CCIE Routing and Switching Exam Certification Guide

Foundational Topics

This section covers Media Access Control (MAC) addresses, the Ethernet family of data-link layer technologies, Token Ring, and wireless LANs (WLANs).

MAC Address Format

Ethernet or Token Ring router interfaces and all device network interface cards (NICs) are identified with a unique burned-in address (BIA). This is the MAC address, which is also called the physical address. It is an implementation of Layer 2 of the OSI reference model—or more specifically, the MAC layer of the IEEE model to identify the station. The MAC address is 48 bits in length (6 octets) and is represented in hexadecimal.

The router output in Example 4-1 shows the MAC address (00-10-7b-3a-92-3c) of an Ethernet interface.

Example 4-1 Router Interface MAC Address

router> show interface
  Ethernet0 is up, line protocol is up
  Hardware is Lance, address is <Anchor0>0010.7b3a.923c (bia 0010.7b3a.923c)

  Internet address is 135.119.110.30/24
  MTU 1500 bytes, BW 10000 Kbit, DLY 1000 usec,
     reliability 255/255, txload 1/255, rxload 1/255
  Encapsulation ARPA, loopback not set
  Keepalive set (10 sec)
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 00:00:00, output 00:00:00, output hang never
  Last clearing of "show interface" counters never
  Input queue: 1/75/0/0 (size/max/drops/flushes); Total output drops: 0
  Queueing strategy: fifo
  Output queue :0/40 (size/max)
  5 minute input rate 1000 bits/sec, 2 packets/sec
  5 minute output rate 1000 bits/sec, 2 packets/sec
     1999164 packets input, 379657585 bytes, 0 no buffer
     Received 1785091 broadcasts, 0 runts, 0 giants, 0 throttles
     1 input errors, 0 CRC, 0 frame, 0 overrun, 1 ignored
     0 input packets with dribble condition detected
     745208 packets output, 82211652 bytes, 0 underruns
     0 output errors, 63 collisions, 16 interface resets
     0 babbles, 0 late collision, 345 deferred
     0 lost carrier, 0 no carrier
     0 output buffer failures, 0 output buffers swapped out

The first three bytes of a MAC address form the Organizational Unique Identifier (OUI), which identifies the manufacturer/vendor. The last three octets are administered by the manufacturer and assigned in sequence.

Canonical Transmission

When converting hexadecimal MAC addresses to binary, each hexadecimal number is represented in its 4-bit binary equivalent. For example, ac-10-7b-3a-92-3c is converted normally to binary as the following:

10101100 00010000 01101011 00111010 01010010 00111100

For Token Ring networks, each octet of this MAC address is transmitted from left to right, from the most significant bit (MSB) to the least significant bit (LSB). This noncanonical transmission is also known as MSB first.

For Ethernet networks, each octet of the previous MAC address is transmitted from left to right, but LSB to MSB. The difference is that for each octet, the LSB is transmitted first and the MSB is transmitted last. This canonical transmission is also known as LSB first. The MAC address AC-10-7b-3a-92-3c is transmitted as the following:

00110101 00001000 11010110 01011100 01001010 00111100

The octet AC is transmitted from left to right as 00110101, the second octet (10) is transmitted from left to right as 000010000, and so on.

For both Ethernet and Token Ring networks, the order of each octet is transmitted the same: from Most Significant Octet to Least Significant Octet. The difference is in the transmission order of the bits of each octet. For Ethernet networks, the LSB of the first octet is transmitted first. This bit is the Individual/Group (I/G) Address Bit. If the I/G is set to 0, it indicates an individual MAC address. If the I/G bit is set to 1, it indicates that the address is a group address. The I/G is set to 1 for broadcast and multicast MAC addresses.

The first (leftmost) bit in the binary representation is the I/G Address Bit. If set to 0, it indicates an individual address. It can be set to 1 in an address allocated by the vendor to indicate that the address is a group address. The second leftmost bit is the U/L bit. If set to 0, it indicates a universally administered address. If set to 1, it indicates that the address is locally administered.

Ethernet

Today's Ethernet networks are based on the Ethernet development by Digital, Intel, and Xerox (DIX). Version 1 of the standard was created in 1980. It used unbalanced signaling, meaning that a 0 is represented by 0 voltage, and a 1 is represented by a positive voltage. In 1982, version 2 of Ethernet was introduced. It added the heartbeat signal to transceivers and moved to balanced signaling. Balanced signaling uses positive and negative voltages, which allow more speed. The heartbeat signal is used as a link test pulse. Ethernet version 2 is the basis of today's Ethernet networks. Ethernet uses carrier-sense multiple access with collision detection (CSMA/CD) as the access method.

