The International Standards Organization (ISO) has developed a reference model for network design called the Open Systems Interconnection (OSI). It proposes a seven-layer architecture for networks. Each layer is characterized by a set of standard protocols which specify its behavior.
These seven layers represent the protocol architecture for the communications component of a host. The nodes in a network implement only the lower three layers. The reason for this is that the upper four layers are irrelevant to the task of communication between the nodes.
In Figure bellow, when host A sends a message to host B, the message moves down the successive layers of host A, from the application layer to the presentation layer, to the session layer, etc., until it reaches the physical layer. It is then transmitted across the communication line between host A and node X, and moves up the three layers of node X and down again. Then it is transmitted to node Y where it goes through the same procedure, and finally is transmitted to host B, where it moves up its seven layers, until it arrives at the application layer of host B.
The terms protocol and layer are often used interchangeably. This is harmless but not entirely accurate. Strictly speaking, protocol refers to the rules and conventions that the functions of a layer should conform to. Layer refers to a set of services and functions and their realization in hardware or software. A layer is therefore characterized by its protocol. A set of network layers is also commonly referred to as a protocol stack.
Each of the seven layers of the OSI model hides the implementation details of the lower layers from the upper layers. Well-defined protocols and interfaces for each of the layers make it possible for the layer to be designed and implemented in isolation from the other layers. Except for the physical layer, which is implemented in hardware, all other layers are implemented in software. For example, each of these layers may be implemented as a set of routines which communicate with the layer above and the layer below it via parameters passed in function calls.
OSI layers as software tasks
Each layer adds an additional piece of information to the message it is transmitting. The same layer removes the additional piece of information on the receiving end. The additional information appears in form of a header (e.g., TH = Transport Header). The data link layer adds a header as well as a trailer to its data.
The Physical Layer
The physical layer is concerned with the transmission of raw data bits over communication lines. Physical layer standards and protocols are concerned with issues such as the following:
• How a physical circuit is established between communicating devices.
• How the circuit is terminated when no longer needed.
• The physical form (e.g., voltages, frequencies, timing) in which data bits (binary values 0 and 1) are represented.
• Whether transmission of data can take place in one or both directions over the same physical connection.
• Characteristics of the physical media that carry the signals (e.g., copper wire, optical fiber, radio waves).
• Characteristics of the connectors used for connecting the physical media.
• How data from a number of sources should be multiplexed before transmission and demultiplexed upon arrival, and the type of multiplexing technique to be used.
• The type of modulation to be used for transmitting digital data over analog transmission lines.
The physical layer accounts for much of the tangible components of a network, including cables, satellites, earth stations, repeaters, multiplexers, concentrators, and modems. Physical layer protocols and standards are of mechanical, electrical, functional, and procedural nature.
The Data Link Layer
The data link layer is concerned with the reliable transfer of data over the communication channel provided by the physical layer. To do this, the data link layer breaks the data into data frames, transmits the frames sequentially over the channel, and checks for transmission errors by requiring the receiving end to send back acknowledgment frames. Data link protocols are concerned with the following issues:
• How to divide the data into frames.
• How to delimit frames by adding special bit patterns to the beginning and end of each frame. This allows the receiving end to detect where each frame begins and where it ends.
• Error detection. Some form of error check is included in the frame header. This is constructed by the transmitting end based on the contents of the frame, and checked for integrity by the receiving end. A change in the frame bits can be detected in this way.
• Error correction. When a frame arrives corrupted or is for any reason lost in the network, it is retransmitted. Lost acknowledgment frames may result in duplicate frames, which need to be detected and corrected as well.
• Flow control. In general, not all communication devices in a network operate at the same speed. Flow control provides a means of avoiding a slow receiver from being swamped by data from a fast transmitter.
The data link layer hides the above details from the higher layers. To the network layer, it appears as a reliable communication channel which can send and receive data packets as frames.
The Network Layer
The network layer is concerned with the routing of data across the network from one end to another. To do this, the network layer converts the data into packets and ensures that the packets are delivered to their final destination, where they can be converted back into the original data. Network layer protocols are concerned with the following issues:
• The interface between a host and the network.
• The interface between two hosts across the network.
• Routing of packets across the network, including the allocation of a route and handling of congestion.
• Correct ordering of packets to reflect the original order of data.
• Collection of statistical information (e.g., number of transmitted packets) for performance measurement and accounting purposes.
• Internetworking: communication between two or more networks.
