| Experiments Home |
|
| Computers |
|
|
| Computer Security |
|
|
|
|
|
|
  |
Computer Security Projects and Experiments
Computer security is a branch of information security applied to both theoretical and actual computer systems. Computer security is a branch of computer science that addresses enforcement of 'secure' behavior on the operation of computers. The definition of 'secure' varies by application, and is typically defined implicitly or explicitly by a security policy that addresses confidentiality, integrity and availability of electronic information that is processed by or stored on computer systems. The traditional approach is to create a trusted security kernel that exploits special-purpose hardware mechanisms in the microprocessor to constrain the operating system and the application programs to conform to the security policy. These systems can isolate processes and data to specifier domains and restrict access and privileges of users. This approach avoids trusting most of the operating system and applications. In addition to restricting actions to a secure subset, a secure system should still permit authorized users to carry out legitimate and useful tasks. It might be possible to secure a computer against misuse using extreme measures:
It is important to distinguish the techniques used to increase a system's security from the issue of that system's security status. In particular, systems which contain fundamental flaws[1] in their security designs cannot be made secure without compromising their usability. Most computer systems cannot be made secure even after the application of extensive "computer security" measures. Furthermore, if they are made secure, functionality and ease of use often decreases. Computer security can also be seen as a subfield of security engineering, which looks at broader security issues in addition to computer security.
Secure operating systemsOne use of the term computer security refers to technology to implement a secure operating system. Much of this technology is based on science developed in the 1980s and used to produce what may be some of the most impenetrable operating systems ever. Though still valid, the technology is almost inactive today, perhaps because it is complex or not widely understood. Such ultra-strong secure operating systems are based on operating system kernel technology that can guarantee that certain security policies are absolutely enforced in an operating environment. An example of such a Computer security policy is the Bell-LaPadula model. The strategy is based on a coupling of special microprocessor hardware features, often involving the memory management unit, to a special correctly implemented operating system kernel. This forms the foundation for a secure operating system which, if certain critical parts are designed and implemented correctly, can ensure the absolute impossibility of penetration by hostile elements. This capability is enabled because the configuration not only imposes a security policy, but in theory completely protects itself from corruption. Ordinary operating systems, on the other hand, lack the features that assure this maximal level of security. The design methodology to produce such secure systems is precise, deterministic and logical. Systems designed with such methodology represent the state of the art of computer security and the capability to produce them is not widely known. In sharp contrast to most kinds of software, they meet specifications with verifiable certainty comparable to specifications for size, weight and power. Secure operating systems designed this way are used primarily to protect national security information and military secrets. These are very powerful security tools and very few secure operating systems have been certified at the highest level (Orange Book A-1) to operate over the range of "Top Secret" to "unclassified" (including Honeywell SCOMP, USAF SACDIN, NSA Blacker and Boeing MLS LAN.) The assurance of security depends not only on the soundness of the design strategy, but also on the assurance of correctness of the implementation, and therefore there are degrees of security strength defined for COMPUSEC. The Common Criteria quantifies security strength of products in terms of two components, security capability (as Protection Profile) and assurance levels (as EAL levels.) None of these ultra-high assurance secure general purpose operating systems have been produced for decades or certified under the Common Criteria. Security by designThe technologies of computer security are based on logic. There is no universal standard notion of what secure behavior is. "Security" is a concept that is unique to each situation. Security is extraneous to the function of a computer application, rather than ancillary to it, thus security necessarily imposes restrictions on the application's behavior. There are several approaches to security in computing, sometimes a combination of approaches is valid:
