Glossary of Information Technology security terms ITSS Legal Issues Working Group 11/07/95 G - 2 Glossary of Information Technology Security Terms Access controls G-2 Advance card technologies (including smart cards) G-3 Audit trails G-4 Authentication G-4 Authorization G-5 Bulletin Board Systems G-5 Clipper Chip: see encryption Communications, Transmission, Computer Security (COMSEC, TRANSEC, COMPUSEC) G-5 Data Matching G-5 Digital Signature G-6 Dynamic Data G-6 EDI and electronic commerce G-6 Encryption G-7 (including algorithm, symmetric and asymmetric encryption, Digital Encryption Standard (DES), public key cryptography, digital signature, hash, confidentiality encryption, RSA algorithm, factoring problem, discrete logarithm problem, public key infrastructure, public key certificates, key escrow (Clipper Chip) ) Firewalls and Gateways G-12 Hacking (examples of network security vulnerability) G-13 Integrity G-15 Internet (including World Wide Web, FTP, TELNET and Gopher) G-15 Legacy system G-15 Non-repudiation G-16 Public key infrastructure, public key cryptography: see encryption Sensitive information G-16 TEMPEST G-16 Tokens G-16 Value Added Networks (VANs) G-17 Virus (including trojan horse, worm, bacteria, logic bomb, time bomb and bugs) G-17 Access controls (see Firewalls and Gateways and Hacking) Access controls refers to the ability of computer systems to be set up so that only specified individuals or computers are permitted to have access to specific parts of the computer system. Passwords is one form of an access control. Tokens, such as a smart card (see Advanced Card Technologies) can also be required before a computer will permit a person to use that computer. Access controls are generally set by the administrator of the computer system, and can include both discretionary and manadatory access controls. Discretionary access controls (DAC) are controls based on the identity of the user, the data requested and access control information. The owner of the data (and privileged system users) can changed the access control information. Because DAC is based solely on user identity, it is vulnerable to certain types of attacks, such as a Trojan Horse. Mandatory access controls (MAC) are based on the clearance of the user, the sensitivity of the data requested, and access control information. Only privileged system users can change the access control information. Because MAC is based on clearance and data sensitivity, it can defeat Trojan Horse attacks. Mode of Operation: There are different modes of operation (or “processing states”) depending on the clearance levels of users and their need-to-know the information in the system. To define the modes of operation, it is necessary to know the clearance of the system users and the sensitivity of the system data, the number of users in any particular mode of operation and how they communicate on the network (e.g., dial-in access, telnet access, direct access). The user profile, system characteristics (e.g., does the system have MAC and DAC), and confidentiality considerations dictate the appropriate mode of operation. There are three modes of operation: dedicated, system-high and multi- level. All users cleared for All users have need to all information on the know all the system information on the system Dedicated Yes Yes System Yes No High Multi- No No level Multi-level requires the most technical security features, but allows the most flexible use of the computer system. It permits users with different security clearance levels and information with different security designations to operate on the same system. Advanced Card Technologies “Advanced Card Technologies” (ACT) refer to any cards used for the purposes of identifying the end-user to gain access to an information system. There are significant privacy issues related to the use of advanced card technologies because potentially an individual could be carry on a single card all of his or her personal data (e.g. medical, criminal, education and employment information and spending habits). In addition to carrying personal information, the card could (when used with a personal identification number) provide access to the individual’s money (e.g. bank card, credit card), home, office car and computers (used as a key or password to gain entrance), and to government and other services used by the individual, and could carry the individual’s secret encryption key which would used to produce electronic proof of identity for many different transactions. Thus, unauthorized access to information on an advanced technology card could result in serious invasions of privacy, and potentially could result in frauds and thefts as well. There are a number of different kinds of advanced card technologies: The magnetic stripe card is undoubtedly the most recognizable ACT. Anyone who uses a credit card or an automated teller banking card is using a magnetic stripe card. This type of card is distinguishable by the dark stripe of magnetic material which is placed on one side of the card. The magnetic stripe is capable of storing 400 bytes (characters) of information. Its very limited information storage capabilities require that systems utilizing this technology store only essential information, such as cardholder identification codes to the card. Although the magnetic stripe card has gained massive acceptance throughout the world, it is widely recognized that its limited data storage capability and the data's susceptibility to fraudulent alteration will severely limit the future uses of the technology. The smart card is perhaps the most promising ACT of all. A smart card is a credit card sized plastic card that contains an imbedded computer chip which possesses computer logic and is capable of processing, storing and retrieving information loaded onto the chip. Although memory storage capacity is currently limited to a maximum of 64 kilobytes, the smart card's microprocessor offers functionality that is not available with any other ACT. The smart card's microprocessor is able to employ sophisticated password and encryption/decryption techniques to prevent unauthorized users from accessing or modifying data stored on the card. Also, the smart card's microprocessor is capable of performing computer logic functions such as