Theoretical Maximum Throughput Of IEEE 802.11 And Its Applications
Theoretical Maximum Throughput of IEEE 802.11 and its Applications Jangeun Jun, Pushkin Peddabachagari, Mihail SichitiuDepartment of Electrical and Computer EngineeringNorth Carolina State UniversityRaleigh, NC 27695-7911{jjun, ppeddab, mlsichit}@eos.ncsu.eduAbstractThe goal of this paper is to present exact formulae forthe throughput of IEEE 802.11 networks in the absence oftransmission errors and for various physical layers, datarates and packet sizes. Calculation of the throughput ismore than a simple exercise. It is a mandatory part of provisioning any system based on 802.11 technology (whether inad-hoc or infrastructure mode). We will discuss the practical importance of theoretical maximum throughput andpresent several applications.1. IntroductionIEEE 802.11 networks are currently the most popularwireless local area network (WLAN) products on the market. The technology has matured, the prices have comedown significantly in the past couple of years and the products fulfill clear needs of many classes of consumers.End consumers use IEEE 802.11 products for mobile networking both in the residential and business markets, enjoying untethered Internet access. Internet ServiceProviders, realizing the significant cost savings that wirelesslinks offer when compared to classical access techniques(cable and xDSL), embraced the technology as an alternative for providing last mile broadband Internet access. Various companies are using IEEE 802.11 off-the-shelf products to provide wireless data access to devices without aneed for special cabling, e.g. remote surveillance cameras,cordless speakers, etc. WLANs make it possible to networkhistorical buildings where it is impossible or impractical touse cables. Researchers in ad-hoc networking are finallyoffered a high data rate, reliable, low cost implementationradio interface for their testbeds.One of the most common misconceptions about 802.11bis that the throughput is 11 Mbps. However, the 11 Mbps so This work was supported by the Center for Advanced Computing andCommunication.hugely advertised on all IEEE 802.11b products only refersto the radio data rate (of only a part) of the packets. Thethroughput offered to a user of IEEE 802.11 technology issignificantly different. For example with no transmissionerrors and 1460 byte sized packets, the throughput of an“1 Mbps” system is just 6.1 Mbps. The efficiency is significantly lower for smaller packet sizes. The efficiencyof IEEE 802.11 is in sharp contrast to wired technologieswhere, for example, a 10 Mbps Ethernet (802.3) link offersthe users almost 10 Mbps.The main contribution of this paper is the exact calculation of the theoretical maximum throughput for 802.11networks, for a variety of technologies (802.11, 802.11b,802.11a) and data rates. All of the information for the calculation of these data rates is available in the IEEE standards [1–3]. However, actually doing it is a laborious procedure requiring data gathering from various standards anda thorough understanding of the mechanisms presented inthe standard. By publishing the calculations in this paper,we hope to spare other research teams and system designers the tedium of wading through the standards to determinethe theoretical maximum throughput. Referenced publications [4–8], concentrate on the analysis of contention window sizes and qualitative performance of the IEEE 802.11standard.To emphasize the importance of the theoretical maximum throughput, we will present several applicationswhich require knowledge of the maximum throughput ifthey are to be designed correctly. The most common useof 802.11 technology is for LAN data access, and correctlyprovisioning such a network implies more than just providing adequate coverage. The theoretical maximum throughput can be used to facilitate optimal network provisioning,both for data as well as multimedia applications. In the caseof ad-hoc networks, it turns out to be a primary factor influencing topological distribution of nodes.Proceedings of the Second IEEE International Symposium on Network Computing and Applications (NCA’03)0-7695-1938-5/03 17.00 2003 IEEE
