Chapter 24 Wireless Network Security Systems

Editor's Notes

• #1: Lecture slides prepared for “Computer Security: Principles and Practice”, 4/e, by William Stallings and Lawrie Brown, Chapter 24 “Wireless Network Security”.
• #2: Wireless networks and communication links have become pervasive for both personal and organizational communications. A wide variety of technologies and network types have been adopted, including Wi-Fi, Bluetooth, WiMAX, ZigBee, and cellular technologies. Although the security threats and countermeasures discussed throughout this book apply to wireless networks and communications links, there are some unique aspects to the wireless environment. This chapter begins with a general overview of wireless security issues. We then focus on the relatively new area of mobile device security, examining threats and countermeasures for mobile devices used in the enterprise. Then, we look at the IEEE 802.11i standard for wireless LAN security. This standard is part of IEEE 802.11, also referred to as Wi-Fi. We begin the discussion with an overview of IEEE 802.11, and then we look in some detail at IEEE 802.11i.
• #3: Wireless networks, and the wireless devices that use them, introduce a host of security problems over and above those found in wired networks. Some of the key factors contributing to the higher security risk of wireless networks compared to wired networks include the following [MA10]: • Channel: Wireless networking typically involves broadcast communications, which is far more susceptible to eavesdropping and jamming than wired networks. Wireless networks are also more vulnerable to active attacks that exploit vulnerabilities in communications protocols. • Mobility: Wireless devices are, in principal and usually in practice, far more portable and mobile than wired devices. This mobility results in a number of risks, described subsequently. • Resources: Some wireless devices, such as smartphones and tablets, have sophisticated operating systems but limited memory and processing resources with which to counter threats, including denial of service and malware. • Accessibility: Some wireless devices, such as sensors and robots, may be left unattended in remote and/or hostile locations. This greatly increases their vulnerability to physical attacks.
• #4: In simple terms, the wireless environment consists of three components that provide point of attack ( Figure 24.1 ). The wireless client can be a cell phone, a Wi-Fi enabled laptop or tablet, a wireless sensor, a Bluetooth device, and so on. The wireless access point provides a connection to the network or service. Examples of access points are cell towers, Wi-Fi hot spots, and wireless access points to wired local or wide-area networks. The transmission medium, which carries the radio waves for data transfer, is also a source of vulnerability.
• #5: [CHOI08] lists the following security threats to wireless networks: • Accidental association: Company wireless LANs or wireless access points to wired LANs in close proximity (e.g., in the same or neighboring buildings) may create overlapping transmission ranges. A user intending to connect to one LAN may unintentionally lock on to a wireless access point from a neighboring network. Although the security breach is accidental, it nevertheless exposes resources of one LAN to the accidental user. • Malicious association: In this situation, a wireless device is configured to appear to be a legitimate access point, enabling the operator to steal passwords from legitimate users and then penetrate a wired network through a legitimate wireless access point. • Ad hoc networks: These are peer-to-peer networks between wireless computers with no access point between them. Such networks can pose a security threat due to a lack of a central point of control. • Nontraditional networks: Nontraditional networks and links, such as personal network Bluetooth devices, barcode readers, and handheld PDAs pose a security risk both in terms of eavesdropping and spoofing. • Identity theft (MAC spoofing): This occurs when an attacker is able to eavesdrop on network traffic and identify the MAC address of a computer with network privileges. • Man-in-the middle attacks: This type of attack is described in Chapter 21 in the context of the Diffie-Hellman key exchange protocol. In a broader sense, this attack involves persuading a user and an access point to believe that they are talking to each other when in fact the communication is going through an intermediate attacking device. Wireless networks are particularly vulnerable to such attacks. • Denial of service (DoS): This type of attack was discussed in detail in Chapter 7. In the context of a wireless network, a DoS attack occurs when an attacker continually bombards a wireless access point or some other accessible wireless port with various protocol messages designed to consume system resources. The wireless environment lends itself to this type of attack, because it is so easy for the attacker to direct multiple wireless messages at the target. • Network injection: A network injection attack targets wireless access points that are exposed to non-filtered network traffic, such as routing protocol messages or network management messages. An example of such an attack is one in which bogus reconfiguration commands are used to affect routers and switches to degrade network performance.
• #6: Following [CHOI08], we can group wireless security measures into those dealing with wireless transmissions, wireless access points, and wireless networks (consisting of wireless routers and endpoints). The principal threats to wireless transmission are eavesdropping, altering or inserting messages, and disruption. To deal with eavesdropping, two types of countermeasures are appropriate: • Signal-hiding techniques: Organizations can take a number of measures to make it more difficult for an attacker to locate their wireless access points, including turning off service set identifier (SSID) broadcasting by wireless access points; assigning cryptic names to SSIDs; reducing signal strength to the lowest level that still provides requisite coverage; and locating wireless access points in the interior of the building, away from windows and exterior walls. Greater security can be achieved by the use of directional antennas and of signal-shielding techniques. • Encryption: Encryption of all wireless transmission is effective against eavesdropping to the extent that the encryption keys are secured. The use of encryption and authentication protocols is the standard method of countering attempts to alter or insert transmissions. The methods discussed in Chapter 7 for dealing with denial of service apply to wireless transmissions. Organizations can also reduce the risk of unintentional DoS attacks. Site surveys can detect the existence of other devices using the same frequency range, to help determine where to locate wireless access points. Signal strengths can be adjusted and shielding used in an attempt to isolate a wireless environment from competing nearby transmissions.
