The authentication of access clients is an important security concern. Authentication methods typically use an authentication protocol that is negotiated during the connection establishment process. IAS also supports unauthenticated connections.
Each authentication method has advantages and disadvantages in terms of security, usability, and breadth of support. The authentication method used is determined by the configuration of the access client, the access server, and the RADIUS server. Consult your access server documentation to determine which authentication methods are supported.
You can configure IAS to accept multiple authentication methods. You can also configure your access servers to attempt to negotiate a connection by using the most secure protocol first, and then the next most secure, and so on down to the least secure. For example, the Routing and Remote Access service tries to negotiate a connection using EAP first, then MS-CHAP v2, then MS-CHAP v1 , then CHAP, and then PAP. When EAP is chosen as the authentication method, the negotiation of the EAP type occurs between the access client and the IAS server.
In addition, you can use remote access policies for various authentication methods, depending on the type of client being authenticated. For example, you can create two remote access policies, one for VPN clients and one for wireless clients—each of which uses a different authentication method. The remote access policy for VPN clients can be configured to use EAP-TLS with smart cards or certificates as the authentication method and authentication type, while the remote access policy for wireless clients can be configured to use PEAP-MS-CHAP v2, which provides secure password authentication.
The entire process of packet encapsulation, transmission, and de-encapsulation is called tunneling. Tunneling is a method of using an internetwork infrastructure of one protocol to transfer a payload. Sometimes, the payload consists of the frames (or packets) of another protocol. Instead of being sent as the originating host produces it, the frame is encapsulated with an additional header. The additional header provides routing information so that the encapsulated payload can traverse an intermediate internetwork (also known as a transit internetwork). The encapsulated packets are then routed between tunnel endpoints over the transit internetwork. After the encapsulated payload packets reach their destination on the transit internetwork, the frame is de-encapsulated and forwarded to its final destination.
The logical path in the transit internetwork through which the encapsulated packets travel is called a tunnel.
With remote access VPN connections, there are two types of tunneling: voluntary and compulsory
Voluntary tunneling
When a client computer is configured with the appropriate tunneling protocol, the user or client computer can issue a VPN request to create and configure a voluntary tunnel from the client computer to the VPN server. In this case, the user’s computer is a tunnel endpoint that acts as the tunnel client, and the VPN server is the tunnel server.
A simple example of voluntary tunneling occurs when a user with a dial-up connection to the Internet through an Internet service provider (ISP) connects to an organization VPN server that is also connected to the Internet. In this scenario, the user has two separate accounts, each account having its own set of credentials:
-
A set of credentials for an account with an ISP
-
A set of credentials for an account with the organization
To connect to the organization VPN server, the client computer uses a dial-up modem to attempt to connect to the ISP dial-up server. The user supplies his ISP account credentials, is authenticated by the ISP RADIUS server against the ISP user accounts database, and is authorized by the ISP RADIUS server to connect to the ISP and the Internet.
After the dial-up connection is established and the client computer is on the Internet, the user initiates a VPN connection to his organization VPN server. The VPN server is configured as a RADIUS client to an IAS server at the organization. The user supplies organization account credentials for the connection attempt, and the organization IAS server uses the credentials to authenticate and authorize the user against an Active Directory user accounts database. After the user is authenticated and authorized, the VPN tunnel is established and the user can access organization resources.
In another example, IAS is used in an outsourced bulk dial scenario for voluntary tunneling. In the outsourced bulk dial scenario, an ISP is providing both dial-up and non-dedicated broadband digital subscriber line (DSL) Internet access for all the employees of an organization. Rather than providing the ISP with a copy of its user accounts database, the organization and the ISP use IAS as a RADIUS proxy so that users, when connecting to the ISP, can be authenticated and authorized using the organization RADIUS servers and Active Directory user accounts database on the organization network. This provides strong security and prevents exposure of sensitive information contained in the user accounts database to the ISP and its employees.
When the organization employee attempts to connect to the organization VPN server, the following process occurs:
-
The client computer establishes the dial-up or DSL connection to the ISP NAS.
