CS 5700 – Computer Security and Information Assurance Section 5: Security in Networks – Part 2 This version of Section 5 includes OPTIONAL slides that.

CS 5700 – Computer Security and Information Assurance Section 5: Security in Networks – Part 2 This version of Section 5 includes OPTIONAL slides that you may SKIP Dr. Leszek Lilien Department of Computer Science Western Michigan University Slides based on Security in Computing. Third Edition by Pfleeger and Pfleeger. Using some slides courtesy of: Prof. Aaron Striegel — course taught at U. of Notre Dame Prof. Barbara Endicott-Popovsky and Prof. Deborah Frincke (U. Idaho) — taught at U. Washington Prof. Jussipekka Leiwo — taught at Vrije Universiteit (Free U.), Amsterdam, The Netherlands Slides not created by the above authors are © by Leszek T. Lilien Requests to use original slides for non-profit purposes will be gladly granted upon a written request.

Covered earlier (in Section 2): Security in Networks – Part 1 – Outline (1)
Outline of Part 1 of Security in Networks (covered in Section 2) 2.1. Network Concepts Introduction The network Media Protocols Types of networks Topologies Distributed systems APIs Advantages of computing networks 2.2. Threats in Networks Network vulnerabilities Who attacks networks? Threat precursors Threats in transit: eavesdropping and wiretapping Protocol flaws

Covered earlier (in Section 2): Security in Networks – Part 1 – Outline (2)
2.2. Threats in Networks - ctd Types of attacks g-1) Impersonation g-2) Spoofing g-3) Message confidentiality threats g-4) Message integrity threats g-5) Web site attacks g-6) Denial of service g-7) Distributed denial of service g-8) Threats to active or mobile code g-9) Scripted and complex attacks Summary of network vulnerabilities

Security in Networks – Part 2 – Outline (1)
Outline of Part 2 of Security in Networks (covered here) 5.1. Network Security Controls Introduction Security threat analysis Impact of network architecture/design and implementation on security Encryption Link encryption vs. end-to-end (e2e) encryption Virtual private network (VPN) PKI and certificates SSH protocol SSL protocol (a.k.a. TLS protocol) IPsec protocol suite Signed code Encrypted Message content integrity controls Error correcting codes Cryptographic checksum Strong authentication One-time passwords Challenge-response systems Digital distributed authentication Kerberos authentication system

5.1. Network Security Controls—ctd. Access controls ACLs on routers Firewalls Intrusion Detection Systems: alarms and alerts Honeypots Traffic flow security Review of network security controls 5.2. Network Security Tools Firewalls Introduction What is a firewall Firewall design Types of firewalls i. Packet filters (i-1) Simple packet filters (i-2) Stateful packet filters ii. Application proxies (ii-1) Guards (“top model” subcategory) iii. Personal firewalls Comparison of firewall types Example firewall configurations What firewalls can—and can’t—block

5.2. Network Security Tools—ctd. Intrusion Detection Systems Introduction Types of IDSs Signature-based IDSs Anomaly-based IDSs Other IDSs Goals for IDSs IDS strengths and limitations Secure Security for Design of PEM (Privacy-enhanced Electronic Mail) Example secure systems i. PGP ii. S/MIME

5. Network Security – Part 2
Part 1 of Security in Networks was covered in lecture Section 2 Part 2 of Security in Networks is covered in this lecture Section 5, as follows: 5.1. Network Security Controls 5.2. Network Security Tools

5.1. Networks Security Controls
Outline Introduction Security threat analysis Impact of network architecture/design and implementation on security Encryption Content integrity Strong authentication Access controls Alarms and alerts Honeypots Traffic flow security Controls review

Introduction There are many security controls:
For program security (incl. s/w engineering issues) For OS security Many of these strategies are useful for network security as well We will now look for security controls designed specifically for computer networks

b. Security threat analysis (1)
Threat analysis steps : 1) Analyze system components and their interactions 2) Analyze possible damage to C-I-A 3) Hypothesize possible kinds of attacks Network elements to be considered: Local elements Nodes / comm links / data storage / processes / devices / LANs Non-local elements Gateways / comm links / control resources / routers / network resources (e.g., databases)

Accessing pgms or data at remote host
Security threat analysis (2) Network threats: Accessing pgms or data at remote host Modifying pgms or data at remote host Running a pgm at a remote host Interception of data in transit Modifying data in transit Insertion of data into communication traffic Incl. replaying previous communication Blocking selected/all traffic Impersonation of entities Attack enablers: Size / anonymity / ignorance / misunderstanding Complexity / motivation / programming skills

c. Impact of network architecture/ design & implement. on security (1)
Security principles for good analysis, design, implementation, and maintenance (as discussed in sections on Pgm Security and OS Security) apply to networks Architecture can improve security by: 1) Segmentation 2) Redundancy 3) Single points of failure 4) Other means

Reducing number of threats
Impact of network architecture/ design & implement. on security (2) Segmentation Architecture should use segmentation to limit scope of damage caused by network penetration by: Reducing number of threats Limiting amount of damage caused by single exploit Enforces least privilege and encapsulation Example 1: component segmentation Placing different components of e-commerce system on different hosts Esp. put on separate host most vulnerable system components E.g., separate host for web server (w/ public access) Exploit of one host does not disable entire system

Example 2: access separation Separating from each other:
Impact of network architecture/ design & implement. on security (3) Example 2: access separation Separating from each other: Production system Testing system Development system E.g., no developer has access to production system and no customer has access to development system

Example: having a redundant web server (WS) in a company
Impact of network architecture/ design & implement. on security (4) 2) Redundancy Architecture should use redundancy to prevent losing availability due to exploit/failure of a single network entity Example: having a redundant web server (WS) in a company Types of redundancy include: Cold spare – e.g., when WS fails, replace it manually with spare WS Warm spare – e.g., failover mode = redundant WSs periodically check each other Hot spare – e.g., 3 WSs configured to perform majority voting

3) Single points of failure (SPF)
Impact of network architecture/ design & implement. on security (5) 3) Single points of failure (SPF) Architecture should eliminate SPFs to prevent losing availability due to exploit/failure of a single network entity Using redundancy is a special case of avoiding SPFs Network designers must analyze network to eliminate all SPFs Example of avoiding SPF (without using redundancy) Distribute 20 pieces of database on 20 different hosts (so called partitioned database) Even if one host fails, 95% of database contents (19/20=95%) still available Elimination of SPFs (whether using redundancy or not) adds cost

4) Other architectural means for improving security
Impact of network architecture/ design & implement. on security (6) 4) Other architectural means for improving security Will be mentioned below as we discuss more network security controls

d. Encryption Arguably most important/versatile tool for network security We have seen that it can be used for: Confidentiality/Privacy Authentication Integrity Limiting data access Kinds of encryption in networks: Link encryption vs. end-to-end (e2e) encryption Virtual private network (VPN) PKI and certificates SSH protocol SSL protocol (a.k.a. TLS protocol) IPsec protocol suite Signed code Encrypted

(i) Link vs. end-to-end encryption (1)
Link encryption = between 2 hosts Data encrypted just before they are placed on physical communication links At OSI Layer 1 (or, perhaps, Layer 2) Fig. 7-20, p. 445 (ed.4) Properties of link encryption (cf. Fig. 7-20) Msgs/pkts unprotected inside S’s/R’s host I.e., unprotected at OSI layers 2-7 of S’s/R’s host (in plaintext) Packets protected in transit between all hosts Pkts unprotected inside intermediate hosts I.e., unprotected at OSI layers 2-3 of interm. hosts => unprotected at data link and network layers at intermediate hosts (if link encryption at Layer 1) Layers 2-3 provide addressing and routing

Encryption service provided by physical (or data) layer
Link vs. end-to-end encryption (2) Link encryption is transparent (invisible) to users, their applications, and their OSs Encryption service provided by physical (or data) layer Can use encryption h/w (link encryption device) Message under link encryption Fig. 7-21, p. 446 (ed.4) See which portions encrypted, which exposed Only part of data link header & trailer created after encryption is exposed Link encryption is useful when transmission line is most vulnerable in a network I.e., when S’s host, intermediate hosts, R’s host are reasonably secure (so msgs/pkts at their Layers 2-7 can be exposed)

