A collision is when the hash value of two different data from one hash function is the same. Hashing is one-way cryptography, which means that the original plaintext is no longer known to be one or the other.

〇：Brewer-Nash Model

The Chinese Wall Model is a security model that focuses on the flow of information within an organization, such as insider trading. Insider trading occurs when inside information leaks to the outside world. In reality, information can spread to unexpected places as it is passed on orally to unrelated parties. In order to take such information flow into account, access privileges are determined in a simulation-like manner. Therefore, the correct answer is the “Chinese Wall Model (Brewer-Nash Model).

×：Lattice-based Access Control

Lattice-based access control is to assume that a single entity can have multiple access rights and to consider access control as all possible relationships under a certain condition.

×：Biba Model

The Biba model is a security model that indicates that data cannot be changed without permission.

×：Harrison-Ruzzo-Ullman Model

The Harrison-Ruzzo-Ullman model is a model that aggregates the eight rules of the Graham-Denning model into six rules using an access control matrix.

〇：Known Plaintext Attacks

A known-plaintext attack is a situation in which a decryptor can obtain plaintext indiscriminately. A ciphertext-alone attack is a situation where a decryptor can acquire ciphertext indiscriminately. A known-plaintext attack acquires the plaintext but does not know what ciphertext it is paired with, meaning that decryption is attempted with only two random ciphertexts. In this situation, it is difficult to decrypt. Therefore, the correct answer is “known-plaintext attack.

×：Selective Plaintext Attack

A choice-plaintext attack is a situation in which the decryptor can freely choose the plaintext to acquire and obtain the ciphertext.

×：Adaptive Choice Plaintext Attack

An adaptive choice-plaintext attack is a situation in which the decryptor can freely choose which plaintext to acquire and acquire the ciphertext, and can repeat the acquisition again after seeing the result.

×：None of the above

It is rare for the answer to be “none of the above” when the choice is “most of the above.

〇：Graham-Denning Model

The Graham-Denning model addresses how access rights between subjects and objects are defined, developed, and integrated. It defines a basic set of rights in terms of the commands that a particular subject can execute on an object. The model has eight basic protective rights or rules on how to safely perform these types of functions

×：Brewer-Nash Model

It is incorrect because its purpose is to provide access control that can be changed dynamically according to the user’s previous actions. The main purpose is to protect against conflicts of interest due to user access attempts. For example, if a large marketing firm provides marketing promotions and materials for two banks, the employee responsible for the Bank A project should not be able to see information about Bank B, the marketing firm’s other bank customer. A conflict of interest could arise because the banks are competitors. If the project manager of the marketing firm’s Project A can see information about Bank B’s new marketing campaign, he may attempt to execute it rather than promote it to please more direct customers. Marketing firms have a bad reputation when internal employees can act irresponsibly.

×：Clark-Wilson Model

The Clark-Wilson model is incorrect because it is implemented to protect data integrity and ensure that transactions are properly formatted within the application. Subjects can only access objects through authorized programs. Segregation of duties is enforced. Auditing is required. The Clark-Wilson model addresses three integrity goals: preventing changes by unauthorized users, preventing inappropriate changes by unauthorized users, and maintaining internal and external consistency.

×：Bell-LaPadula Model

This model was developed to address concerns about the security of U.S. military systems and the leakage of classified information, and is incorrect. The primary goal of the model is to prevent unauthorized access to classified information. It is a state machine model that enforces the confidentiality aspect of access control. Matrices and security levels are used to determine if a subject has access to different objects. Specific rules are applied to control how objects interact with each other compared to the subject’s object classification.

Compromise is when what used to be secure encryption becomes insecure due to the evolution of computers. Cryptography is based on the sharing of a single answer, a key, among those communicating. The key is generated by computer calculations, and a third party must solve a difficult problem that would take several years to derive. However, as the computational power of computers has evolved, it is now possible to solve difficult problems that could not be solved before. In this case, encryption is meaningless. This is the compromise caused by evolution. Therefore, the correct answer is “Compromise.

In an asymmetric algorithm, every user must have at least one key pair (private and public key). In a public key system, each entity has a separate key. The formula for determining the number of keys needed in this environment is by the number N × 2, where N is the number of people to distribute. In other words, 260 x 2 = 520. Therefore, the correct answer is 520.