-
Based on the connection parameters, the NAS sends an Access-Request message to an IAS server that is configured as a RADIUS proxy.
-
The RADIUS proxy processes the Access-Request message and determines where to send it based on the realm portion of the User-Name attribute. The RADIUS proxy forwards the Access-Request message to the organization IAS server, which is accessible on the Internet through a firewall configured to allow traffic on the default RADIUS UDP ports of 1812 (authentication) and 1813 (accounting).
-
The organization IAS server authenticates and authorizes the connection attempt of the client computer and sends an Access-Accept message back to the RADIUS proxy.
-
The RADIUS proxy forwards the Access-Accept message to the ISP NAS and the ISP NAS connects the client computer to the Internet.
-
After it is on the Internet, the client computer initiates a VPN tunnel connection with the organization VPN tunnel server on the Internet.
-
Based on the tunnel connection parameters, the tunnel server sends an Access-Request message to the organization IAS server.
-
The organization IAS server authenticates and authorizes the connection attempt of the tunnel client and sends an Access-Accept message back to the tunnel server.
-
The tunnel server completes the tunnel creation and the tunnel client can now send packets to the organization intranet through the tunnel across the Internet.
The figure, “Voluntary Tunnel Created by a Dial-Up User,” illustrates the establishment of a voluntary VPN tunnel by a dial-up user.
Voluntary Tunnel Created by a Dial-Up User
Voluntary tunneling is not different from other types of network access, and IAS can be used for authentication, authorization, and accounting when the tunnel server is configured as a RADIUS client to the IAS server.
Compulsory tunneling
Compulsory tunneling is the creation of a secure tunnel by another computer or network device on behalf of the client computer. Compulsory tunnels are configured and created automatically for the user without their knowledge or intervention. With a compulsory tunnel, the user’s computer is not a tunnel endpoint. Another device between the user’s computer and the tunnel server is the tunnel endpoint, acting as the tunnel client. The dial-up access server dialed by the client computer is the tunnel endpoint, acting as the tunnel client.
A number of vendors that sell dial-up access servers have implemented the ability to create a tunnel on behalf of a dial-up client. The computer or network device providing the tunnel for the client computer is known as a Front End Processor (FEP) in PPTP, a Layer Two Tunneling Protocol (L2TP), an Access Concentrator in L2TP, or an IP Security Gateway in Internet Protocol Security (IPSec).
In this section, “How IAS Works,” the acronym FEP describes this functionality, regardless of the tunneling protocol. To carry out its function, the FEP must have the appropriate tunneling protocol installed and must be capable of establishing the tunnel when the client computer attempts a connection.
A corporation can contract with an ISP to deploy a nationwide set of FEPs. These FEPs can establish tunnels across the Internet to a tunnel server connected to the private network of the corporation, thereby consolidating calls from geographically diverse locations into a single Internet connection at the corporate network.
The following figure, “Compulsory Tunnel Created by a Tunneling-Enabled NAS,” shows the client computer placing a dial-up call to a tunneling-enabled NAS at the ISP to authenticate against an IAS server on the other side of the tunnel.
Compulsory Tunnel Created by a Tunneling-Enabled NAS
The preceding figure, “Compulsory Tunnel Created by a Tunneling-Enabled NAS,” illustrates how IAS is used in an outsourced bulk dial scenario for compulsory tunneling.
In the outsourced bulk-dial scenario, a dial-up client establishes a dial-up connection to an ISP that is providing tunneled access across the Internet for all the employees of an organization. Based on the dial-up connection parameters, the ISP NAS sends an Access-Request message to an IAS server. The ISP IAS server authorizes, but does not authenticate, the tunnel connection. The ISP IAS server sends an Access-Accept message to the ISP NAS with a series of VPN tunnel attributes. The ISP NAS then acts as a tunnel client and creates a VPN tunnel to the VPN tunnel server of the organization that faces the Internet.
Note
-
Typically, IAS provides both authentication and authorization. However, it is common for the ISP IAS server to provide only authorization when the dial-in user is authenticated by the organization IAS server.