CSMA/CD Media Access Method

CSMA/CD is the media access method on Ethernet networks. In this scheme, hosts listen to the network for activity. If there is no network activity, hosts can transmit a frame onto the network, while transmitting hosts listen to the network for any collisions with other transmitting nodes.

If no collision is detected, the host assumes that the frame transmitted successfully. If a collision is detected, the node waits a random amount of time and listens for traffic on the segment. If there is no traffic, it attempts to send the frame again. It attempts 16 times before sending an error message to the upper-layer protocol (ULP).

Ethernet Encoding

Manchester encoding was selected to code signals on the wire on Ethernet. In Manchester encoding, a 0 is represented as a transition from high to low in the middle of the clocking time interval. A 1 is represented as a transition from low to high in the middle of the time interval. Figure 4-1 shows a sample of Manchester encoding.

Figure 4-1 Manchester encoding

Ethernet Frame Formats

After the Ethernet V2 standard was published, an effort went into producing an IEEE standard for Ethernet. The IEEE 802 committee produced their 802.3 and 802.3 SNAP frame formats, with the 802.2 Logical Link Control (LLC). Novell also produced a frame format for its network operating system. These different groups produced different frame formats for the Ethernet wire, but the signaling, encoding, and frame maximum and minimum sizes remain the same. Therefore, these four frame formats can reside on the same segment. A few differences exist, which are explained.

The four frame formats are as follows:

• Ethernet version 2

• Novell 802.3 Raw

• IEEE 802.3

• IEEE 802.3 Subnetwork Access Protocol (SNAP)

Ethernet Version 2 Frame Format

Figure 4-2 shows the Ethernet version 2 frame format. The frame fields are described in Table 4-1.

Figure 4-2 Ethernet V2 Frame Format

Table 4-1 Ethernet V2 Frame Fields Descriptions

The minimum frame size on Ethernet is 64 bytes, and the maximum is 1518 bytes. When calculating the frame size, do not include the preamble in the summation. The Ethernet V2 frame format conforms to the following specification:

• The minimum frame size is 6 + 6 + 2 + 46 + 4 = 64 bytes

• The maximum frame size is 6 + 6 + 2 + 1500 + 4 = 1518 bytes

Novell 802.3 Raw Frame Format

Prior to the IEEE 802.3 specification, Novell needed a frame format for their NetWare Internetwork Packet Exchange (IPX) network operating system. Novell produced their own frame format to run on CSMA/CD networks.

Figure 4-3 shows the Novell 802.3 raw frame format. The frame field descriptions are displayed in Table 4-2.

Figure 4-3 Novell 802.3 Raw Frame Format@

Table 4-2 Novell 802.3 Raw Frame Fields Descriptions

Novell frames are unique in that the data field begins with an FFFF hex (at the beginning of the IPX protocol header).

IEEE 802.3 Frame Format

The IEEE produced its Ethernet standard in June 1983. The type field changed to a length field and the IEEE 802.2 LLC layer was added.

Figure 4-4 shows the IEEE 802.3 frame format. Table 4-3 displays the IEEE 802.3 frame fields.

Figure 4-4 IEEE 802.3 Frame Format@

Table 4-3 IEEE 802.3 Frame Fields Descriptions

The LLC is specified by IEEE 802.2. The LLC provides connectionless (LLC type 1) and connection-oriented (LLC type 2) services. You use LLC1 with Ethernet networks and LLC2 in IBM SNA environments. The LLC is divided into three fields, as shown in Figure 4-5.

Figure 4-5 LLC Fields

The destination service access point (DSAP) indicates the destination ULP. The source service access point (SSAP) indicates the source ULP. Examples of SAP values are as follows:

• NetBIOS = 0xF0
• Bridge PDU = 0x42
• SNA = 0x04, 0x05, 0x0C
• SNAP = 0xAA
• X.25 = 0x7E
• IP = 0x06

In Ethernet networks, the control field is set to 0x03 to indicate an 802.2 unnumbered format for connectionless service.

