`
`Wideband local access: Wireless LAN and wireless ATM
`
`Article in IEEE Communications Magazine · December 1997
`
`DOI: 10.1109/35.634760 · Source: IEEE Xplore
`
`CITATIONS
`100
`
`3 authors, including:
`
`Kaveh Pahlavan
`Worcester Polytechnic Institute
`
`310 PUBLICATIONS 8,405 CITATIONS
`
`SEE PROFILE
`
`READS
`77
`
`Some of the authors of this publication are also working on these related projects:
`
`WINNER (Wireless INter-technology Networks with optimizEd data Rates) View project
`
`WINGIP (Wireless INdoor Geolocation and IPv6 traffic analysis) View project
`
`All content following this page was uploaded by Kaveh Pahlavan on 03 July 2014.
`
`The user has requested enhancement of the downloaded file.
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 1
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`1(15)
`
`Invited paper IEEE Communication Society Magazine
`
`Wideband Local Access: Wireless LAN and Wireless ATM
`
`Kaveh Pahlavan, Ali Zahedi, and Prashant Krishnamurthy
`
`Center for Wireless Information Network Studies
`Electrical and Computer Engineering Department
`Worcester Polytechnic Institute
`Worcester, MA 01609
`Tel: (508) 831-5634
`Fax: (508) 831-5491
` kaveh@wpi.edu, azahedi@wpi.edu, prashant@wpi.edu
`
`INTRODUCTION
`
`Although Wireless Local Area Networks (WLANs) and Wireless ATM (WATM) both
`provide Wideband Wireless Local Access (WWLA), there are differences and similarities
`among the two. WLAN is a mature technology with available products and market,
`WATM is an evolving technology that has not yet tested the market. WATM is perceived
`to be a service provided by the operating company. WLANs are considered as products
`sold by the manufacturer. WLANs provide an access to legacy LAN applications.
`WATM is expected to provide end-to-end ATM connectivity and Quality of Service
`(QoS) in the wireless channel. In this paper we address these issues in further detail and
`provide an overview of the global WWLA activities.
`
`We are emerging at the beginning of a new and exciting era for the wideband wireless
`local access (WWLA) industry. After a decade of self realization for this industry,
`WLAN and inter-LAN bridges are finding their way into the health care, manufacturing,
`finance, and educational markets. According to the January 1997 Frost & Sullivan’s
`report on the North American Wireless Office Hardware market, the total 1996 revenue
`for wireless offices was $390 million of which $218 million was from WLANs. The
`IEEE 802.11 standard for WLANs is emerging as a mature standard presenting a well
`defined technology that is being adopted by the manufacturers and accepted by the users.
`Chipsets have been developed according to the IEEE 802.11 standard, making software
`creativity easier for developing new applications towards expanding the market. ETSI’s
`RES-10 group has defined another alternative technology, HIPERLAN I, which primarily
`has a focus towards ad-hoc networking applications and supports higher data rates. More
`recently, research around wireless ATM has soared like an epidemic, engaging numerous
`companies in examining the suitability of yet another alternative standard technology for
`WWLA.
`
`1
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 2
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`2(15)
`
`The successful emergence of the market for WLANs operating in the unlicensed ISM
`bands has underlined the need for additional unlicensed bands. The continual demand of
`the WWLA industry for additional unlicensed bands in useful spectrum, initiated by
`WINForum, has resulted in the release of a 20 MHz unlicensed band around 1.9 GHz for
`asynchronous and isochronous applications in 1994 and 300 MHz of unlicensed bands
`earlier this year at 5 GHz referred to as the U-NII bands (formerly SUPERNet). On the
`other hand, the pan-European third generation cellular service (UMTS) is considering
`connectionless packet switched networks as class D bearer services. Under the research
`arm ACTS, the MEDIAN, WAND, SAMBA, and AWACS projects are addressing
`WWLA services. The Japanese are engaged in developing their own WWLA technology
`and at the same time several Japanese companies are involved in developing WLAN
`products for the US market.