addition, subtraction, multiplication, division, conditional branching, etc. In fact, these functions are used by the smart card's microprocessor to monitor access attempts and to shut off the smart card when the number of incorrect access attempts reaches a predefined number. The PCMCIA (Personal Computer Memory Card International Association) is a standard which describes the physical requirements, electrical specifications and software architecture for PC cards. These standards define three physical sizes of cards: Type I, Type II and Type III. All three card types measure the same length and width, roughly the size of a standard credit card, and use the same 68-pin edge connector for attachment to the computer. Where they differ is only in thickness. A Type I Card is typically used for various types of memory enhancements, including RAM, flash memory, one-time programmable (OTP) memory and electronically erasable programmable read only memory (EEPROM). A Type II Card is typically used for memory enhancements and/or for I/O features such as data/fax modems, LANs and host communications. A Type III Card is twice the thickness of the Type II and is used for memory enhancements and/or for I/O features that require a larger size, such as rotating mass storage devices (removable hard disk drives) and radio communication devices. The optical card incorporates optical data storage technology (such as high-capacity optical disks) to portable credit-card sized plastic cards. Data is stored and retrieved using laser beams in essentially the same manner as used for optical disks. Unlike some ACT which allow data to be written, erased and re- written, optical cards employ a write-once technology, i.e., the media can be written to only once and it cannot be altered without destroying the card. Under the proper circumstances, this unerasable write-once system is desirable (e.g., medical records). The main advantage of this technology is its high data storage capacity, typically in the 4 to 6 megabyte range (equivalent to 1800 to 2400 typewritten pages of text). A disadvantage of this technology is its relatively weak capability to prevent unauthorized users from reading data stored on the card, unless the information is stored in an encrypted form. The small memory card's memory storage capacity of 2 kilobytes to 64 kilobytes of memory (2,000 to 64,000 characters) is much more limited than that of the large memory card. The small memory card's credit card like appearance, source of memory and memory capacities are very similar to that of a smart card. It is often referred to as a smart card without the intelligence of a microprocessor. The large memory card is another ACT which under the right circumstances can be a very useful product. Physically, the card's length and width conform to credit card standards, however these cards are considerably thicker than a credit card or other ACT products. The large memory card is basically an electronic computer memory board shrunken to the length and width of a credit card. The memory capacity of this medica can be as high as 2 megabytes (equivalent to 800 typewritten pages of text) depending on the type of memory that is used. A drawback associated with this ACT is that it is not very portable, that is one cannot carry a large memory card in one's wallet. The final type of ACT is classed as other. This refers to card types which do not apply as noted above or combinations of the above card types, for example an ID card which has both a magnetic stripe and a computer chip or a number generating card (e.g., SECURE ID or RACAL) or other token based systems (e.g., KSD-64) Audit Trails Audit trails in computer systems are intended to allow management to investigate events leading up to any incident or to trace or monitor activities or transactions on the system. Audit and monitoring are oftened sacrificed in computer security is because implementing them means dealing with a range of challenging cost, policy and legal issues. Audit functions have a high overhead in processing speed and storage space and are normally turned off except for minimal, critical events, although frequently even these are not audited. audit functions are often sacrificed. The cost implications in terms of storage media, archival facilities, regular review and updating, documenting the system and policies with respect to its use, and dealing with issues relating to personal privacy and public access are very demanding. Authentication (see also authorization and integrity) Authentication means evidence (as conclusive as possible) that the person identified as the send in the electronic message is in actually the person who sent the message. Authorization (see also authentication and integrity) Authorization means evidence (as conclusive as possible) that the person who sent a message (e.g., signed a contract, approved a payment, certify receipt of goods) actually had the authority to send that message. Authorization is based on the individual’s financial authority, security clearance and need-to-know. Bulletin Board System (BBS) An electronic computer dial-in service which provides multiple services to computer users such as e-mail, discussion groups, games, news services. documents, images and data, and access to other computer systems. Certification, Certification Authority, Public Key Certificate (see encryption) Classified Information (see sensitive information) Clipper Chip, Capstone Chip (see encryption) Communications, Transmission, Computer Security COMSEC includes Emission Security (electromagnetic emanations from computers), Cryptographic Security and Transmission Security (telecommunications - TRANSEC). TRANSEC includes measures designed to protect transmissions from intercept and exploitation. TRANSEC does not include protecting encrypted messages from surreptitious decryption (which is Cryptographic Security). TRANSEC measures include adherence to approved operating procedures, use of equipment that encodes messages (i.e., cryptography), frequencies (i.e., frequencies at which signals are sent, various megahertz) and spectrum (across which signals are sent). Computer security is known as COMPUSEC. Confidential information (see sensitive information) Data Matching The comparison of personal data obtained from