Many systems unintentionally result in the first possibility. Approaches one and three lead to failure. Since approach two is expensive and non-deterministic, its use is very limited. Because approach number four is often based on hardware mechanisms and avoid abstractions and a multiplicity of degrees of freedom, it is more practical. Combinations of approaches two and four are often used in a layered architecture with thin layers of two and thick layers of four. There are myriad strategies and techniques used to design security systems. There are few, if any, effective strategies to enhance security after design. One technique enforces the principle of least privilege to great extent, where an entity has only the privileges that are needed for its function. That way even if an attacker gains access to one part of the system, fine-grained security ensures that it is just as difficult for them to access the rest. Furthermore, by breaking the system up into smaller components, the complexity of individual components is reduced, opening up the possibility of using techniques such as automated theorem proving to prove the correctness of crucial software subsystems. This enables a closed form solution to security that works well when only a single well-characterized property can be isolated as critical, and that property is also assessable to math. Not surprisingly, it is impractical for generalized correctness, which probably cannot even be defined, much less proven. Where formal correctness proofs are not possible, rigorous use of code review and unit testing represent a best-effort approach to make modules secure. The design should use "defense in depth", where more than one subsystem needs to be violated to compromise the integrity of the system and the information it holds. Defense in depth works when the breaching of one security measure does not provide a platform to facilitate subverting another. Also, the cascading principle acknowledges that several low hurdles does not make a high hurdle. So cascading several weak mechanisms does not provide the safety of a single stronger mechanism. Subsystems should default to secure settings, and wherever possible should be designed to "fail secure" rather than "fail insecure" (see fail safe for the equivalent in safety engineering). Ideally, a secure system should require a deliberate, conscious, knowledgeable and free decision on the part of legitimate authorities in order to make it insecure. In addition, security should not be an all or nothing issue. The designers and operators of systems should assume that security breaches are inevitable. Full audit trails should be kept of system activity, so that when a security breach occurs, the mechanism and extent of the breach can be determined. Storing audit trails remotely, where they can only be appended to, can keep intruders from covering their tracks. Finally, full disclosure helps to ensure that when bugs are found the "window of vulnerability" is kept as short as possible. Early history of security by designThe early Multics operating system was notable for its early emphasis on computer security by design, and Multics was possibly the very first operating system to be designed as a secure system from the ground up. In spite of this, Multics' security was broken, not once, but repeatedly. The strategy was known as 'penetrate and test' and has become widely known as a non-terminating process that fails to produce computer security. This led to further work on computer security that prefigured modern security engineering techniques producing closed form processes that terminate. Secure codingIf the operating environment is not based on a secure operating system capable of maintaining a domain for its own execution, and capable of protecting application code from malicious subversion, and capable of protecting the system from subverted code, then high degrees of security are understandably not possible. While such secure operating systems are possible and have been implemented, most commercial systems fall in a 'low security' category because they rely on features not supported by secure operating systems (like portability, et al.). In low security operating environments, applications must be relied on to participate in their own protection. There are 'best effort' secure coding practices that can be followed to make an application more resistant to malicious subversion. In commercial environments, the majority of software subversion vulnerabilities result from a few known kinds of coding defects. Common software defects include buffer overflows, format string vulnerabilities, integer overflow, and code/command injection. Some common languages such as C and C++ are vulnerable to all of these defects (see Seacord, "Secure Coding in C and C++"). Other languages, such as Java, are more resistant to some of these defects, but are still prone to code/command injection and other software defects which facilitate subversion. Recently another bad coding practise has come under scrutiny; dangling pointers. The first known exploit for this particular problem was presented in July 2007. Before this publication the problem was known but considered to be academic and not practically exploitable. [2] In summary, 'secure coding' can provide significant payback in low security operating environments, and therefore worth the effort. Still there is no known way to provide a reliable degree of subversion resistance with any degree or combination of 'secure coding.' TermsThe following terms used in engineering secure systems are explained below.
A bigger OS, capable of providing a standard API like POSIX, can be built on a microkernel using small API servers running as normal programs. If one of these API servers has a bug, the kernel and the other servers are not affected: e.g. Hurd.
Secure cryptoprocessors can be used to leverage physical security techniques into protecting the security of the computer system.
