report 07

Upload: manoj212077

Post on 10-Apr-2018

215 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/8/2019 Report 07

    1/30

    1 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    Table of Contents

    CHAPTER1INTRODUCTION ................................ ................................ ................................ .................. 3

    1.0 HIGH PERFORMANCE ................................ ................................ ................................ ............... 4

    Chapter 2 The State of the Standards................................ ................................ ................................ . 5

    2.1 INNOVATIONS IN 802.11N ................................ ................................ ................................ ........ 6

    2.2 802.11n at the RF Layer ................................ ................................ ................................ ............ 7

    2.3 Antenna Diversity ................................ ................................ ................................ ..................... 8

    2.3.1 High Performance 802.11n at the PHY Layer ................................ ................................ ..... 8

    2.4 Channel Bonding ................................ ................................ ................................ ...................... 9

    2.5 More OFDM Subcarriers................................ ................................ ................................ ........... 9

    2.6 Short Guard Interval ................................ ................................ ................................ ............... 10

    2.7 Frame Aggregation and Selective Retransmission ................................ ................................ ... 10

    2.8 Reduced Interframe Spacing................................ ................................ ................................ .. 10

    2.9 DIFFERENCES BETWEEN DRAFT 2.0 AND DRAFT 11.0 ................................ .............................. 11

    2.9.1 Aggregation ................................ ................................ ................................ ..................... 11

    2.9.2 Three Spatial Streams ................................ ................................ ................................ ...... 11

    2.9.3 Coexistence Features ................................ ................................ ................................ ...... 12

    Chapter 3 The State of the Market ................................ ................................ ................................ ... 13

    3.1 802.11N CLIENTS ................................ ................................ ................................ ................... 13

    3.2 Market Size ................................ ................................ ................................ ............................ 13

    3.3 Wireless Only Devices ................................ ................................ ................................ ............ 14

    3.4 INFRASTRUCTURE SIDE ................................ ................................ ................................ .......... 14

    3.4.1 Market Size ................................ ................................ ................................ ..................... 14

    3.4.2 The All-Wireless Edge ................................ ................................ ................................ ...... 15

    Chapter 4 The Business Case for 802.11n ................................ ................................ ........................ 16

    4.1 INCREASED CAPACITY AND PERFORMANCE ................................ ................................ ............ 16

    4.2 Runs All Applications ................................ ................................ ................................ .............. 16

    4.3 Higher User Densities ................................ ................................ ................................ ............. 17

    4.4 Extended Coverage ................................ ................................ ................................ ................ 17

    4.5 LOWER TOTAL COST OF OWNERSHIP ................................ ................................ ..................... 17

    4.6 BATTERY LIFE AND POWER SAVINGS ................................ ................................ ...................... 18

  • 8/8/2019 Report 07

    2/30

    2 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    4.6.1 Better Reception at the Access Point ................................ ................................ ............... 18

    4.6.2 MAC Layer Energy Savings ................................ ................................ ............................... 18

    4.6.3 Spatial Multiplexing Power Save ................................ ................................ ...................... 18

    4.6.4 Power Save Multi Poll ................................ ................................ ................................ ...... 18

    Chapter 5 Issues in 802.11n Network Design................................ ................................ .................... 19

    5.1 DESIGNING THE NETWORK FOR MULTIPATH ................................ ................................ .......... 19

    5.2 Access Point Coverage is Unpredictable ................................ ................................ ................. 19

    5.3 Coverage Planning is Complicated ................................ ................................ .......................... 20

    5.3.1 Microcell ................................ ................................ ................................ ......................... 20

    5.3.2 Virtual Cell ................................ ................................ ................................ ....................... 22

    5.4 MIXED MODE NETWORKS ................................ ................................ ................................ ..... 22

    5.4.1 Airtime Fairness ................................ ................................ ................................ .............. 22

    5.4.2 Fairness Among Clients................................ ................................ ................................ .... 23

    5.4.3 Uplink vs. Downlink fairness ................................ ................................ ............................ 23

    5.5 POWER OVER ETHERNET ................................ ................................ ................................ ........ 23

    5.6 CLIENT DESIGN ISSUES................................ ................................ ................................ ........... 24

    5.7 BAND AND CHANNEL SELECTION ................................ ................................ ........................... 24

    Chapter 6 Real World Experiences ................................ ................................ ................................ .. 28

    6.1 MORRISVILLE STATE COLLEGE: PERVASIVE COVERAGE, HIGH PERFORMANCE ........................ 28

    6.2 HALIFAX HEALTH: 100% RELIABILITY, PREDICTABLE AND TRUSTWORTHY ............................... 28

    6.3 THE WASHINGTON NATIONALS: VERY HIGH USER DENSITY ................................ .................... 29

    Summary ................................ ................................ ................................ ................................ . 29

    References ................................ ................................ ................................ ............................... 30

  • 8/8/2019 Report 07

    3/30

    3 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    CHAPTER1INTRODUCTION In January 2004 IEEE announced that it had formed a new 802.11 Task Group (TGn) to

    develop a new amendment to the 802.11 standard for local-area wireless networks. The real

    data throughput is estimated to reach a theoretical 540 Mbit/s (which may require an even

    higher raw data rate at the physical layer), and should be up to 40 times faster than 802.11b,

    and near 10 times faster than 802.11a or 802.11g. It is projected that 802.11n will also offer a

    better operating distance than current networks.

    There were two competing proposals of the 802.11n standard: WWiSe (World-WideSpectrumEfficiency), backed by companies including Broadcom, and TGn Sync backed by

    Intel and Philips.

    Previous competitors TGnSync, WWiSE, and a third group, MITMOT, said in late July 2005that they would merge their respective proposals as a draft which would be sent to the IEEE

    in September; a final version will be submitted in November. The standardization process is

    expected to be completed by the second half of 2006.

    802.11n builds upon previous 802.11 standards by adding MIMO (multiple-input multiple-output) and orthogonal frequency-division multiplexing (OFDM). MIMO uses multiple

    transmitter and receiver antennas to allow for increased data throughput through spatial

    multiplexing and increased range by exploiting the spatial diversity, perhaps through coding

    schemes like Alamouti coding.

    The Enhanced Wireless Consortium (EWC) was formed to help accelerate the IEEE 802.11ndevelopment process and promote a technology specification for interoperability of next-

    generation wireless local area networking (WLAN) products.

  • 8/8/2019 Report 07

    4/30

    4 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    1.0 HIGH PERFORMANCE

    But 802.11n is more than just another evolutionary step. As Figure 1 shows, it marks the point

    at which wireless LANs can truly equal their wired equivalents in terms of raw performance

    meaning real throughput for TCP applications, not just theoretical data rate. Many users

    already treat wireless as their primary means of connectivity, to such an

    extent that hardware manufacturers have begun to avoid wired Ethernet entirely.