• #7: The main threat involving wireless access points is unauthorized access to the network. The principal approach for preventing such access is the IEEE 802.1X standard for port-based network access control. The standard provides an authentication mechanism for devices wishing to attach to a LAN or wireless network. The use of 802.1X can prevent rogue access points and other unauthorized devices from becoming insecure backdoors.
• #8: [CHOI08] recommends the following techniques for wireless network security: 1. Use encryption. Wireless routers are typically equipped with built-in encryption mechanisms for router-to-router traffic. 2. Use anti-virus and anti-spyware software, and a firewall. These facilities should be enabled on all wireless network endpoints. 3. Turn off identifier broadcasting. Wireless routers are typically configured to broadcast an identifying signal so that any device within range can learn of the router’s existence. If a network is configured so that authorized devices know the identity of routers, this capability can be disabled, so as to thwart attackers. 4. Change the identifier on your router from the default. Again, this measure thwarts attackers who will attempt to gain access to a wireless network using default router identifiers. 5. Change your router’s pre-set password for administration. This is another prudent step. 6. Allow only specific computers to access your wireless network. A router can be configured to only communicate with approved MAC addresses. Of course, MAC addresses can be spoofed, so this is just one element of a security strategy.
• #9: Prior to the widespread use of smartphones, the dominant paradigm for computer and network security in organizations was as follows. Corporate IT was tightly controlled. User devices were typically limited to Windows PCs. Business applications were controlled by IT and either run locally on endpoints or on physical servers in data centers. Network security was based upon clearly defined perimeters that separated trusted internal networks from the untrusted Internet. Today, there have been massive changes in each of these assumptions. An organization’s networks must accommodate the following: • Growing use of new devices: Organizations are experiencing significant growth in employee’s use of mobile devices. In many cases, employees are allowed to use a combination of endpoint devices as part of their day-to-day activities. • Cloud-based applications: Applications no longer run solely on physical servers in corporate data centers. Quite the opposite, applications can run anywhere—on traditional physical servers, on mobile virtual servers, or in the cloud. Additionally, end users can now take advantage of a wide variety of cloud-based applications and IT services for personal and professional use. Facebook can be used for an employee’s personal profile or as a component of a corporate marketing campaign. Employees depend upon Skype to speak with friends abroad or for legitimate business video conferencing. Dropbox and Box can be used to distribute documents between corporate and personal devices for mobility and user productivity. • De-perimeterization: Given new device proliferation, application mobility, and cloud-based consumer and corporate services, the notion of a static network perimeter is all but gone. Now there are a multitude of network perimeters around devices, applications, users, and data. These perimeters have also become quite dynamic as they must adapt to various environmental conditions such as user role, device type, server virtualization mobility, network location, and time-of-day. • External business requirements: The enterprise must also provide guests, third-party contractors, and business partners network access using various devices from a multitude of locations. The central element in all of these changes is the mobile computing device. Mobile devices have become an essential element for organizations as part of the overall network infrastructure. Mobile devices such as smartphones, tablets, and memory sticks provide increased convenience for individuals as well as the potential for increased productivity in the workplace. Because of their widespread use and unique characteristics, security for mobile devices is a pressing and complex issue. In essence, an organization needs to implement a security policy through a combination of security features built into the mobile devices and additional security controls provided by network components that regulate the use of the mobile devices.