`
`Mobile Terminal
`
`Server
`
`Fixed Terminal
`
`Backbone N etwork
`
`Applications
`
`TCP
`
`IP
`
`802.11 M AC
`
`802.11 P HY
`
`Applications
`
`TCP
`
`IP
`
`802.11 M AC
`
`802.11 P HY
`
`Applications
`TCP
`
`IP
`
`AAL
`
`WATM
`
`Access P oint
`
`802.11 M AC
`
`802.3 M AC
`
`802.11 P HY
`
`802.3 P HY
`
`(a)
`
`802.11
` MAC
`
`802.11 P HY
`
`LANE
`
`AAL5
`
`ATM
`
`PHY
`
`(b)
`
`WATM
`
`ATM
`
`Custom P HY
`
`Custom P HY
`
`PHY
`
`Applications
`
`TCP
`
`IP
`
`802.3 M AC
`
`802.3 P HY
`
`Applications
`TCP
`
`IP
`
`AAL
`
`ATM
`
`PHY
`
`Applications
`TCP
`
`IP
`
`AAL
`
`ATM
`
`PHY
`
`(c)
`Figure 1 : WLAN connection to the backbone.
`
`Wireless access cannot be discussed without considering issues related to the backbone.
`There are four options to interconnect the two air interfaces, WLAN and WATM, to the
`two wired backbones, legacy LANs and evolving ATM networks, namely WLAN-LAN,
`
`2
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 3
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`3(15)
`
`WATM-ATM, WLAN-ATM, and WATM-LAN. Figure 1 shows the protocol stack
`needed for the respective implementation of these interconnection techniques. The first
`option shown in Figure 1.a is addressed by the IEEE 802.11 community. The other two
`options are considered in the WATM community. The second approach, shown in Figure
`1.b, has more overhead, is expensive, and less scalable. At this stage this technique is
`considered only as an interim solution for migration and proof of concept [AYA96]. The
`fourth option WATM-LAN is not considered because a WATM air interface assumes
`that the backbone network employs ATM switches. Therefore, we are left with two
`options WATM-ATM and WLAN-LAN which we will refer to as WATM and WLAN in
`the rest of this paper.
`
`SERVICE SCENARIOS
`
`The success of WLANs or WATM depends on the availability of the corresponding
`backbone wired infrastructure and the evolution of the software applications. The
`backbone wired network consists of long-haul and local backbones. Today, it is
`commonly assumed that the future long-haul backbone networks will employ ATM
`transport and ATM will also be the backbone of the third generation wireless
`telecommunication networks. However, there is an on going battle between the
`connection based ATM local backbones versus contention based local legacy LAN
`backbones. Whether the backbone network of the future uses ATM only for long-haul
`and the legacy wired LAN technologies for local access or whether we will have an end-
`to-end wired ATM network will be a major deciding factor on the success of WLANs or
`WATM. The battle between ATM and Gigabit Ethernet for wired local access is not yet
`resolved [MCG96] and an unbiased prediction of the direction of this religious war is
`extremely challenging and beyond the scope of this paper.
`
`Service scenarios for the future WWLA can be categorized into private local networks in
`workplaces, universal access point in homes, and nomadic access in public places
`[GIL96]. The existing WWLA market is almost exclusively for the wireless office
`equipment using TCP/IP based applications over WLANs. The WLAN technology
`provides wireless access to the legacy LANs which support TCP/IP applications with
`minimal overhead. ATM in general has been shown to be inefficient in supporting legacy
`TCP/IP applications [LIP96, BER96]. The ATM local backbone is expected to be suited
`for the future multi-media applications supporting variety of traffic categories with a
`negotiable quality of service (QoS). Using RSVP for the TCP/IP applications the legacy
`LAN local backbone can support quality of service for multi-media applications
`[LAM97]. Potential wireless applications in home include universal wideband access for
`a variety of services such as cordless telephony, Internet access, and flexible positioning
`of audio systems. The WLAN technology can support all these applications but WATM
`could be more suitable for cordless telephone applications that may generate most of the
`in-home wireless traffic. Nomadic public access again depends on the availability of the
`backbone network. If the ATM networks are available in most public places, it may
`appear to be easier to provide traffic policing and charging mechanisms using WATM. If
`
`3
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 4
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`4(15)
`
`legacy wired LAN backbones are available, WLAN technology can also provide charging
`mechanisms but it is rather challenging to enforce traffic policing.
