Skip to content
BoKSA

CYBOK 19.1 3 Network Goals, Applications, and Protocols

CYBOK 19.1-3 Network Goals, Applications, and Protocols

1. The Strategic Framework: Security Goals and Attacker Models

Before you can effectively defend a network, you must define exactly what "secure" means within your specific context. Security is not a binary state—it is a set of orthogonal goals that you must tailor to the requirements of the application and the socio-technical environment. Without a clear definition of these goals, your defense strategy remains reactive rather than architectural. In this guide, we adopt the 4-layer TCP/IP model as our primary reference for examining how these goals are implemented tactically.The CIA Triad and Beyond In your security architecture, you will primarily focus on the CIA Triad :

  • Confidentiality: Ensuring untrusted parties cannot leak or infer sensitive information. You must ensure only the intended recipient can understand the content and, ideally, that no one else can even learn that a communication took place.
  • Integrity: Guaranteeing that untrusted parties cannot alter information without detection. This is critical for everything from financial transactions to safety-critical brake commands in a vehicle.
  • Availability: Ensuring data and services remain accessible to designated users, resisting Denial of Service (DoS) attempts that aim to exhaust resources or disrupt critical paths."So What?" Layer: In professional environments, you must extend these via Authenticity (attributing communication to a specific sender) and Non-repudiation (proving that authenticity to third parties for public verification). You must also evaluate Anonymity . It is a "double-edged sword": while it protects journalists and whistleblowers, it simultaneously shields malicious actors, complicating accountability. As you will see in Section 4, maintaining anonymity against a state-level actor is nearly impossible due to traffic analysis.Understanding the Adversary To build resilient protocols, you must model the capabilities of your potential attackers:
  • Dolev-Yao Model: The "worst-case scenario" benchmark for formal protocol analysis. It assumes the attacker has complete control over the network: they can read, delay, duplicate, or synthesize any message for which they have the keys.
  • Active vs. Passive: Passive attackers (eavesdroppers) observe traffic to extract secrets or perform traffic analysis . Active attackers directly manipulate packets.
  • On-path (PITM) vs. Off-path: A Person-in-the-Middle (PITM) sits directly between parties. Off-path attackers cannot see the traffic but can cause severe harm via IP spoofing or maliciously terminating connections using TCP Reset attacks , which you must mitigate by ensuring strong randomness in sequence number generation."So What?" Layer: You use the Dolev-Yao model as your design benchmark because a protocol that survives an adversary with total network control is fundamentally robust against most real-world threats.Connective Tissue: Once you have defined your goals and modeled your adversaries, you must apply these concepts to the diverse physical and social environments where networking occurs.
2. The Socio-Technical Context: Networking Scenarios

Network security is never one-size-fits-all. The socio-technical context—who uses the network and for what purpose—dictates the vulnerabilities and requirements.Comparative Analysis of Scenarios| Network Type | Primary Characteristics | Key Security Vulnerabilities || ------ | ------ | ------ || Local Area Network (LAN) | Internal home/office systems; traditionally high-trust. | Unauthorized service access, MAC cloning, and BYOD (Bring-Your-Own-Device) bridge risks. || Bus Network (Cyber-Physical) | Shared medium for industrial/vehicular (CAN, Modbus); real-time requirements. | Lack of authentication, message suppression (e.g., compromising an ECU to drop brake commands). || Wireless Network (WLAN) | Broadcast nature; no clear physical boundaries for signals. | Eavesdropping, unauthorized joining, and susceptibility to traffic analysis. || Distributed (P2P/DHT) | Decentralized; no central authority; high scalability. | Sybil attacks (identity overpopulation) and Eclipse attacks (routing table poisoning to isolate nodes). |
"So What?" Layer: When managing a LAN, you must guard against the "default trust" fallacy. The rise of BYOD means untrusted personal devices can bridge your internal network to the outside world. In Bus Networks, "real-time necessity" often conflicts with security; you cannot always add heavy encryption if it delays a safety-critical signal causing it to arrive "too late," potentially leading to physical catastrophe.The Internet and Distributed Systems Connecting local networks via the Internet introduces an insecure end-to-end channel where you lose control of the path. In Fully-Distributed Networks like DHTs, the lack of peer authentication leads to Sybil and Eclipse attacks. While computational puzzles can slow an attacker, they are often ineffective against distributed botnets.Connective Tissue: These diverse scenarios are powered by a standardized set of rules known as the protocol stack, which you must secure through a tactical, layered approach.