2AssumptionsTMTCSMA/CAWe define the upper limit of the throughput that can beachieved in an IEEE 802.11 network as its theoretical maximum throughput (TMT). Since the 802.11 standard covers the medium access control (MAC) and physical layer interms of the OSI reference model [9], we are interested inthe actual throughput provided by the MAC layer. Therefore, the TMT of 802.11 can also be defined as the maximum amount of MAC layer service data units (SDUs) thatcan be transmitted in a time unit. A typical encapsulationbetween application layer and 802.11 is transmission control protocol (TCP) or user datagram protocol (UDP) overthe Internet protocol (IP), over logical link control (LLC).The higher the layer, the lower the maximum throughput ofthat layer, as overhead accumulates at each layer. Also, themaximum throughput at the application layer can be limitedby TCP dynamics as well as overhead due to protocol headers. The effect of TCP dynamics on the maximum throughput is out of the range of this paper. Maximum throughputobserved by an application is described by the followingequation when no fragmentation is involved in the lowerlayers:T M TAP P β T M T802.11 (bps)α β(1)where,T M TAP P is the TMT of the application layer,α is the total overhead above MAC layer,β is the application datagram size andT M T802.11 is the TMT of 802.11 MAC layer.In the rest of this paper the term T M T refers to the TMTof the 802.11 MAC layer (T M T802.11 ), unless explicitlymentioned to the contrary.TMT is defined under the following assumptions: Bit error rate (BER) is zero. There are no losses due to collisions. Point coordination function (PCF) mode is not used. No packet loss occurs due to buffer overflow at the receiving node. Sending node always has sufficient packets to send. The MAC layer does not use fragmentation. Management frames such as beacon and associationframes are not 54MbpsFigure 1. TMT classification based on different MAC and PHY schemes and basic datarates3ClassificationTMT calculation is classified based on different MACschemes, spread spectrum technologies and basic data rates.This classification is required because the standard specifies different values for inter-frame spacing (IFS), minimum contention window size (CWmin ), etc. These parameters substantially affect the calculation of TMT. Although802.11 provides a standard for infrared (IR) medium, weconsider only the radio frequency (RF) medium because IRimplementations are so unpopular.With respect to the MAC schemes, two different sets ofTMTs are calculated - one for CSMA/CA and the other forRTS/CTS. Within those two sets, calculations are groupedbased on different spread spectrum technologies - frequencyhopping spread spectrum (FHSS), direct sequence spreadspectrum (DSSS), high-rate DSSS (HR-DSSS) and orthogonal frequency division multiplexing(OFDM). Finally, theTMT of 802.11 and 802.11b is calculated for different basic data rates - 1 Mbps, 2 Mbps, 5.5 Mbps and 11 Mbps.For 802.11a, mandatory data rates of 6 Mbps, 12 Mbps, 24Mbps and the highest data rate of 54 Mbps are used. All ofthe overheads associated at each sublayer (MAC sublayer,physical layer convergence protocol (PLCP) sublayer andphysical medium dependent (PMD) sublayer) are considered. Fig. 1 illustrates the