The network layer hides the above details from the higher layers. To the transport layer, it appears as a uniform data transfer service, regardless of the location of the communicating devices and how they are connected.
The Transport Layer
The aim of the transport layer is to isolate the upper three layers from the network, so that any changes to the network equipment technology will be confined to the lower three layers (i.e., at the node level). Transport layer protocols are concerned with the following issues:
• Establishment and termination of host-to-host connections.
• Efficient and cost-effective delivery of data across the network from one host to another.
• Multiplexing of data, if necessary, to improve use of network bandwidth, and demultiplexing at the other end.
• Splitting of data across multiple network connections, if necessary, to improve throughput, and recombining at the other end.
• Flow control between hosts.
• Addressing of messages to their corresponding connections. The address information appears as a part of the message header.
• Type of service to be provided to the session layer (e.g., error-free versus error-prone connections, whether messages should be delivered in the order received or not).
The transport layer hides the above details from the higher layers. To the session layer, it appears as a customized data transfer service between two hosts, isolating the underlying network technology from it.
The Session Layer
The session layer provides a structured means for data exchange between user processes on communicating hosts. Session layer protocols are concerned with the following issues:
• Negotiating the establishment of a connection (a session) between user processes on communicating hosts, and its subsequent termination. This includes the setting of various communication parameters for the session (e.g., synchronization and control).
• Correct ordering of messages when this function is not performed by the transport layer.
• Recovery from interrupted transport connections, if necessary.
• Grouping of messages into a larger message, if necessary, so that the larger message becomes available at the destination only when its constituent messages have all been delivered successfully.
The session layer hides the above details from the higher layers. To the presentation layer, it appears as an organized communication service between user processes.
The Presentation Layer
The presentation layer provides a mutually-agreeable binary representation of the application data communicated between two user processes. Since there are many ways of encoding application data (e.g., integers, text) into binary data, agreement on a common representation is necessary. Presentation layer protocols are concerned with issues such as the following:
• Abstract representation of application data.
• Binary representation of application data.
• Conversion between the binary representation of application data and a common format for transmission between peer applications.
• Data compression to better utilize network bandwidth.
• Data encryption as a security measure.
The presentation layer hides the above details from the higher layers. To the application layer, it appears as a universal communication service between user processes, regardless of their system-specific idiosyncrasies, allowing them to converse in a common syntax.
The Application Layer
The application layer is concerned with the semantics of data, i.e., what the data means to applications. The application layer provides standards for supporting a variety of application-independent services. Examples include:
• Virtual terminal standards to allow applications to communicate with different types of terminals in a device-independent manner.
• Message handling system standards used for electronic mail.
• File transfer, access, and management standards for exchanging files or parts thereof between different systems.
• Transaction processing standards to allow different companies with different systems to access each other’s on-line databases (e.g., in banking and airline reservation).
• On-line directory standards for storing details of individuals, organizations, and network components.
• Standards for exchanging formatted documents.
Application layer standards have paved the way for open software systems, in which data can be communicated between incompatible base systems (i.e., different hardware and software architectures) without loss of meaning or usefulness.
Standards
Standards enable equipment from different vendors and with different operating characteristics to become components of the same network. Standards also enable different networks in different geographical locations (e.g., different countries and continents) to be interconnected. From a customer’s point of view, standards mean real cost savings: the same end-user device can be used for access to a variety of networks and services.
Standards are developed by national and international organizations established for this exact purpose. Important standards developed by various organizations, include the following:
• The International Standards Organization (ISO) is a voluntary organization with representations from national standards organizations of member countries, major vendors, and end-users. ISO is active in many areas of science and technology, including information technology. ISO standards are published as ISO serial-no (e.g., ISO 8632).
• The Consultative Committee for International Telegraph and Telephone (CCITT) is a standards organization devoted to data and telecommunication, with representations from governments, major vendors, telecommunication carriers, and the scientific community. CCITT standards are published as Recommendation L.serial-no, where L is a letter of the alphabet (e.g., I.440). These are revised and republished every four years. CCITT standards are very influential in the field of telecommunications and are adhered to by most vendors and carriers.
• The Institute of Electrical and Electronic Engineers (IEEE) is a US standards organization with members throughout the world. IEEE is active in many electric and electronic-related areas. The IEEE standards for local area networks are widely adopted. IEEE standards are published as IEEE serial-no (e.g., IEEE 908).