different sources, including personal information banks, for the purpose of making decisions about the individuals to whom the data pertains. Generally, it applies to comparing computerized sets of data rather than comparing records from two government departments about a single individual. Thus, data matching is generally described as a “program.” In accordance with the Privacy Act, prior to initiating a data-matching program, government institutions must assess the feasibility of the proposed match, analyzing the potential impact on the privacy of individuals and the costs and benefits of the data-matching program. This matching activity may or may not generate a new body of personal information. Data linking or data profiling involves combining personal information from a variety of sources to create a new body of personal information. Matching institution: A government institution that is planning to conduct or is conducting a matching program. Matching source: An organization that discloses personal information to an institution for the purposes of a matching program. A matching source may be within the matching institution, another government institution, or any other organization. Designated Information (see sensitive information) Digital Signature (see encryption for more information) A digital signature is commonly interpreted as involving a mathematical summary (“hash”) of a document, encrypted by an individual’s secret encryption key. (If the message is short, there is no hash: the message itself is encrypted, not for the purpose of confidentiality for the message, but for the purpose of proving who sent the message.) Generally, encryption is tied to a unique encryption key, which only one person has. If a message is encrypted with that key, it proves that the message was sent by the person who purportedly holds that key, and thus serves the same function as a signature, although it is not similar in nature to a hand- written signature. By way of contrast, a digitized signature means reproducing an electronic image of a person’s handwritten signature (such as by faxing a document signed by hand). The received fax will have a digitized signature on it. Dynamic data Data that is constantly changing, such as data in a spreadsheet or database which is constantly being updated. Electronic data interchange (EDI); electronic commerce Electronic data interchange and electronic commerce are terms that refer to electronic transactions and contracts. Generally, EDI involves sending purchase requests and payment invoices electronically, but electronic transactions can cover many other kinds of transactions, such as filing income tax returns, applying for licences, paying tickets. Frequently, there will be a signed, paper agreement between the parties setting out how the electronic transactions will occur (Trading Partner Agreements). EDI makes it possible to order goods and services, invoice them and authorize payment for them without a person actually filling out those forms or even entering that information. For example, a car manufacturer can have a relationship with a steering wheel supplier. The steering wheel supplier might have direct access to the manufacturer’s computer which says how many cars will be manufactured in a given period of time. It is the supplier’s responsibility to ensure enough steering wheels are provided to be put into the car. No purchase order is required. After the cars are built, the manufacturer simply counts the number of cars produced and transfers funds to the supplier for the price of the steering wheels for that number of cars (knowing that the cars could not have been produced without steering wheels and that each car has only one steering wheel). No invoice, certificate of receipt, payment authorization or cheque is required. This kind of electronic commerce requires an established, trusted relationship between supplier and purchaser. Encryption Encryption means transforming plain text into text which is unintelligible to anyone who does not have the “key” (or “code”) which explains the relationships between the plain text characters and the cipher text characters. For example, the key used in the time of Julius Ceasar was simply to replace a letter with the letter which appeared four letters later in the alphabet. Thus, where the letter ‘A’ appeared in plain text, it would be replaced by ‘D’ in cipher text; ‘B’ would be ‘E’, and so on. ‘X, Y, Z’ would be ‘A, B, C’ respectively. Now, encryption involves applying a series of complex mathematical formulae to the characters. Any given set of formulae are known as an algorithm. Mathematical formulae can be applied to letters because in computer language, letters are represented by 0s and 1s. The 0s and 1s which represent a letter also can represent a number. Thus, letters have numerical equivalents and performing numerical calculations with them is relatively straight-forward. There are two kinds of encryption: symmetric and asymmetric. Symmetric encryption means that the same key that is used to encrypt a message is also used to decrypt a message. For almost the entire history of cryptography, encryption has used symmetric keys. The problem with symmetric encryption is that two parties must have access to the key, and those parties must both keep the key secret to ensure that no one else can read their messages or that no one else can impersonate them (if an encryption key is secret and is thought to be known only to two persons, a third person with the key could impersonate either of the other two persons. Getting the key from the sender to the receiver, without anyone else finding out about the key was difficult, and meant that the uses of cryptography were relatively limited. One of the most popularly used symmetric encryption algorithms is the Digital Encryption Standard (DES), which is used on bank cards for Automated Teller Machines, among other uses. The U.S. National Institute for Science and Technology adopted DES as a standard in 1977, and the American National Standards Institute adopted the same encryption algorithm as a standard for commercial purposes, calling it the Digital Encryption Algorithm (DEA). The continued use of DES as an encryption standard is subject to some question as new technology begins to make it