classification of the presentedTMT calculations.In terms of the OSI reference model [9], IEEE 802.11covers the MAC and PHY layers. The PHY layer is againdivided into a PLCP sublayer and a PMD sublayer. A protocol data unit (PDU) at each layer is defined as the length ofthe transmission unit at that layer including the overhead.A service data unit (SDU) is defined as the length of thepayload that a particular layer provides to the layer above.Therefore, when a higher layer pushes a user packet downto the MAC layer as a MAC SDU (MSDU), overheads occur at each intermediate layer. Fig. 2 shows the type ofProceedings of the Second IEEE International Symposium on Network Computing and Applications (NCA’03)0-7695-1938-5/03 17.00 2003 IEEE
CSMA/CALLCDIFS802.11M HDRMAC SDUFCSMACPLCPPMDMAC PDUPreamblePLCP SDUP HDRBODATASIFSACKDIFSBODATAtimeRepeated cycle of CSMA/CARTS/CTSPLCP PDUIFSBitStream (PMD Repeated cycle of RTS/CTSFigure 2. Overhead at different sublayers ofIEEE 802.11Figure 3. Timing diagram for CSMA/CA andRTS/CTSoverheads added at different sublayers when an MSDU istransmitted through an 802.11 interface. At the MAC layer,the MAC layer header and trailer (FCS) are added beforeand after the MSDU, respectively, and form a MAC PDU(MPDU). Similarly, the PLCP preamble and PLCP headerare attached to the MPDU at the PLCP sublayer. DifferentIFSs are added depending on the type of MPDU. The timeconsumed by 802.11’s backoff scheme cannot be neglected.We will consider the IFS and the backoff duration as overhead at the PMD layer.4In order to calculate the TMT, we first convert all of theoverheads at each sublayer into a common unit - time. Toobtain the maximum throughput, we will divide the MACSDU by the time it takes to transmit it:M SDU sizeDelay per M SDUDelay per M SDU (TDIF S TSIF S TBO TRT S TCT S TACK TDAT A ) 10 6 s.(3)The total delay per MSDU is simplified to a function ofthe MSDU size in bytes, x as:Calculation of the TMTTMT technologies. The backoff time is selected randomly following a uniform distribution from (0, CWmin ) giving theexpected value of CWmin /2. Table 1 lists the constant andvarying delay components.The total delay per MSDU is calculated as a summationof all the delay components in Table 1 as follows:(2)The data rate is not always the same even within the samePLCP PDU. The data rate of a MAC PDU is determined byits type. Control frames such as RTS, CTS, and ACK arealways transmitted at 1 Mbps for backward compatibility.When FHSS is used, the number of PLCP frame bits mayincrease because of DC-bias suppression scheme. Fig. 3 illustrates how data packets are transmitted. The same patternwill be repeated with a specific cycle when back-to-backtraffic is offered at the transmitting node. The timing diagram is different for CSMA/CA and RTS/CTS. The exactduration of each block varies for different spread spectrumtechnologies and basic data rates.The duration of each delay component was determinedfrom the standards [1–3]. All delay components vary withthe spread spectrum technology but not with the data rate.The transmission time of an MPDU depends on its size anddata rate. The contention window size (CW ) does not increase exponentially since there are no collisions. Thus,CW is always equal to the minimum contention windowsize (CWmin ), which varies with different spread spectrumDelay per M SDU (x) (ax b) 10 6 s.