Certified Information Systems Security Professional (CISSP) Security Architecture and Engineering Syllabus

What is the Trusted Computing Base (TCB)?

The totality of protection mechanisms (hardware, firmware, software) within a system that enforces the security policy. Its correct operation is what the system's security depends on.

Define the 'security perimeter' and 'reference monitor.'

Security perimeter = the boundary separating the TCB from the rest of the system. Reference monitor = the abstract concept that mediates all access by subjects to objects, enforcing the access control policy.

What is the 'security kernel'?

The hardware, firmware, and software elements of the TCB that implement the reference monitor concept. It must be tamperproof, always invoked (non-bypassable), and small enough to be verified.

What three properties must a reference monitor satisfy?

It must be tamperproof, always invoked (non-bypassable/complete mediation), and small/simple enough to be thoroughly analyzed and verified (verifiable).

What is a Trusted Platform Module (TPM)?

A dedicated hardware chip that provides secure cryptographic functions and tamper-resistant storage for keys, supporting features like full-disk encryption (e.g., BitLocker), remote attestation, and sealed storage.

What is the purpose of CPU protection rings (rings 0–3)?

They enforce a hierarchy of privilege: Ring 0 = OS kernel (most privileged), Rings 1–2 = OS/drivers, Ring 3 = user applications (least privileged). Lower-privilege code cannot directly access higher-privilege resources.

What security capability does 'process isolation' provide?

It keeps each process's memory and resources separate so one process cannot read or modify another's memory, preventing interference and unauthorized data access.

What is a covert channel, and name its two types.

A communication path used to transfer information in a way that violates security policy. Two types: covert storage channel (writing to a shared storage location) and covert timing channel (modulating the timing of events/resource use).

What is a 'maintenance hook' (backdoor) vulnerability?

An undocumented access method left in software (often by developers for testing/maintenance) that bypasses normal authentication and can be exploited if discovered.

What is a TOCTOU (time-of-check to time-of-use) attack?

A race-condition attack that exploits the gap between when a system checks a condition (e.g., permissions) and when it uses the result, altering the resource in between to bypass the control.

What architectural vulnerability does an emanation/TEMPEST attack exploit, and what is a countermeasure?

It exploits electromagnetic emanations leaking from devices that can reveal processed data. Countermeasures include TEMPEST shielding, Faraday cages, white noise, and control zones.

What are the four goals/services cryptography can provide?

Confidentiality (secrecy), integrity (detecting alteration), authentication (verifying identity/origin), and non-repudiation (preventing denial of an action). Availability is sometimes added.

Distinguish a substitution cipher from a transposition cipher.

A substitution cipher replaces each plaintext unit with another (e.g., Caesar shift); a transposition (permutation) cipher rearranges the order of the plaintext units without changing them.

What is the only theoretically unbreakable cipher, and what are its requirements?

The one-time pad. It is unbreakable if the key is truly random, at least as long as the message, used only once, and kept perfectly secret (distributed securely).

What is the difference between confusion and diffusion (Shannon)?

Confusion makes the relationship between the key and ciphertext as complex as possible (via substitution); diffusion spreads the influence of each plaintext bit across many ciphertext bits (via transposition/permutation).

Compare a stream cipher and a block cipher.

A stream cipher encrypts data bit/byte by bit (typically XOR with a keystream, fast, for continuous data like voice); a block cipher encrypts fixed-size blocks (e.g., 128 bits) at a time using rounds of substitution and transposition.

What is the key characteristic of symmetric cryptography?

The same secret (shared) key is used for both encryption and decryption. It is fast and good for bulk data but has a key distribution problem and does not provide non-repudiation.

How many keys are needed for n users to communicate pairwise using symmetric cryptography?

n(n-1)/2 keys. This scalability/key-management problem is a major weakness of symmetric cryptography.

List key facts about DES, 3DES, and AES block/key sizes.

DES: 64-bit block, 56-bit effective key (broken). 3DES: applies DES three times, effective 112-bit (2-key) or 168-bit (3-key) strength. AES (Rijndael): 128-bit block; key sizes 128, 192, or 256 bits.