2) End-to-end encryption = between 2 user applications
Link vs. end-to-end encryption (3) 2) End-to-end encryption = between 2 user applications Data encrypted as „close” to app as possible At OSI Layer 7 (or, perhaps, Layer 6) Fig. 7-22, p.447 (ed.4) Properties of e2e encryption (compare vs. Fig ed.4) Msgs/pkts protected all the way once they „exit” S’s app & before they enter R’s app Msgs/pkts protected (in ciphertext) inside S’s/R’s host Packets protected in transit between S’s & R’s hosts Including protection inside intermediate hosts I.e., protected at OSI layers 1-3 of interm. hosts Layers 1-3 provide physical connectivity, addressing and routing for packets

E2e encryption is visible either to users or their apps
Link vs. end-to-end encryption (4) E2e encryption is visible either to users or their apps Encryption service provided by app or OS Possibly provided only upon explicit user’s request => visible to user Encryption by s/w Message under e2e encryption Fig. 7-23, p.447 (ed.4) (compare vs. Fig ed.4) See which portions encrypted, which exposed Only user’s msg (user’s data) encrypted All headers & trailers exposed (all created after encryption) E2e encryption is useful when transmission lines and intemediate hosts are insecure

Comparison of link vs. e2e encryption
Link vs. end-to-end encryption (5) Comparison of link vs. e2e encryption Encryption of msgs/packets (whether link or e2e encryption) is no silver bullet No guarantees of msg/packet security 1) Link encryption — encrypts all traffic over physical link Typically host H has one link into network => link encryption encrypts all H’s traffic Every H —incl. intermediate hosts— receiving traffic via link encryption must have decryption capabilities Either (pairs of) hosts share symmetric key OR Hosts use asymmetric keys All hosts along a path from S to R must provide link encryption to prevent („partial”) packet exposure => usu. link encryption provided on all network links

Interm. hosts don’t need to decrypt-encrypt pkts
Link vs. end-to-end encryption (6) 2) End-to-end (e2e) encryption — encrypts traffic only between 2 apps („virtual crypto channel between 2 apps”) Interm. hosts don’t need to decrypt-encrypt pkts => interm. hosts don’t need encryption facilities All interm. hosts save time/processing Can encrypts only some msgs between 2 apps If no need to encrypt all msgs => even S’s and R’s hosts save time/processing If needed, can encrypt all msgs Using asymmetric keys requires fewer keys than using symmetric keys (n key pairs vs. n*(n-1)/2 keys)

Comparison conclusions Link encryption: Faster Easier to use
Link vs. end-to-end encryption (7) Comparison conclusions Link encryption: Faster Easier to use Uses fewer keys (1 K pair per host pair vs. 1 K pair per app pair) End-to-end (e2e) encryption: More flexible More selective (can select only some msgs for encryption) User-level, can be integrated with app Optimize whether link or e2e encryption better for you If needed for higher security, use link and e2e encryption together E.g., user not trusting network link encryption can use app with e2e encryption

Physically Protected Network Perimeter
(ii) Virtual private network (VPN) (1) Virtual private network (VPN) = connection over public network giving its user impression of being on private network It could be viewed as „logical link” encryption Could be viewed as e2e encr. between client & server Protecting remote user’s connection with her network Greatest risk for remote connection via public network: Between user’s workstation (client) and perimeter of „home” network (with server) Firewall protects network against external traffic (more later) Physically Protected Network Perimeter Firewall Internal Server User’s Workstation (Client)

Physically Protected Network Perimeter
Virtual private network (VPN) (2) Example VPN connection scenario VPN restricts / filters access to „home” server/network Only „private” accesses allowed => public network access feels like private network Physically Protected Network Perimeter Firewall 1 2 3 Internal Server User’s Workstation (Client) 1 – C authenticates to firewall (firewall passes user’s authentic. data to authentic. server [not shown], which decides whether authentication is OK) 2 – Firewall replies with encryption key (after negotiating with C a session encryption key) 3 – C and S communicate via encrypted tunnel

(iii) PKI and certificates (1)
Public key infrastructure (PKI) = enables use of public key cryptography (asymmetric cryptography) Usually in large & distributed environment Elements of PKI: 1) Policies (higher level than procedures) Define rules of operation E.g., how to handle keys and sensitive info E.g., how to match control level to risk level 2) Procedures (lower level than policies) Dictate how keys should be generated, managed, used 3) Products Implement policies and procedures Generate, store, manage keys

PKI certificate services: 1) PKI creates certificates
PKI and certificates (2) PKI certificate services: 1) PKI creates certificates Certificate binds entity’s identity to entity’s public key Entity = user or system or application or ... 2) PKI gives out certificates from its database 3) PKI signs certificates Adding its credibility to certificate’s authenticity 4) PKI confirms/denies validity of a certificate When queried about it 5) PKI invalidates/revokes certificates For entities that are no longer certified by PKI OR For entities whose private key has been exposed

1) Certificate authorities (CAs) 2) Registration authority
PKI and certificates (3) PKI sets up: 1) Certificate authorities (CAs) 2) Registration authority 1) Certificate authority (CA) CA can be in-house or external E.g., commercial TTP as the trusted third party CA is trusted Entities delegate to CA creation, issuance, acceptance, and revocation of their certificates CA actions: Managing public key certificates (whole life cycle) Issuing certificates by binding entity’s identity to its public key Binding is done via CA’s digital signature Determining expiration dates for certificates Revoking certificates when necessary By publishing revocation lists

Certificate analog: credit card (binds identity to account)
PKI and certificates (4) Example of CA analog: credit card company (CCC) Certificate analog: credit card (binds identity to account) Revocation list analog: lists of invalid credit cards CCC is trusted Customers delegate to CCC creation, issuance, acceptance, and revocation of their credit cards CCC actions (cf. “CA actions” on the previous slide): Managing credit cards (whole life cycle) Issuing credit cards by binding customer’s identity to customer’s account Binding is done via CCC’s protected databases Determining expiration dates for credit cards Revoking credit cards when necessary By checking list of invalid credit cards (before computer-verification transaction era, CCC published booklets of invalidated credit cards)

2) Registration authority (RA) = interface between user and CA Duties:
PKI and certificates (5) 2) Registration authority (RA) = interface between user and CA Duties: Capture and authenticate user’s identity Submit certificate requests to appropriate CA Analog: U.S. Citizen applying for passport and U.S. Postal Service (USPS) Passport (official U.S. authentication) <-> certificate USPS authenticates citizen By verifies citizen’s driver license + other proofs of identity USPS submits passport request forms to appropriate passport office of the U.S. Gov’t Passport office <-> CA USPS brings passport to customer’s home Note: Trustworthiness of USPS authentication determines level of trust that can be placed in passports

PKI efforts stateside and overseas Building PKI for various purposes
PKI and certificates (6) PKI efforts stateside and overseas Building PKI for various purposes E.g., Federal PKI Initiative – to provide secure communication to U.S. gov’t agencies It also specifies how commercial s/w using PKI should operate (so gov’t can use off-the-shelf products) Major PKI product vendors in the U.S.: Baltimore Technologies Northern Telecom/Entrust Identrus Certificates can bind: Identity to public key – classic, most common Other bindings under research E.g., binding financial status to key (credit card companies) Draft standards: ANSI X9.45, Simple Public Key Infrastructure (SPKI)

Still, many points are clear:
PKI and certificates (7) PKI is not yet mature Many outstanding issues Cf. Table 7-6, p. 453 (ed.4) Still, many points are clear: CA should be approved/verfied by independent body CA’s private keys must be stored in tamper-resistant security module (maybe with h/w support) Access to CAs and RAs should be tightly controlled Using strong authentication (e.g., 2FA or 3FA with smart cards)

(iv) SSH protocol (SSH = Secure SHell)
SSH protocol (newer: v.2) – provides authenticated and encrypted communication with shell/OS command interpreter Originally defined for Unix Replaced insecure utilities for remote access Such as Telnet / rlogin / rsh Protects against spoofing attacks (falsifying one end of communication, incl. masquerading, sesssion hijacking, MITM) & message modification / falsification Involves negotiation between local and remote sites Negotiate which encryption algorithm to use E.g., DES? IDEA? AES? Negotiate which authentication technique to use E.g., public key? Kerberos?