IEEE 802.3 SNAP Frame Format

The SNAP field was created to help transition protocols from Ethernet V2 to an IEEE 802.3 compliant frame. It uses the same type values as Ethernet V2, but the EtherType field is preceded by a 3-byte field for protocol family identification (protocol ID).

The SNAP field includes the Ethernet type information in Ethernet V2. Figure 4-6 shows the IEEE 802.3 SNAP frame format. The remainder of this section discusses the fields in this format.

Figure 4-6 IEEE 802.3 SNAP Frame Format

All fields are the same as in the IEEE 802.3 frame format. The SSAP and DSAP of the LLC are set to 0xAA. The SNAP field is added. The SNAP field contains two fields, including a 3-byte vendor code field. This 3-byte vendor code is unique to different vendors. The next field in the SNAP is the 2-byte Ethernet type field in Ethernet V2. The SNAP fields are displayed in Figure 4-7.

Figure 4-7 IEEE SNAP Fields

In the years that followed these standards, there has been a mix of frame formats in the network. IP and DEC protocols use Ethernet V2, Novell's IPX (3.x) uses its 802.3 raw format, and SNA uses 802.3. As most networks migrate to the IP protocol, Ethernet V2 frames are more common on the network.

The remainder of this section covers Ethernet physical specifications.

Ethernet Media Specifications

This section lists the physical media specifications, including range limitations for Ethernet, FE, and GE.

10Base5 Thick Ethernet

Commonly referred to as Thick Ethernet or Thicknet, this specification uses 0.4 inch, 50-ohm coaxial cable. The specifications for Thicknet are as follows:

• 0.4 inch, 50 ohm coax cable.

• Maximum segment length is 500 m.

• Maximum number of attachments per segment is 100.

• Maximum AUI cable length is 50 m.

• Minimum separation between attachments (MAU) is 2.5 m.

• Cable ends terminate with 50 ohm terminators.

• MAU attach workstations.

• Maximum network length is 5 segments and 2500 m.

• Maximum number of stations on the network is 1024.

Figure 4-8 shows a sample 10Base5 Ethernet network.

Figure 4-8 Sample 10Base5 Network

10Base2 Thinnet

Commonly referred to as Thinnet, this specification uses 0.2 inch, 50 ohm coaxial cable. Thinnet specifications are as follows:

• 0.2 inch, RG58-U, 50 ohm coax cable.

• Maximum segment length is 185 m.

• Maximum number of attachments per segment is 30.

• Minimum separation per segment is 0.5 m.

• T-connectors attach workstations.

Figure 4-9 shows a sample 10Base2 Ethernet network.

Figure 4-9 Sample 10Base2 Network

10BaseT Ethernet

UTP has become the defacto standard media for LAN systems. 10BaseT was made an IEEE standard in 1990. The 10BaseT specifications are as follows:

• 24 AWG UTP .4/.6 mm cable.
• Maximum segment length is 100 m.
• 1 device per cable.

Fast Ethernet

The IEEE developed the 802.3U standard in 1995 to provide Ethernet speeds of 100 Mbps over UTP and fiber cabling. The 100BaseT standard is similar to 10 Mbps Ethernet in that it uses CSMA/CD, runs on Category 3, 4, and 5 UTP cable, and the frame formats are preserved. Connectivity still uses hubs, repeaters, and bridges.

The encoding for 100BaseT is 4B/5B with nonreturn to zero (NRZ), the new speed is 100 Mbps, and the media independent interface (MII) was introduced as a replacement to the attachment unit interface (AUI).

The 4B/5B coding takes 4 bits of data and expands it into a 5-bit code for transmission on the physical channel. Because of the 20 percent overhead, pulses run at 125 MHz on the wire to achieve 100 Mbps. Table 4-4 shows how some data numbers are converted to 4B/5B code.

Table 4-4 FE 4B/5B Code

The following specifications are covered in this section:

• 100BaseTX
• 100BaseT4
• 100BaseFX

100BaseTX FE

The 100BaseTX specification uses Category 5 UTP wiring. Similar to 10BaseT, FE uses only two pairs of the 4-pair UTP wiring. If Category 5 cabling is already in place, upgrading to FE only requires a hub or switch and NIC upgrades. Because of the low cost, most of today's installations use switches. The specifications are as follows:

• Transmission over Cat 5 UTP or Cat 1 STP wire.

• RJ-45 connector (same as in 10BaseT).

• Punchdown blocks in the wiring closet must be Category 5 certified.

• 4B5B coding.

