By Steven Totolo and Franca Piccin

Total Voice Control

Imagine the day when you can web surf from your cell phone. Or better yet, your refrigerator automatically reorders your groceries online. Sounds like an episode from your favorite science fiction movie? Hardly! Welcome to the future. These scenarios are possible today. For example, even electrical power companies are communicating with appliances such as air conditioners to reduce electrical system power peaks, and lower consumer energy rates. How are all of these devices talking to each other? Most likely it is TCP/IP. The Transmission Control Protocol (TCP) and Internet Protocol (IP) are but two of the most familiar functions that comprise the complete Internet access suite.

Learning TCP/IP is no small task. One could easily read numerous textbooks or spend weeks of classroom time to do so. The intent of this article, therefore, is not to provide an in-depth technical treatment of this subject, but rather to explain briefly for readers the history and function of TCP/IP.

The TCP/IP protocol suite was first developed for the Defense Advanced Research Project Agency (DARPA) network called ARPANET in the early 1970’s. It was part of a project to connect government-funded research institutions into a homogeneous computer network. By the early 1980’s, the University of California at Berkeley developed the Berkeley Systems Distribution (BSD) version of the UNIX operation system and distributed to all of its in-house computer departments. Soon after, DARPA contracted to have the TCP/IP protocol merged into the Berkley BSD UNIX operating system. This prompted convergence to a standard set of protocols for anyone wanting to connect to the Internet. Since all Unix source code was freely distributed (like Linux is today) most mainframe and minicomputer manufacturers began incorporating TCP/IP technology into their products.

From this point, TCP/IP functioned as the workhorse for network connectivity around the world. The protocol operates independent of any computer hardware or operating system, yet allows both to communicate to one another despite their differences, enabling computers and devices to unite through different hardware and software. This independence stems from the physical characteristics of the network media, enabling TCP/IP to operate over twisted pair wire (Ethernet), phone lines (HomePNA), radio frequency (HomeRF) and power line.

The increase use of computer networking soon led experts to realize that a common frame of reference was necessary for discussing data communication terminology. The Open Systems Interconnect (OSI) Reference Model was developed by the International Standards Organization to describe the structure and function of data communication protocols. The model contains seven layers that define functions performed when data are transferred between devices over a network. Each layer is fundamentally independent and only communicates with neighboring layers directly above and below it through well-defined software interface.

Figure 1 outlines the 7 layers of the OSI model. The top layer, or Application Layer, is where the user accesses the network through an interface such as an Internet browser or remote connection software. The Presentation Layer decides the format to be used for data exchange. Connections between cooperating applications on each end of the network connection are maintained by the Session Layer. The Transport Layer guarantees that the receiver gets the data exactly as it was sent. The Network Layer manages connections across the network and isolates the upper layers from the details of the layers below. The reliable delivery of data across physical media such as twisted pair wire is handled by the Data Link Layer. The Physical Layer defines the characteristics of the hardware needed to carry the data transmission signal. These seven layers define the basic characteristics of a communications protocol.

A protocol such as TCP/IP provides a method for devices to interface through the use of layers. By layering different responsibilities, a protocol can be broken down into a series of independent communication functions. Each function deals with a part of the protocol for which it is responsible, leaving the rest of the information for other layers. As the data is sent from the top layer down, as in Figure 2, each layer will check the data integrity and add its own information to the data set effectively creating a data packet. This process is known as data encapsulation. This added information will be stripped off and used on the receiving end to recover the data sent. This may seem like an inefficient way of sending data, but all the information added is used to guarantee delivery of the data.

Of the seven OSI layers, only four are used in the definition of TCP/IP: the Application Layer which includes the function of the Presentation layer, the Transport Layer which incorporates the Session Layer, the Internet Layer and the Link Layer. The Network Layer defines the datagram or data packet and handles the routing of data. TCP/IP can operate over several different media type and therefore uses the Data Link and Physical layers of existing data transmission standards such as Bluetooth, Ethernet, HomeRF and HomePNA to name a few.

The relationship between the fifteen protocols and services of the Internet suite is illustrated in the Figure 3. The left side indicates the functional equivalence to the OSI layer model.

Included in the Application Layer protocols are Hyper Text Transfer Protocol (HTTP), remote terminal protocol (Telnet), File Transfer Protocol (FTP), Simple Mail Transfer Protocol (SMTP), Simple Network Management Protocol (SNMP), Trivial File Transfer Protocol (TFTP), Dynamic Host Configuration Protocol (DHCP), and Domain Name System (DNS).

The Transport Layer contains the TCP and the User Datagram Protocol (UDP) which are the bases for all application access to the network. UDP is considered a lightweight protocol and is used for passing data that does not need a delivery guarantee. TCP on the other hand does guarantee delivery. This is the reason for its use in file transmissions such as mail and remote control applications. It provides a reliable flow of data between two network devices, dividing the data passed to it from the Application Layer into smaller chunks for the network layer. As pieces of the original data are acknowledged by the other system, more chunks or datagrams are sent. If a communication problem arose, the data would be retransmitted. This process continues until the data is properly received or the transmission exceeds a set time amount.

Functions contained in the Network Layer are the Internet (IP), Internet Group Management (IGMP) and the Internet Control Message (ICMP) protocols. As the diagram reveals, all transmission of data occurs through IP with the exception of two utilities, the Address Resolution Protocol (ARP) and the Reverse Address Resolution Protocol (RARP) which are used by the IP layer.

The IP provides an unreliable, connectionless datagram or packet delivery service. Unreliable in this case means there are no guarantees that an IP datagram is successfully received, but IP will provide the best effort of service. If errors occur in the transmission or reception, IP will throw away any corrupted data and send a message back to the sender that a problem exits. TCP will respond and transmit but UDP will not make an attempt.

The term connectionless means that IP does not maintain any information concerning the datagrams it handles. Each datagram is handled independently enabling IP to send and receive datagrams from TCP or UDP. While that datagrams may be sent out of order, TCP will collect and reassemble them into their original data set order. UDP, in contrast, will not and discards data.

IP recognizes receipt of a datagram only when that datagram is addressed directly to the IP. Located in the IP Header are the IP address of the datagram’s sender, its destination, and which upper level protocol is to receive the data. All devices on the network must have a unique Internet address or IP address. These addresses are 32-bit numbers that are normally written as four decimal numbers, one for each byte of the address. This is called dotted-decimal notation. For example, the address of CABA’s web site is 209.82.5.178.

One central authority, The Internet Corporation for Assigned Names and Numbers (ICANN), allocates addresses for all devices connected to the Internet. ICANN was previously known as the Internet Network Information Centre or InterNIC and was operated by the U.S. Department of Commerce. ICANN, now a non-profit corporation, takes over responsibility for the IP address space allocation, protocol parameter assignment, domain name system management, and root server system management functions. Once a device obtains an IP address, it is eligible to register for the "WWW" type notation. This simplifies website or device recognition, permitting users to access the site by name. For example, users would recall www.CABA.org more easily than 209.82.5.178.

Readers now have a basic understanding of the functions of TCP/IP. Future articles will build on this foundation by discussing other aspects of Internet protocols, contained in the TCP/IP suite. In the interim happy surfing… whether using your cell phone or your refrigerator!
 
 

Steven Totolo is President of tvcAutomation, a home automation specialist and a member of the CABA Standards Committee. He can be reached at (613) 795-7117; fax (613) 737-5323; email: sales@tvcAutomation