• #10:  Mobile devices need additional, specialized protection measures beyond those implemented for other client devices, such as desktop and laptop devices that are used only within the organization’s facilities and on the organization’s networks. NIST SP 800-124 (Guidelines for Managing the Security of Mobile Devices in the Enterprise , June 2013) lists seven major security concerns for mobile devices. We examine each of these in turn. Lack of Physical Security Controls Mobile devices are typically under the complete control of the user, and are used and kept in a variety of locations outside the organization’s control, including off premises. Even if a device is required to remain on premises, the user may move the device within the organization between secure and non secured locations. Thus, theft and tampering are realistic threats. The security policy for mobile devices must be based on the assumption that any mobile device may be stolen or at least accessed by a malicious party. The threat is twofold: A malicious party may attempt to recover sensitive data from the device itself, or may use the device to gain access to the organization’s resources. Use of Untrusted Mobile Devices In addition to company-issued and companycontrolled mobile devices, virtually all employees will have personal smartphones and/or tablets. The organization must assume that these devices are not trustworthy.  That is, the devices may not employ encryption and either the user or a third party may have installed a bypass to the built-in restrictions on security, operating system use, and so on. Use of Untrusted Networks  If a mobile device is used on premises, it can connect to organization resources over the organization’s own in-house wireless networks. However, for off-premises use, the user will typically access organizational resources via Wi-Fi or cellular access to the Internet and from the Internet to the organization. Thus, traffic that includes an off-premises segment is potentially susceptible to eavesdropping or man-in-the-middle types of attacks. Thus, the security policy must be based on the assumption that the networks between the mobile device and the organization are not trustworthy. Use of Untrusted Applications  By design, it is easy to find and install thirdparty applications on mobile devices. This poses the obvious risk of installing malicious software. An organization has several options for dealing with this threat, as described subsequently. Interaction with other Systems  A common feature found on smartphones and tablets is the ability to automatically synchronize data, apps, contacts, photos, and so on with other computing devices and with cloud-based storage. Unless an organization has control of all the devices involved in synchronization, there is considerable risk of the organization’s data being stored in an unsecured location, plus the risk of the introduction of malware. Use of Untrusted Content  Mobile devices may access and use content that other computing devices do not encounter. An example is the Quick Response (QR) code, which is a two-dimensional barcode. QR codes are designed to be captured by a mobile device camera and used by the mobile device. The QR code translates to a URL, so a malicious QR code could direct the mobile device to malicious Websites. Use of Location Services  The GPS capability on mobile devices can be used to maintain a knowledge of the physical location of the device. While this feature might be useful to an organization as part of a presence service, it creates security risks. An attacker can use the location information to determine where the device and user are located, which may be of use to the attacker.
• #11: With the threats listed in the preceding discussion in mind, we outline the principal elements of a mobile device security strategy. They fall into three categories: device security, client/server traffic security, and barrier security (Figure 24.2).
• #12: IEEE 802 is a committee that has developed standards for a wide range of local area networks (LANs). In 1990, the IEEE 802 Committee formed a new working group, IEEE 802.11, with a charter to develop a protocol and transmission specifications for wireless LANs (WLANs). Since that time, the demand for WLANs at different frequencies and data rates has exploded. Keeping pace with this demand, the IEEE 802.11 working group has issued an ever-expanding list of standards. Table 24.1 briefly defines key terms used in the IEEE 802.11 standard.
• #13: The first 802.11 standard to gain broad industry acceptance was 802.11b. Although 802.11b products are all based on the same standard, there is always a concern whether products from different vendors will successfully interoperate. To meet this concern, the Wireless Ethernet Compatibility Alliance (WECA), an industry consortium, was formed in 1999. This organization, subsequently renamed the Wi-Fi (Wireless Fidelity) Alliance, created a test suite to certify interoperability for 802.11b products. The term used for certified 802.11b products is Wi-Fi . Wi-Fi certification has been extended to 802.11g products. The Wi-Fi Alliance has also developed a certification process for 802.11a products, called Wi-Fi5 . The Wi-Fi Alliance is concerned with a range of market areas for WLANs, including enterprise, home, and hot spots. More recently, the Wi-Fi Alliance has developed certification procedures for IEEE 802.11 security standards, referred to as Wi-Fi Protected Access (WPA). The most recent version of WPA, known as WPA2, incorporates all of the features of the IEEE 802.11i WLAN security specification.
• #14: Before proceeding, we need to briefly preview the IEEE 802 protocol architecture. IEEE 802.11 standards are defined within the structure of a layered set of protocols. This structure, used for all IEEE 802 standards, is illustrated in Figure 24.3 . The lowest layer of the IEEE 802 reference model is the physical layer , which includes such functions as encoding/decoding of signals and bit transmission/reception. In addition, the physical layer includes a specification of the transmission medium. In the case of IEEE 802.11, the physical layer also defines frequency bands and antenna characteristics. All LANs consist of collections of devices that share the network’s transmission capacity. Some means of controlling access to the transmission medium is needed to provide an orderly and efficient use of that capacity. This is the function of a medium access control (MAC) layer. The MAC layer receives data from a higher-layer protocol, typically the logical link control (LLC) layer, in the form of a block of data known as the MAC service data unit (MSDU). In general, the MAC layer performs the following functions: • On transmission, assemble data into a frame, known as a MAC protocol data unit (MPDU) with address and error-detection fields. • On reception, disassemble frame, and perform address recognition and error detection. • Govern access to the LAN transmission medium.