    And with 802.11n now the most common wireless connectivity option in new

    client devices, the number of users relying on wireless increases all the time. When

    combined with Merus technology, 802.11n means that almost any organization can become

    an All-Wireless Enterprise. The move to 802.11n is a big one, but it doesnt have to be

    disruptive. For users upgrading from previous wireless technologies, it is designed to be fully

    backwards-compatible although care must be taken to ensure smooth coexistence

    in networks with mixed client types. For those upgrading straight from wired

    Ethernet, there are technologies that can give 802.11n all the reliability, security

    and scalability that users, applications and enterprises expect from wired

    networks.

  • 8/8/2019 Report 07

    5/30

    5 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    Chapter 2 The State of the Standards

    The final 802.11n standard was ratified by the IEEE in September 2009, but products first

    shipped in 2007. This is because the Wi-Fi Alliance, the group that certifies 802.11 wirelessLANs for interoperability, devised a set of tests around a draft of the standard. The initial

    products were tested to conform to Draft 2.0 of the standard, first published in early 2007.

    The specification has gone through many revisions since then, with the full standard based on

    Draft 11.0. But most of these have been very slight, and the final standard is close enough to

    Draft 2.0 that all products certified as compliant with the earlier draft are also interoperable

    with products based on the final standard. In fact, no retesting is necessary for existing

    equipment. As far as the Wi-Fi Alliance is concerned, 802.11n is 802.11n. That isnt just

    good news for early adopters who deployed 802.11n access points before September 2009.

    Its good news for anyone who bought a Wi-Fi laptop, phone or other device in the last three

    years, as theres a good chance that it included a wireless interface based on 802.11n Draft

    2.0. With no changes necessary to support the final standard, this means that a large and

    growing proportion of the installed client base in many organizations is already 802.11n-

    capable. All thats needed to achieve the near-tenfold increase in data rate offered by 802.11n

    is an upgrade of the wireless infrastructure. Draft 2.0 products will be allowed to claim to be

    fully 802.11n certified now. Edgar Figueroa, executive director, Wi-Fi Alliance

  • 8/8/2019 Report 07

    6/30

    6 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    2.1 INNOVATIONS IN 802.11N

    IEEE 802.11n improves on previous 802.11 systems in many ways, primarily designed to

    increase overall throughput but also boosting range and battery life. Because these

    enhancements affect multiple levels of the networking stack, the real gains in performance

    over 802.11g and 802.11a are even greater than implied by the increased headline data rate.

    The move to 802.11n doesnt just increase the total capacity from 54 Mbps to 300 Mbps; it

    reduces the proportion wasted in signaling overhead and error-correction so that more is

    available to real applications. The improvements in 802.11n can be categorized according to

    which of the three lowest levels of the networking stack they affect as shown in Figure

    2.Together, these increase throughput by a factor of about ten. In wireless networks, the

    lowest level of all is the RF layer, covering radio propagation and other phenomena unique to

    wireless networks. Above that is the PHY layer; roughly analogous to Layer 1 in wired

    Ethernet, though the actual technology is very different. Higher still is the MAC layer,

    corresponding to Layer 2 in wired Ethernet. The similarities between 802.11 and Ethernet are

    clearer here: the original 802.11 MAC was based loosely on that of Ethernet, though many

    changes have been made to deal with the unique characteristics of wireless networks,

    culminating in 802.11n. At Layer 3 and above, wired and wireless are identical. TCP/IP

    applications can treat an 802.11n network exactly as they would treat wired Ethernet.

  • 8/8/2019 Report 07

    7/30

    7 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    2.2 802.11n at the RF Layer

    At the very lowest layer, beneath the traditional OSI or TCP/IP network stack models,

    802.11n uses multiple antennas to improve signal diversity and quality. In an all-802.11n link,

    multiple antennas can be used to send multiple data streams at once, increasing bandwidth.

    They also have the useful side effect of improving reception, meaning that a well-designed

    802.11n access point will improve performance even when used with legacy clients. Spatial

    Multiplexing and MIMO The most well-known innovation in 802.11n is MIMO (multiple-

    input, multiple-output), which uses parallel radio streams: sending multiple signals that each

    encode a different set of data and each travel via a different path, shown in Figure 2. Though

    the standard supports up to four streams, only two are needed to reach a data rate of 300

    Mbps. Most equipment supports two, so Wi-Fi Alliance interoperability tests initially covered

    one or two streams. A system using two separate streams can send up to twice as much data

    as a system that uses only one. The receiver needs to recombine the two streams into one,

    similar to how someone listening to music can hear a different instrument in each ear and then

    recombine them into a tune. Just as the music listener needs two ears to do this, the receiver

    needs to have at least as many antennas as there are spatial streams so that it can tell them

    apart. Each stream needs to be transmitted from a separate antenna too, so the number of

    streams is limited by the number of antennas. To support two streams, both the transmitter

    and receiver need to have two antennas.

  • 8/8/2019 Report 07

    8/30

    8 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    2.3 Antenna Diversity

    Adding more antennas doesnt necessarily add more spatial streams. Many access points available

    today include three antennas even though they only support two spatial streams. The extra antenna

    adds diversity, essentially improving reception for better coverage. Because all 802.11 systems drop

    down to lower data rates when signal strength is poor, this also helps improve data rate at most

    distances from an access point. A receiver with more than one antenna can also perform maximum

    ratio combining, which is essentially echo cancellation. This is needed because a signal can take

    different routes from transmitter to receiver, each copy arriving at a slightly different time. The

    receiver listens to the signal on two or three different antennas and calculates the time difference

    between these echoes, then reassemble the signal so that bits are in the correct order. Antenna

    diversity works better for receivers than transmitters, as the transmitter has no easy way to know how

    multipath interference will affect the signal it is sending. And while multiple transmitting antennas

    could theoretically increase the strength of a signal, this is prohibited by FCC regulations (and similar

    rules from regulators in other countries) that limit the transmission power allowed from unlicensed

    radios. Because antenna diversity is usually implemented at the access point, the speed boost it offers

    is asymmetric, improving upstream but not downstream performance.

    2.3.1 High Performance 802.11n at the PHY Layer

    At the physical layer, 802.11n uses several innovations to squeeze more data through the airwaves

    while improving reliability.

  • 8/8/2019 Report 07

    9/30

    9 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    2.4 Channel Bonding

    The 802.11a/g standards used channels that were 20 MHz wide. The 802.11n standard offers

    the option of combining two together into a single 40 MHz channel without needing an

    additional radio. By itself, this doubles the available capacity from each radio in an access

    point. Channel bonding is perhaps the simplest technique used in 802.11n: twice as much

    spectrum leads to twice the performance. However, using larger channels means that fewer

    channels are available, an important consideration in network design. There can be particular

    problems when building out networks based on a micro cell architecture in which adjacent

    access points must use non-overlapping channels, especially in the narrow 2.4 GHz band.

