Metro Wi-Fi, Part 1

Metropolitan based Wi-Fi networks using unlicensed spectrum are now deployed in a large number of major metropolitan centers with many more deployments upcoming. Millions of notebook PCs have Wi-Fi capability built in, and it is embedded in most, if not all, new mobile computing platforms.

More importantly, low cost, long range, indoor and outdoor fixed wireless customer premises equipment (CPE) devices that utilize Wi-Fi technology for network access have also emerged. With new technology developments that combine automated wireless CPE device provisioning and customer billing with outdoor Wi-Fi network hardware, the threat to the cable industry from competitive broadband providers is unmistakable.

The choice facing cable telecommunications providers is stark. Use existing right of ways, HFC infrastructures and unlicensed spectrum to provide your own metro Wi-Fi networks, or risk being overbuilt by municipalities and/or alternative service providers within your licensed service area.

Unlicensed Wi-Fi technology, therefore, can be both friend and foe. For cable operators to befriend Wi-Fi, however, they must first understand it. "They" translates to every member of an operator’s technical team.

The first of this two-part technology discussion provides an education on the Institute of Electrical and Electronics Engineers 802.11 standard. Next month’s installment will take a closer look at the available networking technology, the imminent threats and successful strategies. What is 802.11? Developed by the IEEE and ratified in 1997, 802.11 was the first wireless local area network (WLAN) standard. It was named after the working group assigned to it and is also commonly referred to as Wi-Fi (Wireless Fidelity).

The 802.11 standard specifies "over the air" interfaces between wireless clients and a base station or access point (AP) or other wireless clients. It was initially intended to allow wireless connections of workstations to their "base" LAN using the unlicensed, 2.4 GHz ISM (industrial, scientific, medical) RF band. The original 802.11 standard targeted the case of both the workstation and the LAN being owned by the same entity, thereby providing a wireless extension to an existing wired LAN.

The original 802.11 standard supported a maximum data rate of only 2 Mbps. Because of that slow speed, most ordinary 802.11 wireless products are no longer being manufactured. The original 802.11 standard has now been adopted on a large scale for providing broadband wireless access (BWA) to public LANs, where a service provider owns the "base" LAN and a subscriber owns the workstation. The 802.11 family The 802.11 standard provides specifications for WLANs that cover the data link (Layer 2) and physical (PHY, Layer 1) layers of the open system interconnection (OSI) model.

The 802.11 family currently includes multiple over-the-air, physical, RF modulation techniques that all use the same data link data transfer protocol. The most popular techniques in use today are those defined by the 802.11 b, a, and g amendments to the original standard.

Security was originally included and was later enhanced via the 802.11i amendment. 802.11b was the first widely accepted wireless networking standard, followed (somewhat counter-intuitively) by 802.11a and 802.11g.

802.11n builds upon previous 802.11 standards by the use of what is known as multiple-input multiple-output (MIMO) to further increase data throughput. The standard is still under development, although products based on draft versions of the standard are now being sold. Data link and networks The data link layer in 802.11 defines a set of rules governing the use of a shared physical medium.

For primary data transfer, the 802.11 standard uses carrier sense multiple access/collision avoidance (CSMA/CA) as opposed to wired Ethernet, which uses CSMA/collision detection (CD.)

The most important difference between the data link layer of WLANs and that of most wired networking protocols is the inability of WLAN devices to detect collisions. With a single antenna performing both receiving and transmitting functions in a half duplex manner, a wireless station is unable to see any signal but its own (when transmitting). As a result, a complete packet will be sent before the incorrect checksum reveals that a collision has occurred. In addition, it cannot be assumed that all stations can hear one another in a wireless network area, thereby violating one of the basic requirements of collision detection schemes.

In order to avoid these problems, the 802.11 standard uses a collision avoidance mechanism combined with a positive acknowledgment scheme. The following defines the basic operation of the 802.11 CSMA/CA data transfer function:

1. Listen and wait for free wireless medium.
2. Wait a random time (backoff).
3. Listen again to ensure the wireless medium is still unused.
4. Transmit frame.
5. If a collision occurs, the transmitting station does not notice it.
6. Wait for an acknowledgement (ACK) from the AP.
7. No ACK returned indicates a collision occurred. Try again at step 1.

As stated, the 802.11 data link layer is designed to enable communication over all defined physical layers and is primarily concerned with defining a set of rules for accessing the shared wireless medium. In addition to the data link layer CSMA/CA data transfer rule, the 802.11 standard also defines two primary network architectures: infrastructure and ad-hoc.

An infrastructure network is an architecture that provides communication between wireless clients and wired network resources. Transition of data from the wireless to the wired medium is completed via an access point (AP). The AP and its attached wireless clients define the coverage area, with all devices forming a basic service set (BSS), which is defined by a BSS identifier (BSSID). If defined by a common extended service set (ESS), two adjoining BSSs are indicated with a common ESS identifier (ESSID). If a common ESSID is defined, a wireless client can roam from one BSS to another. This is illustrated in Figure 1. An ad hoc network is an architecture used to support wireless communication between clients. Created spontaneously and specifically designed for client-to-client wireless communication, an ad hoc network does not support access to wired networks. Therefore, ad-hoc networks do not require an AP. This is shown in Figure 2. The 802.11 standard spells out six primary service definitions for the data link layer. These include the following:

Data transfer: Wireless clients use a CSMA/CA algorithm as the access scheme to the physical medium.