• #15: The exact format of the MPDU differs somewhat for the various MAC protocols in use. In general, all of the MPDUs have a format similar to that of Figure 24.4 . The fields of this frame are as follows: • MAC Control: This field contains any protocol control information needed for the functioning of the MAC protocol. For example, a priority level could be indicated here. • Destination MAC Address: The destination physical address on the LAN for this MPDU. • Source MAC Address: The source physical address on the LAN for this MPDU. • MAC Service Data Unit: The data from the next higher layer. • CRC: The cyclic redundancy check field, also known as the Frame Check Sequence (FCS) field. This is an error-detecting code, such as that which is used in other data-link control protocols. The CRC is calculated based on the bits in the entire MPDU. The sender calculates the CRC and adds it to the frame. The receiver performs the same calculation on the incoming MPDU and compares that calculation to the CRC field in that incoming MPDU. If the two values don’t match, then one or more bits have been altered in transit. The fields preceding the MSDU field are referred to as the MAC header, and the field following the MSDU field is referred to as the MAC trailer. The header and trailer contain control information that accompany the data field and that are used by the MAC protocol.
• #16: Figure 24.5 illustrates the model developed by the 802.11 working group. The smallest building block of a wireless LAN is a basic service set (BSS), which consists of wireless stations executing the same MAC protocol and competing for access to the same shared wireless medium. A BSS may be isolated or it may connect to a backbone distribution system (DS) through an access point (AP). The AP functions as a bridge and a relay point. In a BSS, client stations do not communicate directly with one another. Rather, if one station in the BSS wants to communicate with another station in the same BSS, the MAC frame is first sent from the originating station to the AP and then from the AP to the destination station. Similarly, a MAC frame from a station in the BSS to a remote station is sent from the local station to the AP and then relayed by the AP over the DS on its way to the destination station. The BSS generally corresponds to what is referred to as a cell in the literature. The DS can be a switch, a wired network, or a wireless network. When all the stations in the BSS are mobile stations that communicate directly with one another (not using an AP) the BSS is called an independent BSS (IBSS). An IBSS is typically an ad hoc network. In an IBSS, the stations all communicate directly, and no AP is involved. A simple configuration is shown in Figure 24.5 , in which each station belongs to a single BSS; that is, each station is within wireless range only of other stations within the same BSS. It is also possible for two BSSs to overlap geographically, so that a single station could participate in more than one BSS. Furthermore, the association between a station and a BSS is dynamic. Stations may turn off, come within range, and go out of range. An extended service set (ESS) consists of two or more basic service sets interconnected by a distribution system. The extended service set appears as a single logical LAN to the LLC level.
• #17: IEEE 802.11 defines nine services that need to be provided by the wireless LAN to achieve functionality equivalent to that which is inherent to wired LANs. Table 24.2 lists the services and indicates two ways of categorizing them. 1. The service provider can be either the station or the DS. Station services are implemented in every 802.11 station, including AP stations. Distribution services are provided between BSSs; these services may be implemented in an AP or in another special-purpose device attached to the distribution system. 2. Three of the services are used to control IEEE 802.11 LAN access and confidentiality. Six of the services are used to support delivery of MSDUs between stations. If the MSDU is too large to be transmitted in a single MPDU, it may be fragmented and transmitted in a series of MPDUs. Following the IEEE 802.11 document, we next discuss the services in an order designed to clarify the operation of an IEEE 802.11 ESS network. MSDU delivery , which is the basic service, has already been mentioned. Services related to security are introduced in Section 24.3 .
• #18: The two services involved with the distribution of messages within a DS are distribution and integration. Distribution is the primary service used by stations to exchange MPDUs when the MPDUs must traverse the DS to get from a station in one BSS to a station in another BSS. For example, suppose a frame is to be sent from station 2 (STA 2) to station 7 (STA 7) in Figure 24.5 . The frame is sent from STA 2 to AP 1, which is the AP for this BSS. The AP gives the frame to the DS, which has the job of directing the frame to the AP associated with STA 7 in the target BSS. AP 2 receives the frame and forwards it to STA 7. How the message is transported through the DS is beyond the scope of the IEEE 802.11 standard. If the two stations that are communicating are within the same BSS, then the distribution service logically goes through the single AP of that BSS. The integration service enables transfer of data between a station on an IEEE 802.11 LAN and a station on an integrated IEEE 802.x LAN. The term integrated refers to a wired LAN that is physically connected to the DS and whose stations may be logically connected to an IEEE 802.11 LAN via the integration service. The integration service takes care of any address translation and media conversion logic required for the exchange of data.
• #19: The primary purpose of the MAC layer is to transfer MSDUs between MAC entities; this purpose is fulfilled by the distribution service. For that service to function, it requires information about stations within the ESS that is provided by the association-related services. Before the distribution service can deliver data to or accept data from a station, that station must be associated . Before looking at the concept of association, we need to describe the concept of mobility. The standard defines three transition types, based on mobility: • No transition: A station of this type is either stationary or moves only within the direct communication range of the communicating stations of a single BSS. • BSS transition: This is defined as a station movement from one BSS to another BSS within the same ESS. In this case, delivery of data to the station requires that the addressing capability be able to recognize the new location of the station. • ESS transition: This is defined as a station movement from a BSS in one ESS to a BSS within another ESS. This case is supported only in the sense that the station can move. Maintenance of upper-layer connections supported by 802.11 cannot be guaranteed. In fact, disruption of service is likely to occur.