    2.5 More OFDM Subcarriers

    The main innovation in 802.11a and 802.11g was the introduction

    OFDM(orthogonalfrequency division multiplexing), a technique that divides the

    availablechannel into many narrowband subcarriers. The advantage is that if many subcarriers

    are used, each one only has to support a relatively low data rate, improving reliability and

    predictability by making the system more resilient to interference and multipath effects. For

    example, 802.11a/g has a maximum data rate of 54 Mbps, achieved by slicing a 20 MHz.

    channel into 52 subcarriers. Each of these subcarriers only has to carry data at just over 1

    Mbps. In 802.11n, the same size channel is divided into 56 subcarriers but the data rate of

    each subcarrier remains the same, increasing the overall throughput by about 8%. Because the

    40 Hz. channel is twice as wide, it can accommodate 114 subcarriers, an increase of 119% as

    shown in Figure 4.

  • 8/8/2019 Report 07

    10/30

    10 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    2.6 Short Guard Interval

    To avoid confusion between multiple signals sent in a series, the 802.11 standards require that

    a short gap be left between them. In 802.11n, there is the option of shortening this gap (guard

    interval) from the 800 ns used in previous versions of the standard to only 400 ns. Because

    the radio spends less time waiting and more time transmitting, more data can be sent.

    High Performance 802.11n at the MAC Layer

    Although 802.11a/g offered physical layer data rates of 54 Mbps, real throughput in many

    networks is less than half that number thanks to protocol overhead. The 802.11n MAC layer

    is designed to be more efficient, leading to potential TCP throughput of around 200 Mbps.

    2.7 Frame Aggregation and Selective Retransmission

    Previous versions of 802.11need to send a header and a beacon before every data frame, then

    check for an acknowledgement (ACK) that the frame has been received. In 802.11n, up to 64

    frames (or 64 Kbytes) can be sent together in a block called an AMPDU (Aggregated

    ACProtocol Data Unit.)

    Each AMPDU is acknowledged with a single ACK (a block ACK), avoiding the need for up to 63

    headers, beacons and ACKs. Because frame loss is common in all forms of 802.11, the block ACK

    can specify exactly which frames it received, enabling retransmission of only those that were lost.

    Once all frames have been received loud and clear, software in the 802.11n device reassembles them

    into the correct order before passing them off to an application. The benefits of frame aggregation

    depend on the applications running. It is most useful for applications in which large amounts of data

    need to be sent at once such as file transfers. It isnt useful in voice, as VoIP requires that a packet be

    sent every few milliseconds to avoid a noticeable gap in conversation. With the relatively low

    bandwidth requirement of voice, there are frequent small packets and no opportunity for aggregation.

    2.8 Reduced Interframe Spacing

    Frame aggregation requires that all frames are being sent to the same client. When an access point

    needs to transmit multiple frames to different clients, it usually leaves a small gap in between during

    which it checks that it still has access to the airwaves (i.e. that no client is about to transmit.) In

    802.11n, access points are given preferential treatment so that they can avoid this check, transmitting

    multiple frames with a shorter space between them. Because Reduced Interframe Spacing requires

  • 8/8/2019 Report 07

    11/30

    11 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    that clients respect the access points right to send multiple frames in quick succession, it only works

    with 802.11n clients and cannot be used in mixed-mode networks that include legacy clients.

    2.9 DIFFERENCES BETWEEN DRAFT 2.0 AND DRAFT 11.0

    The final IEEE 802.11n standard (Draft 11.0) is fully interoperable with equipment based on the

    earlier Draft 2.0 and products certified for Draft 2.0 are able to claim compliance with the

    standard. However, the two are not entirely identical, as shown in Table 1. The major new

    features in Draft 11.0 are:

    2.9.1 Aggregation

    Although many 802.11n devices have supported frame aggregation from the start, it was not included

    in the Wi-Fi Alliances tests covering Draft 2.0. Some early 802.11n client chipsets only include a 32

    Kbyte buffer, which limits the maximum AMPDU size to this rather than the 64 Kbytes of most

    equipment. Because this is a hardware limitation, it is not something that can be changed without

    replacing the network interface card. Such clients can still use all 802.11ns other performance-

    enhancing features.

    2.9.2 Three Spatial Streams

    The 802.11n standard supports up to four spatial streams, though this is optional and has not been

    implemented. With approval of the final standard, the Wi-Fi Alliance will start testing equipment that

    uses three. However, this too will be optional, while all equipment (except phones) must be able to

    support two. Although many access points include three antennas, most support two streams and use

    the extra antenna for increased diversity (better reception), which leads to improved reliability.

    Supporting three streams will require three antennas on clients, which adds to their cost and physical

    size. Transmitting from extra antennas can also increase power consumption, though this is mitigated

    by sending data faster and so needing to transmit for less time.

  • 8/8/2019 Report 07

    12/30

    12 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    2.9.3 Coexistence FeaturesThe final 802.11n standard also includes two extra features aimed at smoother interoperability

    between different forms of 802.11n. At the RF layer, space-time block coding lets single-

    streamdevices (usually phones) join networks without adversely affecting performance for clients able

    to support two or more streams. At the PHY layer, a bonded 40 MHz channel can drop down to a

    single 20 MHz channel if one part of the channel is blocked by interference.

  • 8/8/2019 Report 07

    13/30

    13 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    Chapter 3 The State of the MarketWith products shipping for more than two years, many wireless devices already support 802.11n. The

    Wi-Fi Alliance has already certified more than 600 products as compliant with Draft 2.0 of the

    standard, none of which will need to be recertified. In addition, the final standard means that many

    more products are expected to ship over the next year. These are not just laptops or other clients

    where wireless replaces Ethernet. They include devices such as cell phones and tablets that previously

    have not offered (wired or wireless) LAN connectivity.

    3.1 802.11N CLIENTS

    Many networks that have not yet deployed 802.11n infrastructure already have a significant

    proportion of 802.11n clients among their installed base, thanks to decisions by laptop makers to

    include 802.11n as standard in new computers. This proportion is growing all the time and will be

    spurred by the ratification of 802.11n

    3.2 Market Size

    As Figure 5 shows, the 802.11n client market overtook the legacy 802.11 client market in mid-2009

    even before the standard was officially ratified. Research firm DellOro Group predicts that legacy802.11 shipments will continue to fall rapidly, with all new wireless network interface cards

    supporting 802.11n by 2012. One important factor to note is that client adaptors only include mini-

    PCI, USB and PC cards that are typically included in devices such as laptops or bought separately for

    upgrades. They do not include devices such as cell phones that include 802.11 radios built-in. This is

    why the total size of the market is projected to fall: 802.11n capability will be standard on so many

    devices that separate wireless network interface cards become unnecessary.