feasible that persons could compromise DES encryption. One alternative is use triple the DES processes (known as ‘Triple DES’) in order to make the encryption stronger. Another alternative is to migrate to different encryption standards. In the mid-1970s, a revolution in cryptography occurred: asymmetric cryptography was discovered. Asymmetric cryptography makes it possible to divide an encryption key into two parts: one part is secret and the other part is public. For this reason, asymmetric encryption is also called private key/public key encryption. Public key encryption allows one person to send an encrypted message to another person without worrying about how to get a secret key to that other person. All the sender has to do is publish the public key, which anyone can see. (Public keys should be notarized by a trusted third party to ensure that the key is unique and that it was issued to the specific individual who claims to hold it.) One of the more popular public key encryption products is PGP (Pretty Good Privacy) which is available for free on the Internet and uses the RSA encryption algorithm (discussed below), although there are intellectual property claims against PGP (thus, PGP should be avoided for commercial uses). Public key encryption can be used for two separate purposes: Digital Signature and Confidentiality Encryption. Digital signature works as follows: Digital signatures involve encryption and hash functions. A ‘hash’ of a message is produced by performing an algorithm on the message. Hash algorithms are different than encryption algorithms. Hash algorithms are designed to produce a very short sequence of bits that are mathematically calculated from the bits in a larger file with the effect that any change in the message will produce a change in the hash (i.e., each bit in the hash has a 50% probability of changing for every change to the original text). There are a number of hash algorithms which are widely known. There is no need to keep a hash algorithm secret as confidentiality is not its purpose. (If the message is short, there is no hash: the message itself is encrypted, not for the purpose of confidentiality for the message, but for the purpose of proving who sent the message.) · A sends a plain text message, and encrypts a hash of that message using A’s private key. · B receives the plain text message and performs a hash of the message, producing the Plain Hash. · B retrieves A’s public key from a public directory, decrypts the encrypted hash that A sent. If A’s public key successfully decrypts the hash, this proves that A actually sent the message, because A’s public key only works in conjunction with A’s private key, and only A has the private key. · B compares the Decrypted Hash with the Plain Hash. If the Plain Hash and Decrypted Hash are identical, it proves that the message has not been altered by anyone. If the two hashes are in any way different, B should disregard the message, inform A of the mismatching hashes and ask A to send the message again and both A and B should report the mismatch to their respective computer security officers. A digital signature does not provide confidentiality because the message itself is not encrypted (only the hash is) and the hash is encrypted using A’s private key, which means that anyone can decrypt the message using A’s public key. The purpose of digital signature is to prove (1) that A sent the message (authentication) and (2) that the message has not been altered in transmission (integrity). Confidentiality encryption works as follows: Public key encryption can be used to ensure confidentiality by reversing which keys are used by the sender and receiver. In Digital Signature, we saw that A used A’s own private key to encrypt the hash. B used A’s public key to decrypt the hash. In Confidentiality Encryption: · A retrieves B’s public key and encrypts a message with it and sends it to B. · B is the only person who can decrypt the message because only B has the private key which is necessary to decrypt to the message. Some public key algorithms can be used only for digital signature (such as the Digital Signature Algorithm (DSA)) but not for confidentiality encryption because part of the algorithm requires knowing the hash of the message. Thus, when an unencrypted message is sent accompanied by an encrypted hash of that message, the only way to decrypt the hash is to make a hash of the unencrypted message and use that in the decryption algorithm. In confidentiality encryption, there is no unencrypted message which can be used to derive the hash. Using symmetric and public key (asymmetric) encryption together Symmetric encryption is faster than public key encryption. Depending the algorithms used, it can be anywhere from 100 to 1,000 times faster. Thus, where symmetric encryption is feasible, it is preferable to public key encryption. The two can be used together. For digital signature, A can encrypt the hash using a symmetric key, and send that symmetric key to B by encryptiing the symmetric key with A’s private key. When B receives the message, B will use A’s public key to decrypt the symmetric key that was sent and then use the symmetric key to decrypt the hash. For confidentiality encryption, A will encrypt a message using A’s symmetric key, then will encrypt that symmetric key using B’s public key. This way, only B will be able to decrypt A’s symmetric key. Once B has decrypted A’s symmetric key, B will decrypt the message. Public Key Revolution The concept of public key cryptography was invented by Whitfield Diffie and Martin Hellman, and independently by Ralph Markle. Their contribution to the field was the notion that keys could come in pairs; a public key and a private key, and that it could be infeasible to generate one key from the other Therefore, one could know how to encrypt without knowing how to decrypt; or verify a signature without knowing how to generate it. Public key cryptography is made possible due to one-way functions associated with modular arithmetic and the properties of large prime numbers. Although there a number of public key algorithms, all of them are based on one of two different mathematical problems associated with prime numbers: the factoring problem and the discrete logarithm problem. Diffie and Hellman first