〇：Statistically predictable

The two main types of symmetric algorithms are block ciphers and stream ciphers. Block ciphers perform a mathematical function on a block of bits at a time. Stream ciphers do not divide the message into blocks. Instead, a stream cipher treats the message as a stream of bits and performs the mathematical function on each bit individually. If it were statistically predictable, it would not be a practical encryption technique in the first place.

×：Statistically Fair Keystreams

Statistically fair keystreams are an element of good stream ciphers. Therefore, it is incorrect. Another way to say a statistically unbiased keystream is that it is a highly random keystream that is difficult to predict.

×：The repetitive pattern of bit strings treated in a keystream is long.

Another way to say the randomness of a keystream is that it is highly random, with long repetitions = rarely repeated = highly random.

×：The keystream is irrelevant to the key.

A keystream that is not related to a key is an element of a good stream cipher. Therefore, it is incorrect. This is important because the key provides the randomness of the encryption process.

〇：One-time pad

Stream ciphers refer to one-time pad technology. In practice, stream ciphers cannot provide the level of protection that one-time pads do, but are practical.

×：AES

AES is incorrect because it is a symmetric block cipher. When a block cipher is used for encryption and decryption purposes, the message is divided into blocks of bits.

×：Block ciphers

Block ciphers are used for encryption and decryption purposes. The message is wrong because it is divided into blocks of bits.

×：RSA

RSA is incorrect because it is an asymmetric algorithm.

〇：Operating system patching is easier.

This is an incorrect choice question. Virtualization does not simplify operating system patching. In fact, it complicates it by adding at least one additional operating system. Each operating system differs from the typical version configuration, adding to the complexity of patching. The server’s own operating system runs as a guest within the host environment. In addition to patching and maintaining the traditional server operating system, the virtualization software itself must be patched and maintained.

For this question, we do not require an understanding of all the technical systems of virtualization. What is required here is a selection of answers based on a process of elimination.

×：I can build a secure computing platform.

Building a secure computing platform may not be a feature of virtualization per se. However, can we build a secure environment? This is not a false choice because it cannot be ruled out.

×：It can provide fault and error containment.

Virtualization can be host independent. In terms of containment, it can be interpreted as being able to provide fault and error containment through independence from physical servers. Therefore, it cannot be denied and is therefore not an incorrect choice.

×：It can provide powerful debugging capabilities.

Virtualization can reproduce a unique environment, not just put up a clean virtual host. Therefore, it is undeniable and therefore out of the wrong choice.

〇：Conflict condition

A race condition is present when a process performs a task on a shared resource and the sequence could be in the wrong order. 2 or more processes can have a race condition if they use a shared resource, like data in a variable. It is important that processes perform their functions in the correct sequence.

×：Backdoors

Backdoors are incorrect because they are “listening” services on certain ports. Backdoors are implemented by attackers to allow easy access to the system without authenticating as a normal system user.

×：Maintenance Hooks

Maintenance hooks are specific software codes that allow easy and unauthorized access to sensitive parts of a software product. Software programmers use maintenance hooks to allow them to get quick access to the code so that they can make fixes in immediate, but this is dangerous.

×：Data validation errors

Data validation errors are wrong because an attacker cannot operate on the process execution sequence.

〇：Digital Signature

A digital signature is a hash value encrypted with the sender’s private key. The act of signing means encrypting a hash value of a message with a private key. A message can be digitally signed, providing authentication, non-repudiation, and integrity. The hash function guarantees the integrity of the message, and the signature of the hash value provides authentication and non-repudiation.

×：Encryption Algorithms

Encryption algorithms are wrong because they provide confidentiality. Encryption is most commonly performed using symmetric algorithms. Symmetric algorithms can provide authentication, non-repudiation, and integrity as well as confidentiality.

×：Hash Algorithms

Hash algorithms are wrong because they provide data integrity. Hash algorithms generate a message digest, which detects whether modifications have been made (also called a hash value). The sender and receiver individually generate their own digests, and the receiver compares these values. If they differ, the receiver can know the message has been modified. Hash algorithms cannot provide authentication or non-repudiation.

×：Encryption paired with digital signatures

This is incorrect because encryption and digital signatures provide confidentiality, authentication, non-repudiation, and integrity. Encryption alone provides confidentiality. And digital signatures provide authentication, non-repudiation, and integrity. The question requires that it can provide authentication, non-repudiation, and integrity. It is a nasty question.