• The Electronic Industries Association (EIA) is a US trade association best known for its EIA-232 standard.
• The European Computer Manufacturers Association (ECMA) is a standards organization involved in the area of computer engineering and related technologies. ECMA directly cooperates with ISO and CCITT.
In addition to these organizations, and because of their global market influence, large vendors occasionally succeed in establishing their products as de facto standards.
Connection Types
Connections between devices may be classified into three categories:
• Simplex. This is a unidirectional connection, i.e., data can only travel in one direction. Simplex connections are useful in situations where a device only receives or only sends data (e.g., a printer).
• Half-duplex. This is a bidirectional connection, with the restriction that data can travel in one direction at a time.
• Full-duplex. This is a bidirectional connection in which data can travel in both directions at once. A full-duplex connection is equivalent to two simplex connections in opposite directions.
Signal Types
All signals are either analog or digital. An analog signal is one in which information appears as a continuous variation of some property. Human speech is an example: it produces a continuous variation of air pressure. A digital signal, on the other hand, is one in which information appears as a sequence of binary values 0 and 1. To represent these two values, a signal is used in which only two wave shapes are allowed, one representing the binary value 0 and the other representing the binary value 1. By definition, therefore, a digital signal is a restricted form of an analog signal. Analog and digital signals can both be transmitted through cable or wirelessly. Wireless internet (Wi-Fi), for instance, uses digital. AM and FM Radio uses analog. Cell phones are digital.
Modulation
Transmission of digital data over an analog line is achieved using a technique called modulation, where the digital bit stream is modulated over an analog carrier signal. A modem (modulator and demodulator) is a commonly used device which employs this technique. As illustrated in Figure bellow, a modem converts the outgoing digital bit stream from a device into an analog signal and converts the incoming analog signal into a digital bit stream.
Three basic types of modulation are possible:
• Amplitude Modulation (AM). In AM, the carrier signal’s amplitude is changed according to the modulating digital signal’s bit value. For example, two amplitude sizes (a small and a large one) may be used to, respectively, represent bit values 0 and 1. AM’s main weakness is its susceptibility to distortion.
• Frequency Modulation (FM). In FM, the carrier signal’s frequency is changed according to the modulating digital signal’s bit value. For example, two frequency values (a low and a high one) may be used to, respectively, represent bit values 0 and 1. FM is more resistant to distortion than AM.
• Phase Modulation (PM). In PM, the carrier signal’s phase is changed according to the modulating digital signal’s bit value. A change in the carrier signal’s phase indicates a change in the modulating digital signal’s bit value from 0 to 1 or from 1 to 0.
Digitization
Digitization is essentially the opposite of modulation. Whereas in modulation a digital signal is modulated over an analog signal for transmission, in digitization an analog signal is converted into digital format through a process of sampling. For example, the analog signal resulting from human speech can be sampled and converted into digital data, transmitted over digital lines, and converted back to analog signal at the other end. These two functions are performed by a device called codec (coder/decoder).
It is worth noting that, unlike modulation (which is an exact process since the digital signal at the source and the digital signal received at the destination are identical), digitization is only an approximate process because of sampling. Figure above illustrates how an analog signal is sampled. Here the time interval for each sample is one millisecond. Each sample (denoted by a small black box) is a real value which is in turn represented by an integer in the range 0-255 so that it can be represented in one byte of data. This process (of representing a continuous value with a discrete value) is called quantization. The relatively small loss of information inherent in the process is called quantization error. The main advantage of digitization is that, due to its resistance to distortion, it is much easier to reliably transmit a digital signal over a long distance than an analog signal.
Transmission Media
Digital data can be transmitted over many different types of media. Selecting a transmission medium is guided by comparing transmission requirements against the medium’s characteristics. Four important criteria influence the choice:
1. Bandwidth. Bandwidth is the maximum frequency range that can be practically supported by a medium. This is usually expressed in kilo Hz (kHz) or mega Hz (MHz). For example, analog transmission of human speech typically requires a bandwidth of 4 kHz. Also related, is the notion of data rate, which denotes the maximum number of bits per second (bps) that can be transmitted. For example, a data rate of 10 mbps means that 10 million bits of data can be transmitted in each second. Because of their obvious relationship, the terms bandwidth and data rate are sometimes used interchangeably. Because of distortion factors, bandwidth and data rate are usually inversely proportional to the communication distance.