(4)We can get the TMT simply by dividing the number ofbits in MSDU (8x) by the total delay (4). Table 2 showsparameters a and b for the TMT formula:T M T (x) 8x 106 bps.ax b(5)When the MSDU size tends to infinity, the TMT isbounded by:lim T M T (x) x 8 106 bps.a(6)Also, as the data rate tends to infinity, parameter a in (4)tends to zero:lima 0,b b T M T (x) 8x 106 bps,b (7)where b is the sum of all the delay components that are notaffected by the data rate. Existence of such a limit is shownby Xiao et al. [4].The use of the parameters a and b in the calculation of theTMT for OFDM technology is based on the assumption thatthe total delay per MSDU is continuous. In fact, the delay isnot continuous due to the ceiling operation in the formulae.However, the approximation error due to this operation isrelatively small - less than 2% in the worst case.Proceedings of the Second IEEE International Symposium on Network Computing and Applications (NCA’03)0-7695-1938-5/03 17.00 2003 IEEE
Table 1. Delay components for different MAC schemes and spread spectrum FDM-545Constant and varying delay components (10 6 s)TRT S TCT S TACK TDAT ATDIF STSIF 4304442322282242128 33/32 8 (34 M SDU )/1128 33/32 8 (34 M SDU )/2192 8 (34 M SDU )/1192 8 (34 M SDU )/2192 8 (34 M SDU )/5.5192 8 (34 M SDU )/1120 4 (16 6 8 (34 M SDU ))/24 20 4 (16 6 8 (34 M SDU ))/38 20 4 (16 6 8 (34 M SDU ))/96 20 4 (16 6 8 (34 M SDU ))/216 128128505050503434343428 328 310 310 310 310 39 39 39 39 0304304304304442322282242128 33/32 8 (34 M SDU )/1128 33/32 8 (34 M SDU )/2192 8 (34 M SDU )/1192 8 (34 M SDU )/2192 8 (34 M SDU )/5.5192 8 (34 M SDU )/1120 4 (16 6 8 (34 M SDU ))/24 20 4 (16 6 8 (34 M SDU ))/38 20 4 (16 6 8 (34 M SDU ))/96 20 4 (16 6 8 (34 M SDU ))/216 AnalysisIn this section we will analyze the behavior of the TMTand spectral efficiency both for single and multiple transmitter systems.5.1. Analysis of TMTUsing (5), we plotted TMT curves for different MACschemes. Fig. 4 and Fig. 5 depict the variation of TMTas a function of MSDU for the CSMA/CA and RTS/CTS,respectively. In each figure a comparison of different datarate and spread spectrum technologies is presented. Sincethe TMT difference between FHSS and DSSS is negligible,both 1 Mbps and 2 Mbps curves are marked by only one label regardless of the spread spectrum technology. The figures show the curve for an MSDU size up to 4095 bytesbecause the 802.11, 802.11b and 802.11a standards specify that maximum MSDU size is 4095 bytes for FHSS andHR-DSSS, and 8191 bytes for DSSS.1TRT S 20 4 16 6 8 20 52, 36, 28 & 24 for each NDBP SN2TCT S20 4 16 6 8 14 44, 32, 28 & 24 for each NDBP SNDBP SDBP S where NDBP S is 24, 48, 96 and 216 for OFDM-6, OFDM-12,OFDM-24 and OFDM-54, respectively. Also note that TCT S TACK .(M SDU in bytes)Fig. 4 and Fig. 5 show that TMT is quite low comparedto the basic data rate. When basic data rate is 11 Mbps,MSDU is 1500 bytes and RTS/CTS scheme is used, TMTis 4.52 Mbps. TMT is higher for CSMA/CA due to fewercontrol frames, and still only 6.06 Mbps (for 1500 byte MSDUs). Therefore, it is almost impossible to see throughputsof over 6.1 Mbps in real deployments where IP packets carrying TCP segments over 1500 bytes are not common. Furthermore, the slope of the curves shows that the higher thebasic data rate is, the more sensitive TMT is to MSDU size.In other words, performance will be substantially degradedwhen small-sized data packets are transmitted especially forhigh data rates. Fig. 4 and Fig. 5 show that the TMTof higher data rates saturates much later than the TMT oflower data rates.The TMT comparison of 802.11a OFDM and 802.11bHR-DSSS is presented in Fig. 6 and Fig. 7 for CSMA/CAand RTS/CTS, respectively. In order to get a clear comparison, the curves are plotted for only the mandatory datarates and the maximum data rate of 802.11a along with thecurve for 11 Mbps of 802.11b. TMT close to 6 Mbps can beachieved in 802.11a when the data rate is 6 Mbps. 802.11asaturates earlier than 802.11b because of smaller inter framespacing and time slot duration.Proceedings of the Second IEEE International Symposium on Network Computing and Applications (NCA’03)0-7695-1938-5/03 17.00 2003 IEEE