`
`MARKET AND PRODUCTS
`
`One of the most challenging issues facing the WLAN industry is expanding the market.
`Those involved in traditional WLAN industry promote privately owned WLAN
`applications, such as a campus area networks - a market for PCMCIA cards and access
`points in large quantities. The more visionary “service providers” are eager to promote
`nomadic WWLA applications in public places, such as airports, to generate a new source
`of income through service charges. Today only WLAN products exist in the market and
`WATM services are expected to appear in the market only by the turn of the century.
`
`In the past, the WLAN industry had a difficult time in predicting the development of the
`market. In 1990 the first generation WLAN products appeared in the market. These
`products, consuming around 20W (not suitable for laptops), were considered as an
`alternative that would avoid the expensive and troublesome installation and relocation
`costs of the coaxial cabled LANs. Under the assumption that WLANs would capture 10-
`15% of the coaxial cabled LAN market, early market predictions for WLANs were
`around $0.5-2 billion for the mid 1990’s. However, by the time WLAN products
`appeared in the market, less troublesome twisted pair wiring technology, similar to
`existing telephone wiring, had already replaced the coaxial cabled LAN technology, so
`that the first generation WLANs did not meet the market predictions.
`
`The second generation WLAN industry evolved in two directions. One group developed
`PCMCIA card WLANs for laptops to address the need for local mobility and its related
`applications. The other group added directional antennas to the first generation shoe-box
`type WLAN products and marketed them as inter-LAN bridges for outdoor applications
`[PAH95a, PAH95b, PAH85]. The existing WLAN products available on PCMCIA
`cards are either direct sequence spread spectrum (DSSS) or frequency hoping spread
`spectrum (FHSS) operating in ISM bands. The diffused IR technology is used for
`nomadic access in shorter distances for applications such as access for laptops to printers
`or in specific areas within the hospitals, such as radiology departments, where using radio
`signal is not encouraged. In addition to spread spectrum technology, other technologies
`such as direct beam IR (DBIR) and traditional radio are used for inter-LAN bridge
`applications. As we mentioned in the introduction the market size for these products is
`around two to three hundred million dollars.
`
`The WLAN market currently aims at four categories of applications [WOZ96]: healthcare
`industry, factory floors, banking industry, and educational institutions. In the healthcare
`market, in addition to traditional equipment such as laptops, notebooks, and hand-held
`terminals, special wireless services such as electronic thermometer and blood pressure
`monitoring devices are expected to be involved in wireless local communications. These
`devices are used to provide mobile access to clinical and pharmaceutical data bases for
`the physician as well as entering personal health data. In manufacturing floors and the
`
`4
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 5
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`5(15)
`
`factory environment, in addition to accessing databases and updating them, wireless
`networks enable rapid modification of the assembly lines and provide instant network
`access, for example, to the delivery trucks at the dock stations. In the banking industry,
`WLANs facilitate extending the network facilities to other branches, upgrading the
`systems without disrupting banking operation, reorganizing and rearranging the
`branches, and the access to Internet. In the educational environment, WLANs facilitate
`distance learning using
` wireless classrooms, provides access
`to
`the Internet,
`computational facilities, and databases servers to students using notebook computers.
`
`STANDARD TECHNOLOGIES
`
`There are three standardization activities for WWLA: the IEEE standard (802.11),
`ETSI’s standard (HIPERLAN), and the ATM Forum’s standard (WATM). We discuss
`these in this section in some detail.
