Lecture 36

Limit the Signal Wire integrity and tapping In a traditional Ethernet, signals do not radiate far beyond the wire, so eavesdropping requires physical proximity to the network cable and physically tapping the line. This is done by splicing into the line (old “thicknet” Ethernet used “vampire” taps to add stations to the network, which actually tapped into the wire), stealing a line from a hub (which splits a network into a star topology), bridging the network, or similar techniques. Tapping can be detected by the change in electrical resistance in the wire. 11/20/2019

Physical limitation The problem then becomes physically limiting who has access to the machines or the endpoints once the problem of tapping the wire between endpoints has been addressed. Machines need to run secure (enough) operating systems and have accounts and secure passwords. 11/20/2019

Encryption One of the most common and useful techniques to protect data is encryption. Encryption has been used for thousands of years; in fact, a simple technique called the Caesar cipher was said to be used by Julius Caesar. Since modern computer speeds have rendered many techniques insecure, we will focus on the essential components of the methods that are used currently 11/20/2019

Private key or symmetric encryption. Public and private key encryption Private key or symmetric encryption. Public key or asymmetric encryption 11/20/2019

Computational and data overhead Private key algorithms generally run significantly faster than public key ones. This computational overhead becomes a factor when significant amounts of data must be encrypted (or decrypted) or if the processing power of the computer is limited, as in palmtop devices o sensor networks. 11/20/2019

Integrity Codes Checksum versus cryptographic hash Checksums are a common type of basic integrity code and are transmitted with the data. The receiver applies a simple formula to the data and compares the result with the integrity code that was transmitted to ensure with high probability that he received the message intact. Parity bits and cyclic redundancy checks (CRCs) are examples of checksums. 11/20/2019

Message authentication code (MAC) A message authentication code (MAC) is a one-way hash function plus a secret key. In the first approach, the function computes the MAC by encrypting the hash of the message using the key; in the second approach, the function computes the MAC by taking the hash of the message and the key concatenated together. MACs protect the message’s authenticity without secrecy. The MACs are attached to the message. 11/20/2019

Payload versus header A packet consists of a header and a payload. The header contains source and destination address information and additional information such as the protocol, sequence number, and special flags. The payload is the data. 11/20/2019

Traffic analysis Even with encryption and integrity codes, packets are still vulnerable to a traffic analysis attack. This is particularly problematic if the header and payload are encrypted together because many parts of the header are predictable. 11/20/2019

Other Security-Related Mechanisms Authentication protocols Authentication protocols provide a mechanism to verify a user’s claimed identity when he establishes a connection to a remote system. Protocols that allow a remote system login or code execution, such as telnet, ssh, and RPC, use the remote operating system’s authentication mechanism (typically a username and password combination). 11/20/2019

AAA Authentication, authorization, and auditing (AAA) help to maintain the security of systems. An IETF Working Group focuses on developing the requirements for AAA and protocols that implement them [AAA]. 11/20/2019

special hardware can be used. One type of hardware is smartcards that contain cryptographic tokens or run algorithms to generate one-time passwords. Another type of hardware is a transmitter that makes it difficult to receive transmissions. Personal and local area networks, as well as cell phones , transmit with only enough power for the corresponding receiver to detect the signal successfully. In addition, they use frequency-hopping and spread-spectrum technology to increase the effective bandwidth, which also has the effect of making it more difficult to reconstruct the signal. 11/20/2019

IPSec IP Security IP Security (IPSec) are protocols that provide security for Internet Protocol (IP) packets. There are three main components to IPSec. The Internet key exchange (IKE) defines a hybrid protocol to negotiate and provide authenticated keying material for security associations in a protected manner. The authentication header (AH) provides message integrity. The encapsulating security payload (ESP) provides confidentiality. 11/20/2019

Authentication header The AH uses a MAC (see Section 13.3.2), referred to as an integrity check value (ICV), to guarantee the integrity of the data. It also can prevent replay attacks by including a sequence number in the header. Similar in ESP, it can operate in transport or tunnel mode. In transport mode, the AH information is added immediately following the IP header information. In tunnel mode, the entire original IP datagram becomes the payload of the IPSec packet, with the AH providing integrity for both its headers and the payload. 11/20/2019

Encapsulating security payload The ESP is a mechanism to provide confidentiality and integrity to data by encrypting the payload. ESP operates in one of two modes, tunnel mode or transport mode, and the packet’s payload consists of either the upper-layer protocol (e.g., TCP, UDP, ICMP, or IGMP) or the entire IP datagram, respectively 11/20/2019

In transport mode, the ESP is added after the IP header, before any upper-layer protocols. The original IP headers are still visible. In tunnel mode, the entire IP datagram is encrypted within the ESP. Tunnel mode can be used by security gateways. The endpoint hosts communicate with the security gateways (through a protected intranet), and the security gateways communicate with each other via IPSec. 11/20/2019