Transport Layer Security and Wireless Protection

Posted Saturday, February 14, 2015 in Network Engineering

Introduction

Discussing security for information systems can be summarized by three concepts: Confidentiality, integrity, and availability. Each offers an avenue for security measures to be taken such as with encryption, authentication, and failover security controls. There are also different scenarios that require different security measures. For confidentiality, these scenarios can be described as data at rest and data in motion. Confidentiality is concerned with the protection of information from unauthorized disclosure. For example, confidentiality for data in disk storage (data at rest) can be ensured through encryption. Confidentiality for TCP/IP traffic (data in motion) can also be ensured through encryption but requires special considerations such as the method of key negotiation. There is also the matter of ensuring integrity, the assurance that data is protected from unauthorized modification, as well as availability. Ensuring confidentiality and integrity would be hardly useful if doing so prevented the use of information systems when needed. Controls such as redundancies and resilience to failures are also an important security consideration when choosing how to protect and information system.

Wireless communications offer a unique set of challenges for ensuring confidentiality, integrity, and availability. First is nature of wireless transmission. Wireless communication that uses omnidirectional antennas or even high-gain antennas that strengthen transmission to a one or more directions can be listened to from a wireless receiver using the same frequency and appropriate digital signal codec (Stallings, 2007). This behavior makes wireless transmissions one of the most easily monitored communication mediums. Wireless transmission protocols apply various security measures to ensure confidentiality, integrity, and availability. In the following, transport layer security and wireless security will be examined in detail.

Encryption Primer

A wide multitude of algorithms exists to perform cryptographic operations such as encryption, hashing, and digital signing. A commonality between cryptographic algorithms is the use of keys as inputs. For example, in a shift cipher, a key designates the number of places a character is shifted to produce the ciphertext. The shift cipher's key usage is one example of a symmetric key stream in which the key is used for both encryption and decryption. Alternately, public-key encryption utilizes a public key for encryption and a private key for decryption.

Symmetric key encryption can be useful for it's simple single key usage. For example, BitLocker, Microsoft's proprietary disk encryption technology, implements the symmetric algorithm Advanced Encryption Standard (AES) to provide data at rest confidentiality (Microsoft, 2015). The simplicity of symmetric key usage can also be a downside. If used for communication symmetric encryption requires both parties to have a copy of the symmetric key (Stallings & Brown, 2008). Exchanging a key between parties that wish to communicate could be intercepted and all subsequent communication decrypted. For this reason symmetric key encryption is inadequate for wireless communication.

Unlike symmetric key encryption, public key encryption allows for encryption of information to be done with a public key and decryption done with a private key. For a communication session, a key pair is used by both parties. The key pair constitutes a public and private key in which the private key is never shared. However, the public key is shared and can be communicated or stored in the open as only the corresponding private key can decrypt data encrypted by its public key (Stallings & Brown, 2008). The public/private dynamic makes public key encryption an ideal solution for presentation and application layer encryption. The presentation and application layer refers to top-level layers of the open system interconnection model used to conceptualize network communication (Stallings & Brown, 2008).

SSL and TLS

Secure socket layer (SSL) and its successor transport layer security (TLS) constitute a public-key encryption protocol. Both SSL and TLS can be seen as different versions of the same function: To negotiate the initiation, management, and termination of a secure channel of communication (Oppliger, 2009). Although, versions of SSL and TLS do have different characteristics which can produce certain security assurances such as forward security but can be considered the same protocol. However, SSLv2 and SSLv3 have been found to be insecure and although still supported in some implementations, newer versions of TLS should be used exclusively (Giesen, Kohlar, & Stebila, 2013).

Establishing a TLS connection begins with a handshake protocol in which the encryption algorithm is chosen and key material is exchanged. The use of TLS doesn't necessarily specify a specific encryption algorithm. Several public key encryption algorithms such as stream, block, and authenticated encryption with associated data (AEAD) ciphers can be used (Dierks & Rescorla, 2008). During the TLS handshake, the client sends which encryption algorithms it can support to the server.

TLS Implementation

Implementation of TLS can and often is applied over transport and session protocols such as TCP and UDP. The application and presentation layer of the OSI model includes protocols such as web and email traffic. For example, web traffic uses TLS to secure data payloads within HTTP packets. Transmission standards such as 802.11 for wireless local area networks can also implement TLS to exchange key material and establish an encrypted channel for all application and presentation layer data. For wireless networks to implement TLS a combined protocol of the extensible authentication protocol and TLS or EAP-TLS is used to establish the TLS tunnel (Liu & Coslow, 2008).

Mobile wireless networks can also implement TLS to both establish a TLS tunnel between device and base station as well as during handoff from one base station to another. Faria, Korhonen, and Souto (2014) found that TLS performed a faster handoff, was less complex to configure, and was more widely used by mobile devices compared to the Internet key exchange (IKEv2) protocol.

Wireless Intrusion Detection and Preventions

One of the security measures that can be taken to counteract the openness of wireless transmissions is the use of a wireless intrusion detection and preventions system. Rogue wireless devices attempting to access a wireless network without authorization can be detected and/or prevented through a system that analyzes wireless traffic to discern malicious traffic from authorized traffic. Unfortunately, wireless intrusion detection cannot detect a passive listening device if it never transmits an intrusion attempt. This allows anyone with physical access to the area of transmission to monitor and capture traffic and highlights the need for an encrypted communications.

There is also the matter of using updated encryption standards when the use of protocols like SSLv3 can be subjected to leaking private keys. Jenkins et al. (2014) found that they were able to evade wireless intrusion detection using a method of crafting frames on the physical layer of wireless communication to successfully fingerprint 802.15.4 devices. The ability to fingerprint a device can also mean the ability to predict the use of outdated encryption standards and use a targeted attack.

Wireless intrusion detection and prevention systems can still prove useful in a layered security approach. A layered approach can be seen as applying several security measures in tandem to provide security fallbacks if one measure should fail. For example, one layered approach may be using physical security in the transmission area to prevent or remove malicious devices, use of wireless intrusion detection, use of TLS for tunneling wireless transmissions, and use of additional public key encryption for application-specific traffic such as web and email data. If the wireless intrusion detection and prevention system failed then data would still be protected by encryption measures.

Conclusion

It has been shown that encryption standards and vulnerabilities to attack is an evolving field. Protecting against evolving attacks and ensuring confidentiality, integrity, and availability should be taken as a continual effort. Network administrators should be familiar with the products that they employ as well as the firmware and configurations that specify their implemented security measures. A continual learning effort may be achieved by network administrators through industry certifications and involvement in continual learning programs such as with the International Information System Security Certification Consortium's continuing professional education system (ISC2, 2015).