The ISP NAS then sends a PPP message to the dial-up client to restart the authentication process so that the dial-up user can be authenticated by the organization IAS server through the tunnel. The dial-up client sends the authentication information to the ISP NAS, which encapsulates the user account credentials and sends them through the tunnel to the organization tunnel server.
After the user account credentials are received by the tunnel server, the tunnel server sends an Access-Request message to the organization IAS server. The organization IAS server authenticates and authorizes the connection of the dial-up client to the tunnel server and sends an Access-Accept message to the tunnel server. The tunnel server then completes the connection to the dial-up client.
All data that is sent by the dial-up client is automatically sent through the tunnel to the tunnel server by the ISP NAS.
This configuration is known as compulsory tunneling because the client is compelled to use the tunnel created by the FEP. After the initial connection is made, all network traffic to and from the client is automatically sent through the tunnel.
In addition, you can configure IAS to instruct an FEP to tunnel different remote access clients to different tunnel servers.
Unlike the separate tunnels created for each voluntary client, a compulsory tunnel between the FEP and organization VPN server can be shared by multiple dial-up clients. When a second client dials into the same access server (the FEP) to reach a destination for which a tunnel already exists, the data traffic for the new client is carried over the existing VPN tunnel.
The following RADIUS Attributes are used to carry the tunneling information from the IAS server to the NAS.
Used in authorization only:
-
Tunnel-Pvt-Group-ID
-
Tunnel-Assignment-ID
-
Tunnel-Preference
-
Tunnel-Password
-
Tunnel-Client-Auth-ID
-
Tunnel-Server-Auth-ID
Used in authorization and accounting:
-
Tunnel-Type (such as PPTP and L2TP)
-
Tunnel-Medium-Type (X.25, ATM, Frame Relay, IP, and so on)
-
Tunnel-Client-Endpt
-
Tunnel-Server-Endpt
Used for accounting only:
The Windows Server 2003 or Windows 2000 Routing and Remote Access service cannot be used as a FEP for compulsory tunneling.
The RADIUS protocol supports several PPP authentication protocols, each with its advantages and disadvantages in terms of security, usability, and breadth of support. The protocol used is determined by the configuration of the NAS device. For information about support for the various authentication protocols, see your NAS documentation, or consult your ISP or other vendor support.
The following sections focus on the advantages and disadvantages of the authentication protocols that are currently supported by IAS, and discuss how to deploy certificate-based authentication methods. The information is also useful in configuring a particular authentication method for network access authentication.
PAP
Password Authentication Protocol (PAP) sends a password as a plaintext string from the user’s computer to the NAS device. When the NAS forwards the password as the User-Password attribute in the Access-Request message, it is encrypted using the RADIUS shared secret as an encryption key. PAP is the most flexible protocol because passing a plaintext password to the authentication server enables that server to compare the password with nearly any storage format. For example, UNIX passwords are stored as one-way encrypted strings that cannot be decrypted. PAP passwords can be compared to these strings by reproducing the one-way encryption method.
Note
-
The use of this authentication method is not recommended for security reasons.
When PAP is enabled as an authentication protocol, user passwords are sent from a client to a NAS in plaintext form. The NAS encrypts the password, using the shared secret, and then sends it in an Access-Request message. Because a RADIUS proxy must encrypt the PAP password by using the shared secret of its forwarding RADIUS server, a RADIUS proxy must decrypt the PAP password, using the shared secret between the RADIUS proxy and the NAS. Because it uses a plaintext version of the password, PAP has a number of security vulnerabilities. Although the RADIUS protocol encrypts the password, it is transmitted as plaintext across the dial-up connection.
PAP-based authentication is enabled:
-
On the appropriate remote access policy. PAP is disabled by default.
-
On a remote access client.
-
As an authentication protocol on the remote access server. For the Routing and Remote Access service, PAP is disabled by default.
For information about a default setting on a particular NAS, see your NAS documentation.
If you set the User passwords to be changed at the next attempt to log on option, users must first log on using a local area network (LAN) connection and change the password before attempting to log on with a remote access connection using PAP. PAP and CHAP do not support changing passwords during the authentication process. If the password is not first changed using a LAN connection, logon fails. An alternative method for changing passwords is to temporarily log on using MS-CHAP to change the password.