(v) SSL protocol (a.k.a. TLS prot.) (1)
SSL protocol (v3) = (approx.) TLS protocol - interfaces betwen app on client C and TCP/IP protocols Provides: (1) server S authentication, (2) optional C authentication, and (3) encrypted communication channel between C and S for session between C and S SSL = Secure Sockets Layer / TLS = Transport Layer Security Originally defined by Netscape to protect browser-to-server communication Simple but effective – most widely used secure commu-nication protocol on Internet (incl. WWW browsers/servers) Involves negotiation between C and S Negotiate which encryption suite to use for session E.g., DES? RC4 w/ 128-bit/40-bit key? RC2? Fortezza? [Bishop] Negotiate which hashing technique to use for session E.g., SHA1 or MD5?

SSL use scenario (handshake protocol)
SSL protocol (a.k.a. TLS protocol) (2) SSL use scenario (handshake protocol) C requests an SSL session by sending: Hello-C, Rand-C (random nr), list of cipher (encryption) algorithms & hash algorithms known to C Hash used to checksum messages S responds with msgs including: Hello-S, Rand-S, cipher & hash algorithm selected by S (from C’s list) , S’s certificate, KPUB-S, [SKIP: request for cert. fr. C] C can use S’s certificate (X.509v3 cert.) to verify S’s authenticity [SKIP: C replies with: C’s certificate] C returns „pre-master secret” encrypted under KPUB-S Pre-master secret - e.g., 48 random bytes if selected cipher is RSA [Bishop-CompSec-A&S, p.296] ...continued...

C and S calculate „master secret” using:
SSL protocol (a.k.a. TLS protocol) (3) ...continued... C and S calculate „master secret” using: „Pre-master secret” Constant strings ‘A’, ‘BB’ and ‘CCC’ Rand-C and Rand-S, SHA hashing algorithm [ibid, p. 294] C and S switch to encrypted communication using „master secret” as session key C and S exchange application data for the duration of the session (i.e., for as long as they stay connected) TLS is potentially vulnerable to MITM attacks [Conklin et al., p.163]

(vi) IPsec protocol suite (1)
IPsec (IP Security Protocol Suite) = standard for securing IP communications by encrypting and/or authenticating all IP packets IPsec is public (published / scrutinized) By design, protects against threats including: spoofing (incl. session hijacking) / eavesdropping Choice of ciphers/hash protocols Communicating parties negotiate which ones to use IPsec defines some ciphers/hash as required in every IPsec implementaion

IPsec provides security at IP layer in IPv6 or IPv4 [Stall.p.499]
IPsec protocol suite (2) IPsec provides security at IP layer in IPv6 or IPv4 [Stall.p.499] IP versions: IPv4 = v.4 — older IP protocol version (still in use) IPv6 = v.6 — newer IP protocol version IPv6 — larger address space Developed ( ) since IPv4 runs out of address space — IPv4 has 32-bit source/destination addresses — IPv6 has 128-bit source/destination addresses IPv6 — also other functional enhancements To accommodate faster networks To accommodate mix of multimedia data streams IPsec protects all layers above IP layer (where it „resides”) In particular, protects TCP or UDP protocols Protects „automatically” Protects transparently (no modifications to TCP, UDP needed)

Encryption algorithm and „mode” Encryption key
IPsec protocol suite (3) Basis of IPsec: security association (SA) = set of security parameters for a secured 1-way communication channel 2 SAs needed for 2-way communication [St,487] Components of SA: Encryption algorithm and „mode” E.g., for DES, mode = CDC – cipher bloc chaining Encryption key Encryption params (e.g., initialization vector for encryption) Authentication protocol and key SA lifespan Allows long-running sessions to select new crypto key Address of opposite end of SA (source <---> destination) Sensitivity level of protected data (e.g., unclassified / restricted / confidential / secret / top secret)

Security parameter index (SPI) - data structure
IPsec protocol suite (4) Security parameter index (SPI) - data structure Resides on each host H running IPsec Used to select 1 of nH SAs that exist on H Different SAs for concurrent communications with different remote Hs Fundamental IPsec data structures / protocols 1) AH = authentication header / AH protocol For authentication-only IPsec service: Authenticates S (sender ) ESP = encapsulated security payload / ESP protocol For encryption-only IPsec service OR For combined encryption/authentication IPsec service

IPsec can be used for various crypto sessions: VPN
IPsec protocol suite (5) IPsec can be used for various crypto sessions: VPN e2e (incl. app 2 app) For network mgmt (e.g., for routing) IPsec scenario TCP layer passes conventional TCP Header & Data down to IP layer IP layer calls upon IPsec to encapsulate conventional TCP Header & Data into ESP (encapsulated security payload) Fig. 7-27, p. 455 (ed.4) IP layer: - adds IP Header - passes packet down to physical layer Physical layer adds Physical Header & Physical Trailer

[---SKIP---] IPsec protocol suite (6)
ESP structure – includes more than just TCP Header & Data Fig. 7-28, p. 456 (ed.4) „authenticated” — portion authenticated by IPsec with Authentication Data (Auth. Data = kind of signature) „encrypted” — portion encrypted by IPsec Fields within ESP: SPI (unencrypted) – index for security association to be used for this pkt Sequence Number (unencrypted) – sequential number, prevents pkt replay attacks Payload Data – (part of) data passed fr. TCP Layer Padding / Padding Length – padding necessary to inflate pkt to required fixed size (e.g., size required by block cipher used for encryption) Next Header – type of data in Payload Data field SKIP: Authentication Data (unencrypted) – authenticates all above fields

ISAKMP is simple, flexible, scalable
IPsec protocol suite (7) ISAKMP (Internet Security Association Key Management Protocol) = key mgmt protocol for IPsec Key mgmt is always a critical element in crypto apps ISAKMP is simple, flexible, scalable Distinct key for each IPsec’s SA (security association) In IPsec, ISAKMP implemented via IKE (ISAKMP Key Exchange) IKE properties Provides ways to agree on protocols, cipher and authentication algorithms, and keys E.g., agree as follows: protocol = EPS, cipher = triple DES; authentication alg. = SHA-1; key used for session Provides ways to manage protocols, cipher and authentication algorithms, keys Uses key exchange protocol (based on Diffie-Hellman scheme)

[---SKIP---] IPsec protocol suite (8)
IKE key exchange scenario (using Diffie-Hellman scheme) X and Y know number g X and Y choose large primes x and y, respectively X send gx to Y / Y send gy to X X calculates (gy)x / Y calculates (gx)y Since (gy)x = (gx)y => X and Y share secret Actual calculation is mod(n) — a bit more complex X and Y exchange identities & certificates Certificates authenticate identities X and Y derive shared crypto key X and Y enter security association => they establish session IKE key exchange is very efficient Can be done in 2 messages if w/o authentication Can be done in 4 messages if w/ authentication

vii. Signed code Problem: malicious active code
E.g., malicious code on a web site for downloads Partial solution: code signed by TTP (trusted third party) TTP appends digital signature to piece of code PKI can be used by prospective code users to validate signature Still code security not guaranteed E.g., March 2001 mistake of Verisign (CA) Erronously issued two code-signing certificates to impostors masquerading as Microsoft employees Verisign detected mistake after almost 2 months Customers who didn’t validate certificate (by checking Verisign’s certificate revocation list) could still trust bad certificates

[---SKIP---] viii. Encrypted e-mail
msgs – like a postard that everybody who handles it between S and R can read People use envelopes for confidentiality (C in C-I-A) We can „envelope” msgs by encrypting them Encryption protects C and can protect I Encryption is easy, establishing good key mgmt is difficult 2 basic key mgmt approaches: 1) Hierarchical certificate-based PKI solution E.g., S/MIME 2) Use of flat, individual-to-individual key exchange E.g., PGP security (incl. PGP and S/MIME) will be discussed soon

