OSI and TCP/IP Reference Models
Several different but similar layered data architectures have been developed to allow reliable data transfer between different computer systems and between different networks. When you understand a little about the service layers and the protocols that these architectures use, you will be in a good position to understand the similarities and the differences between different brands of wireless equipment.
The seven-layer ISO OSI reference model was first proposed around 1983 to allow connectivity (or interworking) between different computer systems. Prior to the OSI reference model, computer systems made by one manufacturer could not easily communicate with computer systems made by other manufacturers. The intent of the OSI reference model was to allow computer systems to successfully communicate with each other even though different vendors manufactured them. Figure 6-1 shows the OSI reference model alongside the TCP/IP architecture.
Figure 6-1 OSI and TCP/IP
Beginning in the 1970s, the United States Department of Defense began promoting computer networking between university research departments and government installations. One of the primary goals of this internetworking effort was to develop a survivable networkone that would be able to continue communicating even if some of the network nodes or some of the communications links were destroyed. This new networking effort was based on two primary protocols: the Transmission Control Protocol (TCP) and the Internet Protocol (IP). TCP performed transport layer functions equivalent to the transport layer in the OSI model. IP performed network layer functions that were equivalent to the network layer (Layer 3) in the OSI model. When application layer (Layer 7) protocols (Telnet, FTP, SMTP, and so on) and physical (Layer 1) and data link layer (Layer 2) protocols were added, the result was an architecture that effectively contained five layers. Figure 6-1 shows the TCP/IP model alongside the OSI reference model for comparison.
The TCP/IP architecture is functionally equivalent to the OSI reference model. The major similarities and differences are as follows:
Both models have an application, a transport, and a network/Internet layer.
The TCP/IP model does not have a session layer (Layer 5 of the OSI reference model) or a presentation layer (Layer 6 of the OSI reference model).
Both models have a lower layer that connects the upper layers to the actual physical network. In the OSI reference model, the lower layer (Layer 1) is called the physical layer. In the original TCP/IP model, the lower layer was called the host-to-network layer. In present-day use, TCP/IP networks use the combination of a Layer 2 sublayer called the medium access control (MAC) sublayer along with Layer 1 to provide connectivity over the wireless link.
Virtually all the wireless equipment features that you evaluate operate at the physical, data link, and network layers of the OSI and TCP/IP reference models. The wireless features and functionality (modulation type, data rate, and so on) take place at the physical layer. Access to (and sharing of) the wireless medium takes place at the data link layer. Routing takes place at the network layer.
Peer protocols run across the Internet but provide communication only between same-layer processes. One example of this same-layer communication process is a Hypertext Transfer Protocol (HTTP) web browser running on the application layer of one network. The HTTP browser retrieves information from its peer web server running on the application layer of another network. Although the HTTP communication is application-layer-to-application layer (peer-to-peer), both networks communicate downward through their lower network layers.
Information is passed from the top (application) layer of one network down through the lower layers. Each layer provides a set of services for the layer just above it and utilizes the services provided by the layer just below it. The set of services between two layers is referred to as the interface between the two layers. For example, Layer 6 provides services for Layer 7; Layer 5 provides services for Layer 6; and so on. In this way, Layer 7 (the application layer) can communicate all the way down to Layer 1 (the physical layer).
The following list illustrates how services and protocols operate. When your web browser uses HTTP over a wireless network, the information flow is as follows:
The HTTP information request originates at the application layer on the originating network.
The HTTP request travels downward from the application layer (using the services provided by all the intermediate layers) to the physical layer on the originating network. The physical layer uses the appropriate wireless protocol (for example, the appropriate direct sequence spread spectrum modulation or DSSS) to communicate the request over the air wirelessly to the physical layer on the other network.
The physical layer on the other network uses the DSSS protocol to receive the request from the originating network. The physical layer then passes the information up through its interface to the data link layer. Using higher and higher layer services, the request passes upward until it eventually reaches the application layer. There, the HTTP protocol processes the request and replies with a response.
Using services of lower and lower layers, the response travels downward to the physical layer. Using the Layer 1 DSSS protocol, the response is transmitted over the air back to the physical layer of the originating network.
Using services, the originating network passes the response upward to the application layer where the HTTP protocol receives the response to its original request.
Basic Packet Structure and Frame Types
Packet switching store-and-forward techniques underlie the operation of layered architectures. Packet switching uses an underlying data structure, called a packet. A packet is like a hamburger on a bun. Although the exact packet structure varies from layer to layer and protocol to protocol, most packets contain a data payload section. The data payload is the hamburger in the middle of the bun. Other fields encapsulate (surround) the data payload and make up the bun. The bun typically provides the following:
Source and destination address information
Packet numbering information
Packet acknowledgment information
Error detection and correction information
Figure 6-2 shows this general packet structure.