• #20: To deliver a message within a DS, the distribution service needs to know where the destination station is located. Specifically, the DS needs to know the identity of the AP to which the message should be delivered in order for that message to reach the destination station. To meet this requirement, a station must maintain an association with the AP within its current BSS. Three services relate to this requirement: • Association: Establishes an initial association between a station and an AP. Before a station can transmit or receive frames on a wireless LAN, its identity and address must be known. For this purpose, a station must establish an association with an AP within a particular BSS. The AP can then communicate this information to other APs within the ESS to facilitate routing and delivery of addressed frames. • Reassociation: Enables an established association to be transferred from one AP to another, allowing a mobile station to move from one BSS to another. • Disassociation: A notification from either a station or an AP that an existing association is terminated. A station should give this notification before leaving an ESS or shutting down. However, the MAC management facility protects itself against stations that disappear without notification.
• #21: There are two characteristics of a wired LAN that are not inherent in a wireless LAN. 1. In order to transmit over a wired LAN, a station must be physically connected to the LAN. On the other hand, with a wireless LAN, any station within radio range of the other devices on the LAN can transmit. In a sense, there is a form of authentication with a wired LAN in that it requires some positive and presumably observable action to connect a station to a wired LAN. 2. Similarly, in order to receive a transmission from a station that is part of a wired LAN, the receiving station also must be attached to the wired LAN. On the other hand, with a wireless LAN, any station within radio range can receive. Thus, a wired LAN provides a degree of privacy, limiting reception of data to stations connected to the LAN. These differences between wired and wireless LANs suggest the increased need for robust security services and mechanisms for wireless LANs. The original 802.11 specification included a set of security features for privacy and authentication that were quite weak. For privacy, 802.11 defined the Wired Equivalent Privacy (WEP) algorithm. The privacy portion of the 802.11 standard contained major weaknesses. Subsequent to the development of WEP, the 802.11i task group has developed a set of capabilities to address the WLAN security issues. In order to accelerate the introduction of strong security into WLANs, the Wi-Fi Alliance promulgated Wi-Fi Protected Access (WPA) as a Wi-Fi standard. WPA is a set of security mechanisms that eliminates most 802.11 security issues and was based on the current state of the 802.11i standard. The final form of the 802.11i standard is referred to as Robust Security Network (RSN). The Wi-Fi Alliance certifies vendors in compliance with the full 802.11i specification under the WPA2 program.
• #22: The 802.11i RSN security specification defines the following services. • Authentication: A protocol is used to define an exchange between a user and an AS (authentication server) that provides mutual authentication and generates temporary keys to be used between the client and the AP over the wireless link. • Access control: This function enforces the use of the authentication function, routes the messages properly, and facilitates key exchange. It can work with a variety of authentication protocols. • Privacy with message integrity: MAC-level data (e.g., an LLC PDU) are encrypted along with a message integrity code that ensures that the data have not been altered. Figure 24.6a indicates the security protocols used to support these services, while Figure 24.6b lists the cryptographic algorithms used for these services.
• #23: The operation of an IEEE 802.11i RSN can be broken down into five distinct phases. The exact nature of the phases will depend on the configuration and the end points of the communication. Possibilities include (see Figure 24.5 ): 1. Two wireless stations in the same BSS communicating via the access point for that BSS. 2. Two wireless stations (STAs) in the same ad hoc IBSS communicating directly with each other. 3. Two wireless stations in different BSSs communicating via their respective APs across a distribution system. 4. A wireless station communicating with an end station on a wired network via its AP and the distribution system. IEEE 802.11i security is concerned only with secure communication between the STA and its AP. In case 1 in the preceding list, secure communication is assured if each STA establishes secure communications with the AP. Case 2 is similar, with the AP functionality residing in the STA. For case 3, security is not provided across the distribution system at the level of IEEE 802.11, but only within each BSS. End-to-end security (if required) must be provided at a higher layer. Similarly, in case 4, security is only provided between the STA and its AP. With these considerations in mind, Figure 24.7 depicts the five phases of operation for an RSN and maps them to the network components involved. One new component is the authentication server (AS). The rectangles indicate the exchange of sequences of MPDUs. The five phases are defined as follows: • Discovery: An AP uses messages called Beacons and Probe Responses to advertise its IEEE 802.11i security policy. The STA uses these to identify an AP for a WLAN with which it wishes to communicate. The STA associates with the AP, which it uses to select the cipher suite and authentication mechanism when the Beacons and Probe Responses present a choice. • Authentication: During this phase, the STA and AS prove their identities to each other. The AP blocks non-authentication traffic between the STA and AS until the authentication transaction is successful. The AP does not participate in the authentication transaction other than forwarding traffic between the STA and AS. Key Management: The AP and the STA perform several operations that cause cryptographic keys to be generated and placed on the AP and the STA. Frames are exchanged between the AP and STA only. • Protected data transfer: Frames are exchanged between the STA and the end station through the AP. As denoted by the shading and the encryption module icon, secure data transfer occurs between the STA and the AP only; security is not provided end-to-end. • Connection termination: The AP and STA exchange frames. During this phase, the secure connection is torn down and the connection is restored to the original state.