[---SKIP---] e) Msg content integrity controls (1)
Content integrity verification provided „for free” with encryption Since can’t perform meaningful data modification w/o decrypting it But attacker can modify encrypted data to make it useless E.g., changing a bit of data in packet Threats to msg content integrity Malicious modification that changes content in a meaningful way 2) Nonmalicious modification that changes content in a way that is not necessarily meaningful 3) Malicious modification that changes content in a way that is not meaningful NOTE: Different cases than in text! Encryption can solve the toughest case: Case (1) above EASIER TO PREVENT OR DETECT

[---SKIP---] Msg content integrity controls (2)
Msg content integrity controls #1: DETECT Error detection and error correction codes As discussed before („Intro to Cryptology” section) Simplest error detection code: even/odd parity check More complex error detection codes: Hash codes or Huffman codes Some detect multiple bit errors Some also pinpoint which bits are incorrect For detecting/correcting nonmalicious changes Does not matter whether accidentally „meaningful” or not Can also stop some malicious attacks Again does not matter whether msg changes „meaningful” or not E.g., attackers not knowing which code used => can’t fool it E.g., attacker manipulating 2 bits & thinking that we have 2–bit correction code fails if we have 3-bit correction code More complex error detection codes can stop tougher attackers

[---SKIP---] Msg content integrity controls (3)
[SKIP] Msg content integrity controls #2: PREVENT or DETECT Cryptographic checksum / msg digest / hash To prevent or malicious attacks on data — As discussed before („Intro to Cryptology” section) Can also detect malicious errors Can also detect nonmalicious errors Note that attacker can modify: Modify only data bits Modify error prevention/detection mechanism

f) Strong authentication
Networked environments as well as both ends of communication need authentication Strong authentication controls include: One-time passwords Challenge-response systems Digital distributed authentication Kerberos authentication system

[---SKIP---] i. One-time passwords (1)
One-time password (OTP) method Controls wiretapping and spoofing Prevents reuse of pwd captured by wiretapper Strong authentication prevents spoofing (incl. masquerading, session hijacking, MITM) Each password used only once „Thrown away” after use OTP uses secretly maintained pwd list OR User has password token (PT) = device to randomly generate new pwd (e.g.) every minute E.g., SecureID device from Security Dynamics – random # every minute

[---SKIP---] One-time passwords (2)
Pwd generation for user U is based on time synchronization with PT of this particular user U Pwd changed in sync by both U’s PT and system authenticating U PTs of different users produce different pwd sequences Problems with OTP 1) Attacker can steal/find PT (device) 2) One-minute window of vulnerability Wiretapper can reuse password for one minute If PT generates new pwd once per minute

[---SKIP---] ii. Challenge-response systems
Challenge-response systems solve OTP problems 1) Solve problem of stolen/lost password tokens (PTs) Solution: PT requires a PIN Even if stolen by attacker, PIN protects „responses” 2) Solve problem of window of vulnerability Solution: New challenge for each use Does not matter whether new use within one minute window or not Challenge-response systems can be viewed as more sophisticated version of OTP Problem: Neither OTP or C-RS protect user from rogue system

[---SKIP---] iii. Digital distributed authentication (1)
„Digital” from „Digital Equipment Corporation” (DEC) Once a powerful computer company Acquired (1998) by Compaq, which subsequently merged with Hewlett-Packard (2002) As of 2004 their product lines were still produced under HP name In 1980s DEC recognized need for authentication of nonhuman entities in distributed environment E.g., mutual authentication of user query processor and DB manager (process) Goals of Digital distributed authentication: Prevent server impersonation (e.g., DB mgr impersonation) Prevent interception/modific. of msgs between entities Prevent replay of previous authentication [Wikipedia]

[---SKIP---] Digital distributed authentication (2)
Architecture of Digital distributed authentication (1989): Each server S has KPRIV-S Corresponding KPUB-S available to any process that might need authenticated channel to S A to B communication uses KPUB-B B to A communication uses KPUB-A A and B use sequence #s to msg replay B can establish secret (symmetric) session key K for communication with A by: Selecting key K Sending E(K, KPUB-A) to A From now on, A and B can exchange data encrypted with K

iv. Kerberos authentication system (1)
Kerberos – system for authentication between intelligent processes in distributed systems (incl. C-S authentication) [K. = guard dog in Hades (Greek mythology)] Developed at MIT (1988+) Design goal: Enable systems to withstand attacks in distributed systems Basic idea of Kerberos: Central server provides tickets to requesting app Ticket is authenticated, non-forgeable, non-replayable token => Implemented as encrypted data structure naming user U and service for which U has access permission (also contains time value, control info)

Server S (with File F that U wants to read)
Kerberos authentication system (2a) Nodes involved: User U (client) Server S (with File F that U wants to read) Kerberos Server KS – distributes “pre-tickets” Having pre-ticket is needed before asking for ticket Ticket-Granting Server TGS – distributes tickets Real-life analogy: U = a fan that wants to watch World Championship Many more Us than available tickets S = game that U wants to watch Pre-ticket = right to buy a ticket KS = lottery that gives a “pre-ticket” to each winner TGS = box-office that sells tickets only to Us that have pre-tickets

U’s workstat’n sends U’s identity to KS
Kerberos authentication system (2b) User’s Step 1: Establishing session w/ Kerberos — Fig. 7-29, p. 462 (ed.4) - TGS = Ticket-Granting Server / KS = Kerberos server U’s workstat’n sends U’s identity to KS KS verifies that U is authorized KS sends 2 msgs: 1) Msg to U, which contains: E(STGS + TTGS , pwd) STGS — U’s session key for session with TGS TTGS — U’s ticket for TGS Enables U to request service from TGS where pwd = user’s pwd (Note: used as encr. key by KS) 2) Msg to TGS, which contains: E(STGS + U’s identity, shared KS/TGS key) (i.e., encrypted under key shared by KS & TGS) STGS — TGS’s session key for session with U (same as U’s session key for session with TGS) If U’s workstation can decrypt E(STGS + TTGS , pwd) using its pwd, then U’s authentication succeeds U U U U U U U U

Using U’s ticket for TGS (received fr. KS in Step 1), U sends to TGS:
Kerberos authentication system (3) Note: KS stores users’ pwds => no need to pass pwds over network between U’s workstation and KS Security advantage! User’s Step 2: U accesses services of distributed system — e.g., access file F — Fig. 7-30, p. 463 (ed.4) Using U’s ticket for TGS (received fr. KS in Step 1), U sends to TGS: E(„request R for U’s ticket for accessing F” , STGS) Note: STGS (U’s session key for session with TGS obtained fr. KS in Step 1) is used to encrypt R TGS verifies U’s access permission ...continued... U U

U’s session key for session with F (really: with F’s file server) = SF
Kerberos authentication system (4) ...continued... If verified OK, TGS returns to U message encrypted under STGS, which contains: U’s session key for session with F (really: with F’s file server) = SF U’s ticket for accessing F Ticket is encrypted under secret key („TGS/F key”) known only to TGS and F’s file server Ticket contains: U’s authenticated identity (TGS knows that STGS is U’s key —and nobody else’s— because KS sent to TGS STGS and U’s identity tied together in Step 1) Identification of file F Access rights (e.g., read permission) F’s session key for F’s session with U = SF (same as U’s session key for session with F) Ticket expiration date U U U U U

Strengths of Kerberos: No pwds communicated over network
Kerberos authentication system (5) Strengths of Kerberos: No pwds communicated over network Pwd sent by user to Kerberos server only once & sent outside the network (e.g., in a letter) User’s pwd is not sent from user’s workstation when it initiates a session User’s pwd stored only at Kerberos server Provides crypto protection against spoofing (e.g., masquearding, session hijacking, MITM) Each access request mediated by a ticket-granting service (TGS) TGS knows user’s identity based on authentication performed initially by the server

4) Prevents replay attacks
Kerberos authentication system (6) Strengths of Kerberos – cont.1 3) Limits period of ticket validity (this disables some long-term attacks—e.g., brute force cryptanalysis) Tickets contain timestamps used by servers to determine ticket’s validity Ticket validity period limits duration of „window of opportunity” for attacker 4) Prevents replay attacks Each user’s request stamped with time of request Servers compare timestamps of requests w/ current time, accept requests only if they are close enough to current time Time-checking prevents most replay attacks Since presentation of tickets by attackers will be delayed more than presentation of tickets by legitimate users