2. Cost. Two types of cost are relevant: (i) the cost of installing the medium, including the medium-specific equipment that may be needed, and (ii) the cost of running and maintaining the medium and its equipment. There is usually a need for tradeoff between cost, bandwidth, and distance.
3. Reliability. Some media, by their physical nature, transmit data more reliably than others. Low reliability translates into a higher number of errors, which needs to be balanced against the potential cost of recovering from the errors (e.g., retransmission, more complex hardware and software).
4. Coverage. The physical characteristics of a medium dictate how long a signal can travel in it before it is distorted beyond recognition. To cover larger areas, repeaters are needed to restore the signal, and this increases the costs.
These seven layers represent the protocol architecture for the communications component of a host. The nodes in a network implement only the lower three layers. The reason for this is that the upper four layers are irrelevant to the task of communication between the nodes.
In Figure bellow, when host A sends a message to host B, the message moves down the successive layers of host A, from the application layer to the presentation layer, to the session layer, etc., until it reaches the physical layer. It is then transmitted across the communication line between host A and node X, and moves up the three layers of node X and down again. Then it is transmitted to node Y where it goes through the same procedure, and finally is transmitted to host B, where it moves up its seven layers, until it arrives at the application layer of host B.
Each of the seven layers of the OSI model hides the implementation details of the lower layers from the upper layers. Well-defined protocols and interfaces for each of the layers make it possible for the layer to be designed and implemented in isolation from the other layers. Except for the physical layer, which is implemented in hardware, all other layers are implemented in software. For example, each of these layers may be implemented as a set of routines which communicate with the layer above and the layer below it via parameters passed in function calls.
Each layer adds an additional piece of information to the message it is transmitting. The same layer removes the additional piece of information on the receiving end. The additional information appears in form of a header (e.g., TH = Transport Header). The data link layer adds a header as well as a trailer to its data.
The Physical Layer
The physical layer is concerned with the transmission of raw data bits over communication lines. Physical layer standards and protocols are concerned with issues such as the following:
• How a physical circuit is established between communicating devices.
• How the circuit is terminated when no longer needed.
• The physical form (e.g., voltages, frequencies, timing) in which data bits (binary values 0 and 1) are represented.
• Whether transmission of data can take place in one or both directions over the same physical connection.
• Characteristics of the physical media that carry the signals (e.g., copper wire, optical fiber, radio waves).
• Characteristics of the connectors used for connecting the physical media.
• How data from a number of sources should be multiplexed before transmission and demultiplexed upon arrival, and the type of multiplexing technique to be used.
• The type of modulation to be used for transmitting digital data over analog transmission lines.
The physical layer accounts for much of the tangible components of a network, including cables, satellites, earth stations, repeaters, multiplexers, concentrators, and modems. Physical layer protocols and standards are of mechanical, electrical, functional, and procedural nature.
The Data Link Layer
The data link layer is concerned with the reliable transfer of data over the communication channel provided by the physical layer. To do this, the data link layer breaks the data into data frames, transmits the frames sequentially over the channel, and checks for transmission errors by requiring the receiving end to send back acknowledgment frames. Data link protocols are concerned with the following issues:
• How to divide the data into frames.
• How to delimit frames by adding special bit patterns to the beginning and end of each frame. This allows the receiving end to detect where each frame begins and where it ends.
• Error detection. Some form of error check is included in the frame header. This is constructed by the transmitting end based on the contents of the frame, and checked for integrity by the receiving end. A change in the frame bits can be detected in this way.
• Error correction. When a frame arrives corrupted or is for any reason lost in the network, it is retransmitted. Lost acknowledgment frames may result in duplicate frames, which need to be detected and corrected as well.
• Flow control. In general, not all communication devices in a network operate at the same speed. Flow control provides a means of avoiding a slow receiver from being swamped by data from a fast transmitter.
The data link layer hides the above details from the higher layers. To the network layer, it appears as a reliable communication channel which can send and receive data packets as frames.
The Network Layer
The network layer is concerned with the routing of data across the network from one end to another. To do this, the network layer converts the data into packets and ensures that the packets are delivered to their final destination, where they can be converted back into the original data. Network layer protocols are concerned with the following issues:
• The interface between a host and the network.
• The interface between two hosts across the network.
• Routing of packets across the network, including the allocation of a route and handling of congestion.
• Correct ordering of packets to reflect the original order of data.
• Collection of statistical information (e.g., number of transmitted packets) for performance measurement and accounting purposes.
• Internetworking: communication between two or more networks.