Some of the following items may belong to the computer insecurity article:
Capabilities vs. ACLsWithin computer systems, the two fundamental means of enforcing privilege separation are access control lists (ACLs) and capabilities. The semantics of ACLs have been proven to be insecure in many situations (e.g., Confused deputy problem). It has also been shown that ACL's promise of giving access to an object to only one person can never be guaranteed in practice. Both of these problems are resolved by capabilities. This does not mean practical flaws exist in all ACL-based systems only that the designers of certain utilities must take responsibility to ensure that they do not introduce flaws. Unfortunately, for various historical reasons, capabilities have been mostly restricted to research operating systems and commercial OSs still use ACLs. Capabilities can, however, also be implemented at the language level, leading to a style of programming that is essentially a refinement of standard object-oriented design. An open source project in the area is the E language. First the Plessey System 250 and then Cambridge CAP computer demonstrated the use of capabilities, both in hardware and software, in the 1970s, so this technology is hardly new. A reason for the lack of adoption of capabilities may be that ACLs appeared to offer a 'quick fix' for security without pervasive redesign of the operating system and hardware. The most secure computers are those not connected to the Internet and shielded from any interference. In the real world, the most security comes from operating systems where security is not an add-on, such as OS/400 from IBM. This almost never shows up in lists of vulnerabilities for good reason. Years may elapse between one problem needing remediation and the next. A good example of a secure system is EROS. But see also the article on secure operating systems. TrustedBSD is an example of an open source project with a goal, among other things, of building capability functionality into the FreeBSD operating system. Much of the work is already done. See also
Notes
References
Further reading Wikibooks has a book on the topic of
This e-primer provides a comprehensive review of the digital and information and communications technology revolutions and how they are changing the economy and society. The primer also addresses the challenges arising from the widening digital divide. External links
Many current computer systems have only limited security precautions in place. This computer insecurity article describes the current battlefield of computer security exploits and defenses. Please see the computer security article above for an alternative approach, based on security engineering principles.
Security and systems designMost current real-world computer security efforts focus on external threats, and generally treat the computer system itself as a trusted system. Some knowledgeable observers consider this to be a disastrous mistake, and point out that this distinction is the cause of much of the insecurity of current computer systems - once an attacker has subverted one part of a system without fine-grained security, he or she usually has access to most or all of the features of that system. Because computer systems can be very complex, and cannot be guaranteed to be free of defects, this security stance tends to produce insecure systems. The 'trusted systems' approach has been predominant in the design of many Microsoft software products, due to the long-standing Microsoft policy of emphasizing functionality and 'ease of use' over security. Since Microsoft products currently dominate the desktop and home computing markets, this has led to unfortunate effects. However, the problems described here derive from the security stance taken by software and hardware vendors generally, rather than the failing of a single vendor. Microsoft is not out of line in this respect, just far more prominent with respect to its consumer marketshare. It should be noted that the Windows NT line of operating systems from Microsoft contained mechanisms to limit this, such as services that ran under dedicated user accounts, and Role-Based Access Control (RBAC) with user/group rights, but the Windows 95 line of products lacked most of these functions. Before the release of Windows 2003 Microsoft has changed their official stance, taking a more locked down approach. On 15 January 2002, Bill Gates sent out a memo on Trustworthy Computing, marking the official change in company stance. Regardless, Microsoft's operating system Windows XP is still plagued by complaints about lack of local security and inability to use the fine-grained user access controls together with certain software (esp. certain popular computer games). Financial costSerious financial damage has been caused by computer security breaches, but reliably estimating costs is quite difficult. Figures in the billions of dollars have been quoted in relation to the damage caused by malware such as computer worms like the Code Red worm, but such estimates may be exaggerated. However, other losses, such as those caused by the compromise of credit card information, can be more easily determined, and they have been substantial, as measured by millions of individual victims of identity theft each year in each of several nations, and the severe hardship imposed on each victim, that can wipe out all of their finances, prevent them from getting a job, plus be treated as if they were the criminal. Volumes of victims of phishing and other scams may not be