100BaseT4 FE

The 100BaseT4 specification was developed to support UTP wiring at the Category 3 level. This specification takes advantage of higher speed Ethernet without recabling to Category 5 UTP. This implementation is not widely deployed. The specifications are as follows:

• Transmission over Cat 3, 4, or 5 UTP wiring.

• Three pairs are for transmission, and the fourth pair is for collision detection.

• No separate transmit and receive pairs are present, so full-duplex operation is not possible.

• 8B6T coding.

100BaseFX

The 100BaseFX specifications for fiber are as follows:

• Operates over two strands of multimode or single-mode fiber cabling

• Can transmit over greater distances than copper media

• Uses media interface connector (MIC), ST, or SC fiber connectors defined for FDDI and 10BaseFX networks

• 4B5B coding

1000 Mbps GE

Gigabit Ethernet is specified by two standards: IEEE 802.3z and 802.3ab. The 802.3z standard specifies the operation of GE over fiber and coaxial cable and introduces the Gigabit MII (GMII). The 802.3z standard was approved in July 1998. The 802.3ab standard specifies the operation of GE over Category 5 UTP, as approved in June 1999. GE still retains the frame formats and frame sizes of 10 Mbps Ethernet, along with the use of CSMA/CD in shared segments. Similar to Ethernet and FE, full-duplex operation is possible. Differences can be found in the encoding; GE uses 8B/10B coding with simple NRZ. Because of the 20 percent overhead, pulses run at 1250 MHz to achieve 1000 Mbps. GE includes the following methods to achieve 1 Gbps speed:

• 8B/10B Coding.

• Bytes are encoded as 10-bit symbols.

• Run-length limited (no long sequences of 1s or 0s).

• Pulses on the wire run at 1250 MHz to achieve 1000 Mbps speed.

Table 4-5 shows how data is converted into 8B/10B code for transmission.

Table 4-5 8B/10B Encoding

The following specifications are covered in this section:

• 1000BaseLX
• 1000BaseSX
• 1000BaseCX
• 1000BaseT

000BaseLX Long Wavelength GE

The IEEE 1000BaseLX uses long wavelength optics over a pair of fiber strands. The specifications are as follows:

• Uses long wave (1300 nm).

• Use on multimode or single-mode fiber.

• Maximum lengths for multimode fiber are as follows:

  • 62.5 um fiber: 440 m

  • 50 um fiber: 550 m

• Maximum length for single-mode fiber is 9 um: 5 km.

• Uses 8B/10B encoding with simple NRZ.

1000BaseSX Short Wave GE

• The IEEE 1000BaseSX uses short wavelength optics over a pair of multimode fiber stands. The specifications are as follows:

• Uses short wave (850 nm).

• Use on multimode fiber.

• Maximum lengths are as follows:

  • 62.5 um: 260 m

  • 50 um: 550 m

• Uses 8B/10B encoding with simple NRZ.

1000BaseCX GE over Coaxial Cable

• The IEEE 1000Base-CX standard is intended for short copper runs between servers. The specification is as follows:

• Used on short run copper.

• Runs over a pair of 150 ohm balanced coaxial cable (twinax).

• Maximum length is 25 m.

• Mainly for server connections.

• Uses 8B/10B encoding with simple NRZ.

1000BaseT GE over UTP

The IEEE standard for 1000 Mbps Ethernet over Category 5 UTP is IEEE 802.3ab; it was approved in June 1999. This standard uses the 4 pairs in the cable. (100BaseTX and 10BaseT Ethernet only use 2 pairs.) The specifications are as follows:

• Category 5, 4-pair UTP.

• Maximum length is 100m.

• Encoding defined is a 5-level coding scheme.

• 1 byte is sent over the 4 pairs at 1250 MHz.

10 GE

Although not a test topic, the CCIE candidate should be familiar with the developing IEEE 802.3ae 10 GE technology. The draft standard mentions that it only functions on optical fiber and operates in full-duplex mode. It can provide media solutions for metropolitan-area networks (MANS) and wide-area networks (WANS). The standard is to be adopted in mid 2002. More information is at the 10 Gigabit Alliance web site at www.10gea.org.

Token Ring

Token Ring was developed by IBM for the forwarding of data on a logical unidirectional ring. Token Ring is implemented in the data-link layer. Token Ring networks move a small frame, called a token, around the network. Possession of the token grants the right to transmit data. After a station has the token, it modifies it into a data frame, appends the data for transmission, and sends the frame to the next station. No token is on the ring until the data frame is received by the source station marked as read and copied, and releases a token back into the ring.