Name common symmetric algorithms besides DES/AES.

IDEA (128-bit key), Blowfish (variable up to 448-bit), Twofish, RC4 (stream), RC5, RC6, CAST, Skipjack, Serpent.

What is the foundation of asymmetric (public-key) cryptography?

A mathematically related key pair: a public key (shared openly) and a private key (kept secret). What one key encrypts only the other can decrypt. Based on hard math problems (factoring, discrete logs).

In public-key crypto, which key provides confidentiality and which provides non-repudiation/digital signature?

Encrypt with the recipient's public key for confidentiality (only their private key decrypts). Encrypt/sign with the sender's private key for authentication/non-repudiation (anyone can verify with the sender's public key).

What hard problem underlies RSA, and what underlies Diffie-Hellman/ElGamal/ECC?

RSA relies on the difficulty of factoring large composite numbers. Diffie-Hellman, ElGamal, and ECC rely on the difficulty of computing discrete logarithms (ECC over elliptic curves).

What is the main advantage of Elliptic Curve Cryptography (ECC)?

ECC provides equivalent security with much smaller key sizes (e.g., a 256-bit ECC key ≈ 3072-bit RSA), making it efficient for low-power and mobile devices.

What does the Diffie-Hellman algorithm accomplish, and what is its main vulnerability?

It allows two parties to securely establish a shared secret key over an insecure channel without prior secrets. It is vulnerable to man-in-the-middle attacks because it provides no authentication by itself.

What is a hybrid cryptosystem and why is it used?

It uses asymmetric crypto to securely exchange a symmetric session key, then uses fast symmetric crypto to encrypt the bulk data. It combines secure key distribution with speed (e.g., SSL/TLS, PGP).

What properties must a cryptographic hash function have?

Fixed-length output (digest), one-way (preimage resistant), deterministic, fast to compute, and collision resistant (computationally infeasible to find two inputs with the same hash) including second-preimage resistance.

Give the digest sizes for MD5, SHA-1, SHA-256, and SHA-512.

MD5 = 128 bits (broken). SHA-1 = 160 bits (deprecated, collisions found). SHA-256 = 256 bits. SHA-512 = 512 bits (both part of SHA-2). SHA-3 (Keccak) is the newest standard.

What is a MAC (Message Authentication Code) and what does it provide?

A MAC is a keyed checksum (e.g., HMAC, CBC-MAC, CMAC) that provides both integrity and authentication of a message using a shared secret key. It does not provide non-repudiation.

How is a digital signature created and what services does it provide?

Hash the message, then encrypt the hash with the signer's private key. It provides integrity, authentication, and non-repudiation (but not confidentiality unless the message itself is also encrypted).

What is the difference between a MAC and a digital signature regarding non-repudiation?

A MAC uses a shared symmetric key, so either party could have created it — no non-repudiation. A digital signature uses the signer's unique private key, providing non-repudiation.

What is a digital certificate and which standard defines it?

A digital certificate binds a public key to an identity, vouched for by a Certificate Authority's digital signature. It follows the X.509 standard (containing version, serial number, subject, issuer, validity period, public key, signature).

What are the main components/roles in a PKI?

Certificate Authority (CA, issues/signs certs), Registration Authority (RA, verifies identities), the certificate repository, and CRL/OCSP services for revocation; plus end entities and the certificate policy/CPS.

What is the difference between a CRL and OCSP?

A Certificate Revocation List (CRL) is a periodically published list of revoked certificate serial numbers (can be stale). OCSP (Online Certificate Status Protocol) provides real-time, on-demand certificate validity status.

What is key escrow and what is the M of N control concept?

Key escrow stores a copy of keys with a trusted third party for recovery/lawful access. M of N control requires that at least M of N total key custodians cooperate to reconstruct/recover a key, preventing single-person abuse.

What are the typical stages of the key management lifecycle?

Generation, distribution, storage, use, rotation/change, archival/escrow, recovery, and destruction. Keys should be protected throughout and securely destroyed at end of life.

What is the difference between a known-plaintext and a chosen-plaintext attack?