• #24: We now look in more detail at the RSN phases of operation, beginning with the discovery phase, which is illustrated in the upper portion of Figure 24.8 . The purpose of this phase is for an STA and an AP to recognize each other, agree on a set of security capabilities, and establish an association for future communication using those security capabilities. During this phase, the STA and AP decide on specific techniques in the following areas: • Confidentiality and MPDU integrity protocols for protecting unicast traffic (traffic only between this STA and AP) • Authentication method • Cryptography key management approach Confidentiality and integrity protocols for protecting multicast/broadcast traffic are dictated by the AP, since all STAs in a multicast group must use the same protocols and ciphers. The specification of a protocol, along with the chosen key length (if variable), is known as a cipher suite . The options for the confidentiality and integrity cipher suite are: • WEP, with either a 40-bit or 104-bit key, which allows backward compatibility with older IEEE 802.11 implementations • TKIP • CCMP • Vendor-specific methods The other negotiable suite is the authentication and key management (AKM) suite, which defines (1) the means by which the AP and STA perform mutual authentication and (2) the means for deriving a root key from which other keys may be generated. The possible AKM suites are • IEEE 802.1X • Pre-shared key (no explicit authentication takes place and mutual authentication is implied if the STA and AP share a unique secret key) • Vendor-specific methods The discovery phase consists of three exchanges. • Network and security capability discovery: During this exchange, STAs discover the existence of a network with which to communicate. The AP either periodically broadcasts its security capabilities (not shown in figure), indicated by RSN IE (Robust Security Network Information Element), in a specific channel through the Beacon frame; or responds to a station’s Probe Request through a Probe Response frame. A wireless station may discover available access points and corresponding security capabilities by either passively monitoring the Beacon frames or actively probing every channel. • Open system authentication: The purpose of this frame sequence, which provides no security, is simply to maintain backward compatibility with the IEEE 802.11 state machine, as implemented in existing IEEE 802.11 hardware. In essence, the two devices (STA and AP) simply exchange identifiers. • Association: The purpose of this stage is to agree on a set of security capabilities to be used. The STA then sends an Association Request frame to the AP. In this frame, the STA specifies one set of matching capabilities (one authentication and key management suite, one pair-wise cipher suite, and one group-key cipher suite) from among those advertised by the AP. If there is no match in capabilities between the AP and the STA, the AP refuses the Association Request. The STA blocks it too, in case it has associated with a rogue AP or someone is inserting frames illicitly on its channel. As shown in Figure 24.8, the IEEE 802.1X controlled ports are blocked, and no user traffic goes beyond the AP. The concept of blocked ports is explained subsequently. As was mentioned, the authentication phase enables mutual authentication between an STA and an authentication server located in the DS. Authentication is designed to allow only authorized stations to use the network and to provide the STA with assurance that it is communicating with a legitimate network.
• #25: IEEE 802.11i makes use of another standard that was designed to provide access control functions for LANs. The standard is IEEE 802.1X, Port-Based Network Access Control. The authentication protocol that is used, the Extensible Authentication Protocol (EAP), is defined in the IEEE 802.1X standard. IEEE 802.1X uses the terms supplicant , authenticator , and authentication server. In the context of an 802.11 WLAN, the first two terms correspond to the wireless station and the AP. The AS is typically a separate device on the wired side of the network (i.e., accessible over the DS) but could also reside directly on the authenticator. Until the AS authenticates a supplicant (using an authentication protocol), the authenticator only passes control and authentication messages between the supplicant and the AS; the 802.1X control channel is unblocked, but the 802.11 data channel is blocked. Once a supplicant is authenticated and keys are provided, the authenticator can forward data from the supplicant, subject to predefined access control limitations for the supplicant to the network. Under these circumstances, the data channel is unblocked. As indicated in Figure 24.9, 802.1X uses the concepts of controlled and uncontrolled ports. Ports are logical entities defined within the authenticator and refer to physical network connections. For a WLAN, the authenticator (the AP) may have only two physical ports: one connecting to the DS and one for wireless communication within its BSS. Each logical port is mapped to one of these two physical ports. An uncontrolled port allows the exchange of PDUs between the supplicant and the other AS, regardless of the authentication state of the supplicant. A controlled port allows the exchange of PDUs between a supplicant and other systems on the LAN only if the current state of the supplicant authorizes such an exchange. The 802.1X framework, with an upper-layer authentication protocol, fits nicely with a BSS architecture that includes a number of wireless stations and an AP. However, for an IBSS, there is no AP. For an IBSS, 802.11i provides a more complex solution that, in essence, involves pairwise authentication between stations on the IBSS.
