Wireless Networking

IEEE 802.11 Standards

 

The IEEE 802.11 standard defines the MAC layer (sublayer of Data Link layer) and Physical layer specifications for Wireless LANs with data rates of 1 and 2 Mbps. It was first completed in 1997 but went through several changes until it was finalized in 1999. The MAC layer specifications are concerned with how the network devices access the medium, in this case, the air. 802.11 uses the 2.4 GHz frequency band. The Physical layer specifications in 802.11 define standards for three different radio technologies: Direct Sequence Spread Spectrum (DSSS), Frequency Hopping Spread Spectrum (FHSS), and InfraRed (IR).

In 2001, the IEEE approved two new amendments for the original 802.11 standard, but with additions to the Physical layer specifications: the 802.11a and 802.11b standards. The term 802.11x is sometimes used to refer to the entire group of 802.11 WLAN standards, of which some are still under development. It includes the standards outlined above, as well as several others addressing the need for speed, region specific regulations, and security. Do not confuse 802.11x with 802.1x, the layer 2 port-based authentication protocol that provides authenticated access to 802.11 wireless networks and wired Ethernet networks.

802.11b

The first wireless networking products that became widely available are based on the extended IEEE 802.11b standard. Because of the availability and affordability of 802.11b equipment, it has become popular especially in small and home networks. According to the standard, 802.11b provides data rates of 5.5 and 11Mbps, and is backwards compatible with the 1 and 2 Mbps data rates of 802.11. An organization called Wi-Fi Alliance, formerly known as the Wireless Ethernet Compatibility Alliance (WECA), is concerned with the compatibility of 802.11b equipment from different manufacturers. When products based upon the 802.11b standard pass the compatibility tests performed by the Wi-Fi Alliance, they are awarded the WiFi (Wireless Fidelity) logo. 802.11b operates on the 2.4 GHz frequency band just like 802.11.

802.11a

It took several years before a wide range of products based upon the 802.11a standard became available. When they finally did around 2002, more companies became interested in wireless networking. The primary reason for this is that the 802.11a standard increases the maximum data throughput to 54 Mbps. However, 802.11a is not backward compatible with 802.11 and 802.11b because it uses the 5 GHz frequency band instead of 2.4 GHz, and a different modulation scheme (OFDM instead of QPSK).

802.11g

The 802.11g standard also allows data transfer rates up to 54 Mbps, but is backward-compatible with both 802.11 and 802.11b, supporting both their data rates (1, 2, 5.5, and 11 Mbps) and modulation scheme (QPSK). 802.11g also supports the modulation scheme used by 802.11a (OFDM), but is not compatible with 802.11a because 802.11g uses the 2.4 GHz frequency band. While both standards define 54 Mbps as the maximum throughput, in reality it is closer to 50% of that.

802.11 Network Operation

802.11 can be considered the wireless equivalent of 802.3 wired Ethernet, but there are some major differences. The first one is obviously the media, which carries the network traffic. For wired Ethernet networks, the medium is the cable, a copper or fiber cable for example, and the network traffic is a collection of electrical signals or light pulses. For a wireless network, the media is the air, and the network traffic is a radio wave on a particular frequency in the Radio Frequency (RF) spectrum.

Not just the media, but also the access method of an Ethernet network and an 802.11 wireless network is different. Ethernet networks use the Carrier Sense Multiple Access/Collision Detection (CSMA/CD) access method. This means a station listens to check if the cable is currently accessed before starting its own data transmission. When two stations both determine the media is available and start sending the data simultaneously, a collision occurs, meaning the electrical signals on the cable collide with each other. When the collision is detected, both stations will retransmit the data after a different and random amount of time determined by a backoff algorithm.

In a wireless 802.11 network, the access method is Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA). This means that if a wireless station determines the network is busy, it will also back off for a random amount of time, but because collisions between wireless signals cannot be detected, acknowledgments are used to inform stations of the availability of the media. If an acknowledgment is not received for a packet, the sender will assume that a collision occurred and the packet was lost. CSMA/CA in wireless networks does not operate exactly the same as CSMA/CA in some wired networks such as LocalTalk, in which broadcasts are used to notify other stations the network is busy and thereby avoiding collisions.

Ad Hoc and Infrastructure Mode

Wireless stations can be interconnected in a peer-to-peer network, also known as an ad-hoc network or Independent Basic Service Set (IBSS). In this type of wireless LAN clients communicate directly with each other and do not use an access point. Wireless stations need to be configured with a matching Service Set Identifier (SSID) and a channel. The SSID is a case sensitive, alphanumeric value of 2-32 characters long, used to define a wireless network. The channel defines a specific range in the RF spectrum, and only a portion of the available frequency range. This allows multiple wireless networks to coexist in the same area as long as they use a different channel. To be able to communicate with each other through TCP/IP they must of course have an IP address in the same IP subnet.