Known-plaintext: attacker has some plaintext-ciphertext pairs and uses them to deduce the key. Chosen-plaintext: attacker can encrypt plaintexts of their choosing and analyze the resulting ciphertext (e.g., differential cryptanalysis).

What is a ciphertext-only attack?

The attacker has access only to ciphertext (no plaintext) and tries to recover the key or plaintext, typically via frequency analysis or brute force. It is the hardest attack to perform.

What is a birthday attack and what does it target?

An attack based on the birthday paradox that finds hash collisions much faster than brute force (roughly 2^(n/2) for an n-bit hash). It targets the collision resistance of hash functions.

What is a meet-in-the-middle attack and which algorithm did it weaken?

An attack that encrypts from one end and decrypts from the other to match in the middle, reducing the effective key strength of double encryption. It is why Double DES offers little more security than single DES (driving the use of 3DES).

What is a side-channel attack? Give examples.

An attack that exploits physical information leaked by a cryptosystem's implementation rather than the math. Examples: timing analysis, power analysis (SPA/DPA), electromagnetic emanation, acoustic, and cache attacks.

What is a replay attack and a common countermeasure?

Capturing valid data/authentication and retransmitting it to gain access. Countermeasures: timestamps, nonces, sequence numbers, and session tokens that make captured data invalid for reuse.

What is differential vs linear cryptanalysis?

Differential cryptanalysis studies how differences in plaintext affect differences in ciphertext (uses chosen plaintext) to find the key. Linear cryptanalysis uses linear approximations of the cipher's behavior across many known plaintexts to recover key bits.

What is a fault-injection (implementation) attack?

Deliberately inducing errors in a device (via voltage glitches, clock manipulation, heat, or radiation) to cause faulty computations that reveal secret keys or bypass security checks.

Why is quantum computing a concern for cryptography, and what is the response?

Shor's algorithm could break RSA/ECC/DH by efficiently factoring and solving discrete logs; Grover's algorithm halves symmetric key strength. The response is post-quantum (quantum-resistant) cryptography, such as lattice-based algorithms.

What is quantum key distribution (QKD)?

Using quantum mechanics (e.g., photon polarization, BB84 protocol) to securely exchange keys; any eavesdropping disturbs the quantum states and is detectable, guaranteeing detection of interception.

What security concern arises from homomorphic encryption, and what is its benefit?

Benefit: computations can be performed on encrypted data without decrypting it, preserving confidentiality in cloud/processing. Concern: it is computationally expensive/immature and still emerging, requiring careful validation before reliance.

What is CPTED and its three main strategies?

Crime Prevention Through Environmental Design — using the physical environment to reduce crime. Three strategies: natural surveillance, natural access control, and natural territorial reinforcement.

When selecting a facility site, what location-based security factors should be evaluated?

Visibility/profile of the site, surrounding area crime rate and neighbors, proximity to emergency services (police/fire/hospital), accessibility/transportation, and exposure to natural disasters (floods, earthquakes).

What is a security control zone / facility layering approach?

Concentric layers of protection from the perimeter inward (fence, building, internal rooms, secure server room), with increasing controls toward critical assets — applying defense in depth physically.

What CCTV factors affect effective surveillance?

Field of view, focal length/lens, depth of field, lighting/illumination (lux), resolution, and whether the camera is for detection vs identification. PTZ cameras add pan-tilt-zoom flexibility.

What deters vehicles and controls vehicular approach to a facility?

Bollards, planters, and barriers that stop or slow vehicles (anti-ram protection), plus controlled gates and serpentine/curved approach roads.

What is a mantrap (access control vestibule) and what attack does it prevent?

A small space with two interlocking doors where only one opens at a time, allowing one person through at a time. It prevents tailgating/piggybacking by trapping unauthorized followers.

What fence heights are recommended for different deterrence levels?

3–4 ft (1 m) deters casual trespassers; 6–7 ft (2 m) is too high to climb easily; 8 ft (2.4 m) plus three strands of barbed wire deters determined intruders.

What is the difference between tailgating and piggybacking?

Tailgating: an unauthorized person follows an authorized person through a door without their knowledge/consent. Piggybacking: the unauthorized entry happens with the authorized person's consent/knowledge.