The IEEE standard for Token Ring is IEEE 802.5; the differences with IBM's specification are minor. Table 4-6 shows the similarity and difference between the specifications.

Table 4-6 IBM and IEEE 802.5 Token Ring Specification Similarities

The physical implementation topology can be a ring or more commonly a star. When connected as a ring, devices connect to a multistation access unit (MSAU). MSAUs can be connected together with patch cables to form a ring. The MSAU can also bypass stations that are defective on the ring. Figure 4-10 shows the connectivity in the Token Ring network. The MSAUs are connected in a physical ring.

Figure 4-10 Token Ring MSAU

Token Ring Coding

Token Ring uses differential Manchester as the encoding scheme. In differential Manchester, a 0 is represented as a transition at the beginning of the clock time interval. A 1 is indicated as an absence of a transition.

Token Ring Operation

Access is controlled by using a token. A token is passed along the network from station to station. Stations with no data to transmit forward the token to the next station. If the station wants to transmit data, it seizes the token, produces a data frame by appending data, and sends it to the destination. The receiving station reads the frame and forwards it along the ring back to the source station. The receiving station also sets the address-recognized and framed-copied bits on the forwarded frame. The source station verifies that the data frame was read and releases a token back onto the network. The default operation permits the use of the Token in round-robin fashion.

Token Ring Priority

Token Ring includes an optional priority system that permits stations configured with a higher priority value to use the network more frequently than permitted by the default round-robin operation. Eight levels of priority are provided using a 3-bit reservation field and a 3-bit priority field. As an information frame passes, a station sets a higher priority in the reservation field, which reserves the token. The transmitting station then sends a token out with the higher priority set. After the high priority station completes sending its frame, it releases a token with the normal or previous priority.

Active Monitor (AM)

One station on the Token Ring is selected to be the AM. This station performs a variety of ring-maintenance functions. The AM removes continuously circulating frames that are not removed by a failed transmitting station. As a frame passes the AM, the monitor count bit is set. If a frame passes with the monitor count bit set, the AM assumes that the original sender of the frame was unable to remove the frame from the ring. The AM purges this frame, sends a Token Soft Error message to the Ring Error Monitor, and generates a new token.

The AM provides timing information to ring stations. The AM inserts a 24-bit propagation delay to prevent the end of a frame from wrapping onto the beginning of the frame, and also confirms that a data frame or token is received every 10 milliseconds.

Standby and Ring Error Monitors are also on Token Ring networks. Standby Monitors take over as AM if the primary AM is removed from the ring or no longer performs its functions. Ring Error Monitors can also be present on the ring to collect ring status and error information.

Beaconing

A beacon frame is sent by a station that does not receive any more frames—either a data frame or a token—from its upstream neighbor. An adapter keeps beaconing until it begins to receive frames again. A beacon MAC frame includes the beaconing station's MAC address and the address of the station's nearest active upstream neighbor (NAUN), indicating that the problem lies between the two stations.

Early Token Release (ETR)

In a normal token ring operation, the station that transmitted the data frame removes it and then generates a free token.

With ETR, a token is released immediately after the sending station transmits its frame. The sending station does not wait for the data frame to circle the ring. ETR is only available on 16 Mbps rings. Stations running ETR can coexist with stations not running ETR. With ETR, a free token can circulate the ring with multiple data frames.

On supported router platforms, early token release is configured as follows:

interface token 0
 early-token-release

Ring Insertion

The process for a station to insert into the token ring follows five phases:

• Phase 0—Lobe media test
• Phase 1—Physical insertion
• Phase 2—Address verification
• Phase 3—Participation in ring poll
• Phase 4—Request initialization

Phase 0: Lobe Media Test

The first step when a token ring device is inserted is the lobe media test. The transmitter and receiver of the adapter and the cable between the adapter and the MSAU are tested in this phase.

Phase 1: Physical Insertion

In phase 1, the adapter opens a relay on the MSAU. After the MSAU opens, the adapter determines an AM is present on the ring, which indicates successful completion of phase 1.

Phase 2: Address Verification

This phase verifies that the MAC address is unique to the ring. This phase can detect if two Locally Administered Addresses (LAAs) are configured with the same MAC address. This phase is also called the duplicate address test.