5) Provides mutual authentication
Kerberos authentication system (7) Strengths of Kerberos – cont.2 5) Provides mutual authentication Service user can be assured of any server’s authenti-city by requesting an authenticating response from S 6) Uses public key technology for key exchange

Weaknesses of Kerberos system
Kerberos authentication system (8) Weaknesses of Kerberos system 1) Requires continuous availability of trusted ticket-granting server (TGS) 2) Server S’ authenticity requires trust between TGS & S 3) Requires timely transactions (too quick ticket expiration will result in rejecting legitimate requests) 4) Subverted workstation can replay user pwds 5) Pwd guessing works (attacker can send initial —Step 1— authentication request to Kerberos server, receive response, try to decrypt response by guessing at pwd) 6) Kerberos does not scale well (due to system size might need > 1 KS and/or TGS server; coordination and security problems if more than one KS and/or more than one TGS is needed; cf. Fig. 7-31, p.450-ed.4) 7) Use of Kerberos requires compatibility of all apps in a given computing environment (to date few apps are compatible with Kerberos; modifying apps to make them compatible is not feasible)

g) Access controls (1) Before user is allowed access to network resources, must know: Who needs access => authentication What and how will be accessed => access controls Access controls include: ACLs (Access Control Lists) on router Firewalls

1) ACLs on routers (ACL = Access Control List) Router directs traffic:
Access controls (2) 1) ACLs on routers (ACL = Access Control List) Router directs traffic: To subnetworks it controls OR To other routers (for delivery to other subnetworks) Routers convert external (network-wide)IP address to internal (subnetwork-wide) MAC address Recall that MAC address is unique physical address of device’s NIC—network interface card Can put ACL on a router to deny access to particular host D from particular host S E.g., to prevent spam (flooding) of D with packets from S, router can delete all packets from S to D It’s OK if router uses ACLs in a limiteded way Use sparingly: only for specific & known threats BUT...

... Problems with putting too many ACLs on routers:
Access controls (3) ... Problems with putting too many ACLs on routers: (i) Packet-checking overhead for router Router must check each packet against each ACL – a lot of work => degraded performance More ACLs on router => more work Routers are already busy just routing all packets ingoing/outgoing to/from their subnets (ii) Logging overhead for router To be able to detect spam, router must log source addresses of packets Then can analyze to see which source addresses produce floods Routers are designed to do only essential work — anything else is inefficient => logging on router is inefficient => adds workload

(iii) Inability of router to detect all spams
Access controls (4) ... Problems with putting too many ACLs on routers-CONT. (iii) Inability of router to detect all spams Because source addresses in datagrams (UDP packets) can be easily forged (by attacker using UDP protocol) If attacker sends many datagrams with the same (repeated) forged address, router with ACL can detect & block them Otherwise (i.e., if attacker sends datagrams with few repeated forged addresses), router with ACL will not even detect being flooded => can not block flooding datagrams

Designed to do screening that routers can’t do efficiently
Access controls (5) 2) Firewalls Designed to do screening that routers can’t do efficiently Because routers designed for routing (of course!) Firewalls designed for access filtering AND auditing AND examining whole packets (not only source/destination IP/ MAC addresses—which is what routers do) Firewalls will be discussed in detail later (but very soon)

h) Intrusion Detection Systems: Alarms & Alerts
Example of 2-layer network protection Provided by router (Layer 1) AND firewall (Layer 2) Fig. 7-32, p. 469 (ed.4) We can add one more layer of protection: intrusion detection systems (IDS) = device placed within protected network for monitoring for illegitimate actions in order to detect attacks in progress (beginning, advanced) or after they have occurred E.g.: Can detect reconaissance & alert sysadmin or secadmin, raise alarm, thus preventing „real” attack OR Can detect that attack has already occurred & raise alarm, starting system recovery actions IDS is a.k.a. IPS = intrusion protection system A marketing gimmick? IDS can be Layer 3 of layered network protection To be discussed in detail soon

i) Honeypots Honeypot – system built as a bait attracting attackers
Once attackers take the bait: They are observed to learn how they behave/operate New attacks / Prefered targets / ... They are traced to catch them or scare them off Or at least trace enough to be able to threaten them with identifying them if they don’t stop They are diverted from really valuable attack targets E.g., diverted to phony credit card database while real credit card database remains obscure to them User lessons learned (thanks to honeypots) to build better countermeasures

j) Traffic flow security (1)
Threat: attacker infering occurrence/location of some event / structure from intensity of encrypted network traffic (If not encrypted, no need to infer – attacker can simply read all) Example 1: Inference that traffic between Thinges Corp. and bankruptcy lawyer precedes declaration of bankruptcy by Thinges Example 2 (old): Battlefield network: Busiest network node is at enemy’s HQs Solution 1: Masking by steady traffic volume X and Y always send the same volume of encrypted traffic between them If X has nothing to communicate to Y, X sends meaningless padding packets to Y (Y behaves analogously)

Onion-like packet sent by W to X
Traffic flow security (2) Solution 2: Masking by onion routing Example: W wants to send packet to Z in a hidden way W wraps „real” packet to Z into packet addressed to Y, which asks Y to send it toZ W wraps packet to Y into packet addressed to X, which asks X to send it to Y Full route: W  X  Y  Z W sends green packet to X X unwraps it, gets red packet X sends red packet to Y Y unwraps it, gets blue packet Y sends blue packet to z Z unwraps it, gets blue packet Send packet to Z Send packet to Y Onion-like packet sent by W to X

Note: (Recall the full route: W  X  Y  Z )
Traffic flow security (3) Why „onion” routing? Layers of wraps around „real” packet to Y– like layers of an onion Note: (Recall the full route: W  X  Y  Z ) X knows that packet came from W & should be forwarded to Y But X does not know if W is source or intermediate host, does not know if Y is destination or intermediate host Y knows that packet came from X & should be forwarded to Z But Y does not know if X is source or intermediate host, does not know if Z is destination or intermediate host Z knows that packet came directly from Y & knows that W is its source Z knows that Y is just an intermediate host => Intermediate nodes do not know source/destination They only know host that forwarded packet to them & know host to which they should forward packet

k) Review of network security controls
Table 7-4, p. 439 (ed.4) provided classification of network vulnerabilities (during our earlier discussion of threats) Table 7-7, p. 471 (ed.4) lists controls for each of these classes of network vulnerabilities — it shows that: There are many great network security controls Most are used also in environments other than networks (including applications and OSs) Three of these controls are specific to networks: Firewalls / IDSs / encrypted We shall discuss them in some detail next Table 7-7 is a great reference for network security controls! Use it often If you can get copyright permission from publisher, I’d advise you to copy it and post above your desk Otherwise, make your own notes based on it

5.2. Network Security Tools
Firewalls Intrusion Detection Systems Secure [Fig: B. Endicott-Popovsky]

5.2.1. Firewalls Outline Introduction What is a firewall
Firewall design Types of firewalls Packet filters (i-1) Simple packet filters (i-2) Stateful packet filters Application proxies (ii-1) Guards (“top model” subcategory) iii. Personal firewalls Comparison of firewall types Example firewall configurations What firewalls can—and can’t—block

a. Introduction Firewalls Invented in the early 1990’s
But idea related to reference monitors from 1970’s What is reference monitor? OS includes kernel / core / nucleous – responsible for lowest-level functions Incl. synchronization, inter-process communication, msg passing, interrupt handling Security kernel – provides security mechanisms for entire OS Incl. security interfaces among h/w, OS, other parts of computing system Typically, security kernel is a part of OS kernel Reference monitor is portion of security kernel that controls access to objects (controls „references” to objects)

b. What is a firewall (1a) Firewall = device (h/w), or software, or combination of both designed: 1) to prevent unauthorized outside users from accessing network and/or single workstation 2) to prevent inside users from xmitting sensitive information or accessing insecure resources It is a wall between protected local (sub)net & outside global net Inspect each individual inbound or outbound packet of data sent to / from protected system Check if it should be blocked or allowed to enter

Firewalls keep „bad things” out, keep sensitive info in
b. What is a firewall (1b) Firewalls keep „bad things” out, keep sensitive info in Security policy specifies what are „bad things” E.g., requires that traceroute & ping -o can't see internal hosts F. protect against security threats from external network F. are effective in protecting local subnet incl. its sensitive info