8711Mbps (HR DSSS)6TMT (Mbps)55.5Mbps (HR DSSS)432Mbps (FHSS, DSSS)211Mbps (FHSS, DSSS)005001000150020002500MSDU size (bytes)3000350040004500Figure 5. TMT curve for RTS/CTS - FHSS,DSSS, HR-DSSS454054Mbps (OFDM)3530TMT (Mbps)Table 2. TMT parameters for different MACschemes and spread spectrum technologiesSchemeData Rate abCSMA/CAFHSS1 Mbps8.251179.52 Mbps4.1251039.25DSSS1 Mbps811382 Mbps41002HR-DSSS 5.5 Mbps1.45455 915.4511 Mbps0.72727 890.73OFDM6 Mbps1.33333 223.512 Mbps0.66667 18724 Mbps0.33333 170.7554 Mbps0.14815 159.94RTS/CTSFHSS1 Mbps8.251763.52 Mbps4.1251623.25DSSS1 Mbps818142 Mbps41678HR-DSSS 5.5 Mbps1.45455 1591.4511 Mbps0.72727 1566.73OFDM6 Mbps1.33333 337.512 Mbps0.66667 27324 Mbps0.33333 244.7554 Mbps0.14815 225.942524Mbps (OFDM)201512Mbps (OFDM)1011Mbps (HR DSSS)506Mbps (OFDM)05001000150020002500MSDU size (bytes)3000350040004500Figure 6. TMT curve for CSMA/CA - 11 MbpsHR-DSSS, OFDM9811Mbps (HR DSSS)4073554Mbps (OFDM)305.5Mbps (HR DSSS)5254TMT (Mbps)TMT (Mbps)632Mbps(FHSS, DSSS)224Mbps (OFDM)201512Mbps (OFDM)1101Mbps (FHSS, DSSS)005001000150020002500MSDU size (bytes)30003500400011Mbps (HR DSSS)450050Figure 4. TMT curve for CSMA/CA - FHSS,DSSS, HR-DSSS6Mbps (OFDM)05001000150020002500MSDU size (bytes)3000350040004500Figure 7. TMT curve for RTS/CTS - 11 MbpsHR-DSSS, OFDMProceedings of the Second IEEE International Symposium on Network Computing and Applications (NCA’03)0-7695-1938-5/03 17.00 2003 IEEE
1001001Mbps (FHSS, DSSS)905.5Mbps (HR DSSS)802Mbps (FHSS,DSSS)7060Efficiency (percent)Efficiency (percent)7011Mbps (HR DSSS)5040305.5Mbps (HR DSSS)605011Mbps (HR DSSS)403020201001Mbps (FHSS, DSSS)902Mbps (FHSS, DSSS)801005001000150020002500MSDU size (bytes)30003500400004500Figure 8. Bandwidth efficiency curve forCSMA/CA - FHSS, DSSS, HR-DSSS05001000150020002500MSDU size (bytes)3000350040004500Figure 9. Bandwidth efficiency curve forRTS/CTS - FHSS, DSSS, HR-DSSS1006Mbps (OFDM)5.2. Analysis of bandwidth efficiency9012Mbps (OFDM)54Mbps (OFDM)80where R is the basic data rate.Fig. 8 and Fig. 9 show the bandwidth efficiency forCSMA/CA and RTS/CTS, respectively. From the formula,bandwidth efficiency is inversely proportional to basic datarate. Bandwidth efficiency is only 41% when the data rateis 11 Mbps and RTS/CTS is used, and it is 55% whenCSMA/CA is used. In the bandwidth curves, we can observe the saturation tendency more clear than in the TMTcurve. It is also evident that bandwidth efficiency increasesas MSDU size is increased.The bandwidth efficiency comparison of 802.11a OFDMand 802.11b HR-DSSS is presented in Fig. 10 and Fig. 11for CSMA/CA and RTS/CTS, respectively. The higher datarates are compared in a separate figure and low data ratessuch as 1 Mbps and 2 Mbps with FHSS and DSSS are notincluded. CSMA/CA performs better than RTS/CTS because of less control frames. For the same MAC scheme,802.11a outperforms 802.11b in terms of bandwidth efficiency.24Mbps (OFDM)Efficiency (percent)706011Mbps (HR DSSS)5040302010005001000150020002500MSDU size (bytes)3000350040004500Figure 10. Bandwidth efficiency curve forCSMA/CA - 11 Mbps HR-DSSS, OFDM100906Mbps(OFDM)12Mbps (OFDM)8024Mbps (OFDM)70Efficiency (percent)As a measure of spectral utilization, we define bandwidthefficiency ε:TMT,(8)ε R54Mbps (OFDM)6011Mbps (HR DSSS)504030206Applications100In this section we discuss the practical utility of the TMTcalculations and present an application that uses these values to measure the bandwidth utilization at any given point(on a particular channel) in an 802.11 network.05001000150020002500MSDU size (bytes)3000350040004500Figure 11. Bandwidth efficiency curve forRTS/CTS - 11 Mbps HR-DSSS, OFDMProceedings of the Second IEEE International Symposium on Network Computing and Applications (NCA’03)0-7695-1938-5/03 17.00 2003 IEEE