3. Tactical Implementation: The Layered Protocol Stack

The Application Layer (Layer 4) Security here is often "wrapped" around legacy protocols.

  • HTTPS: Wraps HTTP in TLS to provide confidentiality, integrity, and server authentication.
  • Email: SMTP is secured end-to-end via PGP/S/MIME (authenticity/non-repudiation) or hop-by-hop via TLS.
  • DNS: You must defend against cache poisoning using DNSSEC . To prevent resolvers from linking your IP to your queries, evaluate Oblivious DNS over HTTPS (ODoH) , which adds a trusted proxy between the client and resolver to solve the "centralization" privacy problem.The Transport Layer (Layer 3) This layer manages end-to-end communication, primarily via TCP and UDP.
  • TLS 1.2 vs. 1.3: TLS 1.3 is more efficient (1-RTT) and removes insecure RSA key exchanges in favor of Diffie-Hellman for perfect forward secrecy. It also introduces 0-RTT (zero round-trip time) for resumed connections.
  • "So What?" Layer: You must understand the 0-RTT trade-off: while it maximizes speed, it weakens forward secrecy and increases susceptibility to replay attacks.
  • TCP/UDP Security: Mitigate SYN Flooding via SYN Cookies , which delay resource allocation until the client's IP is verified. Note that UDP's connectionless nature makes it the primary tool for amplification DDoS attacks .The Internet Layer (Layer 2)
  • IPsec: Secure traffic in Transport mode (payload only) or Tunnel mode (entire packet, including headers). Tunnel mode is essential for VPNs to hide internal addressing from the public Internet.
  • BGP and RPKI: BGP prefix hijacking is a critical threat. While RPKI (Resource Public Key Infrastructure) allows for Route Origin Validation (ROV), it only detects if the AS owning the prefix is on the path. It cannot detect "bogus advertisements" where a malicious AS is on-path as an intermediary to reroute traffic.The Link Layer (Layer 1)
  • 802.1X (EAP): Port-based access control. You should favor EAP-TLS over EAP-MD5 to avoid dictionary attacks.
  • VLAN Hopping: Attackers use switch spoofing (impersonating a trunking switch) or double tagging (inserting two VLAN tags) to jump segments.
  • Wireless: Transition from WEP/WPA2 toward WPA3 , which uses the Dragonfly key exchange (Simultaneous Authentication of Equals) for forward secrecy.Points of Attention (Quality Check)
  • Is EAP-TLS enforced to prevent credential-harvesting PITM attacks?
  • Is VLAN hopping mitigated by disabling automatic trunk negotiation (DTP) and explicitly assigning ports?
  • Is ingress filtering enabled at the edge to drop spoofed VXLAN packets from the Internet?
  • Are you using Opportunistic Wireless Encryption (OWE) for open networks to provide DHKE-based session secrets?
4. Advanced Notes (For 3rd & 4th Year Students)

As you advance, you must move beyond tool configuration to understanding architectural trade-offs.

  • QUIC Protocol: QUIC integrates the TLS 1.3 handshake directly into its UDP-based connection setup. To prevent being used as a DDoS reflector, QUIC limits the amplification factor to three prior to address verification.
  • Formal Verification: TLS 1.3 is the first major protocol to be formally proven secure using mathematical methods, a milestone in protocol design.
  • Onion Routing (Tor): While Tor provides sender/recipient anonymity via three-layer encryption, you must recognize its limitations against "state-level" adversaries. By correlating traffic analysis patterns (inter-arrival times and packet sizes) at both entry and exit nodes, an attacker can deanonymize users regardless of encryption strength.For deeper mathematical and operational contexts, refer to the Applied Cryptography and Security Operations Knowledge Areas. Multi-layered defense-in-depth is the only way to protect these modern, interconnected systems.