The network layer hides the above details from the higher layers. To the transport layer, it appears as a uniform data transfer service, regardless of the location of the communicating devices and how they are connected.
The Transport Layer
The aim of the transport layer is to isolate the upper three layers from the network, so that any changes to the network equipment technology will be confined to the lower three layers (i.e., at the node level). Transport layer protocols are concerned with the following issues:
• Establishment and termination of host-to-host connections.
• Efficient and cost-effective delivery of data across the network from one host to another.
• Multiplexing of data, if necessary, to improve use of network bandwidth, and demultiplexing at the other end.
• Splitting of data across multiple network connections, if necessary, to improve throughput, and recombining at the other end.
• Flow control between hosts.
• Addressing of messages to their corresponding connections. The address information appears as a part of the message header.
• Type of service to be provided to the session layer (e.g., error-free versus error-prone connections, whether messages should be delivered in the order received or not).
The transport layer hides the above details from the higher layers. To the session layer, it appears as a customized data transfer service between two hosts, isolating the underlying network technology from it.
The Session Layer
The session layer provides a structured means for data exchange between user processes on communicating hosts. Session layer protocols are concerned with the following issues:
• Negotiating the establishment of a connection (a session) between user processes on communicating hosts, and its subsequent termination. This includes the setting of various communication parameters for the session (e.g., synchronization and control).
• Correct ordering of messages when this function is not performed by the transport layer.
• Recovery from interrupted transport connections, if necessary.
• Grouping of messages into a larger message, if necessary, so that the larger message becomes available at the destination only when its constituent messages have all been delivered successfully.
The session layer hides the above details from the higher layers. To the presentation layer, it appears as an organized communication service between user processes.
The Presentation Layer
The presentation layer provides a mutually-agreeable binary representation of the application data communicated between two user processes. Since there are many ways of encoding application data (e.g., integers, text) into binary data, agreement on a common representation is necessary. Presentation layer protocols are concerned with issues such as the following:
• Abstract representation of application data.
• Binary representation of application data.
• Conversion between the binary representation of application data and a common format for transmission between peer applications.
• Data compression to better utilize network bandwidth.
• Data encryption as a security measure.
The presentation layer hides the above details from the higher layers. To the application layer, it appears as a universal communication service between user processes, regardless of their system-specific idiosyncrasies, allowing them to converse in a common syntax.
The Application Layer
The application layer is concerned with the semantics of data, i.e., what the data means to applications. The application layer provides standards for supporting a variety of application-independent services. Examples include:
• Virtual terminal standards to allow applications to communicate with different types of terminals in a device-independent manner.
• Message handling system standards used for electronic mail.
• File transfer, access, and management standards for exchanging files or parts thereof between different systems.
• Transaction processing standards to allow different companies with different systems to access each other’s on-line databases (e.g., in banking and airline reservation).
• On-line directory standards for storing details of individuals, organizations, and network components.
• Standards for exchanging formatted documents.
Application layer standards have paved the way for open software systems, in which data can be communicated between incompatible base systems (i.e., different hardware and software architectures) without loss of meaning or usefulness.
Standards
Standards enable equipment from different vendors and with different operating characteristics to become components of the same network. Standards also enable different networks in different geographical locations (e.g., different countries and continents) to be interconnected. From a customer’s point of view, standards mean real cost savings: the same end-user device can be used for access to a variety of networks and services.
Standards are developed by national and international organizations established for this exact purpose. Important standards developed by various organizations, include the following:
• The International Standards Organization (ISO) is a voluntary organization with representations from national standards organizations of member countries, major vendors, and end-users. ISO is active in many areas of science and technology, including information technology. ISO standards are published as ISO serial-no (e.g., ISO 8632).
• The Consultative Committee for International Telegraph and Telephone (CCITT) is a standards organization devoted to data and telecommunication, with representations from governments, major vendors, telecommunication carriers, and the scientific community. CCITT standards are published as Recommendation L.serial-no, where L is a letter of the alphabet (e.g., I.440). These are revised and republished every four years. CCITT standards are very influential in the field of telecommunications and are adhered to by most vendors and carriers.
• The Institute of Electrical and Electronic Engineers (IEEE) is a US standards organization with members throughout the world. IEEE is active in many electric and electronic-related areas. The IEEE standards for local area networks are widely adopted. IEEE standards are published as IEEE serial-no (e.g., IEEE 908).