`
`The IEEE standard for WLANs started in 1988 as IEEE 802.4L, a part of the IEEE 802.4
`Token Bus wired LAN standard. In 1990 the IEEE 802.4L changed its name to IEEE
`802.11 to form a stand alone WLAN standard in the IEEE 802 LAN standards
`organization. The technical aspects of this standard were completed this year.
`Throughout this unexpectedly long endeavor, the 802.11 group developed a framework to
`incorporate wireless specific issues such as power control, frequency management,
`roaming, and authentication in a LAN standard. The IEEE 802.11 standard was
`developed based on existing products in the market and so it addresses both technical and
`marketing issues.
`
`The 802.11 standard [IEE96] specifies data rates upto 2 Mbps using spread spectrum
`technology in the 2.4 GHz ISM bands. As wider U-NII bands with no restriction on
`modulation become available it is possible to update the physical layer of the standard to
`support higher data rates and other modulation techniques. The 802.11 standard considers
`two network topologies: infrastructure based and ad hoc (see Figure 2). In an
`infrastructure network (Figure 2a), mobile terminals communicate with the backbone
`network through an Access Point (AP). The AP is a bridge interconnecting the 802.11
`network to the backbone wired infrastructure. In this configuration, a distribution system
`interconnects multiple Basic Service Sets (BSS) through access points to form a single
`infrastructure network named an Extended Service Set (ESS). A mobile terminal can
`roam among different BSSs in one ESS without losing the connectivity to the backbone.
`In ad-hoc configuration (Figure 2b), the mobile terminals communicate with each other in
`an independent BSS without connectivity to the wired backbone network.. In this case
`some of the functions of the AP, such as release of a beacon with a defined ID and timing
`reference, that are needed to form and maintain a BSS are provided by one of the mobile
`terminals. The cells in both configurations shown in Figure 1 can overlap with one
`another. In addition to the contention based CSMA/CA access method suited for
`asynchronous data applications, the IEEE 802.11 also supports a contention free
`prioritized point coordination function (PCF) mechanism to support time bounded
`
`5
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 6
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`6(15)
`
`isochronous applications. The PCF mechanism supports limited assurance for providing
`quality of service. The MAC services in the IEEE 802.11 support authentication,
`encryption, frequency management and power conservation mechanisms that are not
`available in the MAC layer of other 802 standards such as 802.3.
`
`Infrastructure Network
`
`BSS 1
`
`AP
`
`Existing Wired LAN
`
`AP
`
`BSS 2
`
`Ad Hoc Networks
`
`BSS 1
`
`(a)
`
`(b)
`
`ESS
`
`AP
`
`BSS 3
`
`BSS 2
`
`Figure 2: Network topologies supported by IEEE 802.11 standard
`
`The HIPERLAN standard was developed by the RES-10 group of the ETSI as a pan-
`European standard for high speed wireless local networks. The so called HIPERLAN I,
`the first defined technology by this standard group, was started in 1992 and completed
`this year. It supports data rates of 2-23 Mbps using traditional radio modulation
`techniques in the 5.2 GHz band. The HIPERLAN I network [ETS96, ETS94] supports a
`multi-hop ad-hoc configuration (Figure 3). The multi-hop routing extends the
`HIPERLAN communication beyond the radio range of a single node. Each HIPERLAN
`node is either a forwarder or a non-forwarder. A non-forwarder node simply accepts the
`packet that is intended for it. A forwarder node retransmits the received packet, if the
`packet does not have its own node address, to other terminals in its neighborhood. Each
`non-forwarder node should select at least one of its neighbors as a forwarder. Inter-
`HIPERLAN forwarding needs bilateral cooperation and agreement between two
`HIPERLANs. To support routing and maintain the operation of a HIPERLAN, the
`forwarder and non-forwarder nodes need to periodically update six and four databases
`respectively. These databases are identified in Figure 3. The Non-Preemptive Multiple
`Access (NPMA) protocol used in HIPERLAN is a listen before talk protocol that supports
`both asynchronous and isochronous transmissions. The HIPERLAN defines a priority
`
`6
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 7
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`7(15)
`
`scheme and a lifetime for each packet which facilitates the control of the quality of
`service. In addition to the routing, the MAC layer also handles the encryption and power
`conservation.