Association: This enables the establishment of wireless links between wireless clients and APs in infrastructure networks.

Re-association: This takes place when a wireless client moves from one BSS to another. If a common ESSID is defined, a wireless client can roam from one area to another. Although re-association is defined in the 802.11 standard, the mechanism that allows AP-AP coordination to handle roaming is not specified.

Authentication: This is a process of proving a client identity, which takes place prior to a wireless client associating with an AP. By default, 802.11 devices operate in an open system where any wireless client can associate with any AP without the checking of credentials. True authentication is provided with 802.11 options known as wired equivalent privacy (WEP) and Wi-Fi protected access versions 1 and 2 (WPA and WPA2), where a shared encryption key is configured into the AP and its wireless clients.

Privacy/security: By default, data is transmitted "in the clear," and any 802.11-compliant device can eavesdrop on 802.11 traffic (with a similar PHY) that is within range. The WEP, WPA, and WPA2 options encrypt data before it is sent wirelessly using multiple bit encryption algorithms. Only devices with the correct shared encryption key used in authentication can decipher the data.

Power management: Two power modes are defined. In the active power mode, a wireless client is powered to transmit and receive. In the power save mode, a wireless client cannot transmit or receive, but consumes less power. Physical layer The original 802.11 PHY layer defined both optical and radio technologies for the Layer 1 transmission of wireless signals through the air.

The optical PHY layer for 802.11 is known as "Diffused Infrared." The radio technologies defined in the original 802.11 standard use the unlicensed 2.4 GHz ISM frequency band and are based on spread spectrum transmission, which originated in the military. Two types of spread spectrum are defined for the 802.11 PHY layer. These include direct sequence spread spectrum (DSSS) and frequency hopping spread spectrum (FHSS).

Spread spectrum uses more bandwidth but has lower power density than traditional digital transmission techniques. This is shown in Figure 3. Spread spectrum transmission provides many advantages, including:

• Low power density
• Inherent transmission security
• Resistance to interference from other radio sources
• Redundancy
• Resistance to multi-path interference and fading effects
• Ability to co-exist with other radio sources

Because of their elegant behavior and tolerance for interference from other equipment using the same frequency bands, spread spectrum systems may be operated without the need for a license. This is why they were chosen by the IEEE 802.11 working group for license-free WLAN and BWA operation using the ISM frequency band. 2.4 GHz ISM RF band The ISM radio bands were originally reserved internationally for the use of RF electromagnetic fields for ISM purposes other than communications. In general, communications equipment must accept any interference generated by ISM equipment.

These bands are generally confined to the 900 MHz and 2.4 GHz range and are currently used by 802.11b, 802.11g and Bluetooth devices. They are also used by a variety of noncomputing devices, such as cordless phones, low-power light bulbs and microwave ovens. All in all, they are busy portions of RF spectrum.

In the United States, ISM uses of the ISM bands are governed by Part 18 of the FCC rules, while Part 15 Subpart B contains the rules for unlicensed communication devices.

The legacy 802.11 standard defines 14 fixed 22 MHz-wide channels in the 2.4 GHz ISM band as displayed in Figure 4. It should be noted that only the first 11 channels are available for use in North America by Wi-Fi devices. Therefore, only three, nonoverlapping channels can be achieved where multiple systems are collocated – channels 1, 6 and 11. If more than three systems are collocated, their fixed bandwidth, 22 MHz channels will overlap, forcing users to share the spectrum, resulting in reduced data throughput. Actual behavior, as well as interference, is a function of overlapping channel size and signal strength. 802.11b In 1999, the IEEE expanded on the original 802.11 standard and created the 802.11b specification. 802.11b supports improved data rates of 5.5 Mbps and 11 Mbps, comparable to traditional wired Ethernet of the time.

The data link layer definitions remained the same from the legacy 802.11 standard, and 802.11b also uses the same unlicensed RF band and ISM channel definitions – 2.4 GHz/14 channels – as the legacy standard. The major change occurred in the defined RF transmission scheme (complementary code keying, CCK, vs. DSSS) for achieving improved data rates. This is similar to DOCSIS and its support of multiple RF modulation formats to achieve higher data throughput in virtually the same amount of RF spectrum (for example, 64-QAM/27 Mbps vs. 256-QAM/38 Mbps in the same 6 MHz channel).

802.11b cards operate at 11 Mbps, but will scale back to 5.5 Mbps, then 2 Mbps, then 1 Mbps (also known as adaptive rate selection). The data rate and modulation format selected is dynamic and a function of both signal strength and quality.

802.11b was the first widely accepted wireless networking standard, followed by 802.11a, and is the most widely deployed to date. 802.11a In 1999 (at the same time as the 802.11b standard was created) the IEEE also expanded on the original 802.11 standard and created the 802.11a specification.