6.1. Importance of TMT100908070Link utilization (%)TMT is important to researchers as well as system designers. It is a strict barrier that cannot be overcome by anymeans while remaining standard-compliant. It is a numerical upper bound on the throughput given the MAC scheme,spread spectrum technology, basic data rate and packet size.It can be used to derive any one of the parameters that describe the performance of a network (maximum allowableMSDU size, delay, throughput or number of users) giventhe others.TMT can be used in call admission and control procedures for QoS schemes to determine accurate upper boundson available bandwidth. For instance, consider ARME [10]and DIME [11] - protocols that aim to provide throughput guarantees in a wireless LAN based on the Differentiated Services architecture. A node running these protocolswould require the knowledge of current link utilization andthe maximum throughput that can be achieved at any givenpoint in order to perform accurate statistical bandwidth allocation.The knowledge of the bandwidth efficiency curves(Figs. 8-10) enables an application protocol designer to observe the effects of a trade off between the size of the dataunit passed to the MAC layer and the delay in generatingthe data unit on the bandwidth efficiency. This is especiallyuseful to minimize jitter in multimedia applications.As demonstrated in [12], TMT is vital in the estimationof the maximum number of voice channels that can be accommodated in a wireless LAN. Voice and video applications can use TMT to calculate the optimum MSDU size tomaximize throughput and, hence, determine the amount ofbuffering required for a communication link. TMT formulae can be used to validate and check the sanity of networksimulators that model 802.11 protocols.One of the most important aspects of designing the layout of a wireless LAN is provisioning. Extensive trafficmodeling and workload analysis have to be carried out tocorrectly estimate the infrastructure needs of any given location. Over-provisioning in a wireless LAN is just as damaging as under-provisioning as noted in the comprehensivestudy done on a campus-wide wireless network at Dartmouth [13] and also in [14]. They observed that unnecessary handoffs between access points that are placed tooclose to each other result in considerably lower throughput.Also, it is straightforward to see that when we consider anetwork where each node is within the transmission rangeof every other node, the sum of the throughputs achieved byall the nodes in the network cannot exceed the TMT of thenetwork. Thus, the ability to accurately measure the linkutilization at various locations in order to perform provisioning is extremely useful. We have implemented an application called WeNoM (Wireless Network Monitor) that6050403020100050010001500Time (s)200025003000Figure 12. Bandwidth utilization at MobiCom2002does exactly that.6.2. Wireless Network Monitor - WeNoMWeNoM was implemented on a Redhat Linux system using the libpcap library from the tcpdump project [15]. It waslater ported to an Intel StrongArm based HP iPAQ runningFamiliar Linux so that it could be used as a handy mobilenetwork monitor. The source code and documentation forthe application is available [16].The principle behind the application is to use the values presented in Table 2 to calculate the actual transmissiontime of an 802.11 frame given the length of the frame andthe rate at which it was transmitted. WeNoM passively listens to the traffic in the network on a single channel andgathers from each frame the transmission rate, length of theMSDU and the time at which