What are the four main classes of locks/access devices used for physical entry?

Mechanical (key/warded, pin-tumbler), combination/cipher locks, electronic/smart locks, and biometric locks; plus device locks (cable locks, port controls). Locks are considered delay devices, not absolute prevention.

What is the recommended guideline for guard and dog use in physical security?

Security guards provide discernment/judgment and can adapt to situations (best where reasoning is needed); guard dogs provide deterrence and detection but raise cost, care, and liability concerns.

What lighting standard is commonly cited for perimeter/critical area illumination?

Critical areas should be illuminated to about 2 foot-candles (≈ 8 feet of light, e.g., the NIST guidance of 2 ft-candles at 8 ft height). Continuous, glare, and standby lighting are common types.

What humidity range is recommended for a data center, and why do extremes matter?

Roughly 40–60% relative humidity. Too low causes static electricity (ESD risk to components); too high causes condensation and corrosion. (Older guidance cited 20–80%, but tighter ranges are preferred.)

What is the ideal temperature range for a server/computer room?

Approximately 64–81°F (18–27°C) per ASHRAE guidance; older CISSP references cite ~60–75°F. Excessive heat damages equipment and components have varying damage thresholds.

Define brownout, blackout, sag, spike, and surge in power terms.

Blackout = total loss of power; Brownout = prolonged low voltage; Sag/Dip = momentary low voltage; Spike = momentary high voltage; Surge = prolonged high voltage. (Fault = momentary loss; inrush = initial surge at startup.)

What is the difference between a UPS and a generator for power continuity?

A UPS provides immediate, short-term battery power to bridge brief outages and allow safe shutdown or transfer; a generator provides long-term backup power for extended outages but takes time to start.

What are the classes of fire and the correct suppression agent for each?

Class A (common combustibles): water/soda acid; Class B (liquids): CO2/foam/dry chemical; Class C (electrical): CO2/non-conductive agent; Class D (combustible metals): dry powder; Class K (kitchen/oils): wet chemical.

What replaced Halon as a clean fire-suppression agent and why?

Halon was banned (Montreal Protocol) for depleting the ozone layer. Replacements include FM-200, FE-13, Inergen, Argon, and CO2 — clean agents that suppress fire without harming electronics.

Compare wet pipe, dry pipe, deluge, and pre-action sprinkler systems.

Wet pipe: water always in pipes (fast, leak risk). Dry pipe: pipes hold air, water released when triggered (delay). Deluge: open heads dump large water volume. Pre-action: combines dry pipe with a two-stage trigger (detector + head) to reduce accidental discharge — best near electronics.

What are the four stages of a fire, and which detector type catches the earliest stage?

Incipient (no visible smoke), Smoke (visible), Flame (visible fire), Heat (intense heat). Ionization/air-aspirating detectors catch the earliest incipient stage; photoelectric detectors sense smoke; heat detectors trigger latest.

Why is a data center typically built on a raised floor, and what hazard does that introduce?

A raised floor routes cabling and conditioned/cooling air underneath. The hazard is water accumulation beneath the floor, so water/moisture detection sensors should be placed under raised floors.

What is the purpose of a positive pressure HVAC system in a data center?

Positive air pressure pushes air outward when doors/openings occur, preventing contaminants, dust, and smoke from being drawn into the room. HVAC should also have dedicated power and emergency shutoff.

What is the difference between a 'hot aisle/cold aisle' arrangement and why is it used?

Server racks are arranged so cold-air intakes face cold aisles and hot exhausts face hot aisles, separating cool supply from hot exhaust air to improve cooling efficiency and prevent equipment overheating.

What grounding and static-control measures protect equipment in a data center?

Proper electrical grounding, anti-static flooring/mats, ESD wrist straps, humidity control (40–60%), and avoiding carpet — all to prevent electrostatic discharge that can damage sensitive components.

What is the security purpose of a Faraday cage in facility design?

A Faraday cage (conductive enclosure) blocks electromagnetic fields, preventing emanation/eavesdropping (TEMPEST) of signals out and shielding equipment from external EMI/RFI interference.