Network security covers a wide range of issues. Such issues can include protecting the network and all associated devices, protecting the information stored in the system and traveling over the network, and protecting against users who don't have access permissions.
Many network users equate security with secrecy. Secrecy (also called confidentiality) is only one part of the network security puzzle. Indeed, there are three general types of network security: secrecy, integrity (accuracy), and availability. Depending on the organization's network environment, one type of security might be more important than another.
A secure network means that the systems on the network must be able to do all of the following:
There are three basic steps that must be taken to provide network security. System Managers need to:
The needs of individual network environments determine exactly which threats are worth protecting against and how expensive or disruptive that protection can be.
For example, a breach in network security in a bank could result in the loss of substantial amounts of money. A breach in network security in a research group could result in the loss of files, loss of time, and substantial inconvenience for the researchers. In both cases, system management would need to make a policy decision about the type of security service to implement.
The U.S. National Computer Security Center (NCSC) publishes the Trusted Computer System Evaluation Criteria (known as the Orange Book), which contains commonly used classifications for standalone computer systems. Trusted Network Interpretation (TNI), published by the same organization, interprets the policies in the Orange Book for systems attached to networks.
Various kinds of threats to network security can occur. Most are from unauthorized external users, but some can be from an internal source also. Common kinds include:
On a network, an observer might be able to correctly interpret that a major activity is underway by observing the number, types, or frequency of messages on the network.
After identifying potential threats to a network, you can determine the types of services to use to protect your network against the threats. You can use a variety of available services individually or in any combination. Some of these services include:
These services are the single most important type of security service because they prove the identity of the user. There are a number of authentication variants. One is data origin authentication that determines the origin of data being received.
Other variants are single-entity authentication and mutual-entity authenticating. In single-entity authentication, one entity authenticates itself to another, such as when a user enters a password to log in.
In mutual entity authentication, both entities authenticate themselves to the other.
For example, when a user enters a password and the system to which the user logs in displays a password.
These services control what a user can do on the system, such as access certain resources. Access control services are usually implemented after users have proved their identity through authentication.
Unauthorized users are prevented from reading data by these services. They can be of two types: connection-oriented confidentiality services and connectionless confidentiality services.
Connection-oriented services guaranty the confidentiality of all data traveling across a specific connection. Connectionless services provide the confidentiality of a single message going across a network using a connectionless protocol.
It’s also possible to have selective field confidentiality for certain fields in a message.
These services guard against modification or replay of data as it travels across the network.
These service prevent denial of participation in a communication. You can do this by recording proof of origin, proof of receipt, or both.
Illegal users are prevented from inferring information from the activities on the network when these services are in place. For example, false information can be transmitted during slack periods to prevent an illegal observer from guessing that network activity has changed.
A single service can provide protection against a single threat to network security, however, it is more likely that a single service provides only partial protection against a threat. In some cases, more than one service is needed. Figure 1 summarizes the types of services you can combine to use against specific threats.
Security services can counter threats. The methods you use to implement security services do the actual work.
Different methods can frequently provide the same services. A single method can sometimes provide for multiple services. For example, you can implement access control in a number of different ways. A single method for encryption can be useful in a number of different services.
Some methods for providing security services include:
This requires the exchange of single or mutual authentication information. The exchange protocols define what that information looks like. The Kerberos protocol (described later) is one example of Authentication exchange.
Control information or data is encoded in a form that only authorized persons or processes can read. There are two major types of encryption algorithms: private key and public key.
This method provides security by granting or denying designated users access to designated resources. A user could be granted access to one group of files and denied access to another. One way to grant or deny access to a user is to create one or more Access Control Lists (ACLs) for a resource. An ACL indicates which users are allowed access to certain resources and just how they can use the resources.
Frequently used to provide nonrepudiation services, Digital Signatures are similar to physical signatures on paper documents. This method typically uses encryption techniques to provide the Digital signature and for verification. A Digital signature can also provide data integrity services guaranteeing that the data in question was not modified.
Protects against traffic analysis. For example, false information can be inserted into the stream of transmitted data. In successful traffic padding mechanisms, attackers cannot tell the difference between false and real information.
Matching method types with security services is not always clear. Sometimes a single method might provide all or part of several services. Implementing a single service might require several methods.
Figure 2 shows the relationship between different services and the methods to implement the services. For example, you might use encryption to provide for authentication and data confidentiality.