  • 8/8/2019 Report 07

    14/30

    14 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    3.3 Wireless Only Devices

    An increasing number of devices now lack wired Ethernet entirely. The Apple MacBook is the best

    example among laptops, but many netbooks are also relying exclusively on wireless network

    connections. Wireless links are practical for classes of client that have never included Ethernet

    connectivity, from cell phones to locator badges. The wireless LAN is extending the network out to

    make it truly ubiquitous, a trend that will be accelerated by 802.11n.

    3.4 INFRASTRUCTURE SIDE

    Meru Networks shipped the first enterprise 802.11n infrastructure products in early 2007. Since then,

    many other enterprise vendors have shipped access points supporting Draft 2.0 of the standard. Take-

    up has been even more rapid in the home market, which has helped to drive client adoption and user

    expectations of high-performance wireless connectivity.

    3.4.1 Market Size

    As figure 6 shows, the enterprise 802.11n access point market continued to grow despite the

    recession. Although the overall market for wireless LAN hardware shrank in 2009, a result of the

    same economic forces that have affected the rest of the networking industry, the growing proportion

    of access points featuring 802.11n has more than compensated for this.

  • 8/8/2019 Report 07

    15/30

    15 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    As with the figures for client devices, the dollar numbers on the access point market do not tell the

    whole story. This is because much of the intelligence in most enterprise wireless LAN systems resides

    in controllers and management software rather than access points. The market share of controller-

    based systems is projected to increase as wireless access becomes increasingly critical, as independent

    access points are generally not suitable for large scale networks. The overall size of the enterprise

    wireless LAN market is about twice that for access points alone. Most controllers are not 802.11n-

    specific, so 802.11n controllers do not make up a separate category. However, the move to 802.11n is

    a significant boost to the controller market, as both the higher data rates of 8022.11n and the increased

    demand for wireless access will necessitate faster (or more) controllers.

    3.4.2 The All-Wireless Edge

    DellOro group projects that unlike the market for network interface cards, the market for enterprise

    access points will grow for the foreseeable future. Much of this growth will come at the expense of

    wired networks, as 802.11n allows an increasing number of people to use wireless as their primary

    means of network connectivity. Few organizations will actually rip out their wires. In many cases,

    upgrading to 802.11n will be an alternative to upgrading Ethernet switches and cables, as wireless is

    the mode of access seeing an increase in traffic while wired traffic remains constant or even declines.

    Other organizations will forego cabling in new buildings, choosing instead to rely on wireless. The

    latter strategy is already popular in some industries such as higher education and healthcare, and

    likely to extend elsewhere as the benefits of 802.11n become clearer.

  • 8/8/2019 Report 07

    16/30

    16 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    Chapter 4 The Business Case for 802.11n

    With IT budgets under constant pressure, many organizations are resistant to the idea of upgrading or

    installing a new wireless network. However, the 802.11n market is still growing, as even more

    organizations recognize that it can actually save money. The reason is that when properly designed

    and built, an 802.11n wireless network makes upgrades to edge Ethernet unnecessary.

    4.1 INCREASED CAPACITY AND PERFORMANCE

    The most obvious reason to upgrade to 802.11n is its increased data rate. For the first time, wireless

    networks are faster than their wired counterparts, opening up the possibility of replacing wires inalmost all applications. Many users had already switched away from Ethernet to wireless links based

    on older versions of 802.11, but this often meant compromising performance. With 802.11n, such

    compromises are no longer necessary. In tests conducted by independent research firm Novarum1 a

    single 802.11n radio link using a Meru access point consistently offered real TCP throughput of more

    than 190 Mbps about twice that of switched Ethernet. In addition to support for more bandwidth-

    hungry applications like video, this high performance enables greater user density and increased

    reliability.

    4.2 Runs All Applications

    Many applications require more bandwidth than 802.11a/g can comfortably provide. While the

    maximum data rate from an 802.11a/g network is typically about 20 Mbps, applications requiring

    lower data rates can still be impacted by insufficient bandwidth. This is because all 802.11 systems

    drop down to lower data rates whenever the signal to noise ratio is low, for example when the client is

    far from an access point or in the presence of interference. With 802.11ns maximum of nearly 200Mbps, there is much more bandwidth to spare, increasing reliability. In addition, antenna diversity

    means that the system is less likely to need to drop to lower data rates in a well-architected network.

    This opens up many new applications such as:

    Electronic medical record systems that need at least 24 Mbps

    High-definition video streaming, important in schools as well as in the home market

  • 8/8/2019 Report 07

    17/30

    17 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    Two-way video communication, used by some hospitals for real-time interpreting between spoken

    language and sign language

    4.3 Higher User Densities

    Even applications that never need to consume more than 10 Mbps can benefit from 802.11n, as the

    greater overall capacity enables more users to share the same access point. The most important such

    example is voice, whose bandwidth requirements rarely exceed a few kbps yet is very demanding in

    terms of latency and quality-of-service. Because each voice client consumes less airtime using

    802.11n, networks using 802.11n can support more users per access point. They can also offer each

    packet a high quality of service while ensuring minimal impact on data applications. In tests by

    Novarum, a single 802.11n radio on a Meru access point comfortably handled ten legacy voice clients

    with each call consistently sounding better than toll quality, all while maintaining high speeds for data

    users. When other vendors access points were similarly loaded, voice was unintelligible or silentaltogether.

    4.4 Extended Coverage

    Equally important for many applications, an 802.11n network improves range and coverage compared

    to legacy 802.11g and 802.11a. This is due mostly to the multiple antenna systems used at both ends

    of the wireless link, and means that an 802.11n network can offer higher performance even when used

    with legacy equipment. Because the data rate in 802.11 systems depends on the clients distance from

    the access point, improving range can also improve capacity. The area over which the maximum data

    rate is available will generally be greater in an 802.11n system than in a legacy system.

    4.5 LOWER TOTAL COST OF OWNERSHIP

    According to independent consultancy Network Strategy Partners2, an 802.11n wireless LAN based

    on Meru technology generally costs less than one third of a wired Ethernet system that provides

    similar capacity. For example, in an office measuring 250,000 square feet, wireless LAN network

    infrastructure would cost 57 cents per square foot compared to $2.07 per square foot for wired

    Ethernet. These savings have always been possible. However, previous wireless technologies could

    not always match the performance and predictability of wired Ethernet, meaning that a direct wired

    vs. wireless comparison was not possible for organizations with very demanding applications.