Wireless LAN in Ad-Hoc Mode

In a wireless LAN in i nfrastructure mode , network stations are interconnected through one or more Wireless Access Points (WAPs), which results in a star network topology similar to a wired Ethernet network using switches and hubs. The w ireless stations are configured with an SSID that matches the SSID configured on the WAP. The channel is configured on the WAP only.

Wireless LAN in Infrastructure Mode

Wireless Access Point (WAP)

A wireless access point (WAP) is essentially a hub equipped with one or more, fixed or removable antennas to provide wireless capabilities. Just as with a wired hub, all of the attached clients share the available medium. A major difference with a wired hub is that WAPs have Layer 2 functionality. As you may remember from the first paragraph in the IEEE 802.11 Standards section above, the 802.11 standards includes specifications for the Physical and the MAC layer (sublayer of Data Link layer). When a wireless client wants to ‘attach’ to a WAP to send and receive data through it, it first needs to be associated and authenticated with the WAP. The client initiates this process by broadcasting its own MAC address to identify itself to the WAP. So a WAP, unlike a hub, reads the MAC addresses in data frames, hence operates at the bottom two layers of the OSI model: the Physical layer and the Data Link layer. We will cover this process in more detail in the Wireless Network Security section below.

Wireless Access Point (WAP)

Besides interconnecting wireless network stations, a WAP is typically used to integrate wired and wireless stations into a single network. Actually, most WAPs offer additionally functionality, such as bridging/switching and routing (Layer 3 / Network Layer) even. Many consumer WAPs have several built-in wired switch ports as USB and UTP interfaces, but the enterprise WAPs used in commercial applications do not. Instead, it usually provides a single bridge port to connect the WAP to a wired network, allowing it to function as a gateway bridge for wireless clients.

Remember that a switch is just a multiport bridge, so if you interconnect LAN segments with a switch, you are effectively bridging. For example, a WAP can function as a bridge between a wired Ethernet 802.3 and a wireless 802.11b network. As depicted in the image below, the stations on the wired and the wireless network become part of the same LAN, and use an IP address from the same IP subnet. In other words, they become part of the same broadcast domain. However, the stations on the wireless segment are all in the same collision domain, while every wired connection on the WAP (and each connection behind the switch) is its own collision domain.

WAP functioning as bridge between wired and wireless segment

Almost every wireless network connects to a wired network at some point. In a SOHO network, the wired network is often the Internet connection. The following diagram shows a typical example in which the WAP functions as a router, also referred to as a wireless access router.

WAP functioning as router

The wireless access router appears as the only client to the ISP, and usually performs NAT for the connected wireless clients instead of actually created a routed connection. A DHCP server, running on the WAP, assigns the clients an IP address from a private address range. A WAP functioning as a router, operates also on the Network layer of the OSI model an separates broadcasts and collision domains just like a router between two wired networks would.

Another use of a wireless access point is to extend network connection by functioning as a repeater. Remember that a repeater amplifies the signal to allow a connection to span a larger distance. The following diagram depicts a simple network with a wireless access point repeating the signal to allow it to cross the distance. The preferred method would be to use a more power antenna.WAP functioning as repeater

Read the Internet Connections TechNotes for more information about routed and translated connections, and read the Network Components TechNotes for more information about bridges, switched, routers, and broadcast and collision domains.

Modern WAPs often support multiple 802.11 standards, often referred to as ‘modes’. For example, B-only Mode (802.11b), G-only Mode (802.11g), and B/G (mixed) Mode. The latter allows for a mixed environment, which sounds convenient but should only be used temporarily, e.g. when migrating from 802.11b to 802.11g, but preferably be avoided entirely. Allowing both 802.11b and 802.11g clients to connect to the same access point has significant negative impact on the performance. In B-only mode, the AP only uses DSSS allowing up to 11 Mbps of bandwidth (in reality about 5.5 Mbps of data throughput). In G-only mode, the AP only uses OFDM, and only 802.11g clients can connect to the AP, allowing up to 54 Mbps of bandwidth (in reality about 20 Mbps of data throughput). In mixed mode, the AP uses both DSSS and ODFM, supporting both 802.11b and 802.11g, but reducing the total realistic data throughput from 20 Mbps to 8 Mbps or less, when both 802.11b and 802.11g clients are connected.