〇：The expiration date should be set short.

Key management is critical for proper protection. Part of key management is to determine the key’s period of validity, which would be determined by the sensitivity of the data being protected. For sensitive data, periodic key changes are required and the key’s expiration date will be shortened. On the other hand, for less secure data, a key with a longer expiration date is not a problem.

×：Keys should be deposited in case of backup or emergency.

This is incorrect because it is true that keys must be deposited in the event of a backup or emergency situation. Keys are at risk of being lost, destroyed or damaged. Backup copies must be available and readily accessible when needed.

×：Keys must not be made public.

Of course. It is a key.

×：Keys should be stored and transmitted by secure means.

Wrong, since it is true that keys should be stored and transmitted by secure means. Keys are stored before and after distribution. If keys are distributed to users, they must be stored in a secure location in the file system and used in a controlled manner.

〇：Data Encryption

The Advanced Encryption Standard (AES) is a data encryption standard developed to improve upon the previous de facto standard, Data Encryption Standard (DES). As a symmetric algorithm, AES is used to encrypt data. Therefore, the correct answer is “data encryption.

There are other situations where AES is used in the other choices, but encrypting data is the most focused or better answer. Thus, there are cases where all of the choices are correct.

×：Data integrity

This is a characteristic of digital signatures.

×：Key recovery

It is a property of decryption and key escrow.

×：Symmetric key distribution

Using symmetric keys for AES distribution lowers the key delivery problem.

〇：A virtual firewall in bridge mode allows the firewall to monitor individual traffic links, while a firewall integrated into the hypervisor can monitor all activity taking place within the host system.

Virtual firewalls can be bridge-mode products that monitor individual communication links between virtual machines. They can also be integrated within a hypervisor in a virtual environment. The hypervisor is the software component that manages the virtual machines and monitors the execution of guest system software. When a firewall is embedded within the hypervisor, it can monitor all activities that occur within the host system.

×：A virtual firewall in bridge mode allows the firewall to monitor individual network links, while a firewall integrated into the hypervisor can monitor all activities taking place within the guest system.

A virtual firewall in bridge mode is incorrect because the firewall can monitor individual traffic links between hosts and not network links. Hypervisor integration allows the firewall to monitor all activities taking place within the guest system rather than the host system.

×：A virtual firewall in bridge mode allows the firewall to monitor individual traffic links, while a firewall integrated into the hypervisor can monitor all activities taking place within the guest system.

A virtual firewall in bridge mode is wrong because the firewall can monitor individual traffic links, and the hypervisor integration allows the firewall to monitor all activity taking place within the host system, but not the guest system. The hypervisor is the software component that manages the virtual machines and monitors the execution of the guest system software. A firewall, when embedded within the hypervisor, can monitor all activities taking place within the system.

×：A virtual firewall in bridge mode allows the firewall to monitor individual guest systems, while a firewall integrated into the hypervisor can monitor all activities taking place within the network system.

A virtual firewall in bridge mode allows the firewall to monitor individual traffic between guest systems, and a hypervisor integrated allows the firewall to monitor all activity taking place within the host system, not the network system, so Wrong.

A key escrow system is one in which a third-party organization holds a copy of the public/private key pair. If the private key is stolen, all ciphers can be decrypted. Conversely, if it is lost, all ciphers cannot be decrypted. Therefore, you want to have a copy. However, if you have it yourself, it may be stolen if a break-in occurs, so you leave it with a third-party organization.

Cryptographic algorithms refer to the calculations to be encrypted, and even if the cryptographic algorithms were publicly available, it would take an enormous amount of effort to decipher them. cryptographic algorithms that provide modern cryptography, such as AES, are publicly available. On the other hand, in-house development is not recommended because, although it has the security of concealment, it requires a great deal of resources to be allocated.

〇：The sender encrypts the message digest with his/her private key.

A digital signature is a hash value encrypted with the sender’s private key. The act of digitally signing means encrypting the hash value of the message with his/her private key. The sender would encrypt that hash value using her private key. When the recipient receives the message, she performs a hash function on the message and generates the hash value herself. She then decrypts the hash value (digital signature) sent with the sender’s public key. The receiver compares the two values and, if they are the same, can verify that the message was not altered during transmission.