presented the concept of public key cryptography in 1976 at the (U.S.) National Computer Conference in the paper entitled “New Directions in Cryptography,: which was published in the Institute of Electronic and Electronic Engineers (IEEE) periodical Transactions on Information Theory. This prototype algorithm makes use of the discrete log problem. Three professors at the Massachussets Institute of Technology (Rivest, Shamir and Adleman), inspired by the article written by Diffie and Hellman on one-way functions, developed their public key algorithm in 1977 (known as RSA). RSA’s strength is based on the difficulty of factoring the modulus that is used in the trap- door one-way function. Simply stated, it is very easy to multiply two large prime numbers together, P1*P2=C, but very difficult to take C and calculate P1 and P2 (so long as P1 and P2 are large). Every private key/public key pair uses different combinations of prime numbers. As computers become more powerful and perhaps become able to factor large numbers into their primes, the solution is simply to use larger prime numbers, to continually keep ahed of the power and speed of computers. In symmetric encryption, the most popular algorithms, such as Digital Encryption Standard (DES) use keys that are 56 bits long (bits are 0s and 1s, and generally, a numerical digit is equal to approximately three and a third (3.322) bits). The new International Data Encryption Algorithm (IDEA) will use a 128 bit key. Public key cryptography, by contrast, uses much longer keys to produce the same level of protection as DES. For example, the Rivest, Shamir, Adleman (RSA) public key algorithm uses a 512 bit key. The difference in length of keys is based on the technical design of the algorithms and the methods that are available for unauthorized persons to discover the keys through mathematical calculations. Although the RSA algorithm (on which PGP is based) uses the factoring problem, there are public key systems whose strength is based on the difficulty of evaluating discrete logarithms (e.g., the Diffie-Hellman key exchange algorithm). The discrete log problem is based on the difficulty of evaluating x, when given x mod p ( is publicly known and so is p (a large prime)). Protocols based on the discrete log problem can be used for key management, signatures, encryption, etc., just like RSA. The components of the public key and private key When public key encryption is based on the factoring problem, public key is actually two numbers, and a private key is one number. (The public key is three numbers for the discrete log problem.) These three numbers are derived from the public key algorithm. The RSA algorithm is as follows: RSA algorithm To encrypt: To decrypt: AB C = D, remainder E EF C = G, remainder A “A” is the original text “E” is the cypher text. “C” is the modulus, produced by multiplying the two large prime numbers together Not only are the two prime numbers used to create the modulus, they are also used to determine the values of “B” and “F”. “B” can be any number that meets a few limiting parameters. “B” is usually a relatively small number, often only 5 bits long. Rather than calculating “B”, one chooses a value for “B” and then calculates the value for “F”, using a fairly complicated mathematical formula that requires knowing the two prime numbers used to generate the modulus (F=B-1 mod (P1-1)(P2-1)). Many encryption schemes will use the same “B”. · The public key is two numbers: “B” and the modulus “C”. · The private key is the number “F”. The prime numbers needed to create the modulus “C” or the private key “F” are not in either the public or private key, but if a person had one of the prime numbers, it would be possible to calculate the numbers needed to learn the private keys. Public key infrastructure and key management: For public key encryption to work, your public key must be public. This is done by posting your key on an electronic directory. It is important to have confidence that your public keys has not been tampered with because: · If A’s public key has been altered in any way, it will be impossible for anyone to use the altered public key to decrypt A’s messages. · If C replaces A’s public key with C’s public key (but the key continues to be publicly identified as A’s public key), then when B intends to send a confidential message to A, it will be C that is able to decrypt the message and not A. Further, using his own private key, C will be able to digitally sign documents and convince B that it was A who signed them. It is also important to keep the private key secret. If C has A’s private key, C will be able to digitally sign documents and convince B that it was A who signed them and will be able to decrypt messages intended for A that are encrypted using A’s public key. To ensure there is confidence that the public and private keys which are purported to be associated A are in fact associated with A, the key management services are provided by a trusted third party: a public key infrastructure or PKI. Typically, generating the public keys, posting them, and preventing tampering with them is provided as a central computer service. The human involvement are the technology specialists who monitor the operation of the computer that manages the public keys. The administrator of the computer managing the public keys “certifies” that the public key you want to use has not been tampered with by attaching its own digital signature to your document. Thus, when A ‘signs’ a document by encrypting the hash of the message with A’s private key, the trust third party, the ‘Certification Authority’ (CA) will encrypt its own digital signature to certify that A’s public key in fact belongs to A. To decrypt A’s encrypted hash, B will have to retrieve A’s public key, check to see if it is the correct key by retrieving the CA’s public key and decrypting the CA’s certificate for A (this process will compare A’s public key with the public key the CA says it issued to A). After B has confirmed that A’s public key is the correct public key, B can then decrypt A’s digital signatures. There can also be heirarchies of notaries, where one level is responsible for certifying the public keys of the certifiers one level below. Public key certificates The public key certificate has a number of fields in it. They include the following: 1. Subject (J. Doe) 2. Public key numbers 3. Public key algorithm (there are a number of public key algorithms. It is important to identify which one is being used to make it possible to know how to use the public key numbers. 