Antennas

Although most WAPs and wireless network cards have integrated antennas, many of them allow an external antenna to be connected. The main advantage is that a different, more powerful antenna can be connected to increase the maximum range. The latter is important to provide proper connectivity to clients or other wireless devices. The further a client is located from a WAP, the weaker the signal it receives will be. If a client is entirely out of range, the signal will be too weak, preventing the client from connecting. Another advantage of using an external antenna is that it can be placed outdoors while connected to a WAP indoors. The cable running from the WAP or WLAN NIC to the antenna is often a proprietary cable, also referred to as a pigtail. The connectors for these pigtails differ per brand. High-end antennas typically use a coaxial cable with an F-type connector. The quality and length of the cable have a significant influence on the signal power.

Following are the three main types of antennas, omni, semi, and highly-directional, and Each type has its own characteristics and suitable purposes and are discussed below.

Omni-directional

This is the most common type of antenna on WAPs and wireless network cards, usually referred to as a dipoleantenna or just dipole (similar to a magnet having two poles). The ‘rabbit ears’ on older TVs is a classic example of a dipole antenna. The following image depicts a simple dipole antenna; the donut-shaped wireframe show how the signal propagates.

Omni antenna – Side view	Omni antenna – Top view

Omni-directional antennas are particularly suitable for point-to-multipoint connections and a re often used in conjunction to create a large wireless network with to a cell topology as depicted below. At the center of these cells is an omni antenna of which the signal coverage overlaps slightly with adjacent cells to provide full coverage. The cell topology with omni antennas is used both indoors and outdoors. The most know outdoor example is of course the cell phone network using GSM or UMTS for example.

Cells in a wireless network

Semi-Directional

The Yagi and the Patch antenna are two common types of semi-directional antennas. They each use a similar method to force the signal to propagate into a certain direction. The Yagi antenna is typically used for point-to-point and point-to-multipoint connections outdoors. The element with the black protective cover on the Yagi antenna below is the one connected to the cable. The element left from it reflects the signal while the elements on the right help propagate the signal in the right direction.Directional Yagi Antenna

A directional patch antenna contains a back panel and a ‘patch’ inside the box as depicted in the image below on the right. The back panel performs the same function as the reflective element in the Yagi antenna, it reflects the signal originating from the patch and forces it to propagate into a certain direction. Patch antennas are typically used for point-to-multipoint connections outdoors and point-to-point connections indoors.

Directional Patch Antenna

Highly-Directional

Highly-directional antennas are most suitable for outdoor long distance point-to-point connections. The p arabolic dish and the parabolic grid are the most common examples of highly-directional antennas. The main difference between them is the structure of the antenna, which enables the parabolic grid to withstand strong winds.

Parabolic Dish Antenna	Parabolic Grid Antenna

The following image depicts the RF beam of a highly-directly antenna. The dish or grid bundles the signal and forces the signal to propagate in a specific direction . An important requirement for point-to-point connections using highly-directional antennas is a clear line of sight. If a building or other large obstacle blocks the visibility between two highly-directional antennas, the narrow beam will neither be able to avoid the obstacle.

Highly-directional antenna

Besides point-to-point connections, directional antennas are often used in combination with omni antennas to create point-to multipoint connections. For example, the main building on a campus could have an omni antenna on the roof and several smaller buildings each use a directional antenna to connect to the main building.

Environmental Factors and Interference

Radio Frequency Interference (RFI) is one of the major challenges for wireless networks. RFI refers to interference of other devices that operate on the same radio frequency, which can cause delays, hence reduced data throughput and even loss of data and connectivity. Common sources of RFI in 802.11b and 802.11g networks are cordless phones, microwave ovens, neighboring wireless LANs, and Bluetooth devices, which like the former, operate in the 2.4-GHz band.

One of the major factors on the signal quality in wireless networks is the environment. Walls, plants, windows, office equipment and human beings are examples of obstacles in indoor wireless networks that can have a negative impact. Outdoors, trees, mountains, lakes, buildings and other structures absorb or reflect the signal causing undesired results.

The primary cause of signal power in a wireless network is path loss, which refers to the signal power decreasing by distance. This is commonly compared to a flashlight: as the distance increases, the beam gets wider shining less light on its target. The same thing happens with an RF wave, and it is also known as dispersion. Eventually the signal will be to low for the receiving antenna to separate data from noise.