known. Individuals who have been infected with spyware or malware likely go through a costly and time-consuming process of having their computer cleaned. Spyware and malware is considered to be a problem specific to the various Microsoft Windows operating systems, however this can be explained somewhat by the fact that Microsoft controls a major share of the PC market and thus represent the most prominent target. ReasonsThere are many similarities (yet many fundamental differences) between computer and physical security. Just like real-world security, the motivations for breaches of computer security vary between attackers, sometimes called hackers or crackers. Some are teenage thrill-seekers or vandals (the kind often responsible for defacing web sites); similarly, some web site defacements are done to make political statements. However, some attackers are highly skilled and motivated with the goal of compromising computers for financial gain or espionage. An example of the latter is Markus Hess who spied for the KGB and was ultimately caught because of the efforts of Clifford Stoll, who wrote an amusing and accurate book, The Cuckoo's Egg, about his experiences. For those seeking to prevent security breaches, the first step is usually to attempt to identify what might motivate an attack on the system, how much the continued operation and information security of the system are worth, and who might be motivated to breach it. The precautions required for a home PC are very different for those of banks' Internet banking system, and different again for a classified military network. Other computer security writers suggest that, since an attacker using a network need know nothing about you or what you have on your computer, attacker motivation is inherently impossible to determine beyond guessing. If true, blocking all possible attacks is the only plausible action to take. VulnerabilitiesTo understand the techniques for securing a computer system, it is important to first understand the various types of "attacks" that can be made against it. These threats can typically be classified into one of these seven categories: ExploitsSoftware flaws, especially buffer overflows, are often exploited to gain control of a computer, or to cause it to operate in an unexpected manner. Many development methodologies rely on testing to ensure the quality of any code released; this process often fails to discover extremely unusual potential exploits. The term "exploit" generally refers to small programs designed to take advantage of a software flaw that has been discovered, either remote or local. The code from the exploit program is frequently reused in trojan horses and computer viruses. In some cases, a vulnerability can lie in certain programs' processing of a specific file type, such as a non-executable media file. EavesdroppingAny data that is transmitted over a network is at some risk of being eavesdropped, or even modified by a malicious person. Even machines that operate as a closed system (ie, with no contact to the outside world) can be eavesdropped upon via monitoring the faint electro-magnetic transmissions generated by the hardware such as TEMPEST. The FBI's proposed Carnivore program was intended to act as a system of eavesdropping protocols built into the systems of internet service providers. Social engineering and human errorA computer system is no more secure than the human systems responsible for its operation. Malicious individuals have regularly penetrated well-designed, secure computer systems by taking advantage of the carelessness of trusted individuals, or by deliberately deceiving them, for example sending messages that they are the system administrator and asking for passwords. This deception is known as Social engineering. Denial of service attacksDenial of service (DoS) attacks differ slightly from those listed above, in that they are not primarily a means to gain unauthorized access or control of a system. They are instead designed to render it unusable. Attackers can deny service to individual victims, such as by deliberately guessing a wrong password 3 consecutive times and thus causing the victim account to be locked, or they may overload the capabilities of a machine or network and block all users at once. These types of attack are, in practice, very hard to prevent, because the behavior of whole networks needs to be analyzed, not only the behaviour of small pieces of code. Distributed denial of service (DDoS) attacks are common, where a large number of compromised hosts (commonly referred to as "zombie computers") are used to flood a target system with network requests, thus attempting to render it unusable through resource exhaustion. Another technique to exhaust victim resources is through the use of an attack amplifier - where the attacker takes advantage of poorly designed protocols on 3rd party machines, such as FTP or DNS, in order to instruct these hosts to launch the flood. There are also commonly vulnerabilities in applications that cannot be used to take control over a computer, but merely make the target application malfunction or crash. This is known as a denial-of-service exploit. Indirect attacksAttacks in which one or more of the attack types above are launched from a third party computer which has been taken over remotely. By using someone else's computer to launch an attack, it becomes far more difficult to track down the actual attacker. There have also been cases where attackers took advantage of public anonymizing systems, such as the tor onion router system. BackdoorsMethods of bypassing normal authentication or giving remote access to a computer to somebody who knows about the backdoor, while intended to remain hidden to casual inspection. The backdoor may take the form of an installed program (e.g., Back Orifice) or could be in the form of an existing "legitimate" program, or executable file. A specific form of backdoors are rootkits, which replaces system binaries and/or hooks into the function calls of the operating system to hide the presence of other programs, users, services and open ports. It may also fake information about disk and memory usage. Direct access attacks