    Wireless was effectively a drain on resources an extra edge network that had to be maintained andsupported in addition to wired Ethernet. The increased performance of 802.11n changes things.

    Wireless becomes a cost saver, able to rival and then replace Ethernet. Enterprises can move to

    wireless confident that all their existing applications will run as well as on a wired network, saving

    money while users gain the added benefits of mobility.

  • 8/8/2019 Report 07

    18/30

    18 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    4.6 BATTERY LIFE AND POWER SAVINGSThe extended range of an 802.11n network cuts power consumption compared to a legacy network,

    reducing cost and environmental impact at the same time as increasing the uptime offered from a

    UPS system of a given capacity. But the most significant energy-saving benefits are on the client side.

    For a device to be truly wireless, it needs to be able to run without a power cord for an extended

    period of time so battery life is critically important. Faster Transmissions to Clients The faster a

    client can transmit data, the less time it needs to spend transmitting and thus the longer its radio can

    spend switched off. Higher data rates for downstream (access point to client) traffic also help, as

    actively receiving data uses more than passively waiting for it.

    4.6.1 Better Reception at the Access Point

    Many access points include extra antennas for increased gain through maximal ratio combining. This

    allows the client to transmit at a lower power, further reducing power consumption.

    4.6.2 MAC Layer Energy Savings

    In addition to speed improvements, the 802.11n MAC includes two mechanisms directly designed to

    reduce power consumption, increasing client battery life and reliability.

    4.6.3 Spatial Multiplexing Power Save

    Using multiple antennas can increase the power drain compared to using just one, even if they are set

    to listen only. Spatial Multiplexing Power Save allows the system to turn off all but one antenna so

    that the client can still listen for transmissions without wasting energy. The antennas that are powered

    down can rapidly switch back on when needed.

    4.6.4 Power Save Multi Poll

    If a client knows that it will not be receiving a transmission for a given amount of time, it can switch

    off all its radios and antennas until that time for dramatic power savings. Previous 802.11 standards

    were based on random access to the airwaves, with one device having no way of knowing when

    another would try to contact it. In 802.11n, clients can reserve a transmission at a specific time,

    making the transmission process more predictable and allowing them to switch off.

  • 8/8/2019 Report 07

    19/30

    19 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    Chapter 5 Issues in 802.11n Network DesignBecause 802.11n is so different from previous wireless technologies, network architecture principles

    that worked for 802.11a/b/g cannot necessarily be applied to 802.11n. Multipath effects and MIMO

    change signal propagation dramatically, making planning more difficult. The need for backward

    compatibility complicates things, as many networks will need to ensure that older clients are

    supported without slowing down 802.11n clients. The increased demands on 802.11n networks also

    entail new considerations in design and planning. With wireless a critical, primary network

    connection, it needs to match Ethernet in all ways, not just performance. While 802.11n itself

    provides increased wirelike speed, other techniques are needed to ensure that the network offers

    wirelike stability, security and scalability.

    5.1 DESIGNING THE NETWORK FOR MULTIPATH

    Multipath effects are a result of the different routes that a signal can take between a transmitter and

    receiver. They have traditionally been regarded as a problem because signals that take different routes

    can interfere with each other. MIMO turns them from a bug into a feature by using the multiple paths

    to carry different streams of data. However, multipath effects can still cause issues. Because paths

    depend on obstacles between the transmitter, they are difficult to predict and fluctuate from moment

    to moment as people or objects move. This makes network design much more complicated.

    5.2 Access Point Coverage is Unpredictable

    In 802.11a/b/g networks, maps showing the radio footprint of each access point are relatively simple.

    An access points coverage area is a contiguous blob, with stronger signals available closer to the

    access point. Coverage can be represented as a series of concentric circles emanating from each of the

    APs, with higher data rate circles closer to the AP not an entirely accurate representation, but oneclose enough for rough planning purposes. The main factor affecting coverage is distance from access

    point. Some types of obstacles can partially block radio waves, but these simply reduce the coverage

    area. In 802.11n, coverage is much less predictable. It depends on the way in which radio waves

    reflect off, refract through or diffract around obstacles such as walls, cube dividers and even people.

    Figure 7 shows the actual coverage of an 802.11n AP in an office building, higher data rates indicated

    by darker shades. Some areas of high performance coverage are very far away from the AP and not

    contiguous with those closer to it.

  • 8/8/2019 Report 07

    20/30

    20 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    5.3 Coverage Planning is Complicated

    To ensure that a wireless network is available over a large area, the radio footprints of access points

    need to overlap so that clients can move from one to another without any interruption.However,

    overlapping coverage areas usually cause interference, regardless of which version of 802.11 thenetwork is using. There are two different ways to solve this problem:

    5.3.1 Microcell

    Most 802.11a/b/g networks still use a microcell architecture, so-called because it is essentially a

    scaled-down version of the design pattern used in early cellular systems. Each access point is tuned to

    a different channel from its neighbors to avoid interference. The advantage of this approach is that it

    allows a network to be deployed with little coordination between access points, an important

    consideration when they were standalone devices that had to be managed independently. The

    traditional disadvantages are that it requires a lot of radio spectrum at least three non-overlapping

    channels and that it forces clients to retune to a new channel as they move between cells, guessing

  • 8/8/2019 Report 07

    21/30

    21 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    for themselves which access point they should connect to.

    The move to 802.11n highlights another problem of the microcell architecture: the difficulty in

    planning the coverage pattern. As Figure 8 illustrates, three channels are sufficient to provide wide

    area coverage without interference in a network where all access points have a roughly circular

    coverage area such as one based on 802.11a/b/g. If the same AP placement is tried with an 802.11n

    network, it cannot operate at full power because the extended range results in increased interference.

    Turning down the power introduces dead zones areas with no signal because the spiky coverage

    areas are harder to fit together. Trying to fill these dead zones with new access points causes more

    interference, forcing a new channel plan to be created in which some access points power output is

    further reduced, perhaps causing new dead zones. Even introducing more channels is not always

    helpful, as shown in Figure 9.