Examples of security policy requirements w.r.t. firewalls:
What is a firewall (2) Examples of security policy requirements w.r.t. firewalls: Block any access from the outside, allow all accesses to the outside Allow”from” accesses only for certain activities OR only to/from certain subnets/hosts/apps/users E.g., prevent outside access to subnet hosts except for mail server accesses Choice of default firewall behavior 1) Default permit „That which is not expressly forbidden is allowed” 2) Default deny „That which is not expressly allowed is forbidden” Users prefer default permit, security experts prefer default deny Sysadmin must make the choice

c. Firewall design (1) Firewall design principles:
Small / simple enough for rigorous analysis KISS principle (= „Keep It Simple, Stupid”) Simple firewall functionality Tamperproof Typically well isolated (=> highly immune to modifications) On a separate computer With direct connections only to the outside networks and to the inside network Designed to be always invoked Efficient enough not too be a bottleneck Placed strategically All network accesses that we want to control pass through it

General firewall techniques: 1) Service control
Firewall design (2) General firewall techniques: 1) Service control Type of service: inbound or outbound 2) Traffic filtering — based on IP address & TCP port nr Provide proxy software to receive or interpret service request before passing it on Could also host server software (e.g. Web or mail service) Not recommended Complicates it (more code => more vulnerabilities) 3) User Control Control access to service using ACLs 4) Behavior Control E.g. filter for spam

Basic firewall characteristics
++SKIP++ Firewall design (3) Basic firewall characteristics All traffic (incoming / outgoing) must pass thru firewall Only authorized traffic allowed to pass Firewall itself must be immune to penetration I.e. must use trusted system w/ secure OS (min. size/complexity) Usually implemented on dedicated device Dedicated = only firewall functions performed on this device Firewall code must be very well protected Basic kinds of firewalls: Hardware firewalls More common Implemented on router level More expensive / more difficult to configure Software firewalls Used in single workstations Less expensive / Easier to configure

d. Types of firewalls (1) Types of firewalls
i. Packet filters / packet filtering firewalls Control packets based on packet’s IP adresses and port #s (i-1) Simple packet filters / (simple, stateless) packet filtering gateways / screening routers (stateless = ignore history) (i-2) Stateful packet filters / stateful inspection firewalls ii. Application proxies / proxy firewalls / application-level gateways More sophisticated than packet filters. Control packets based on any data within a packet (not only based on packet’s IP adresses and port # as packet filetres do) => => can detect/derail more sophisticated attacks (e.g., can filter out harmful commands in pkt stream) (ii-1) Guards (a special case of app proxies) iii. Personal firewalls Protect single hosts (not subnetworks R like regular firewalls do) For small business / home office / home

++SKIP++ Firewall properties: Packet filter properties: Transparent
Types of firewalls (2) ++SKIP++ Firewall properties: Packet filter properties: Transparent Does not change traffic, only passes it (or blocks some) Proxy properties: Active Intercepts traffic & acts as an intermediary Different firewall types needed for different needs „Different strokes for different floks” — e.g.: Simple packet filters / screening routers – implement simplistic security policies Simple is best if sufficient to counter exisiting threats well App proxies – much richer capabilities

Firewall is a type of host
Types of firewalls (3) Firewall is a type of host Even some routers are host-based To have better tools available (editors, programming tools) Programmable Minimal functionality Reduces vulnerabilities Small = > less complex => fewer vulnerabilities Reduces motivation for attacks No password files to steal, etc.

(i) Packet filters (1) Packet filters — a.k.a. packet filtering firewalls (i-1) Simple packet filters („memoryless”) (i-2) Stateful packet filters (with „memory”) Basis for packet filtering 1) Packet IP addresses Filtering based on both source/destination addresses 2) Port number determines TCP transport protocol type Recall “portprotocol” mapping in TCP: 21FTP, 23Telnet, 25SMTP, 80HTTP, 110POP, 161 SNMP, etc. Filtering based on port nr Packet filtering firewalls do not „see” other packet fields See only IP address ‘ transport protocol type E.g., can not allow only some Telnet commands OR exclude only some other Telnet commands

Examples of packet filtering – see text
(i) Packet filters (2) Examples of packet filtering – see text 1) Packet filtering based on packet address 1a) Can block and/or allow traffic from/to specific subnet, i.e., all its addresses (cf. Fig. 7-34, p. 476-ed.4) 1b) Can block and/or allow traffic from/to specific IP addresses (cf. Fig. 7-35, p. 476-ed.4) 2) Packet filtering based on transport protocol type – e.g., HTTP or Telnet (cf. Fig. 7-34, p. 476-ed.4) Can block traffic using Telnet protocol (port 23) but allow HTTP traffic (port 80) To avoid overburdening router, firewall can run on device behind router (on subnet side)

(i-1) Simple packet filters (1)
Simple packet filters / (simple) packet filtering gateways / screening routers — simplest firewall type Simple packet filters (PFs) are memoryless => can not perform attack detections that require remembering state (history/context) of ≥ N last pkts ++SKIP++ what follows E.g., can not see that prev. & curr. pkt indicate attack “Attack signature” (i.e., attack pattern) would be clearly visible if both pkts were put together Example: Certain attack script known to use Telnet (port 23) and then SNMP (port 161) The same source address in previous pkt, using port 23, and in current packet, using port 161, constitutes attack signature Why need to remember ≥ N last pkts? TCP pkts arrive in order different than sending order => remembering ≤ N last pkts is not enough

Cheating simple (memoryless) PF:
++SKIP++ (i-1) Simple packet filters (2) Cheating simple (memoryless) PF: Attacker divides pkt (including attack signature) into many v. short pkts Attack signature (pattern) would be visible in original long pkt Even memoryless simple PF would detect it Pattern of attack is completely invisible in any single short pkt => memoryless simple PF is unable to detect attack Additionally, TCP pkts arrive in order different than their sending order => remembering just last packet would not be enough – must remember N last packets

Inside hosts trust more other inside host
++SKIP++ (i-1) Simple packet filters (3) One very important task for simple packet filtering gateways: Validating inside IP addresses Inside hosts trust more other inside host Simple filtering assures that no external source can masquerade as internal source Blocks any packet coming from outside network that claims to be sent by internal host Cf. Fig. 7-36, p. 477 (ed.4)

Simplicity of inspection is both disadvantage & advantage
++SKIP++ (i-1) Simple packet filters (4) Simplicity of inspection is both disadvantage & advantage Disadvantage because of high granularity E.g., can block all Telnet coomands, but can not block only selected telnet commands Advantage beacuse reduces complexity Filtering rules to block, e.g., only selected Telnet traffic would have to be much more detailed => more complex => more vulnerable

(i-2) Stateful packet filters
Stateful packet filters — a.k.a. stateful inspection firewalls Keep state/history/context of  N previously seen pkts => stateful packet filters have memory This allows detection of some attacks that simple PFs can not detect Still limited to detection based on IP address & TCP transport protocol type (port nr)

(ii) Application proxies (1)
Application proxies / proxy firewalls / application-level gateways / application proxy gateways Note: The term bastion host (used in text) should not be used as a synonym. Bastion host is a host that serves as a platform for app proxy or circuit-level proxy [Stallings, Crypto&Net.Sec, p.625]. Application proxies include — as a special case (ii-1) Guards App proxy firewalls fix basic problem with packet filtering firewalls because they: See all pkt data (not just IP adresses and port #s) (In addition, they are stateful => can analyze multiple pkts) => can detect/derail more sophisticated attacks Can filter out harmful commands in pkt stream

For example, app proxies can prevent:
++SKIP++ (ii) Application proxies (2) For example, app proxies can prevent: Use of back door open to pkts inbound to SMTP port (Port 25) Flawed application run by user U (e.g., an agent) has all U’s privileges => can cause damage

Ex. scenario of using app proxy gateway G: [cf. ibid, p.624]
++SKIP++ (ii) Application proxies (3) Act as mediators/censors (!) of app-level traffic – like benevolent „woman-in-the middle”  (not an official term!) They “censor” insecure actions Maybe a rare case of a truly benevolent censor Ex. scenario of using app proxy gateway G: [cf. ibid, p.624] Extern. user U tries to Telnet to host H protected by G G intercepts U’s packets G acts as H would: asks U for id+pwd U replies w/ id+pwd G logs in into H on behalf of U G relays H’s msgs to U Etc., etc.