• #26: The lower part of Figure 24.8 shows the MPDU exchange dictated by IEEE 802.11 for the authentication phase. We can think of authentication phase as consisting of the following three phases. • Connect to AS: The STA sends a request to its AP (the one with which it has an association) for connection to the AS. The AP acknowledges this request and sends an access request to the AS. • EAP exchange: This exchange authenticates the STA and AS to each other. A number of alternative exchanges are possible, as explained subsequently. • Secure key delivery: Once authentication is established, the AS generates a master session key (MSK), also known as the Authentication, Authorization, and Accounting (AAA) key, and sends it to the STA. As explained subsequently, all the cryptographic keys needed by the STA for secure communication with its AP are generated from this MSK. IEEE 802.11i does not prescribe a method for secure delivery of the MSK but relies on EAP for this. Whatever method is used, it involves the transmission of an MPDU containing an encrypted MSK from the AS, via the AP, to the AS.
• #27: During the key management phase, a variety of cryptographic keys are generated and distributed to STAs. There are two types of keys: pairwise keys used for communication between an STA and an AP and group keys used for multicast communication. Figure 24.10, based on [FRAN07], shows the two key hierarchies, and Table 24.3 defines the individual keys.
• #28: Pairwise keys are used for communication between a pair of devices, typically between an STA and an AP. These keys form a hierarchy beginning with a master key from which other keys are derived dynamically and used for a limited period of time. At the top level of the hierarchy are two possibilities. A pre-shared key (PSK) is a secret key shared by the AP and a STA and installed in some fashion outside the scope of IEEE 802.11i. The other alternative is the master session key (MSK) , also known as the AAAK, which is generated using the IEEE 802.1X protocol during the authentication phase, as described previously. The actual method of key generation depends on the details of the authentication protocol used. In either case (PSK or MSK), there is a unique key shared by the AP with each STA with which it communicates. All the other keys derived from this master key are also unique between an AP and an STA. Thus, each STA, at any time, has one set of keys, as depicted in the hierarchy of Figure 24.10a , while the AP has one set of such keys for each of its STAs. The pairwise master key (PMK) is derived from the master key. If a PSK is used, then the PSK is used as the PMK; if a MSK is used, then the PMK is derived from the MSK by truncation (if necessary). By the end of the authentication phase, marked by the 802.1x EAP Success message ( Figure 24.8 ), both the AP and the STA have a copy of their shared PMK. The PMK is used to generate the pairwise transient key (PTK), which in fact consists of three keys to be used for communication between an STA and AP after they have been mutually authenticated. To derive the PTK, the HMAC-SHA-1 function is applied to the PMK, the MAC addresses of the STA and AP, and nonces generated when needed. Using the STA and AP addresses in the generation of the PTK provides protection against session hijacking and impersonation; using nonces provides additional random keying material. The three parts of the PTK are as follows. • EAP Over LAN (EAPOL) Key Confirmation Key (EAPOL-KCK): Supports the integrity and data origin authenticity of STA-to-AP control frames during operational setup of an RSN. It also performs an access control function: proof-of-possession of the PMK. An entity that possesses the PMK is authorized to use the link. • EAPOL Key Encryption Key (EAPOL-KEK): Protects the confidentiality of keys and other data during some RSN association procedures. • Temporal Key (TK): Provides the actual protection for user traffic. Group keys are used for multicast communication in which one STA sends MPDUs to multiple STAs. At the top level of the group key hierarchy is the group master key (GMK). The GMK is a key-generating key used with other inputs to derive the group temporal key (GTK). Unlike the PTK, which is generated using material from both AP and STA, the GTK is generated by the AP and transmitted to its associated STAs. Exactly how this GTK is generated is undefined. IEEE 802.11i, however, requires that its value is computationally indistinguishable from random. The GTK is distributed securely using the pairwise keys that are already established. The GTK is changed every time a device leaves the network.
• #29: The upper part of Figure 24.11 shows the MPDU exchange for distributing pairwise keys. This exchange is known as the 4-way handshake. The STA and AP use this handshake to confirm the existence of the PMK, verify the selection of the cipher suite, and derive a fresh PTK for the following data session. The four parts of the exchange are as follows. • AP → STA: Message includes the MAC address of the AP and a nonce (Anonce) • STA → AP: The STA generates its own nonce (Snonce) and uses both nonces and both MAC addresses, plus the PMK, to generate a PTK. The STA then sends a message containing its MAC address and Snonce, enabling the AP to generate the same PTK. This message includes a message integrity code (MIC) using HMAC-MD5 or HMAC-SHA-1-128. The key used with the MIC is KCK. • AP → STA: The AP is now able to generate the PTK. The AP then sends a message to the STA, containing the same information as in the first message, but this time including a MIC. • STA → AP: This is merely an acknowledgement message, again protected by a MIC. For group key distribution, the AP generates a GTK and distributes it to each STA in a multicast group. The two-message exchange with each STA consists of the following: • AP → STA: This message includes the GTK, encrypted either with RC4 or with AES. The key used for encryption is KEK. A MIC value is appended. • STA → AP: The STA acknowledges receipt of the GTK. This message includes a MIC value.