it was received. Using thisdata and the appropriate constants from Table 2 in (4) (Section 4) for the transmission time, an accurate estimate forthe time taken to transmit each frame is obtained. The ratio of the transmission time to the inter-arrival time betweenframes gives the instantaneous link utilization at the placeof measurement. The sensitivity of the measurements canbe controlled by using either a weighted average of the cumulative and instantaneous utilization values, or a runningaverage of the utilization for a certain number of consecutive frames.We used WeNoM to measure the WLAN traffic at theMobiCom 2002 conference at Atlanta for a period of about40 minutes. Given that there were 2 access points and over200 users, one would expect the network to be fairly saturated. The plot in Fig. 12, depicts the bandwidth utilizationtoward the end of the day. One can observe the link utilization decrease as the participants leave the conference.Proceedings of the Second IEEE International Symposium on Network Computing and Applications (NCA’03)0-7695-1938-5/03 17.00 2003 IEEE
7ConclusionIn this paper we presented the calculation of the theoretical maximum throughput of 802.11 networks. To broadenthe applicability of the results, many physical layer andMAC layer variations were considered. To illustrate howto apply the results presented in this paper, we presented anapplication which monitors the link utilization of an 802.11network. We hope that the results of this paper will help researchers and system designers to easily and correctly provision systems based on IEEE 802.11 technology.References[1] Wireless LAN medium access control (MAC) andphysical layer (PHY) specification, IEEE Standard802.11, June 1999.[2] Wireless LAN medium access control (MAC) andphysical layer (PHY) specification: High-speed physical layer extension in the 2.4 GHz band, IEEE Standard 802.11, Sept. 1999.[3] Wireless LAN medium access control (MAC) andphysical layer (PHY) specification: High-speed physical layer in the 5 GHz band, IEEE Standard 802.11,Sept. 1999.[4] Y. Xiao and J. Rosdahl, Throughput and delay limits of IEEE 802.11, IEEE Communications Letters,vol. 6, no. 8, pp. 355–357, Aug. 2002.[5] B. Bing, Measured performance of the IEEE 802.11wireless LAN, in Local Computer Networks - LCN’99, 1999, pp. 34–42.[6] F. Cali, M. Conti, and E. Gregori, IEEE 802.11 wireless LAN: capacity analysis and protocol enhancement, in Proc. of INFOCOM ’98, Seventeenth AnnualJoint Conference of the IEEE Computer and Communications Societies, vol. 1, 1998, pp. 142–149.[7] Y. Tay and K. C. Chua, A capacity analysis for IEEE802.11 MAC protocol, Wireless Networks, vol. 7, pp.159–171, 2001.[8] J. C. Chen and J. M. Gilbert, Measured performance of 5-GHz 802.11a wireless LAN Paper.pdf.[9] A. S. Tanenbaum, Computer Networks, 4th ed.Prentice-Hall, 2002.[10] A. Banchs and X. Perez, Providing throughput guarantees in IEEE 802.11 wireless LAN, in Proc. ofIEEE Wireless Communications and Networking Conference - WCNC2002, vol. 1, 2002, pp. 130–138.[11] A. Banchs, M. Radimirsch, and X. Perez, Assuredand expedited forwarding extensions for IEEE 802.11wireless LAN, in Proc. of the Tenth IEEE International Workshop on Quality of Service, 2002, pp. 237–246.[12] M. Veeraraghavan, N. Cocker, and T. Moors, Supportof voice services in IEEE 802.11 wireless LANs, inProc. of INFOCOM ’01, Twentieth Annual Joint Conference of the IEEE Computer and CommunicationsSocieties, vol. 1, 2001, pp. 488–497.[13] D. Kotz and K. Essien, Analysis of a campus-widewireless network, in Proc. of Mobicom, Sept. 2002.[14] A. Balachandran, G. M. Voelker, P. Bahl, and P. V.Rangan, Characterizing user behavior and networkperformance in a public wireless LAN, in Proc of.ACM SIGMETRICS, vol. 30, no. 1, June 2002.[15] 2,[16] Wirelessnetworkmonitor,http://www4.ncsu.edu/ mlsichit/Software/Wenom/wenom release 1.0.tar.gz, 2002.Proceedings of the Second IEEE International Symposium on Network Computing and Applications (NCA’03)0-7695-1938-5/03 17.00 2003 IEEE