4. Certificate expiry date (public key certificates have time limits, after which new keys are issued and certified) 5. Certification Authority (Big Company or Government Office) 6. Certificate Serial Number. The Certification Authority makes a hash of the above information and encrypts the hash using its private key. Users obtain the CA’s public key, decrypt the hash to ensure that the certificate information has not been altered, then uses the public key numbers found in the certificate. (The algorithm used to produce the hash of a message can be part of the certificate or can be communicated in other ways.) Generally, the computer will do all of the steps of encryption without the user being aware that these steps are taking place. Ultimately, encryption will be as easy as clicking on a box in a dialogue box asking whether the user wants to encrypt or decrypt a message. Key escrow Escrow means a third party holding something not to be released until specific conditions are met. It is commonly used in the phrase “key escrow encryption”, which in turn refers to special encryption technologies such as the Clipper Chip (for use in telephones) and Capstone Chip (for use in computers). Both of these chips use the Skipjack encryption algorithm and include a ‘Law Enforcement Access Field’ (LEAF). When U.S. law enforcement agencies wish to decrypt an encrypted communication that uses a device with a Clipper or Capstone Chip in it, they can obtain the unique identifier for that chip (held in the LEAF). The encryption key for that chip is in two parts, one held by the U.S. National Institute of Standards and Technology (NIST), which is part of the U.S. Dept. of Commerce, the other held by the Treasury Department. These agencies are not to turn over the two parts of the encryption keys unless a court warrant authorizes them to. Once the two parts are put together, it becomes possible to decrypt the messages. NIST has established Clipper and Capstone Chips as voluntary standards for U.S. federal government agencies,, thereby encouraging them but not requiring them to use this technology. This is called NIST’s 1994 Escrow Encryption Standard (EES). Firewalls and Gateways The purpose of firewalls and gateways is to make it possible to connect a computer with confidential information on it to another network which is not confidential and where the various users are not known to the first network. Basically, a firewall is an access control, which controls who on a given network is permitted to send information to other networks, and who from the outside world is permitted to send information to that given network. Generally, the access depends on the user having been given the necessary privileges. Access by an approved user is then obtained by the user providing the necessary information, such as the user’s network address, user ID and password. Generally, firewalls are configured to protect against unauthenticated interactive logins from the "outside" world. This, more than anything, helps prevent vandals from logging into machines on your network. Normally if a user queries a system to find out who or what is connected to that system, all workstations or servers connected to the network would be identified. A firewall ensures that when a query is made in a public network, that the connection to the organization’s network will not be made known in the response to the query, except to authorized users. Some firewalls permit only e-mail traffic through them, thereby protecting the network against any attacks other than attacks against the e-mail service. Other firewalls provide less strict protections, and block only those services that are known to be problems. More elaborate firewalls block traffic from the outside to the inside, but permit users on the inside to communicate freely with the outside. Firewalls are also important since they can provide a single "choke point" where security and audit can be imposed. Thus, the firewall can act as an effective "phone tap" and tracing tool. Firewalls cannot protect against attacks that do not go through the firewall. Users have been known, however, to bypass firewalls by installing a modem into their PC to allow remote dial-in. Firewalls cannot protect very well against things like viruses. There are too many ways of encoding binary files for transfer over networks, and too many viruses to try to search for them all. In other words, a firewall cannot replace security- consciousness on the part of users. In general, a firewall cannot protect against a data-driven attack — attacks in which something is mailed or copied to an internal host where it is then executed. A secure gateway (or Trusted Guard) goes beyond the function of a firewall in that not only does it act as a firewall but has added features involving looking into each packet of information passing through it to or from an authorized user to ensure that no unauthorized information is enclosed in the packet. Current technology limits this to e-mail or EDI because specific information would reside in a specific format in a specific location. Artificial Intelligence must become much more advanced before this feature would be able to address general communications packets. An unsecure gateway may be thought of simply as an access point to a computer system. Hacking (examples of vulnerability and attacks) (much of the information below is drawn from “Network Confidential”, Joe Flower, New Scientist, October 1994, p. 27) Any person who attempts to enter a computer or communications network without authorization is known as a hacker. Before personal computers became popular, “hackers” were known as “phreaks,” persons who attempted unauthorized use of telephone networks. At that time, “hackers” simply referred to persons who spent a lot of time working with computers, generally with no formal training in the computer operations they were running (there was no implication of illicit behaviour in the word). The basic source of network vulnerability is