Networks are implemented to enable the sharing of resources and the exchange of information between users. As the number of resources, users, and connections increases, most networks must be routinely modified to accommodate growth ideally without any reduction in the features and performance levels users have come to expect.
This makes network architecture and designs a dynamic and continuous process within most organizations. Network architecture can be described as the structured grouping of hardware, software, and applications.
The successful integration of these three elements allows for the transfer of all types of messages between users, administrators, and devices. Networking or LAN technologies for PCs were introduced in the 1980s to manage the rapid growth of desktop computing within many organizations.
Problems associated with stand alone devices include the following:
Each PC stores data locally, unlike mainframes or minicomputers, both of which use centralized storage. As a result, faulty reporting and decision making can occur if the same data is updated on some PCs, but not on others.
Security is a major concern, since every PC can potentially contain sensitive data, making it easier for unauthorized individuals to gain access to valuable organizational information.
Backups may not exist for critical data if users do not duplicate their files on a regular basis. The failure of a single storage device can result in significant disruption and costs in terms of time, resources, and money.
Various users or groups are free to install and use different software applications on their PCs, making file sharing difficult within the organization.
It is more difficult to justify the purchase of expensive devices or services, since only one PC at a time can benefit. For example, a dedicated high rate Internet connection is typically not economically feasible for every user in an organization with tens or hundreds of stand-alone PCs.
Implementing a network makes it possible to solve these problems in the following ways:
A network makes it possible to centralize data. All files shared by users are stored in a central location, which guarantees consistency and simplifies the update process.
Multiple levels of security can be implemented on a network, making it more difficult to obtain unauthorized access to data.
A network can be equipped with a backup system that runs at specific intervals, ensuring that critical data is available from a secondary source if needed.
In addition to user-created files, software applications can be installed on centralized storage, accessible from any station connected to the network. This accelerates the deployment process, since any application or update needs to be installed only once on the network versus on each stand-alone PC.
All users can access any resource connected to the network (e.g., high-speed copiers or Internet links). Physically, a network can be as small as two connected PCs and the media used to enable the connections (i.e., cabling, a wireless link, or a telecommunications circuit). In addition to wired or wireless connectivity, a network must be equipped with services designed for resource sharing, including:
Access control, which is necessary in cases where two or more devices attempt to use a shared resource (e.g., telecommunications circuit between two sites, a printer) at the same time.
Synchronization, which is necessary to ensure that a receiving device is listening when a sending device is transmitting to that device.
Flow control, which is necessary to monitor and adjust the rate at which data is transferred from sender to receiver in order to minimize transfer time and data loss. For example, if a receiving device is occupied with other tasks, it will use flow control to ask the sending device to pause transmission.
Error control, which is necessary to verify that a message was transferred successfully between a sender and a receiver (or to request a retransmission if the transfer was not successful).