  • 8/8/2019 Report 07

    22/30

    22 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    5.3.2 Virtual Cell

    A Virtual Cell tightly coordinates the transmissions of adjacent access points so that all can use the

    same channel (or the same channels, as most modern access points have more than one radio.) The

    advantages of this approach are that less radio spectrum is needed for each later of coverage, as in the

    single-channel architecture of 3G and CDMA cellular networks, and that client devices are not

    responsible for connectivity decisions. As with 3G and CMA networks, this architecture lets multiple

    channels be used in the same physical space so that multiple Virtual Cells can coexist. The difference

    is that whereas 3G and CDMA use the multiple channels to support multiple competing operators, the

    Virtual Cell uses them to add capacity and redundancy to the network for increased performance and

    reliability, an architecture known as Channel Layering. The disadvantage is that it requires a lot of

    intelligence within the network. It is currently only available from Meru Networks. Although initially

    used with 802.11a/b/g, the Virtual Cells advantages are greatest in an 802.11n network. This is

    because it avoids the problems caused by co-channel interference automatically, allowing all access

    points to operate at full power. A dead zone can be filled by a new access point without causing

    problems for others. Rather than causing problems for each other, adjacent access points augment

    each other like light bulbs of the same color.

    5.4 MIXED MODE NETWORKSOne important feature of 802.11n is full backward compatibility with previous wireless standards:

    legacy clients can connect to an 802.11n network, and the improved radio reception enabled by

    antenna diversity means that they should perform at least as well (if not better) than when connected

    to a legacy network. However, supporting legacy clients has its drawbacks. The diversity of wireless

    devices and drivers already causes issues with existing 802.11g networks, which can sometimes be

    held back by 802.11b clients. If the network is not well-designed, a single legacy client can slow

    down the network for all users. This issue is likely to get worse with 802.11n. The standard offers so

    many options that the difference between the fastest and slowest client is much more dramatic than in

    802.11g. Data rates range all the way from 300 Mbps down to 1 Mbps. Because slower clients take

    longer than faster clients to send the same amount of data, they tend to dominate networks that allow

    random access. If all clients are allowed an equal chance to send a packet, they send equal numbers ofpackets on average and so an 802.11n network could spend nearly all its time listening to slow

    802.11b transmissions. The worst client dominates the airwaves and the performance of the entire

    network suffers due to the slower clients.

    There are two effective ways of dealingwith this: Airtime Fairness and Channel Stacking .

    5.4.1 Airtime Fairness

    Airtime fairness is based on bit fairness, a concept originally proposed in the 1980s for wired

    networks. Now implemented extensively in switched Ethernet, bit fairness ensures that all stations

    connected to a network receive an equal share of the networks capacity. Wireless networks are morecomplex because different devices have different data rates and different loss rates. This is true even

    in single-mode networks: One 802.11n client connected to a network might see no packet loss and

    experience a full 200 Mbps, while another might be in a more noisy environment and suffer a lot of

    packet loss. The unicast MAC layer of 802.11n automatically retransmits lost packets so that

    applications dont notice them, but these retransmissions take time, lowering the data rate. For

    example, if half the packets are lost, the data rate would fall by about a third So that slow clients do

    not hog the airwaves, fairness in wireless networks is better measured by time than data quantity.

    Instead of letting each client at a particular QoS level transmit the same amount of data, each is given

  • 8/8/2019 Report 07

    23/30

    23 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    access to the airwaves for the same amount of time. Thus, a fast 802.11n client can transmit about ten

    times as much data as a slower 802.11g client, just as each would in an all-802.11n or all-802.11g

    network. Meru Networks first introduced Airtime Fairness into production wireless LANs in 2003,

    used to prevent a legacy 802.11b device from taking over a channel in 802.11g networks. The

    introduction of 802.11n makes it even more critical. Although its benefits are most important in

    mixed-mode networks that combine 802.11n with legacy clients, Airtime Fairness is needed in every

    wireless network. It enables reliable service in two other ways:

    5.4.2 Fairness Among Clients

    If multiple 802.11n clients need to send or receive traffic, all should receive the same amount of

    airtime. This ensures predictable data rates: There is no use having a 200 Mbps network if one client

    takes all the bandwidth and leaves others with nothing. The important metric here is the minimum

    data rate of an 802.11n client. Ideally, this should be as close as possible to the average (and to the

    maximum.)

    For example, Novarum tests measured a total data rate of 180 Mbps when ten clients were connected

    to a Meru access point, making the average throughput 18 Mbps. The minimum was 15 Mbps so the

    network was fair and predictable. With another vendors access point, the total throughput was 140

    Mbps but this was distributed very unfairly. Instead of each client getting roughly 14 Mbps, some got

    a lot more and one was denied any bandwidth at all.

    5.4.3 Uplink vs.Downlink fairness

    As the number of devices connected to a network increases, the number of clients contending for

    uplink capacity far exceeds the number of access points contending for downlink capacity. WLAN

    infrastructure must be sophisticated enough to manage this ratio, ensuring that an access point is given

    enough airtime to transmit packets to all clients. A system that gives the access point the same airtime

    as an individual client will result in much lower downlink than uplink data rates.

    5.5 POWER OVER ETHERNET

    Most 802.11n access points incorporate multiple radios, each of which transmits on more than one

    spatial stream. Because of the multiple transmissions, an 802.11n access point requires more power

    than an 802.11a/b/g access point for the same number of radios, resulting in potential power supply

    issues. Traditionally, wireless access points have been powered through Power Over Ethernet (PoE) to

    avoid the need for a separate cable. The industry standard for PoE is IEEE 802.3af, in which DC

    power is injected by a switch or a separate inline device and carried over a twisted-pair cable. The

    nominal limit for 802.3af is 12.95W, too low for many 802.11n APs. Many vendors will claim that

    their 802.11n access points can be powered by standard 802.3af. However, its important to check the

    detail of these claims. Some may not operate at full power because the maker is assuming that they

    will usually be part of a microcell network in which most access points power needs to be turned

    down anyway. Others may not be able to use all of their antennas or support the maximum number of

    spatial streams when powered by 802.3af.

    For access points that require more power than 802.3af such as those with more than two radios, the

    other options are typically:

    A local DC power supply. This will generally provide the exact amount of power necessary, but it

    requires an AC outlet and means that remote power management is not possible.

  • 8/8/2019 Report 07

    24/30

    24 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    The new 802.3at standard for PoE. This aims to deliver at least 30W, but is not finalized yet and

    requires new LAN switches or injectors.

    Multiplexed 802.3af. This requires an access point that can support more than one cable uplink,

    often for a redundant backend data connection as well as additional power.

    5.6 CLIENT DESIGN ISSUESThe greatest barrier to 802.11n clients is the need for multiple antennas. As well as increasing the

    BoM (bill of materials) for building each client, they also increase the physical size: the antennas used

    to support MIMO must be separated by at least half a wavelength 3 to 6 cm. at the frequencies used

    by 802.11. This is simply too large for many cell phones. As a result, the Wi-Fi Alliance is certifying

    phones for 802.11n support even if they include only one antenna. These do not benefit from the

    performance gains of MIMO but still support other performance improvements such as channel

    bonding and short guard intervals. Though they dont run at 300 Mbps, they offer data rates at least

    twice as high as those of 802.11a/g devices without any increase in size or cost. Without MIMO, the

    maximum data rate is about 150 Mbps. This increase over 802.11/b/g is obviously useful for smart

    phones with data capability, but also brings benefits in networks with many devices used only for

    voice. Though the actual throughput requirement of most codecs is measured in kbps, the high datarate helps voice networks by allowing a greater user density per access point, extending battery life on

    the client and reducing the impact of voice on data users.