Wireless Network Security

WEP (Wired Equivalent Privacy)

The Wired Equivalent Privacy (WEP) protocol is part of the 802.11 standard and is developed as an effort to provide privacy in wireless networks similar to privacy in wired networks. Intercepting traffic (eavesdropping) on a wireless network is very easy compared to wired networks, so it is essential that data frames are encrypted.

When a station powers up, and attempts to establish a wireless connection, it will first be associated with an access point. When the station is associated, it will attempt to authenticate itself to the access point. The original IEEE 802.11 standards provide the following two types of authentication:

• Open System Authentication -The client broadcasts its MAC address to identify itself, an AP replies with an authentication verification frame. Although its name implies differently, no actual authentication occurs when Open System Authentication is used.
• Shared Key Authentication - The client will be authenticated only if it is configured with a preshared key. This means that the same key must be configured on both the client station and the AP. The AP sends a challenge text to the client requesting authentication, which is encrypted using WEP and the shared key at the client, and then send back to the AP where it is decrypted again to see if it matches the original challenge. If it matches, the client will be able to start transmitting and receiving data and participate in the network.

The WEP key used for authentication can also be used to encrypt the wireless traffic. Although this produces a significant amount of overhead on low-speed wireless networks such as 802.11b, it should be enabled if no other, better, security method is used instead of WEP. WEP offers rather weak security that can be relatively easily cracked using widely available tools that can be found free on the Internet. WEP uses the RC4 cipher, with key lengths of 64 and 128-bits. The key includes a 24-bit Initialization Vector (IV), which is sent in clear text format, so actually the secret portion of the WEP keys are 40 and 104-bits.

In 2004 the 802.11 standard was updated by the IEEE 802.11i workgroup in an effort to increase security in wireless networks. This resulted in WPA and WPA2, which should be used instead of WEP whenever available.

WPA (Wi-Fi Protected Access)

Wi-Fi Protected Access (WPA) is a standard released by the Wi-Fi Alliance to provide more advanced security for 802.11 wireless networks, including stronger encryption and authentication methods. It includes the Temporal Key Integrity Protocol (TKIP), which offers dynamic key distribution. RC4 is still used for the actual data encryption, but they key used to encrypt the data is changed periodically. This makes it harder to crack than WEP. WPA also supports authentication through the IEEE 802.1x port-based authentication protocol.

While WPA and TKIP were designed as a more secure replacement for WEP, partly because RC4 was still used several flaws were quickly discovered. The Wi-Fi Alliance released WPA2 in 2004, which adds support for the Advanced Encryption Standard (AES) to allow for much stronger encryption.

The best approach to secure wireless networks is to implement multiple layers of security. A secure wireless network employs security technologies such as IPSec, 802.1x authentication, and strong encryption protocols. A c ostly but secure option is using Virtual Private Networks (VPN), and basically treat the wireless network as if it were a public network like the Internet.

InfraRed

The Infrared Data Association (IrDA) creates standards for short-distance infrared communication. It defines data transmission rates from 9600 bps with primary speed/cost steps of 115 Kbps, a maximum of 4 Mbps for Fast IrDA and a maximum of 16 Mbps for Very Fast IrDA . Infrared is commonly used to exchange data between mobile devices and other devices such as printers. The primary limitation of infrared communication is that it requires a clear line-of-sight. IR waves are easily absorbed and reflected by obstacles, which is the main reason why network communication using infrared light is not very popular.

Bluetooth

Bluetooth is a wireless networking technology that is very common in Personal Area Networks (PANs). A PAN covers the area directly surrounding the user, and typically includes handheld devices such as PDAs and smartphones. Wireless connectivity allows users to synchronize files, email, and connect to printers and other network devices. Bluetooth is also popular for providing connectivity between non-network devices, such as connecting hands-free sets to phones or controllers to videogame systems. The specified maximum data transfer rate is 1 Mbps, but in reality it is much lower. Bluetooth with Enhanced Data Rate offers transfer rates up to 3 Mbps, but again the effective rates are much lower.

Bluetooth operates in the unlicensed ISM band at 2.4GHz and uses the FHSS (Frequency Hopping Spread Spectrum). It avoids interference from other signals by hopping to a new frequency after transmitting or receiving a packet. Compared to other systems in the same frequency band, the Bluetooth radio hops faster and uses shorter packets. Bluetooth's hop rate of 1,600 hops per second over 79 channels means the chance of other signals interfering is very low, but the hopping also limits the maximum transfer rates. The main reason why Bluetooth uses FHSS instead of DSSS is that FHSS is much simpler, requiring less powerful chips, using less power. The latter is obviously very important for handheld devices that run on batteries.