`
`RIB : Routing Info. Base
`
`NIB : Neighborhood Info. Base
`
`HIB : Hello Info. Base
`
`AIB : Alias Info. Base
`
`SMRIB : Source MultiPoint Relay Info. Base
`
`TIB : Topology Info. Base
`
`3
`
`RIB
`NIB
`HIB
`AIB
`
`is S
`
`MR of
`
`6
`
`F
`
`RIB
`NIB
`HIB
`AIB
`SMRIB
`TIB
`
`2
`
`RIB
`NIB
`HIB
`AIB
`
`is S
`
`MR of
`
`1
`
`F
`
`RIB
`NIB
`HIB
`AIB
`SMRIB
`TIB
`
`R of
`M
`is S
`
`5
`
`HIPERLAN 1
`
`RIB
`NIB
`HIB
`AIB
`
`4
`
`F
`
`RIB
`NIB
`HIB
`AIB
`SMRIB
`TIB
`
`Neighborhood
`
`HIPERLAN 2
`
`Figure 3: HIPERLAN ad-hoc network configuration
`
`Other versions of HIPERLAN for accommodating wireless ATM and other alternatives
`are currently under consideration [WIL96]. During the standardization process a couple
`of HIPERLAN I prototypes were developed but as of now, no manufacturer has adopted
`this standard for product development. The bands assigned for HIPERLAN in the
`European Community (EC) was one of the motives of the FCC for releasing the U-NII
`bands discussed in the next section.
`
`In the past couple of years several companies and projects have developed infrastructure
`based prototypes to implement WATM technology [AGR96, AYA96, ENG95, RAY97,
`WAN96]. Of these, three major prototypes were developed at Lucent, NEC and the
`Magic WAND project of the ACTS research program in Europe. Table 1 summarizes the
`technical aspects of IEEE 802.11 and HIPERLAN I with the three major WATM
`prototypes. To merge these WATM studies, last year the WATM working group of the
`ATM Forum was formed to define a standard for WWLA using WATM technology.
`The work of this group started recently and it is planning to develop specifications for
`radio access, MAC layer and mobility support for WATM. The standard is aiming for
`completion in 1999 so that products could appear in the market by the turn of the
`century.
`
`7
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 8
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`8(15)
`
`Frequency Band Modulation
`Technique
`Spread Spectrum Direct
`Sequence: DBPSK,
`DQPSK; Spread
`Spectrum Frequency
`Hopping: 2GFSK,
`4GFSK BT=0.5; Diffused
`Infrared: 16 and 4 PPM
`
`Spread Spectrum
`Direct Sequence: 2.4-
`2.4835 GHz,
`Frequency
`Hopping: 2.4-2.4835
`GHz; Diffused
`Infrared: 850-950
`nanometer
`
`Data Rate Access Method
`
`Topologies MAC Services QoS
`
`Availibility
`
`1 and 2 Mbps Basic CSMA/CA,
`RTS/CTS, PCF with
`polling list, 20 frames
`
`Ad-hoc, Infra-Authentication,
`
`structure
`Encryption,
`Power
`conservation,
`Time bounded
`services
`
`No explicit support for
`QoS, but includes
`infrastructure topology
`and priority scheme in
`PCF that are useful for
`quality assurance.
`
`Technial standard finalized.
`Final administrative approval
`under progress. Products (e.g
`DEC Roamabout) and chipsets
`(e.g. Harris PRISM and
`Raytheon RAYLINK) are
`available.