    5.7 BAND AND CHANNEL SELECTION

    Previous 802.11n standards were limited to either 2.4 GHz or 5 GHz, giving users no choice of band.

    Either can be used in 802.11n, and not all clients support both. Combined with the need for legacy

    client support, this means many deployments will need to support both.

    2.4 GHz: Limited Channel Availability

    The 2.4 GHz band is overwhelmingly the most popular, used in 802.11g and 802.11b. Itraditionally

    had two huge benefits: lower cost client radios and longer range. The cosdifference in radios is now

    disappearing, but the range difference results from the laws ophysics and so remains. Though 802.11n

    increases the range and data rate at bothfrequencies, a 2.4 GHz signal will always go further than a 5

    GHz signal. The big disadvantage of the 2.4 GHz band is that it is narrow and crowded, with room for

    onlythree non-overlapping 20 MHz. channels. With channel bonding in 802.11n, that number

    isreduced to one a big problem for microcell architectures that require non-overlappinghannels to

    avoid interference. As a result, most microcell vendors recommend that 802.11nonly be deployed at 5

    GHz. The Virtual Cell can support a 40 MHz. channel at 2.4 GHz, allowing full data rate

    deploymentsof 802.11n without interference. Because this only requires two of the three available

    non

    overlapping channels, an 802.11b/g network can also be deployed in the remaining 20 MHz band.

    5 GHz: Dynamic Frequency Selection

    When the 802.11 standards were originally defined, they were restricted to a few channels a2.4 GHz.

    and 5 GHz. Since then, the FCC and other national regulators have opened up morechannels around 5

  • 8/8/2019 Report 07

    25/30

    25 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    GHz, expanding the spectrum available to 802.11 networks. However, manyof these channels are also

    shared with other users, notably radar systems. Radar takes priorityso wireless LANs need to move

    away from a channel when they detect a radar system using ita process called Dynamic Frequency

    Selection. In a July 2008 research note3 analyst firm Gartner found several problems with most

    vendors implementations of DFS. Many access points ship with it switched on by default on all

    channels, even those where it is not required. By using it too much, networks risk becoming

    overwhelmed by frequent changes that can cascade across access points.

    Switched on By Default

    Though the FCC only requires DFS in certain 5 GHz channels, some access points apply it

    everywhere even at 2.4 GHz. Because DFS can degrade performance, customers need to ensure that

    it is switched off except where really needed.

    Ripple Effect

    In networks based on a microcell architecture, retuning one access point can force all of its neighbors

    to retune too because adjacent APs must use non-overlapping frequencies. In turn, all its neighbors

    neighbors will also need to retune, ultimately causing a cascade of changes throughout the network.

    To make these changes, the network management system often attempts to calculate a new channel

    plan in real time, a process that tends to result in coverage holes and increased interference. These

    problems are more likely to occur in 802.11n than legacy networks because of the unpredictable

    coverage caused by multipath effects. Another potential problem with the newly-available channels at

    5 GHz is that not all equipment supports every channel. This applies to both clients and access points,

    so it is critical to check specific channel support when choosing equipment. The newer channels may

    not be a good choice when guest access or other clients not owned by the IT department need tosupported. As in 2.4 GHz, the Virtual Cell architecture avoids most of the problems with DFS

    because it requires only one channel for each layer of coverage. Multiple channels can be used to

    provide multiple layers, either network-wide or only in areas where increased coverage is needed as

    shown in Figure 10.

  • 8/8/2019 Report 07

    26/30

    26 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    Figure 10: Two additional channels are in use in the area on the right, while a third provides network

    wide coverage.

    SECURITY

    The 802.11n specification itself does not define any new security technologies as it is intended mainly

    to boost data rates. However, previous standards such as 802.11i have largely solved the earlier

    security issues found in wireless networks. With the right approach, wireless can actually be more

    trustworthy than wired. To ensure that no holes are left open in the network, 802.1n makes the

    advanced security of 802.11i mandatory. In addition, all Wi-Fi certified 802.11n products Draft 2.0

    and Draft 11.0 alike must support WPA2, the set of interoperability tests covering 802.11i. The

    former assures users that their links are secure; the latter assures buyers that equipment can be secured

    easily. However, this can cause issues when supporting legacy applications.

    WPA2 Mandatory

    All Wi-Fi Certified 802.11n equipment supports Wi-Fi Protected Access 2 and 802.11i, with

    mandatory AES encryption for all secured links. In addition, most equipment aimed at business users

    supports WPA2 Enterprise, with authentication and key exchange via 802.1x. The enterprise version

    is more secure than the home version for two reasons: strong per-user authentication and new keys

    generated on the fly. Without WPA2 Enterprise, a permanent key is generated and shared between

    many users, making the key more likely to be compromised. Because 802.11n deliberately drops

    support for the older RC4 encryption algorithm, if offers only two encryption modes:

    CLEARTraffic is not encrypted. This is useful in guest access and open networks. Organizations and end

    users can still encrypt data using software-based techniques such as TLS or IPsec VPNs, but

    encryption is not provided by the wireless network itself.

  • 8/8/2019 Report 07

    27/30

    27 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    AESEncrypted

    AES is the algorithm specified in 802.11i. Encryption is completely transparent to the user and to the

    IT department, while authentication can normally use the same client- and server- side software as on

    the wired network.

    Supporting Legacy Clients

    Most newer clients support AES, as do all Wi-FI certified 802.11n devices. However, manylegacy

    applications and devices do not. If a device does not support AES, it will be unable toconnect to an

    802.11n network using its built-in encryption. This is true even if a device isphysically capable of

    supporting 802.11n data speeds, and it doesnt just exclude thenotoriously insecure WEP. It also

    excludes TKIP, the form of RC4 specified by the original WPAcertification.Non-AES devices

    connecting to an 802.11n network thus have two options: Dont use encryption at all, or drop down to

    802.1a/g. Because backward-compatibility is built into the standard, an 802.11n access point will still

    support all the encryption options offered by the older standards, including RC4-based WEP and

    TKIP. However, it will only support them at 802.11a/g data rates. A bigger problem for most

    enterprises is supporting legacy 802.11a/b/g devices that lack the capability to handle 802.1x and thuscannot use the enterprise version of WPA2. Because a network is only as secure as its weakest link,

    allowing these devices to connect using pre-shared keys (or TKIP) and without enterprise-grade

    authentication may open up a security hole that could be exploited by an attacker. To prevent such

    exploits, enterprises need additional lines of defense beyond that provided by

    Per-User Firewalls

    A per-user firewall enables fine-grained control over the access rights of each device on a network,

    limiting it to specific authorized activities. For example, many older 802.11 phones lack 802.1x

    capability. An application firewall can ensure that devices which access the network as a legacy

    phone without data capability are only able to send VoIP packets, not access other systems. For

    maximum security, a per-user firewall needs to use flow signatures in addition to contents of

    encrypted packets.