×：The sender encrypts the message digest with his/her public key.

The sender is wrong because if the message encrypts the digest with his/her public key, the recipient cannot decrypt it. The recipient needs access to the sender’s private key, which must not occur. The private key must always be kept secret.

×：The receiver encrypts the message digest with his/her private key.

The receiver is wrong because the message must decrypt the digest with the sender’s public key. The message digest is encrypted with the sender’s private key, which can only be decrypted with the sender’s public key.

×：The receiver encrypts the message digest with his/her public key.

The receiver is wrong because the message must decrypt the digest with the sender’s public key. The message digest is encrypted with the sender’s private key, which can only be decrypted with the sender’s public key.

〇：Primary entry doors should have controlled access via swipe card or cryptographic locks. Secondary doors should not be secured from the inside and allowed entry. 

Data centers, server rooms, and wiring closets should be located in the core areas of the facility, near wiring distribution centers. Strict access control mechanisms and procedures should be implemented for these areas. Access control mechanisms can lock smart card readers, biometric readers, or a combination of these. These restricted areas should have only one access door, but fire code requirements typically dictate that there must be at least two doors in most data centers and server rooms. Only one door should be used for daily entry and exit and the other door should be used only in case of an emergency, i.e., if a fire breaks out in a data center or server room, the door should be locked. This second door should not be an access door, meaning people should not be able to come through this door. It should be locked, but should have a panic bar that will release the lock if it is used as an exit, pushed from the inside.

×：The primary and secondary entry doors must have control access via swipe cards or cryptographic locks.  

This is incorrect because even two entry doors should not be allowed to pass through with the identification, authentication, and authorization process. There should only be one entry point into the server room. No other door should provide an entry point, but can be used for an emergency exit. Therefore, secondary doors should be protected from the inside to prevent intrusion.

×：The primary entry door should have controlled access via a guard. Two doors should not be secured from the inside and allowed entry.

The main entry door to the server room is incorrect as it requires an identification, authentication, and authorization process to be performed. Swipe cards and cryptographic locks perform these functions. Server rooms should ideally not be directly accessible from public areas such as stairways, hallways, loading docks, elevators, and restrooms. This helps prevent foot traffic from casual passersby. Those who are by the door to the area to be secured should have a legitimate reason for being there, as opposed to those on the way to the meeting room, for example.

×：The main entry door must have controlled access via swipe card or crypto lock. Two doors must have security guards.  

Two doors should not have security guards, because it is wrong. The door should be protected from the inside simply so it cannot be used as an entry. Two-door must function as an emergency exit.

〇：Execution Domain Switching

Execution domain switching occurs when the CPU needs to move between executing instructions for a more trusted process versus a less trusted process. Trusted Computing Base (TCB) allows processes to switch domains in a secure manner to access different levels of information based on sensitivity. Execution domain switching occurs when a process needs to invoke a process in a higher protection ring. The CPU executes the user-mode instruction back into privileged mode.

At first glance, this is a geeky problem that does not make sense. But don’t give up. Since there is no such thing as skipping, you can only get a right or wrong answer when the question is posed, so it is preferable to answer the question with some degree of prediction.

From this point on, let’s consider how to answer the questions. If you look at the question text and read it to the point where it reads, “You moved from one area to the other, and that was a security breach?” If you can read to that point, then you have two choices: deny or “stop the process,” or change or “switch the domain of execution. Next, the question text reads “if you need to access it,” which is asking how to accomplish this objective, not whether or not you should.

×：Execution of I/O operations

This is incorrect because input/output (I/O) operations are not initiated to ensure security when a process in one domain needs to access another domain in order to retrieve sensitive information. I/O operations are performed when input devices (such as a mouse or keyboard) and output devices (such as a monitor or printer, etc.) interact with an application or applications.

×：Stopping a Process

A process deactivation is one that occurs when a process instruction is fully executed by the CPU or when another process with a higher priority calls the CPU, which is incorrect. When a process is deactivated, new information about the new requesting process must be written to a register in the CPU. The TCB component must ensure that this is done, since the data replaced in the registers may be confidential.

×：Mapping from virtual memory to real memory

Incorrect because memory mapping occurs when a process needs its instructions and data processed by the CPU. The memory manager maps logical addresses to physical addresses so that the CPU knows where to place the data. This is the responsibility of the operating system’s memory manager.