`
`5.15-5.30 GHz
`
`Low bit rate: FSK; High
`bit rate:GMSK (BT=0.3)
`
`1.47 and
`23.53 Mbps
`
`Non-Preemptive
`Multiple Access
`(NPMA), 10 PDU
`
`Ad-hoc
`
`Advanced user priority
`scheme and packet
`lifetime mechanism to
`support QoS
`
`Encryption,
`Power
`conservation,
`Routing and
`forwarding, Time
`bounded services
`
`Standard is finalized. No
`product in the market. Two
`prototypes: HIPERION, fully
`standards compliant, and
`LAURA, not fully compliant
`[Wil96].
`
`900 MHZ (Proposed
`5 GHz U-NII Bands)
`
`OFDM or GMSK with
`LMS or RLS Equalization
`
`2-20 Mbps
`between
`laptop and
`PBS, and
`Gbps
`between
`PBSs
`
`Distributed Queue
`Reservation Updated
`Multiple Access
`(DQRUMA):
`Reservation and
`Piggybacking
`
`Infrastructure
`, ad hoc base
`station
`placement
`(optional)
`
`Scheduling,
`piggybacking etc.
`
`Base station responsible
`for checking and
`guaranteeing QoS,
`connections with or
`without QoS guarantees
`possible.
`
`Prototype at Bell labs in Lucent
`Technologies
`
`2.4 GHz ISM Bands
`
`p /4 - QPSK with decision
`feedback equalization
`
`8 Mbps
`
`TDMA/TDD with
`Slotted ALOHA
`
`Infrastructure
`based
`
`Scheduling,
`multiplexing and
`demultiplexing of
`VCs
`
`ABR, UBR, VBR and
`CBR slots are available
`but QoS support is not
`finalized
`
`Prototype at NEC USA's C&C
`Research Laboratories,
`Princeton, NJ.
`
`5.2 GHz
`
`16 Channel OFDM
`
`> 24 Mbps
`
`Reservation, Slotted
`ALOHA: Mobile
`Access Scheme based
`on Contention and
`Reservation
`(MASCARA)
`
`Infrastructure
`Based
`
`Scheduling, radio
`resource
`management and
`under further
`study
`
`Worst case QoS
`estimate (cell delay or
`cell loss) to be used for
`determining the
`connection
`
`Prototyping under the European
`ACTS AC085 project
`
`802.11
`
`HIPERLAN
`
`MII Bahama
`
`W I R E L E S S
`
`L A N
`
`W I R E L E S S
`
`NEC
`
`Magic
`WAND
`
`A T M
`
`Table 1: Comparison of WLAN standard technologies
`
`There are significant challenges for employing ATM in a mobile wireless link that has to
`be resolved before this technology identifies itself as a legitimate standard. ATM was
`essentially designed for high bandwidth, highly reliable and static optical fiber channels.
`Wireless channels are inherently unreliable and time varying with limited available
`bandwidth. ATM is a connection oriented transmission scheme operating based on an
`initial negotiation with the network for a virtual channel with a specific QoS. In a mobile
`environment management of the virtual channel and the QoS is not simple because the
`route has to be continually modified as the terminals move during the lifetime of a
`connection.
`
`Several proposals have been submitted for consideration by WATM working group
`[HUL96, DEA96, ACH96, BAR96, VEE96, ACH97] but there is no official technical
`specification released by the group. We refer to the overview material provided by the
`chairman of the committee in [DEL96]. Figure 4 provides an overall architecture
`perceived for the WATM services of the future. Similar to the IEEE 802.11 standard, the
`architecture is perceived to support both infrastructure and ad-hoc networking. Unlike
`the IEEE 802.11 standard, the wireless connection is expected to be performed through
`wired ATM switches. WATM access points are expected to communicate with one
`another to provide facilities to support a dynamic topology. WATM in this figure is
`shown to be a part of an integrated multimedia network multimedia over a variety of links
`that includes standard wired networks as well as dedicated microwave and satellite links.
