Lesson 4.1: Introduction to Computer Networks Flashcards
PAN
A PAN is normally used for short-range communications, within a few feet, such as wireless headphones to a smartphone or a wireless mouse to a PC.
LAN
LANs consist of a collection of computers in a single building or building complex. For example, the computers in a manufacturing building may be connected via a LAN.
WAN
WANs link systems over a greater distance, including machines on the opposite sides of the world.
Bus
Yet another classification of networks considers the topology of the network or the pattern in which the machines are connected. Four common topologies are bus, star, ring, and mesh. Bus and star network topologies are the most common. In a bus network, the machines are connected to a common communication line called a bus.
Star
Star networks have a single machine serving as a central point to which all others are connected.
Ring
Ring topologies connect directly to each other as a peer.
Mesh
Mesh connects every device in a network to every other device; mesh networks enable redundancy while also introducing significantly more network traffic.
Alternatively, a network could use a hybrid of these topologies.
The bus topology became popular in the 1990s when it was implemented under a set of standards/protocols known as the Ethernet. The star topology has roots as far back as the 1970s when it evolved from the idea of a large central computer system servicing many users at once. The star topology is used today in wireless networks where communication is carried out via radio broadcast to a central machine called the access point (AP).
Other classifications
Other classifications of networks consider whether the network’s internal workings are based on designs that are in the public domain (open network) or on designs owned or controlled by third parties (closed or proprietary networks).
Open network designs are freely circulated and are often more popular than proprietary designs that are restricted by license fees and contracts. The internet is an example of an open network. Communication over the internet is governed by an open collection of standards known as the TCP/IP protocol. Anyone is free to use these protocols without paying fees or signing license agreements.
In contrast, an IT company might develop proprietary systems for which it chooses to maintain ownership rights, allowing the company to draw income from selling or leasing its system.
Client/Server Model
The client/server model is a popular convention used for interprocess communication. The basic roles played by the processes are categorized as either a client making requests or a server satisfying client requests. An early application of the client/server model appeared in networks connecting clusters of offices with a single printer available to all computers. The printer (also known as the print server) is the server, and the machines are clients requesting printed documents. Early networking systems also used file servers that accepted requests for company data that were centrally stored.
Peer-to-Peer (P2P)
Peer-to-peer (P2P) is another model for interprocess communication. In this model, processers both request and provide service to each other. Instant messaging and interactive games played by users on multiple machines are both examples of the P2P model.
A new generation of P2P services arose to fill the void, expanding the range of sharable file types and further decentralizing networks. The Gnutella protocol operates without any centralized server and allows for numerous software clients to be used for access, which makes it nearly impossible to shut down. BitTorrent, used commonly for distributing large video files, employs a “swarm” model, whereby files are downloaded in simultaneous pieces from multiple host computers. Newer services have established degrees of encryption and anonymity to protect users from legal action by copyright holders.
Distributed Systems
Interactions between computers via networks have become commonplace and multifaceted. Many modern systems, such as global information retrieval systems or computer games, are designed as distributed systems. Distributed systems execute software as processes on more than one computer.
There are several types of distributed systems.
Cluster computing uses many independent computers to provide computation or services comparable to those of a larger machine. The cost of several individual machines can be less than a higher-priced supercomputer, with comparable performance. Cluster computing provides high availability as it is likely that at least one computer in the cluster will be able to answer a request even when others in the cluster are unavailable or broken down. In addition, clusters can balance loads by automatically shifting requests among the cluster members.
Grid computing is a type of distributed system that is more loosely coupled than clusters but still works together as a system to complete large tasks. Grid computing typically includes specialized software to make it easier to distribute the workload and data among the machines in the grid.
Cloud computing provides large pools of shared computers that can be allocated to clients as needed. Services such as Amazon’s Elastic Compute Cloud allow clients to rent virtual computers by the hour no matter where the associated computer hardware is located. Services such as Google Cloud and Google Apps allow users to collaborate or build web services without needing to know how many computers are working on the problem or where the relevant data is stored. Cloud computing provides reasonable guarantees for reliability and scalability while raising concerns about security and privacy.