Someone gaining physical access to a computer can install all manner of devices to compromise security, including operating system modifications, software worms, key loggers, and covert listening devices. The attacker can also easily download large quantities of data onto backup media, for instance CD-R/DVD-R, tape; or portable devices such as keydrives, digital cameras or digital audio players. Another common technique is to boot an operating system contained on a CD-ROM or other bootable media and read the data from the harddrive(s) this way. The only way to defeat this is to encrypt the storage media and store the key separate from the system. See also: Category:Cryptographic attacks Reducing vulnerabilitiesComputer code is regarded by some as just a form of mathematics. It is theoretically possible to prove the correctness of computer programs though the likelihood of actually achieving this in large-scale practical systems is regarded as unlikely in the extreme by some with practical experience in the industry -- see Bruce Schneier et al. It's also possible to protect messages in transit (ie, communications) by means of cryptography. One method of encryption the one-time pad has been proven to be unbreakable when correctly used. This method was used by the Soviet Union during the Cold War, though flaws in their implementation allowed some cryptanalysis (See Venona Project). The method uses a matching pair of key-codes, securely distributed, which are used once-and-only-once to encode and decode a single message. For transmitted computer encryption this method is difficult to use properly (securely), and highly inconvenient as well. Other methods of encryption, while breakable in theory, are often virtually impossible to directly break by any means publicly known today. Breaking them requires some non-cryptographic input, such as a stolen key, stolen plaintext (at either end of the transmission), or some other extra cryptanalytic information. Social engineering and direct computer access (physical) attacks can only be prevented by non-computer means, which can be difficult to enforce, relative to the sensitivity of the information. Even in a highly disciplined environment, such as in military organizations, social engineering attacks can still be difficult to foresee and prevent. In practice, only a small fraction of computer program code is mathematically proven, or even goes through comprehensive information technology audits or inexpensive but extremely valuable computer security audits, so it's usually possible for a determined cracker to read, copy, alter or destroy data in well secured computers, albeit at the cost of great time and resources. Extremely few, if any, attackers would audit applications for vulnerabilities just to attack a single specific system. You can reduce a cracker's chances by keeping your systems up to date, using a security scanner or/and hiring competent people responsible for security. The effects of data loss/damage can be reduced by careful backing up and insurance. Security measuresA state of computer "security" is the conceptual ideal, attained by the use of the three processes:
Today, computer security comprises mainly "preventive" measures, like firewalls or an Exit Procedure. A firewall can be defined as a way of filtering network data between a host or a network and another network, such as the Internet, and is normally implemented as software running on the machine, hooking into the network stack (or, in the case of most UNIX-based operating systems such as Linux, built into the operating system kernel) to provide realtime filtering and blocking. Another implementation is a so called physical firewall which consists of a separate machine filtering network traffic. Firewalls are common amongst machines that are permanently connected to the Internet (though not universal, as demonstrated by the large numbers of machines "cracked" by worms like the Code Red worm which would have been protected by a properly-configured firewall). However, relatively few organisations maintain computer systems with effective detection systems, and fewer still have organised response mechanisms in place. Difficulty with responseResponding forcefully to attempted security breaches (in the manner that one would for attempted physical security breaches) is often very difficult for a variety of reasons:
Further readingThere are operating systems designed specifically with security in mind, such as the operating system OpenBSD, which is widely considered one of the most heavily code-audited operating systems available. There is an extensive culture associated with electronic security; see electronic underground community. See alsoLists and categories
Individual articles
References
External links
This article is licensed under the GNU Free Documentation License. It uses material from Wikipedia Encyclopedia article "Computer Security" |