`
`8
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 9
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`9(15)
`
`Figure 4: Wireless ATM Architecture proposed in [DEL97]
`
`As far as the specifications of the radio access are concerned, the WATM community
`argues that the short packet size (53bytes), high data rates (~20Mbps), and the short
`header (5bytes) in WATM ideally suites the on-off fading characteristics of the local
`radio channels providing an efficient usage of the band [DEL96]. It should be noted that
`the IEEE 802.11 MAC layer provides packet segmentation facilities that shortens the
`packets and if this is utilized properly, the same advantages can be gained. High data
`rates can also be provided by the IEEE 802.11 MAC layer if we migrate to higher
`frequency bands (e.g. U-NII) and adopt other modulation techniques. However, in any
`case, WLAN or WATM, migration to higher frequencies will reduce the coverage of the
`access point and increases the cost of the infrastructure. In the past decade products with
`smaller coverage have not survived the market test.
`
`GLOBAL ACTIVITIES
`
`Today’s global market for WLANs primarily belongs to the US. For manufacturers
`aiming at the WWLA market, the two most important issues are the availability of
`unlicensed bands for future product development and a spectrum etiquette that secures a
`minimum bandwidth availability for users in different scenarios. The relation between a
`licensed band and an unlicensed band is similar to the relation between a privately owned
`backyard and a public park. If you can afford a backyard you can have a barbecue there;
`otherwise you may go to a public park and depending on the availability of space, enjoy
`similar benefits without any investment on the property. WLANs are new products
`developed by small companies in the midst of large companies. None of them can afford
`backyards in the expensive neighborhoods of 1-2 GHz where a big market for voice
`
`9
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 10
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`10(15)
`
`applications exists. As increasing number of people realize the benefits of a public park
`and use it, there increases a need for an “etiquette” to control the crowd of barbecue fans
`and prevent the smoke from overwhelming everyone.
`
`In 1994, the FCC released 20 MHz of spectrum between two parts of licensed bands in
`the 1.9 GHz for operation of unlicensed personal communications services (U-PCS). The
`20 MHz of spectrum allocated for U-PCS is shown in Figure 5.
`
`Figure 5: Spectrum allocated for Unlicensed PCS [STE94]
`
`The unlicensed nature of the frequency bands necessitated a “coexistence etiquette”
`which was developed by WINTech, the technical subcommittee of the WINForum
`industry association. This spectrum etiquette forms the basis for the rules adopted by the
`FCC for operation in the U-PCS bands. The interesting point about the WINForum
`etiquette is the assignment of two separate bands for the asynchronous and isochronous
`transmissions. This separation implies that, in the view of WINForum, voice and bursty
`data transmissions are incompatible enough requiring the use of separate channels in the
`air. This separation is in contrast with the view of those working in wired
`communications (such as ISDN and ATM) where the trend is towards the integration of
`voice and bursty data on the same channel. There are three basic principles adopted for
`the spectrum etiquette in the U-PCS bands: (a) listen before talk (or transmit) (LBT)
`protocol (b) low transmitter power and (c) restricted duration of transmissions. The
`requirements on timing accuracy is quite rigid making this protocol unsuitable for ATM
`transport. Further details about this protocol are available in [STE94].
`
`The U-NII initiative started with WINForum filing a petition before the FCC requesting
`allocation of 250 MHz of spectrum for high speed devices providing 20 Mbps in May
`1995. This was followed by a petition by Apple Computer requesting allocation of 300
`MHz of spectrum in the same frequency bands for promoting full deployment of the so-
`called National Information Infrastructure (NII)1. These bands and devices were referred
`to as “Shared Unlicensed PErsonal Radio Network” (SUPERNet) bands and devices. On
`January 9th 1997, the FCC released 300 MHz of unlicensed spectrum around 5 GHz for
`use by a new category of radio equipment called “Unlicensed National Information
`Infrastructure” or U-NII devices. This decision of the FCC makes a large amount of
`
`1 The FCC Describes the NII as a group of networks including the public switched telephone network
`(PSTN), radio and television networks and other networks yet to be built, which will together serve the
`communications and information processing needs of the people of the US in the future.