Phase 3: Participation in Ring Poll

In this phase, the station learns its upstream neighbor's address and informs its downstream neighbor of the inserting adapter's address and produces a station list. If the adapter successfully participates in a ring poll, it proceeds into the final phase of insertion.

Phase 4: Request Initialization

Phase 4 is the final phase of ring insertion. The adapter sends request initialization MAC frames to the functional address of the Ring Parameter Server (RPS). The RPS responds with information such as the ring number and speed. The adapter uses its own default values and reports successful completion of the insertion process if no RPS is present.

Token Ring Frame Format

The two types of frame formats are tokens and data/command frames, as displayed in Figure 4-11 and Figure 4-12, respectively. Tokens are 3 bytes in length and consist of a start delimiter, an access control byte, and an end delimiter. Data/command frames vary in size, depending on the size of the Information field. Command frames contain control information and do not carry upper-layer protocols.

Figure 4-11 Token Frame (No Data)

Figure 4-12 Data/Command Frame

Table 4-7 contains an explanation of the fields in Figures 4-11 and 4-12.

Table 4-7 Token Ring Frame Fields Descriptions

Token Ring Interface

As shown in Example 4-2, the Token ring interface maximum transmission unit (MTU), speed, MAC, and other information can be checked with the show interface tokenring 0 command. From the router output, you can see that the encapsulation is SNAP, the ring speed is 16 Mbps, and the interface MTU is 4464 bytes.

Example 4-2 Token Ring Interface

router> show interface tokenring 0
TokenRing0 is up, line protocol is up
  Hardware is TMS380, address is 0007.0d26.7612 (bia 0007.0d26.7612)
  Internet address is 133.5.88.8/24
  MTU 4464 bytes, BW 16000 Kbit, DLY 630 usec,
     reliability 255/255, txload 1/255, rxload 1/255
Encapsulation SNAP, loopback not set
  Keepalive set (10 sec)
  ARP type: SNAP, ARP Timeout 04:00:00
  Ring speed: 16 Mbps
  Duplex: half
  Mode: Classic token ring station
  Group Address: 0x00000000, Functional Address: 0x08000000
  Ethernet Transit OUI: 0x000000
  Last input 00:00:00, output 00:00:00, output hang never
  Last clearing of "show interface" counters never
  Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
  Queueing strategy: fifo
  Output queue :0/40 (size/max)
  5 minute input rate 0 bits/sec, 0 packets/sec
  5 minute output rate 0 bits/sec, 0 packets/sec
     1278 packets input, 55402 bytes, 0 no buffer
     Received 1007 broadcasts, 0 runts, 0 giants, 0 throttles
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored, 0 abort
     331 packets output, 26189 bytes, 0 underruns
     0 output errors, 0 collisions, 1 interface resets
     0 output buffer failures, 0 output buffers swapped out
     3 transitions

Wireless LANs

WLANs provide the capability to access internetworking resources without having to be wired to the network. WLAN applications include inside-building access, LAN extension, outside building-to-building communications, public access, and small-office home-office (SOHO) communications. For the CCIE written test, focus on the IEEE 802.11b standard.

The first standard for WLANs is IEEE 802.11, approved by the IEEE in 1997. IEEE 802.11 implemented WLANs at speeds of 1 Mbps and 2 Mbps using Direct Sequence Spread Spectrum (DSSS) and Frequency Hopping Spread Spectrum (FHSS) on the physical layer of the OSI model. DSSS divides data into separate sections, and each section is sent over different frequencies at the same time. FHSS uses a frequency hopping sequence to send data in bursts. With FHSS, some data is transmitted at frequency 1; then the system hops to frequency 2 to send more data, and so on, returning to transmit more data at frequency 1.

Current implementations use the IEEE 802.11b standard. The IEEE 802.11b standard is referred to as high-rate and provides speeds of 11, 5.5, 2, and 1 Mbps. An interoperability certification exists for IEEE 802.11b WLANs called Wireless Fidelity (Wi-Fi). The Wi-Fi certification is governed by the Wireless Ethernet Compatibility Alliance (WECA). IEEE 802.11b uses DSSS and is backward-compatible with 802.11 systems that use DSSS. The modulation techniques used by IEEE 802.11b are as follows:

• Complementary Code Keying (CCK) at 5.5 and 11 Mbps

• Differential Quadrature Phase Shift Keying (DQPSK) at 2 Mbps

• Differential Binary Phase Shift Keying (DBPSK) at 1 Mbps

A description of each modulation technique is outside the scope of the CCIE written test and is not covered in this book.