The data link layer definitions remained the same from the legacy 802.11 standard, but 802.11a uses a different RF band from the original 802.11 standard and a different RF transmission scheme (that is, orthogonal frequency division multiplexing, OFDM).

802.11a supports data rates up to 54 Mbps in the unlicensed 5 GHz U-NII (Unlicensed National Information Infrastructure) frequency band. The data rate is adaptive and is dynamically reduced to 48, 36, 24, 18, 12, 9, then 6 Mbps if required based on signal strength and quality. The high data rate is achieved by combining many lower speed sub-carriers to create one high-speed channel, as with channel bonding in DOCSIS 3.0. One of four modulation formats – binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), 16-QAM (quadrature amplitude modulation) and 64-QAM – is used on each sub-carrier to achieve different data rates. The one used is dynamically selected based on signal strength and interference.

Because 802.11a and 802.11/802.11b utilize different frequencies, the two technologies are incompatible with each other. This is like the difference between AM and FM radio systems using different RF bands and different modulation formats. Some vendors offer hybrid 802.11a/b network gear, but these products simply implement the two standards side by side.

A lot of companies use 802.11a systems for higher-throughput applications and for increased security, since this system operates separate from the more available 802.11b and 802.11g systems. While 802.11a (U-NII) offers a greater throughput than 802.11b (ISM) devices, 802.11a systems tend to suffer more when faced with line-of-sight obstructions. These can be office walls, cubicles, desks or other obstructions. 5.0 GHz U-NII RF band The U-NII band covers the higher 5.15-5.35 GHz and 5.725-5.825 GHz range, as displayed in Figure 5, and is designed to allow for higher data rates (up to 54 Mbps) via the 802.11a standard. The 802.11a standard defines 12 fixed, nonoverlapping channels for use in the 5.0 GHz U-NII band. 802.11g In 2003, the IEEE again expanded on the original 802.11 standard and created the 802.11g specification. 802.11g combines the best of both 802.11a and 802.11b – with data rates up to 54 Mbps (using OFDM from 802.11a) in the 2.4 GHz ISM band (from 802.11b) providing greater range and resiliency.

As with 802.11a, the transmission scheme used in 802.11g is OFDM for the data rates of 6, 9, 12, 18, 24, 36, 48, and 54 Mbps. In addition, 802.11g devices will dynamically revert to (like the 802.11b standard) CCK for 5.5 and 11 Mbps and DSSS for 1 and 2 Mbps, making 802.11g fully backward compatible with the 802.11b and 802.11 legacy standards.

See Figure 6 for a view of the 802.11 family of standards and their relationship to the OSI model. Industry trends No one can doubt the ubiquity and increasing, daily acceptance of 802.11 technology. Ordinary people are becoming accustomed to wireless broadband access when they leave their homes. Why should it have to be from an alternative service provider?

Metropolitan based Wi-Fi networks using unlicensed spectrum are now deployed in a large number of major metropolitan centers and strategic locations (airports, restaurants, coffee shops) with many more deployments committed to and upcoming. Visit www.muniwireless.com/reports/docs/June-1-2007summary.pdf to find out if your service area is, or is about to be, overbuilt by the municipality. The number of municipalities getting into the Wi-Fi game has grown tremendously, from 122 in July 2005 to 385 in July 2007.

In July, AT&T announced free nomadic wireless access for their broadband subscribers to more than 10,000 Wi-Fi hotspots. Whereas cable operators have been kings of bundling services, they appear to be losing out at this game where nomadic Wi-Fi access is concerned. Yet with infrastructure and rights of way, as well as indoor and outdoor cable-friendly Wi-Fi technology available, why not compete?

Or preempt the competition? Over the past two years, one Canadian cable operator has done just that by deploying more than 400 outdoor DOCSIS hotspots on its existing HFC rights of ways in the most strategic locations of several cities. The service model is clear: free nomadic Wi-Fi access for existing triple-play customers and fee-based access for all others. By utilizing the unlicensed Wi-Fi spectrum for itself and positioning devices at the most strategic locations in its service territory, the cable operator has severely damaged the business case of an alternative provider wireless overbuild, occupied the limited Wi-Fi public use spectrum, provided another sticky service to its customer base, and achieved a new revenue stream from fee-based wireless users. This is clearly a win-win-win-win strategy.

The result is that alternate broadband service providers – utility company, wireless Internet service provider (WISP), municipality, competitive local exchange carrier (CLEC), etc.-have not wirelessly overbuilt this operator in any area where it has employed this strategy. The same cannot be said for neighboring cable operators.

The threat is clear for both rural, one-way cable operators and larger cable operators with two-way HFC plant in large metropolitan centers. On the flip side, the opportunity is also great. Wi-Fi is not going away any time soon. It can be both our friend and foe. It all depends on how the cable industry deals with this technology. Christopher Skarica is vice president, engineering, Lindsay Broadband. Chris Busch is vice president, broadband technologies, Incognito Software. Reach them at [email protected] and [email protected].