`
`10
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 11
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`11(15)
`
`spectrum available for wireless LAN applications which can provide data rates of upto
`several tens of Mbps needed for multimedia type of applications. The frequencies
`allocated are compatible with those specified earlier by the CEPT for HIPERLAN. The
`WINForum spectrum etiquette was also suggested by the FCC for U-NII devices, but was
`not adopted because of concerns about its suitability to accommodate WATM services.
`The FCC believes that this spectrum would enable opportunities for providing advanced
`telecommunications services to educational institutions, health care providers, libraries
`etc [FCC96] thereby significantly assisting in meeting universal service goals [HUD94]
`as set forth in the Telecommunications act of 1996 [TEL96].
`
`The frequency bands and their respective technical restrictions are shown in Table 2. To
`encourage maximum flexibility in developing new technologies minimum technical
`restrictions have been adopted. Detailed restrictions or standards can only delay the
`implementation of the appropriate products that are able to attract commercial
`applications.
`
`Table 2: FCC Requirements for the U-NII Frequency Bands
`
`Band of
`operation
`
`Maximum
`Tx Power
`
`5.15 - 5.25 GHz
`
`50 mW
`
`Max. Power
`with antenna
`gain of 6 dBi
`200 mW
`
`Maximum PSD
`
`2.5 mW/MHz
`
`5.25 - 5.35 GHz
`
`250 mW
`
`1000 mW
`
`12.5 mW/MHz
`
`5.725-5.825 GHz
`
`1000 mW
`
`4000 mW
`
`50 mW/MHz
`
`Applications:
`suggested and/or
`mandated
`Restricted to
`indoor
`applications
`Campus LANs
`
`Community
`networks
`
`Other Remarks
`
`Antenna must be
`an integral part
`of the device
`Compatible with
`HIPERLAN
`Longer range in
`low-interference
`(rural) environs.
`
`Although no devices have been developed yet for U-NII bands, several activities for
`developing standards and products in these bands have already started. The Mobile
`Information Infrastructure (MII), an ongoing research project at Bell Labs which is
`partially funded by the National Institute of Standards and Technology is considering
`wireless ATM applications in this band [AYA96]. The IEEE 802.11 has set up a study
`group for higher speed physical layer standards development in the 5 GHz band with the
`same MAC protocol used currently for the 2 Mb/s standard using spread spectrum
`technology in the ISM bands. Data rates beyond 20 Mbps as well as the capability of the
`MAC layer to support data, voice and video services are to be examined by this study
`group. In the U-NII, as opposed to ISM bands, there is no restriction to use only spread
`spectrum technology and other modulation techniques such as OFDM and GMSK have
`been suggested. At higher U-NII frequencies the LAN adapter is expected to be more
`expensive, but the cost/performance ratio is expected to improve.
`
`Although the WLAN market is primarily in the US, there are significant WWLA related
`activities in the EC and Japan. In the EC under the ACTS projects several research
`programs are devoted towards examining a variety of technologies for wideband local
`applications. The intention is to integrate these technologies into the evolving third
`
`11
`
`IPR2020-00910
`Garmin, et al. EX1009 Page 12
`
`
`
`WORCESTER POLYTECHNIC INSTITUTE
`Center for Wireless Information Network Studies
`
`10/09/97
`
`12(15)
`
`generation cellular systems. Also, ETSI has announced a new standardization project
`BRAN (Broadband Radio Access Networks) which will take over the current
`HIPERLAN activities. The BRAN project will address issues related to wireless access
`systems with bit rates larger than 25 Mbps. Close relationships are being established with
`the ATM Forum, IEEE 802.11, the Internet Engineering Task Force and ITU-R to ensure
`overall coherence with other existing and emerging technologies.
`
`The ACTS (Advanced Communications Technologies and