• The Electronic Industries Association (EIA) is a US trade association best known for its EIA-232 standard.
• The European Computer Manufacturers Association (ECMA) is a standards organization involved in the area of computer engineering and related technologies. ECMA directly cooperates with ISO and CCITT.
In addition to these organizations, and because of their global market influence, large vendors occasionally succeed in establishing their products as de facto standards.
Connection Types
Connections between devices may be classified into three categories:
• Simplex. This is a unidirectional connection, i.e., data can only travel in one direction. Simplex connections are useful in situations where a device only receives or only sends data (e.g., a printer).
• Half-duplex. This is a bidirectional connection, with the restriction that data can travel in one direction at a time.
• Full-duplex. This is a bidirectional connection in which data can travel in both directions at once. A full-duplex connection is equivalent to two simplex connections in opposite directions.
Signal Types
All signals are either analog or digital. An analog signal is one in which information appears as a continuous variation of some property. Human speech is an example: it produces a continuous variation of air pressure. A digital signal, on the other hand, is one in which information appears as a sequence of binary values 0 and 1. To represent these two values, a signal is used in which only two wave shapes are allowed, one representing the binary value 0 and the other representing the binary value 1. By definition, therefore, a digital signal is a restricted form of an analog signal. Analog and digital signals can both be transmitted through cable or wirelessly. Wireless internet (Wi-Fi), for instance, uses digital. AM and FM Radio uses analog. Cell phones are digital.
Modulation
Transmission of digital data over an analog line is achieved using a technique called modulation, where the digital bit stream is modulated over an analog carrier signal. A modem (modulator and demodulator) is a commonly used device which employs this technique. As illustrated in Figure bellow, a modem converts the outgoing digital bit stream from a device into an analog signal and converts the incoming analog signal into a digital bit stream.
• Amplitude Modulation (AM). In AM, the carrier signal’s amplitude is changed according to the modulating digital signal’s bit value. For example, two amplitude sizes (a small and a large one) may be used to, respectively, represent bit values 0 and 1. AM’s main weakness is its susceptibility to distortion.
• Frequency Modulation (FM). In FM, the carrier signal’s frequency is changed according to the modulating digital signal’s bit value. For example, two frequency values (a low and a high one) may be used to, respectively, represent bit values 0 and 1. FM is more resistant to distortion than AM.
• Phase Modulation (PM). In PM, the carrier signal’s phase is changed according to the modulating digital signal’s bit value. A change in the carrier signal’s phase indicates a change in the modulating digital signal’s bit value from 0 to 1 or from 1 to 0.
Digitization
Digitization is essentially the opposite of modulation. Whereas in modulation a digital signal is modulated over an analog signal for transmission, in digitization an analog signal is converted into digital format through a process of sampling. For example, the analog signal resulting from human speech can be sampled and converted into digital data, transmitted over digital lines, and converted back to analog signal at the other end. These two functions are performed by a device called codec (coder/decoder).
Transmission Media
Digital data can be transmitted over many different types of media. Selecting a transmission medium is guided by comparing transmission requirements against the medium’s characteristics. Four important criteria influence the choice:
1. Bandwidth. Bandwidth is the maximum frequency range that can be practically supported by a medium. This is usually expressed in kilo Hz (kHz) or mega Hz (MHz). For example, analog transmission of human speech typically requires a bandwidth of 4 kHz. Also related, is the notion of data rate, which denotes the maximum number of bits per second (bps) that can be transmitted. For example, a data rate of 10 mbps means that 10 million bits of data can be transmitted in each second. Because of their obvious relationship, the terms bandwidth and data rate are sometimes used interchangeably. Because of distortion factors, bandwidth and data rate are usually inversely proportional to the communication distance.
2. Cost. Two types of cost are relevant: (i) the cost of installing the medium, including the medium-specific equipment that may be needed, and (ii) the cost of running and maintaining the medium and its equipment. There is usually a need for tradeoff between cost, bandwidth, and distance.
3. Reliability. Some media, by their physical nature, transmit data more reliably than others. Low reliability translates into a higher number of errors, which needs to be balanced against the potential cost of recovering from the errors (e.g., retransmission, more complex hardware and software).
4. Coverage. The physical characteristics of a medium dictate how long a signal can travel in it before it is distorted beyond recognition. To cover larger areas, repeaters are needed to restore the signal, and this increases the costs.
Im a Software Engineer from Tanzania specializin in Web developer, design , programming web applications.