that if a computer is attached to a wire, anything else that can connect to that wire can connect to that computer. Thus, if a computer is attached to a telephone wire, it is potentially attached to every telephone wire in the world. Getting to the computer is one problem, getting into the computer is a separate challenge. There are many different ways to get to a computer that is connected to a telephone line. Trying to figure out which computer to go to can be done in a number of ways. One way is to assemble lists of people who participate in any of the Internet discussion groups. It is simple to capture the the login identifiers (the electronic “address”) of everyone participating in the discussion. A program called “Finger” can then be used to gather details about the persons attached to the address, such as the user’s real name, and, if the person is using a computer service and it is a personal account, Finger can find other information associated with the account, such as the home phone number and address. If a person wants to take part in a discussion group, it will be necessary to type in the person’s ID and password, which is transmitted unencrypted through dozens of different computers until they arrives at the computer where the discussion is actually hosted. Hackers can sneak into one of dozens of computers (Internet “nodes”) “and set up a “sniffer” program that examines the address header accompanying each message that passes through the node. The sniffer copies the login ID, searches the message for anything that looks like a password, and sends the lists of login IDs and passwords back to the hacker. There are also programs which can track everyone who reads particular newsgroups or conferences, even if they never contributed to them. So if a person even glances at a controversial discussion group out of one-time-only curiousity, the person could be recorded on a list somewhere. Once a hacker identifies a target, the hacker will try to enter the system. The hacker can try to obtain a password in a number of ways, such as by tricking persons into revealing their codes, by finding passwords in trash bins, or by using automatic dialers to discover passwords. Dialers work by randomly generating possible passwords until the correct one is found. Generally, the dialers only work on relatively simple passwords. Many computers comprise three separate systems: the mini-computer (and the data and programs stored there); the local area network (and the file servers located there); and the organization’s mainframe computers. Most end-users need to get at applications on all three systems. (Increasingly, for reasons of security, license requirements and quality and productivity control, it is desirable to make software available on a central computer or to distribute it to users from a central computer.) There is little or no correspondence between security on one system and another. “The only way to assign access levels system-wide is for the network manager to go in and do so on a user-by-user basis – a full-time career even in a mid-sized organization.” On any system, there will always be a person with access rights to instal and alter the security software itself. This is known as the ‘systems administrator’ who has ‘super-user’ rights. Thus, hackers frequently seek to obtain ‘super-user’ rights. Typically, there are two ways that hackers attempt to circumvent an access control package: they either alter the AUTOEXEC.BAT file or CONFIG.SYS files (which launch the security program) or start the workstation from a system disk in the A drive (thus bypassing the security system on the network or mainframe; using the A drive this assumes the individual is physically present with the computer). Once entered through the A drive, it is possible to switch to the C drive to gain access to the network. Some organizations take away disk drives from individual workstations to prevent this problem. It is also possible to add hardware to the workstation that prevent starting the computer from the A drive, but the hardware costs hundreds of dollars each and are notorious for causing conflicts with other PC components. Another approach is to program the computer not to start from the A drive. Most access control programs generate and store audit trails. However, the audit trail is not a deterent for persons who use another user’s name, easily done if a person walks away from their machine while they are logged on (a common practice). To guard against this, most access control packages lock the computer after a period of inactivity and require a password to reactivate the computer. Passwords are frequently easy to guess, or to discover. Hash (see encryption) Integrity (see authentication and authorization) Integrity of information means the information is accurate, complete and dependable. It is especially important to prove that a message sent from A to B has not been altered during transmission; that is, that the integrity of the message has been maintained. Internet An international network of computer systems which originally started out as a network of university and military research computers in the USA and which has rapidly grown to a worldwide network of unknown numbers of computers and networks. It was originally funded by the U.S. Department of Defence’s Advanced Research Projects Agency (ARPA). The network was originally called ARPANET. Eventually, the military part of the network (MILNET) was split off and the remaining part of ARPANET was decommissioned in 1990. However, its functionality continued under the National Science Foundation’s NSFNET, which became a prominent backbone of the Internet. Because the Internet is a “network of networks”, major networks within the overall Internet are called “backbones” of the Internet. World Wide Web is an information (data, sound, graphics, etc) search and retrieval service which allows a user to select key- words or phrases and then provides all matching information locations on the Internet. FTP (file transfer protocol), allows one to transfer files to and from any computer linked to the Internet. TELNET allows oneself to be teleported (so to speak) to a specific computer somewhere on the Internet and then to use that computer as if one where a user on site rather than