Service Set Identifier (SSID)

WLANs use a SSID to identify the network name of the WLAN. The SSID can be 2 to 32 characters in length. All devices in the WLAN must have the same configured SSID to communicate.

WLAN Access Method

The IEEE 802.11 MAC layer implements carrier sense multiple access with collision avoidance (CSMA/CA) as an access method. With CSMA/CA, each WLAN station listens to see if a station is transmitting. If there is no activity, the station transmits. If there is activity, the station uses a random countdown timer. When the timer expires the station transmits.

WLAN Modes

WLAN architecture has three modes of operation. The first mode is the Basic Service Set (BSS). In BSS mode, all stations communicate with the Access Point (AP). The AP provides communication between clients and connects the WLAN network with the wired LAN. As shown in Figure 4-13, in BSS mode, clients do not communicate directly with each other; all communication is through the AP. BSS is also referred to as infrastructure mode.

Figure 4-13 BSS (Infrastructure) Mode

The second mode is the Independent Basic Service Set (IBSS). In IBSS mode, stations communicate directly with each other without using an AP. IBSS is also known as Ad hoc mode. Figure 4-14 displays IBSS (Ad-hoc) mode.

Figure 4-14 IBSS (Ad-hoc) Mode

The third mode is the Extended Service Set (ESS). As shown in Figure 4-15, ESS is a set of BSS where APs have connectivity in the WLAN, which provides a distribution system for roaming capabilities.

Figure 4-15 Wireless ESS Mode

Frequencies Used by WLANs

The IEEE 802.11b standard uses the 2.4 GHz band of the Industrial, Scientific, and Medical (ISM) frequencies. The Federal Communication Commission (FCC) authorizes ISM frequencies for unlicensed use in the United States. The three ISM frequency bands are as follows:

• 902 to 928 MHz
• 2.4000 to 2.5000 GHz
• 5.725 to 5.875 GHz

IEEE 802.11, 802.11b, and 802.11g standards all use the 2.4 ISM band.

The IEEE 802.11a standard uses the 5 GHz bands of the Unlicensed National Information Infrastructure (UNII) frequencies. The three UNII bands are as follows:

• 5.15 to 5.25 GHz (lower band)
• 5.25 to 5.35 GHz (middle band)
• 5.75 to 5.85 GHz (upper band)

WLAN Security

WLANs without any encryption present a security risk because the SSIDs can be snooped by using publicly available software. The IEEE 802.11 standard specifies the use of the Wired Equivalency Privacy (WEP) for encryption. WLANs use two types of WEP keys: 64-bit and 128-bit. Although WEP provides additional security, it has some weaknesses that provide security risks. By gathering (snooping) traffic, hackers can obtain the WEP keys by using freeware software.

Some APs can implement MAC address and protocol filtering to enhance security or limit the protocols over the WLAN. Again, MAC address filtering can be hacked.

To enhance security, WLANs can be implemented with Virtual Private Network (VPN) software or use the IEEE 802.1x port-based access control protocol. IEEE 802.1x is covered in the LAN security section in this chapter.

Cisco also provides dynamic per-user, per-session WEP keys to provide additional security over statically configured WEP keys, which are not unique per user. For centralized user-based authentication, Cisco developed the Cisco Extensible Authentication Protocol (LEAP), which uses mutual authentication between the client and the network server and uses IEEE 802.1x for 802.11 authentication messaging. LEAP uses a RADIUS server to manage user information.

New and Future WLAN Standards

The IEEE 802.11a standard provides an increase of throughput from 802.11b with speeds up to 54 Mbps. IEEE 802.11a uses the 5 GHz bands of the UNII frequencies. For this reason, it is not backward-compatible with 802.11b WLANs.

IEEE 802.11g is an emerging standard that provides faster WLAN speeds in the ISM 2.4 GHz band. IEEE 802.11g is backward-compatible with 802.11b WLANs.

IEEE 802.11d provides specifications for WLANs in markets not served by the current 802.11, 802.11b, and 802.11a standards.

IEEE 802.11i provides enhancements to the security and authentication protocols for WLANS.

The emerging IEEE 802.15 standard provides specifications for Wireless Personal Area Networks (WPANs). The emerging IEEE 802.16 standard provides specifications for fixed Broadband Wireless Access.