Figure 6-2 General Packet Structure
Packets prepared by Layer 2 (the data link layer) are called frames. Not all frames contain payload data. Wireless APs and wireless stations exchange three types of frames, each with the following functions:
Data frames carry user payload data (the hamburger) between different wireless network nodes.
Control frames carry information such as request-to-send (RTS) and clear-to-send (CTS) messages as well as frame acknowledgments (ACK).
Management frames carry association and authentication requests and responses in addition to beacon information.
Application Layer Functions and Protocols
The application layer is where the end user programs run. Telnet, Simple Mail Transfer Protocol (SMTP), File Transfer Protocol (FTP), and HTTP are examples of application layer protocols. Wireless equipment that you evaluate will likely have network management software that operates at the application layer level.
Transport Layer Functions and Protocols
The transport layer's job is to provide reliable communications from application to application regardless of the lower-layer protocols and communications links. The transport layer encapsulates data from the application layer (and the session layer, if used) and passes it down to the network layer.
Typical transport layer protocols are TCP and User Datagram Protocol (UDP). Wireless equipment that you evaluate does not usually have features that operate at the transport layer level.
Network Layer Functions and Protocols
The essential network layer protocol is IP. In addition to IP, the network layer often utilizes other routing protocols such as Routing Information Protocol (RIP) and Border Gateway Protocol (BGP).
The network layer encapsulates data (the hamburger) from the transport layer between IP source, IP destination, and IP routing information. Packet routing typically goes from intermediate network to intermediate network before the packets finally arrive at their destination network.
Data Link Layer Functions and Protocols
The data link layer includes the logical link control (LLC) sublayer and the MAC sublayer. The data link layer normally performs a wide variety of functions, including segmenting the bit stream into frames, error handling, flow control, and access control.
Examples of data link layer protocols include Point-to-Point Protocol (PPP) and Spanning Tree Protocol.
LLC Sublayer Functions and Protocols
The LLC sublayer makes up the top half of the data link layer and interfaces to the network layer (above) and the MAC sublayer (below). The LLC Sublayer encapsulates the Layer 3 data by adding sequence and acknowledgment numbers. The LLC Sublayer might provide different service options, depending on the network software.
MAC Sublayer Functions and Protocols
The MAC sublayer makes up the bottom half of the data link layer. The MAC sublayer interfaces to the physical (wireless) layer and provides the following functions:
Reliable deliveryThe MAC sublayer provides a reliable delivery mechanism that looks for an acknowledgment for every frame that is sent. If an acknowledgment is not received, the MAC sublayer retransmits the frame.
Access controlThe MAC sublayer controls access to the wireless channel. The two basic types of access control are carrier sense multiple access with collision avoidance (CSMA/CA) and polling. CSMA/CA is a distributed coordination function (DCF) because the decision about when to transmit is distributed to all wireless stations. Each wireless station listens before transmitting. If a station hears that the frequency is busy, it backs off (waits) a random amount of time and tries again. When the frequency is clear, the station proceeds to transmit. In addition, a request-to-send/clear-to-send (RTS/CTS) mechanism can be enabled. Large packets are more likely to collide; therefore, stations that have packets larger than the RTS/CTS threshold must request and receive clearance from the AP MAC before they can transmit their packets. Finally, in networks that have heavy traffic and many end users, a point coordination function (PCF) can be used. One single point (the MAC in the AP) coordinates transmissions from all stations. The PCF polls each wireless station and then tells each end user when it can transmit.
You might have heard of wireless networks with a hidden node problem. This problem can occur in a network that uses DCF. In most wireless networks, certain wireless stations cannot hear all the other wireless stations. Under heavy traffic loads, several stations might try to transmit at the same time. This can happen even when the stations are using RTS/CTS. When stations transmit at the same time, collisions occur and network throughput drops drastically. The solution to hidden-node problems is to use wireless equipment that can support PCF.
EncryptionThe MAC also provides encryption. The most frequently used encryption method is wired equivalent privacy (WEP).
MAC control frame subtypes include RTS, CTS, and ACK. Examples of management frame subtypes include Beacon, Probe Request, Authentication, and Association Request.
Physical Layer Functions and Protocols
The physical layer transports encapsulated data from the data link layer and transmits it wirelessly to the distant network. There are several physical layer wireless standards. There are also many proprietary physical layer wireless protocols. In addition to your wireless feature evaluation, you will evaluate physical layer wired-interface features such as Ethernet and serial port features.