    Physical Security

    A physical barrier can prevent radio waves from leaking outside a perimeter, making a wireless

    network completely undetectable. This was traditionally accomplished using radio jamming systems

    or Faraday cages large metallic walls but is now possible using selective signal blocking

    technology built into wireless networks themselves.

  • 8/8/2019 Report 07

    28/30

    28 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    Chapter 6 Real World ExperiencesProducts based on IEEE 802.11n Draft 2.0 draft have been shipping since September 2007, giving

    enterprises across all industries nearly three years to prove the technology in real deployments. It is

    already in use by thousands of users worldwide, with more adopting it every day

    6.1 MORRISVILLE STATE COLLEGE: PERVASIVE COVERAGE, HIGH

    PERFORMANCE

    Morrisville State College in New York deployed the worlds first all-802.11n network two years ago.

    It is still one of the worlds largest, providing coverage in 43 buildings as well as outdoor areas like

    courtyards and the football field. It is the primary network connection for more than 3000 students,

    part of a generation for whom phones have always been mobile and computers have always been

    laptops.The network now handles more than 1500 simultaneous users, carrying a mix of data, voice

    an video traffic. It sees speeds right up to the maximum 300 Mbps, averaging download times nin

    times faster than with 802.11g. Whereas uploading a 50MB file from a laptop used to take 3 m51

    seconds, it now takes only 26 seconds.

    6.2 HALIFAX HEALTH: 100% RELIABILITY, PREDICTABLE AND

    TRUSTWORTHY

    Halifax Health has been using wireless LANs since the early 2000s, successfully covering its flagship

    764-bed hospital in Datona Beach with an 802.11a/b/g network used for voice and data applications

    such as barcode scanning for drug dose verification. But the legacy network proved insufficient for a

    move to full electronic medical records which Halifax was implementing in a new ten-storey tower.

    Though the legacy technology could theoretically reach the 24 Mbps that the EMR applications need,

    real bandwidth was often much less due to microcell interference or contention for access from

    multiple users. Halifax upgraded to an 802.11n network based on a Virtual Cell architecture,blanketing the 500,000 square feet of the facility with pervasive coverage. The old network is still in

    use in older parts of the building, the two coexisting with no problems. The new 802.11n network

    supplies enough bandwidth for the EMR application as well as Lifelinks remote interpreting, a video

    system that connects a hearing-impaired patient to a live interpreter who translates a healthcare

    workers speech into sign language in real-time. Both the EMR application and the Lifelinks system

    are hosted on Workstations on Wheels, carts that staff move around to wherever needed. The

    hospital is also expanding its use of wireless telephony, using both Siemens VoIP phones and Vocera

    badges, as well as exploring new applications such as a semi-autonomous janitorial robot. But the

  • 8/8/2019 Report 07

    29/30

    29 802.11n : NEXT GENERATION NETWORK

    SOE,CUSAT - IT

    major benefit of 802.11n in a Virtual Cell is reliability: If the wireless EMR system went down,

    doctors and nurses would have to resort to a time-consuming manual process. With it, they can devote

    themselves fully to patient care.

    6.3 THE WASHINGTON NATIONALS: VERY HIGH USER DENSITYOpened in March 2008, Nationals Stadium is more than just the first Major League Baseball park to

    offer 802.11n coverage in all areas. It is a showcase for innovation, relying on its wireless network

    both to make its own operations more efficient and to improve the experience for all fans of the

    Washington Nationals whether they are on the bleachers, in the park's hospitality suites or at home

    watching games through the media. The stadiums network is available throughout the 41,88-seater

    park, offering Internet access to guests as well as multiple services to the stadiums own staff who use

    Avaya phones and Symbol handheld ticket readers that can be carried to the most crowded entry

    points. The readers verify each ticket's authenticity with a database at Tickets.com to eliminate

    forgeries and scalping, so even a fraction of a second's delay in checking each ticket could add up to

    long lines. The network is also made available to the media and to the stadiums more than 200 food

    concessions who need always-on on connectivity when verifying customer payment information.Future plans call for "room service"-style ordering, allowing fans to buy food or drink over the

    network and have it delivered to their seats so that they don't risk missing a crucial part of the game.

    Summary

    With thousands of customers already using it for critical applications, 802.11n is a mature and reliable

    technology. Its high-performance is proven in industries including education, healthcare,

    manufacturing, retail and hospitality, running data, voice and video simultaneously over networks

    spanning hundreds of access points. The technology is used both indoors and outdoors, serving

    laptops, phones, locator badges and client devices of all kinds. When combined with the appropriate

    architectural choices, it gives wireless the speed, security and scalability of wired Ethernet all whilerealizing cost savings over legacy wired or wireless systems. With official ratification, 802.11n is

    already taking over the wireless networking market on both the infrastructure and the client side. But

    its true implications are broader, with continued growth likely at the expense of wired Ethernet. As

    wireless now offers all the benefits of wires but with added mobility and reduced cost, an increasing

    number of enterprises move to an all-wireless edge.Deploying 802.11n is more than just a matter of

    replacing legacy radios. To maximize its benefits, organizations need to design networks for 802.11n

    from the ground up. This means taking account of multipath effects, bonded channels and the need to

    support legacy clients, as well as the standards full security implications and the increased demands

    likely on the network. Doing so will make the network simple and trustworthy, assuring predictable

    service levels for all users. By using an architecture purpose-built for 802.11n, IT departments can

    ensure that users and applications receive the same performance and reliability that they expect fromwired Ethernet.

  • 8/8/2019 Report 07

    30/30

    30 802.11n : NEXT GENERATION NETWORK

    References

    y Meru Networks Powering the All-wireless Enterprise White paper :The state of 802.11nDated : September 2009

    y Wi-Fi vs. WiMAX Comparison of the basic features of Wi-Fi (802.11b/g) - WiMAX (802.16)y IEEE PPT : History and Status of IEEE 802.11.n standardy

    WikiPedia : IEEE 802.11N-2009y MIMO : http://www.timeatlas.com/term_to_learn/general/what_is_mimoy VIDEO : http://www.youtube.com/watch?v=cDmWytRB3go