mahdavi, J. and T. Ott, "The Macroscopic Behavior of the TCP Congestion Avoidance Algorithm", Computer Communication Review, Vol. 27, number 3, July 1997 Padhye, J., Firoiu, V., Towsley, D. and J. Kurose, "Modeling TCP Throughput: a Simple Model and its Empirical Validation", UMASS CMPSCI Tech Report TR98-008, February 1998 iPerf, NLANR/DAST Source: iperf.sourceforge.net Quick ‘n Easy FTP Server, “FTP Server is a multi threaded FTP server for Windows”, Source: www.pablosoftwaresolutions.com VLC Media Player, “The cross-platform media player and streaming server” Source: www.videolan.org Bandwidth Meter, “Software for bandwidth usage monitoring, reporting, and notification”, Source: www.bandwidth-meter.net Heuel, S., “Development of an optimized data throughput measurement technique for OFDMA WiMAX interfaces”, Rohde & Schwarz and University of Siegen, 2009
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
46
Additional Information Additional Information
10 Additional Information This application note is likely to be extended for future data throughput applications. Please visit out web site www.rohde-schwarz.com in order to download updated or related application notes. Please send any comments or suggestions about this application note to
[email protected]. Further information on R&S®CMW270 or R&S®CMW500 can be obtained at www.wimax.rohde-schwarz.com.
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
47
Ordering Information Additional Information
11 Ordering Information Ordering Information Communication Testers Designation
Type
Order No.
Selection: Wideband Communication Tester
R&S®CMW500
1201.0002K50
Mainframefor CMW500 frequency range 70 MHz to 3,3 GHz
R&S®CMW-PS502
1202.5408.02
Selection: Front Panel without Display/Keypad (contains DVI interface)
R&S®CMW-S600A
1201.0102.02
Selection: Front Panel with Display/Keypad
R&S®CMW-S600B
1201.0102.03
Selection: Wireless Connectivity Tester
R&S®CMW270
1201.0002K75
Mainframefor CMW270 frequency range 70 MHz to 6 GHz
R&S®CMW-PS272
1202.9303.02
Selection: Front Panel without Display/Keypad (contains DVI interface)
R&S®CMW-S600C
1201.0102.04
Selection: Front Panel with Display/Keypad
R&S®CMW-S600D
1201.0102.05
Designation
Type
Order No.
RF Frontend Module
R&S®CMW-S590A
1202.5108.02
Signaling Unit, Universal
R&S®CMW-B200A
1202.6104.02
WiMAX™ Extension Module for R&S®CMW-B200A Option
R&S®CMW-B270A
1202.6504.02
Option Carrier for Ethernet Switch Board
R&S®CMW-B660A
1202.7000.02
Ethernet Switch Board
R&S®CMW-B661A
1202.7100.02
Application Enabler, extension convergence sublayer, IPv4
R&S®CMW-KA700
1202.6904.02
TX Measurement, Mobile WiMAX™ (IEEE 802.16e)
R&S®CMW-KM700
1202.6604.02
TX Measurement, Mobile WiMAX™ (graphical results)
R&S®CMW-KM701
1202.6610.02
Signaling, Mobile WiMAX™ (IEEE 802.16e), SISO
R&S®CMW-KS700
1202.6710.02
Message Analyzer, Mobile WiMAX™ (IEEE 802.16e), online
R&S®CMW-KT700
1202.6804.02
Designation
Type
Order No.
RF Converter Module (TRX)
R&S®CMW-B570B
1202.8659.03
RF Frontend Module
R&S®CMW-B590A
1202.8707.02
Signaling, Mobile WiMAX™ (IEEE 802.16e), R&D extension
R&S®CMW-KS701
1202.6710.02
Signaling, Mobile WiMAX™ (IEEE 802.16e), MIMO extension
R&S®CMW-KS702
1202.6640.02
Frequency Range 3.3 GHz to 6 GHz
R&S®CMW-KB036
1203.0851.02
Required options
Options for second channel
Note: minimum requirements marked
02
Rohde & Schwarz © Mobile WiMAX™ Throughput Measurements
48
About Rohde & Schwarz Rohde & Schwarz is an independent group of companies specializing in electronics. It is a leading supplier of solutions in the fields of test and measurement, broadcasting, radio monitoring and radio location, as well as secure communications. Established 75 years ago, Rohde & Schwarz has a global presence and a dedicated service network in over 70 countries. Company headquarters are in Munich, Germany. Regional contact Europe, Africa, Middle East +49-1805-124242* or +49-89-4129-13774
[email protected] North America 1-888-TEST-RSA (1-888-837-8772)
[email protected] Latin America +1-410-910-7988
[email protected] Asia/Pacific +65-65-13-04-88
[email protected]
This application note and the supplied program may only be used subject to the conditions of use set forth in the download area of the Rohde & Schwarz website.
ROHDE & SCHWARZ GmbH & Co. KG Mühldorfstraße 15 | D - 81671 München Phone + 49 89 4129 - 0 | Fax + 49 89 4129 – 13777 www.rohde-schwarz.com
