Possible attacks on the TCP/IP protocol stack and countermeasures

Let’s look at what types of threats each layer of the TCP/IP protocol stack may be susceptible to.

The task of a computer security system is to safeguard the information transmitted over the network and to adequately preserve the data stored in it. 

Excluding in this discussion threats due to natural disasters, we can classify the man-made risk, to which an information system is subject, into intentional threats or unintentional threats due to negligence or inexperience.

Businesses need to protect themselves from these threats, which can put both applications and assets at serious risk.

Intentional human threats can come from individuals with an interest in acquiring information or limiting the operation of business processes, driven by the pursuit of financial or political gain, or simply for fun.

An intentional attack can come from individuals outside the organisation or from internal staff such as ex-employees, disgruntled employees or malicious actors. In fact, personnel who are familiar with the security systems and the structure of the information system and who have the authorisation to access the system itself, can get hold of information or insert malicious code more easily.

The development of the Internet and the distributed processing of information over shared lines has certainly made security a necessary duty. Therefore, the corporate network, if not adequately protected, could be subject to unauthorised access with possible network compromise and information theft.

Network communication vulnerabilities

Network communication on the Internet follows a layered approach, where each layer adds to the activity of the previous layer according to the TCP/IP implementation paradigm. The TCP/IP protocol stack has only 4 layers compared to the standard ISO/OSI protocol (Application, Presentation, Session, Transport, Network, Data link, Physical), namely the Application, TCP, IP and Network Access layers.

According to this model, any data, divided by the host into packets on the 4 layers, is sent to the remote recipient host along a network path that includes many intermediate nodes. Each of these layers can be susceptible to certain types of attacks that exploit system, controls, and security policies vulnerabilities. 

Let’s look at what types of threats each layer may be susceptible to.

The Application Layer which is the highest layer of the stack and closest to the users and its function is to interface and provide services for application processes. It also contains standard and native applications such as Telnet, SMTP, FTP. For these reasons, this layer can be vulnerable to phishing (scammers can pass themselves off as a legitimate contact trying to steal information) and hijacking (a valid user session is exploited to gain unauthorized access to the system).

The TCP layer (Transport layer) is responsible for transporting data, dividing it into packets and handling transmission errors. A classic attack to which this layer can be subject is the Denial of Service that is implemented to disable, interrupt or damage a website, a service or the communication of an entire network by saturating the transmission bandwidth. On the other hand, a typical attack scenario to which the IP layer can be subject, which is in charge of managing data addressing through their transmission on the network, is the so called Man in the Middle attack.

Finally, the network access layer incorporates two ISO/OSI layers, the data link and the physical channel, respectively. The data link, which is the interface between the network card and the transmission medium, is responsible for specifying how data should be sent regardless of the network type. Spoofing that masks an unknown source as legitimate and trusted in a network communication could pose a real threat to this layer: ARP spoofing, for example, allows manipulation at the data link layer to fool higher layer protocols. For the physical channel that is responsible for converting digital data over the transmission medium (e.g., copper cable/electrical pulses or fiber optics/light modulation), packet sniffing techniques, by allowing the capture of data circulating over the network, can pose a real threat to the confidentiality of communications.

How mitigate the damage (countermeasures)

Defence tools cannot provide total and impenetrable protection, but they can certainly make it more difficult and costly to carry out an attack, detect its execution and mitigate the damage.

Here are some:

• Firewall. The Firewall is a hardware/software tool whose purpose is to protect a host or a network segment from potentially harmful traffic coming from the external network (e.g. Internet). This tool can operate at different levels of the ISO/OSI protocol stack: Application (Application Layer Firewall), TCP (Stateful Firewall), IP (Network Layer Firewall) and Data Link layers (Bridge Firewall).

• Intrusion Detection System (IDS). An IDS is a tool to detect possible unwanted manipulation of a particular system or network. It can be implemented at different levels of the ISO/OSI protocol stack to detect, for instance, sniffing attempts (Network access Layer) or DoS attacks (TCP Layer).

• Protocol security. In general, it’s very difficult to ensure that traffic on a network is not intercepted. Therefore, one solution might be to encrypt the transmitted data so that it’s not intelligible in case of sniffing. In this case, a VPN (Virtual Private Network) can be used to create a secure communication network through the Internet, which is by definition not secure. Basically, after an authentication phase, the encapsulated network packets, which travel along a virtual tunnel, are encrypted and decrypted at both ends of the VPN network. Encapsulation and tunneling can be implemented on different layers of the ISO/OSI stack, and each protocol with its own specifics can help protect transmitted data. Among the most common protocols are PPTP and L2PT (Data link layer), IPSEC (Network layer), SSL/TLS (Transport layer) and HTTPS (Application layer).

• Security analysis. It’s possible to monitor and evaluate the policies adopted by a company through a security audit.  To carry out this operation, which can only be performed from within the company network, it’s necessary to have a complete knowledge of all the network resources to be examined. Without having to know the characteristics of the infrastructure in advance, as an alternative or in conjunction with the security audit, a penetration test can be implemented to simulate an attack from outside the network.

• Security awareness. Raising awareness of the various aspects of IT security, to increase the level of user awareness, is more essential than ever to improve the organisation’s level of protection. The purpose of an adequate security awareness programme is to provide users (employees and managers) with basic skills, establishing appropriate prevention guidelines and rules of conduct, also with internal/external training, newsletters and company intranets.

About the author: Salvatore Lombardo

IT officer, ICT expert, Clusit member

If you want to receive the weekly Security Affairs Newsletter for free subscribe here.

Follow me on Twitter: @securityaffairs and Facebook

Pierluigi Paganini

(SecurityAffairs – hacking, TCP/IP)

The post Possible attacks on the TCP/IP protocol stack and countermeasures appeared first on Security Affairs.