an Internet user somewhere else in the world. The Gopher service is a text search and retrieval service for the Internet, first developed at the University of Minnesota (whose sports team name is the Golden Gophers). It will go to multiple sites to search for particular subjects being sought. It finds the sources and tells you where the sources are. You then have to go each location to retrieve the information. World Wide Web provides a graphical user interface, search and retrieval of text and other kinds of electronic information, provides lists of sources, links files (hypertext) so you can go directly from one to another related file, and takes you directly to the source where the file is located. Legacy System An computer system using old technology which, for some reason, cannot immediately be updated or replaced. Consequently, provision must be made for it within the current or proposed IT system. Non-repudiation Proof that a message was sent by a particular person at a particular time, without being altered since being sent. This proof should be verifiable by any third party at any time. Without non-repudiation, persons could deny having sent certain messages, which could cause important problems, particularly in financial transactions. Thus, an essential part of any signature (digital or otherwise) is the difficulty that a person would have in repudiating what appears to be his or her signature. Sensitive Information There are two kinds of government information: sensitive and non- sensitive. There are two types of sensitive information: classified and designated. Classified information is information whose disclosure could harm the national interest and designated information is information whose disclosure could cause other kinds of harm. All other information is non-classified and non- designated. Classified and designated information is divided into three categories reflecting the seriousness of harm that could result if the information is disclosed. For classified information, the three categories are Top Secret, Secret and Confidential. For designated information the three categories are extremely sensitive, particularly sensitive and low-sensitive for designated information (sometimes called Protected C, B and A, respectively). The primary guide for determining whether information should be classified or designated is by referring to the various provisions of the Access to Information Act and the Privacy Act. The Security Policy provides an Invasion of Privacy test to determine which level of sensitivity should be given to particular kinds of personal information. The Security Policy also states that one of its fundamental principles is the “need- to-know” principle: classified and designated information should only be disclosed to individuals who have a legislated right to access the information or who have a need to access the information to perform their public service duties. Smart cards (see advanced card technologies) TEMPEST One security officer has suggested that TEMPEST is an acronym for Transient Electro-Magnetic Pulse and Emissions Suppression Testing. Computers and keyboards give off electromagnetic and acoustic emanations which can be heard or monitored using appropriate equipment, and from the sounds or signals it can be determined what information is being communicated. TEMPEST refers to the tests to see where this might be a particular problem. Generally, it would only be a problem where there is a strong likelihood that the information being communicated is of special interest to those few people (or countries) with the necessary equipment and staff resources to try to intercept such signals and sounds. Tokens (see advanced card technologies)Anything that can be used to identify the user. A smart card is a token. Some encryption devices, because they contain a mathematical formula or number which is unique to that device, can be a token. Value Added Networks (VANs) A third party communications network which arbitrates the flow of messages (usually commercial) between parties. The services provided typically include receiving, storing and forwarding messages on behalf of a party. VANs are generally used to overcome communications difficulties resulting from different time zones, communications protocols, technologies and to reduce the need for one party to establish large numbers of direct connections to other parties. A party can simply direct all its commercial partner to direct their messages to one location: its VAN. Virus A Trojan Horse that attaches itself to legitimate programs in a computer system and reproduces itself. Viruses have a mission component, a trigger component and a self-propagating component. They can damage a system simply by filling the memory due to having too many reproduced copies or requiring the system to do too many reproduction commands. Trojan Horse: A computer program with an apparently or actually useful function that contains additional, hidden functions that operate once inside the computer system. Generally, Trojan Horses refer to hidden operations that cause damage to computer systems or that compromise the security of the system. Worm: A program that creates a complete copy of itself. Unlike a virus, a worm does not need to attach to a host program. Like a virus, a worm may access, alter, destroy, or consume system resources using the rights and privileges of its host program or user. Bacteria (see virus, Trojan Horse, worm, bugs): A sub-class of viri; malicious software that does not attach itself specifically to other programs, does not use network resources, but still has a damaging effect. Logic bombs: A computer program that triggers an unauthorized act when particular circumstances on the system occur. For example, an individual working on a payroll program could instruct the computer to delete the hard drive when that individual’s name no longer appears on the payroll. Time-Bomb: A computer program that triggers an unauthorized act at a particular time. A Time-Bomb is one kind of Logic Bomb. Many viruses include time-bombs so that damage is caused to stored data on a given date. Bugs: A defect in a computer program. Most computer programs have bugs and most of the time spent developing a program is spent finding and eliminating the bugs.