Examples of app proxy activities
++SKIP++ (ii) Application proxies (4) Examples of app proxy activities Preventing outsiders from modifying company’s online price list More - see bulleted list on p. 478 – 479 (ed.4) App proxy must implement code for given app (e.g., for Telnet) to be able to perform service to this app Netadmin can configure app proxy to support only selected features of an app Unsupported features are considered too risky => not available

App proxies provide higher level of security than packet filters (PFs)
++SKIP++ (ii) Application proxies (5) App proxies provide higher level of security than packet filters (PFs) PFs try to deal with all potentially deployable applications that could use TCP/IP (default permit philosophy) App proxy considers only few allowable apps among ones actually deployed in a given system (default deny philosophy) App proxy can easily log/audit traffic at app level (vs. transport level for PFs) Prime disadvantage of app proxies: Processing overhead for each app-level connection 1 connection split into 2 logical connections With “woman-in-the-middle”  Circuit-level gateways (another proxy subcategory) splits 1 TCP connection into 2 TCP connections

(ii) Application proxies (6)
(ii-1) Guards = most sophisticated category of app proxies (“top model”) Limited only by what is computable (& by human creativity) No sharp boundary between app proxies and guards At some point of upgrading app proxy, it becomes a guard --SKIP > Examples of guard activities Limiting nr of msgs (or nr of msg characters) that a student may per week Easiest if done by gurad monitoring mail transfer protocol More - see bulleted list on p. 480 (ed.4)

(iii) Personal firewalls
Regular firewalls protects subnetworks Personal firewalls protect single hosts For small business / home office / home Can be used to complement conventional firewall Next line of defense Customized to user(s) of particular host Firewall capabilities at a lower price Personal firewall is application program Products include: Norton Personal Firewall (Symantec), McAfee Personal Firewall, Zone Alarm (Zone Labs) Personal firewall also enforces certain security policy E.g., if you’re using default personal firewall’s policy on your computer, see its rules Combine it with antivirus software for more effective protection & with automatic (or very frequent manual) OS and antivirus s/w updates

++SKIP++ e. Comparison of firewall types
See Table 7-8, p. 482 (ed.4) Criteria: Complexity Part of packets visible to firewall Difficulty of auditing Basis for screening Difficulty of configuring

f. Example firewall configurations
Subnet with screening router (simple packet filtering) — Fig. 7-38, p. 482 (ed.4) Subnet with proxy gateway (app proxy) — Fig. 7-40, p. 483 (ed.4) Subnet with screening router (a simple packet filter) & proxy firewall (app proxy) Note: The LAN between the outer firewall (“screening router” in the fig) and the inner firewall (“proxy firewall” in the fig) constitutes DMZ (demilitarized zone)

g. What firewalls can—and can’t—block
Firewalls are not a panacea - only a perimeter protection Points 2 remember about firewalls — see text, p Can protect environment only if control its whole perimeter Do not protect data outside the perimeter Are most visible subnet component – attractive attack targets Must be correctly configured, & config must be periodically updated Firewall platforms should not have any s/w that could help attacker who penetrates firewall in subsequent exploits Firewalls exercise very limited control over content they let in Other means of verifying/enforcing accuracy/correctness must be used inside perimeter

5.2.2. Intrusion Detection Systems
Outline Introduction Types of IDSs Signature-based IDSs Anomaly-based IDSs Other IDSs Goals for IDSs IDS strengths and limitations

a. Introduction (1) It is better to prevent attack than to detect it after it succeeds Unfortunately, not all attacks can be prevented Some attackers become intruders — succeed in breaking defenses Intrusion prevention — first line of defense Intrusion detection — second line of defense Intrusion detection system (IDS) - a device (typically a seprate computer) monitoring system activities to detect malicious / suspicious events IDSs attempt to detect Outsiders breaking into a system OR Insiders (legitimate users) attempting illegitimate actions Accidentally OR deliberately

IDS runs constantly in the background
Introduction (2) IDS runs constantly in the background Alarms when it detects something suspicious We’ll be talking only about real-time IDSs Ignoring off-line IDSs (early IDSs reviewed logs after th fact) IDS should operate in stealth mode — be invisible to outside world IDS interfaces (cf. Fig. 7-42, p.488-ed.4) 1) Alarm interface through which raises alarm On a completely separate control network 2) Monitoring interface to protected (sub)network Never sends msgs on this interface! Preferably, IDS has no published address through the monitored subnetwork => router does not even know IDS platform exists => router can not send anything to IDS directly

Anomaly — abnormal behavior
Introduction (3) IDS terminology Anomaly — abnormal behavior Might either be still legitimate OR illegitimate Misuse — activity that violates the security policy (subset of “anomaly” – anomaly that is illegitimate) Intrusion — misuse by outsiders or insiders Audit — activity of looking at user/system behavior, its effects, or collected data Profiling — looking at users or systems to determine what they usually do

Superset of IDS functions Monitoring users and system activity
---[SKIP]--- Introduction (4) Superset of IDS functions Monitoring users and system activity Auditing system configuration for vulnerabilities & misconfigurations Assessing integrity of critical system and app files Recognizing known attack patterns Identifying abnormal activities through statistical analysis Managing audit trails (logs) and highlighting misuses or anomalies in user activities Correcting system config errors Installing & operating traps (honeypots) to learn about intruders

b. Types of IDSs (1a) IDS types w.r.t. scope: Host-based
Runs on a host Monitors activities on this host only Network-based Stand–alone device Monitors entire (sub)network

IDS types w.r.t. their operation
b. Types of IDSs (1b) IDS types w.r.t. their operation i. Signature-based IDSs (optimistic: “block only ‘blacklisted’ behavior”) Models & looks for unacceptable system activities (= an attack) Each known attack characterized by its „signature” (pattern) To detect attack, matches current activities to known attack signatures Problem: Unable to detect new attacks (unknown signatures!) ii. Anomaly-based (heuristic) IDSs (pessimistic: “allow only permitted behavior”) Solves the above problem (but might generate more false alarms) Uses model of acceptable user activities Not models (signatures) of unacceptable system activities Raises alarm upon detection of deviation from model behavior Alarm even if this deviation is legitimate iii. Other IDS types E.g., hybrid IDSs (combining signature- and anomaly-based IDSs), immune-system-based IDSs

Looks for unacceptable system activities (considered an attack)
Types of IDSs (2) i. Signature-based IDS Looks for unacceptable system activities (considered an attack) Each known attack characterized by its „signature” (pattern) To detect attack, matches current activities to known attack signatures Even attack with signature „similar” to signature of a known attack can be detected Example signature for port scan attack Syn to port 80 Syn to port 25 Syn to port 161 ...

Problems with signature-based IDSs
Types of IDSs (3) Problems with signature-based IDSs (1) Can not detect attack with unknown signature Must know (exact or very similar) attack signature to detect attack (2) Attacker can „pad” attack with packets that break known signature into unknown patterns (3) False alarms occur When legitimate data accidentally match attack signature Solution to (1) and (2): anomaly-based IDSs

ii. Anomaly-based IDS (heuristic IDS)
Types of IDSs (4) ii. Anomaly-based IDS (heuristic IDS) Uses model of acceptable user behavior Recall that signature-based IDS uses models (signatures) of unacceptable system activities (system behavior) Watches user activities Watches for deviations from model behavior Raises alarm upon detection of suspicious behavior If false alarm, admin can „teach” IDS to accept such behavior in the future

Example model of acceptable user’s behavior
Types of IDSs (5) Example model of acceptable user’s behavior Logs in Reads Use word processor ... Note: Use of sensitive system management utilities by a user would be a deviation from the above model of acceptable user’s behavior, causing an alarm - Even if such use is legitimate

Uses model of unacceptable user behavior
Types of IDSs (6) (ii-1) Misuse IDS — can be considered a subset of anomaly detection IDS (since misuse is a subset of anomaly – see above) Uses model of unacceptable user behavior Recall that „regular” anomaly-based IDS uses models of acceptable user behavior Recall that signature-based IDS uses models (signatures) of unacceptable system activities (system behavior) Could misuse IDS be considered a subset of signature-based IDS? Only if consider user behavior a subset of system behavior Example model of unacceptable user’s behavior Attempt to copy pwd file Attempt to open pwd file ...