ers the tedium of wading through the standards to determine the theoretical maximum throughput. Referenced publica-tions [4-8], concentrate on the analysis of contention win-dow sizes and qualitative performance of the IEEE 802.11 standard. To emphasize the importance of the theoretical max-imum throughput, we will present several applications
IEEE 3 Park Avenue New York, NY 10016-5997 USA 28 December 2012 IEEE Power and Energy Society IEEE Std 81 -2012 (Revision of IEEE Std 81-1983) Authorized licensed use limited to: Australian National University. Downloaded on July 27,2018 at 14:57:43 UTC from IEEE Xplore. Restrictions apply.File Size: 2MBPage Count: 86Explore furtherIEEE 81-2012 - IEEE Guide for Measuring Earth Resistivity .standards.ieee.org81-2012 - IEEE Guide for Measuring Earth Resistivity .ieeexplore.ieee.orgAn Overview Of The IEEE Standard 81 Fall-Of-Potential .www.agiusa.com(PDF) IEEE Std 80-2000 IEEE Guide for Safety in AC .www.academia.eduTesting and Evaluation of Grounding . - IEEE Web Hostingwww.ewh.ieee.orgRecommended to you b
Standards IEEE 802.1D-2004 for Spanning Tree Protocol IEEE 802.1p for Class of Service IEEE 802.1Q for VLAN Tagging IEEE 802.1s for Multiple Spanning Tree Protocol IEEE 802.1w for Rapid Spanning Tree Protocol IEEE 802.1X for authentication IEEE 802.3 for 10BaseT IEEE 802.3ab for 1000BaseT(X) IEEE 802.3ad for Port Trunk with LACP IEEE 802.3u for .
Signal Processing, IEEE Transactions on IEEE Trans. Signal Process. IEEE Trans. Acoust., Speech, Signal Process.*(1975-1990) IEEE Trans. Audio Electroacoust.* (until 1974) Smart Grid, IEEE Transactions on IEEE Trans. Smart Grid Software Engineering, IEEE Transactions on IEEE Trans. Softw. Eng.
effort to get a much better Verilog standard in IEEE Std 1364-2001. Objective of the IEEE Std 1364-2001 effort The starting point for the IEEE 1364 Working Group for this standard was the feedback received from the IEEE Std 1364-1995 users worldwide. It was clear from the feedback that users wanted improvements in all aspects of the language.File Size: 2MBPage Count: 791Explore furtherIEEE Standard for Verilog Hardware Description Languagestaff.ustc.edu.cn/ songch/download/I IEEE Std 1800 -2012 (Revision of IEEE Std 1800-2009 .www.ece.uah.edu/ gaede/cpe526/20 IEEE Standard for SystemVerilog— Unified Hardware Design .www.fis.agh.edu.pl/ skoczen/hdl/iee Recommended to you b
IEEE 802.1Q—Virtual LANs with port-based VLANs IEEE 802.1X—Port-based authentication VLAN Support IEEE 802.1W—Rapid spanning tree compatibility IEEE 802.3—10BASE-T IEEE 802.3u—100BASE-T IEEE 802.3ab—1000BASE-T IEEE 802.3ac—VLAN tagging IEEE 802.3ad—Link aggregation IEEE
IEEE 1547-2003 IEEE P1032 IEEE 1378-1997 Controls IEEE 2030-2011 IEEE 1676-2010 IEEE C37.1 Communications IEC 61850-6 IEC TR 61850-90-1 & IEEE 1815.1-2015 IEC TR 61850-90-2 Cyber & Physical Security IEEE 1686-2013 IEEE 1402-2000
IEEE 610-1990 IEEE Standard Glossary of Software Engineering Terminology, IEEE, 1990 IEEE 829-2008 IEEE Std 829 IEEE Standard for Software and System Test Documentation, IEEE, 2008 IEEE 1012-2016 IEEE Standard for System, Software, and Hardware
be a concise, but not superficial, treatment on GUI programming. Part III contains information on the features of Python that allow you to accomplish big things with surprisingly little code. In preparing this book the Python documentation atwww.python.orgwas indispensable. This book was composed entirely in LATEX.