• #30: IEEE 802.11i defines two schemes for protecting data transmitted in 802.11 MPDUs: the Temporal Key Integrity Protocol (TKIP) and the Counter Mode-CBC MAC Protocol (CCMP). TKIP is designed to require only software changes to devices that are implemented with the older wireless LAN security approach called Wired Equivalent Privacy (WEP). TKIP provides two services: • Message integrity: TKIP adds a message integrity code to the 802.11 MAC frame after the data field. The MIC is generated by an algorithm, called Michael, that computes a 64-bit value using as input the source and destination MAC address values and the data field, plus key material. • Data confidentiality: Data confidentiality is provided by encrypting the MPDU plus MIC value using RC4. The 256-bit TK ( Figure 24.10 ) is employed as follows. Two 64-bit keys are used with the Michael message digest algorithm to produce a message integrity code. One key is used to protect STA-to-AP messages, and the other key is used to protect AP-to-STA messages. The remaining 128 bits are truncated to generate the RC4 key used to encrypt the transmitted data. For additional protection, a monotonically increasing TKIP sequence counter (TSC) is assigned to each frame. The TSC serves two purposes. First, the TSC is included with each MPDU and is protected by the MIC to protect against replay attacks. Second, the TSC is combined with the session TK to produce a dynamic encryption key that changes with each transmitted MPDU, thus making cryptanalysis more difficult.
• #31: CCMP is intended for newer IEEE 802.11 devices that are equipped with the hardware to support this scheme. As with TKIP, CCMP provides two services: • Message integrity: CCMP uses the cipher-block-chaining message authentication code (CBC-MAC), described in Chapter 12. • Data confidentiality: CCMP uses the CTR block cipher mode of operation with AES for encryption. CTR is described in Chapter 20. The same 128-bit AES key is used for both integrity and confidentiality. The scheme uses a 48-bit packet number to construct a nonce to prevent replay attacks.
• #32: At a number of places in the IEEE 802.11i scheme, a pseudorandom function (PRF) is used. For example, it is used to generate nonces, to expand pairwise keys, and to generate the GTK. Best security practice dictates that different pseudorandom number streams be used for these different purposes. However, for implementation efficiency we would like to rely on a single pseudorandom number generator function. The PRF is built on the use of HMAC-SHA-1 to generate a pseudorandom bit stream. Recall that HMAC-SHA-1 takes a message (block of data) and a key of length at least 160 bits and produces a 160-bit hash value. SHA-1 has the property that the change of a single bit of the input produces a new hash value with no apparent connection to the preceding hash value. This property is the basis for pseudorandom number generation. The IEEE 802.11i PRF takes four parameters as input and produces the desired number of random bits. The function is of the form PRF(K , A , B , Len ), where K = a secret key A = a text string specific to the application (e.g., nonce generation or pairwise key expansion) B = some data specific to each case Len = desired number of pseudorandom bits For example, for the pairwise transient key for CCMP: PTK = PRF(PMK, “Pairwise key expansion”, min(AP-Addr, STA-Addr)|| max (AP-Addr, STA-Addr)|| min(Anonce, Snonce)|| max(Anonce, Snonce), 384) So, in this case, the parameters are K = PMK A = the text string “Pairwise key expansion” B = a sequence of bytes formed by concatenating the two MAC addresses and the two nonces Len = 384 bits Similarly, a nonce is generated by Nonce = PRF(Random Number,"Init Counter", MAC || Time, 256) Where Time is a measure of the network time known to the nonce generator. The group temporal key is generated by: GTK = PRF(GMK, "Group key expansion", MAC || Gnonce, 256) Figure 24.12 illustrates the function PRF( K , A , B , Len ). The parameter K serves as the key input to HMAC. The message input consists of four items concatenated together: the parameter A , a byte with value 0, the parameter B , and a counter i . The counter is initialized to 0. The HMAC algorithm is run once, producing a 160-bit hash value. If more bits are required, HMAC is run again with the same inputs, except that i is incremented each time until the necessary number of bits is generated. We can express the logic as PRF(K, A, B, Len) R ← null string for i ← 0 to ((Len + 159)/160 – 1) do R ← R||HMAC-SHA-1(K, A||0||B||i) Return Truncate-to-Len(R, Len)
• #33: Chapter 24 summary.