Tripwire program — cf. p. 488 (ed.4)
Types of IDSs (7) Other IDSs E.g., hybrid IDSs (combining signature- and anomaly-based IDSs), immune-system-based IDSs ++SKIP++ Under a more liberal definiton of IDS, the following are called IDSs: Tripwire program — cf. p. 488 (ed.4) Produces hash value for each file when installed, values saved in secure place Later can be called to verify that file was not compromised Vulnerability scanners Honeypots

c. Goals for IDSs (1) IDS goals 1) Detect all attacks correctly
Avoid false positives (false alarms) False alarms annoy sysadmins, users, ... Avoid false negatives (not recognizing attacks) 2) Little overhead / performance impacts ++SKIP++ Range of IDS alarms Write a record to audit log ... Page security administrator Urgency

Categories of IDS responses
++SKIP++ Goals for IDSs (2) Categories of IDS responses 1) Monitor, collect data, collect more data For attacks with modest impact Invisible to attacker Can be used to study attacker’s actions (like honeypot) 2) Protect, reduce exposure E.g., making certain files unavailable to any access May be very visible to attacker 3) Call a human For attacks with highest impact potential Calling a human usually preceded by some automatic actions Since humans take very long time (compared to computer time) to respond Any combination of (1), (2) , (3) into single response is possible

--SKIP-- d. IDS strengths and limitations
Becoming more effective (e.g., DBs of attack signatures grow) Becoming cheaper Becoming easier to administer IDS limitations Attackers use avoidance strategies to avoid detection by IDS IDS sensitivity is difficult to measure and adjust Must strike a balance false alarms and missing attacks Only as good as the process/people using it

5.2.3. Secure E-Mail Outline Introduction Security for e-mail
Design of PEM (Privacy-enhanced Electronic Mail) Example secure systems i. PGP ii. S/MIME

Introduction E-mail is the most heavily used network-based application
Yet, ordinary is very public, exposed It has no C / I (confid./integ) Unencrypted message contents can be peeked at either in transit or by privileged users at destination host

b. Security for e-mail (1)
--SKIP-- Threats to Controls Msg interception (confidentiality) encryption Msg interception (blocked delivery) no S/R controls Msg interception+replay encryption (+seq.nr) Msg content modification encryption Msg origin modification encryption Msg content forgery by outsider encryption Msg origin forgery by outsider encryption Msg content forgery by recipient (R) publ.key encr. Msg origin forgery by recipient publ.key encr. Repudiation (denial) by sender (S) publ.key encr.

Secure e-mail requirements:
Security for (2) Secure requirements: Msg confidentiality (protection from disclosure) Msg integrity (protection from modification) Sender authentication Non-repudiation (preventing denial by sender) Not every msg requires all 4 capabilities but all 4 needed to cover requirements of all kinds of msgs

c. Design of PEM (1) Standard for encrypted e-mail:
Privacy-enhanced Electronic Mail (PEM) [Bishop-CS-A&S, p.286] Developed by Internet Society ( ) A design goal: Allowing PEM msgs travel as ordinary mail msgs For legacy reasons Protection occurs within msg body We’ll look in turn at: 1) Confidentiality in PEM 2) Other security features in PEM

1) Confidentiality in PEM Scenario: S sends msg M to R (using PEM)
Design of PEM (2) 1) Confidentiality in PEM Scenario: S sends msg M to R (using PEM) S picks random symm.encr. key K (corr. to session key) S creates new cleartext header [a] S encrypts K with KPUB-R: E(K, KPUB-R) [b] Additionally, unprintable characters are converted into „regular” characters Since unprintable characters used as control signals by handlers S encrypts M with K: E(M, K) [c] Entire M, with original header is encrypted => Summarizing, „enveloping” msg sent by S to R is: [a] = new „external” header (plaintext) [b] = encrypted symm. key K [c] = encrypted msg M (cf. Fig. 7-43, p. 492-ed.4)

---SKIP--- Design of PEM (3)
Example of subfields within field [b]: (M = original msg) Proc-Type: indicates applied privacy-enhancement services (e.g., „ENCRYPTED”) DEK-Info: (DEK=data encipherment key, corr. to session key) includes symmetric key K (used to encr. M) AND indicates kind of corresponding symmetric key encr. alg. (e.g., DES-CBC) Actually not just K but whole field [b] (including K) is encrypted under KPUB-R,that is, E([b], KPUB-R ) Variant of PEM standard: Using only symmetric key encryption Requires establishing common secret before sending

2) Other security features in PEM
Design of PEM (4) 2) Other security features in PEM PEM msgs always carry a digital signature in [b] Provide S’s authentication & non-repudiation Hash within digital signature assures msg integrity S can be anonymous in external header [a] S’s identity encrypted in [c] Multiple encryption algorithms are supported by PEM „Unenhanced” & and PEM msgs can be mixed by sender at will

Problem 1: key mgmt (as always with public key exchange)
Design of PEM (5) Problems with PEM Problem 1: key mgmt (as always with public key exchange) Solution 1a: Using certificate scheme Solution 1b: Using PGP (to be discussed next) Problem 2: endpoint vulnerability (S and R ends) Attacker penetrating S’s or R’s host can subvert PEM code or install Trojan for leaking keys

++SKIP++ d. Example secure e-mail systems (1)
Many encrypted systems exist (both academic and commercial) Two popular encrypted systems: i. PGP (Pretty Good Privacy) ii. --SKIP-- S/MIME

PGP (Pretty Good Privacy) Widely used, de facto secure email standard
++SKIP++ Example secure systems (2) PGP (Pretty Good Privacy) Widely used, de facto secure standard Available on most operating systems Originally free Commercial versions available Confidentiality — IDEA encryption (IDEA = International Data Encryption Algorithm; popular in Europe) Integrity – RSA-encrypted MIC (Message Integrity Check/hash, MD5) Authentication & non-repudiation – RSA-encrypted MIC Uses common key distribution channels Trusted „introducers” used to validate keys No CA (certification authority) hierarchy needed

The application must be integrated into existing email
++SKIP++ Example secure systems (3) PGP in practice The application must be integrated into existing Each user exchanging keys with all potential recipients Establishing a keyring - a ring of trusted recipients User’s keyring of known keys Containing public and private keys (protected by a password) Public keys given to you directly by a friend Public keys signed by trusted introducers Keys used for signing or encrypting messages to be sent and validate messages received

--SKIP-- Example secure e-mail systems (4)
ii. S/MIME Internet standard for secure attachments Very much like PGP Principal difference: method of key exchange PGP: user exchanging keys with all potential recipients Establishing ring of trusted recipients S/MIME: uses hierarchically validated certificates for key exchange Usually in X.509 format => S and R do not have to exchange keys in advance if have a common trusted certifier Supports many ciphers, incl. DES, AES, RC2 Integrated into many commercial packages => likely to dominate the secure market

The End of Section 5 (Ch. 7 - Part 2): Network Security (incl
The End of Section 5 (Ch. 7 - Part 2): Network Security (incl. OPTIONAL details that you may SKIP)

CS 5700 – Computer Security and Information Assurance Section 5: Security in Networks – Part 2 This version of Section 5 includes OPTIONAL slides that.

Similar presentations

Presentation on theme: "CS 5700 – Computer Security and Information Assurance Section 5: Security in Networks – Part 2 This version of Section 5 includes OPTIONAL slides that."— Presentation transcript:

Similar presentations

About project

Feedback

Log in

Auth with social network:

CS 5700 – Computer Security and Information Assurance Section 5: Security in Networks – Part 2 This version of Section 5 includes OPTIONAL slides that.

Similar presentations

Presentation on theme: "CS 5700 – Computer Security and Information Assurance Section 5: Security in Networks – Part 2 This version of Section 5 includes OPTIONAL slides that."— Presentation transcript:

Similar presentations

About project

Feedback