Note
-
A malicious user or process on a RADIUS proxy can record user names and passwords for PAP connections while the password is in its plaintext form. For this reason, the use of PAP is highly discouraged, especially for virtual private network connections.
CHAP
Challenge Handshake Authentication Protocol (CHAP) is designed to address the concern of passing passwords in plaintext. By using CHAP, the NAS sends a random number challenge to the user’s computer. The challenge and the user’s password is then hashed by using Message Digest 5 (MD5). The client computer then sends the hash as a response to the NAS challenge and the NAS forwards both the challenge and response in the RADIUS Access-Request message.
When the authenticating server receives the RADIUS message, it uses the challenge and the user’s password to create its own version of the response. If the version of the server matches the response supplied by the user’s computer, the access request is accepted.
CHAP responses cannot be reused because NAS devices send a unique challenge each time a client computer connects to them. Because the algorithm for calculating CHAP responses is well known, it is very important that passwords be carefully chosen and sufficiently long. CHAP passwords that are common words or names are vulnerable to dictionary attacks if they can be discovered by comparing responses to the CHAP challenge with every entry in a dictionary. Passwords that are not sufficiently long can be discovered by brute force by comparing the CHAP response to sequential trials until a match to the user’s response is found.
Historically, CHAP is the most common dial-up authentication protocol used. When the server does not store the same password that was used to calculate the CHAP response, it cannot calculate an equivalent response. Because standard CHAP clients use the plaintext version of the password to create the CHAP challenge response, passwords must be stored in a reversibly encrypted form on the server to validate the CHAP response of the client.
Although the IAS server supports CHAP, a Windows NT 4.0–based domain controller cannot validate CHAP requests without support for storing reversibly encrypted passwords. This support is available in Windows Server 2003 and Windows 2000; in Windows NT 4.0, this support is available after installing Service Pack 3 or later.
CHAP-based authentication is enabled when you have done the following:
-
Enabled CHAP on the appropriate remote access policy. CHAP is disabled by default.
-
Enabled storage of a reversibly encrypted form of the user’s password. For a stand-alone server, use computer Group Policy to enable storage of reversibly encrypted passwords for all users of the computer. For Active Directory domains, Group Policy at the domain or Organizational Unit (OU) level can be used.
-
Forced a reset of users' passwords so that the new password is in a reversibly encrypted form. When you enable passwords to be stored in a reversibly encrypted form, the current passwords are in a non-reversibly encrypted form and are not automatically changed. You must either reset user passwords or set user passwords to be changed the next time you log on. After the password is changed, it is stored in a reversibly encrypted form.
-
Enabled CHAP on the remote access client.
-
Enabled CHAP as an authentication protocol on the remote access server. For the Routing and Remote Access service, CHAP is disabled by default.
For information about a default setting on a particular NAS, see your NAS documentation.
If you set the User passwords to be changed at the next attempt to log on option, the user must first log on using a LAN connection and change the password before attempting to log on with a remote access connection using CHAP. CHAP and PAP do not support changing passwords during the authentication process. If the password is not first changed by using a LAN connection, logon fails. As an alternative to this method for changing passwords, you can temporarily log on by using MS-CHAP to change the password.
MS-CHAP v1
MS-CHAP v1, is a variant of CHAP that does not require a plaintext version of the password on the authenticating server. In MS-CHAP v1, the challenge response is calculated with a hashed version of the password and the NAS challenge. This enables authentication over the Internet to a Windows Server 2003 or Windows 2000 domain controller.
Note
-
The use of this authentication method is not recommended for security reasons.
MS-CHAP v1 passwords are stored more securely at the server but have the same vulnerabilities to dictionary and brute force attacks as CHAP. If you deploy MS-CHAP v1, use strong passwords for all user accounts. A strong password includes the following:
-
Has at least seven characters.
-
Does not contain your user name, real name, or company name.
-
Does not contain a complete dictionary word.
-
Is significantly different from previous passwords. Incremental passwords (Password1, Password2, Password3) are not strong.
-
Contains characters from each of the following four groups: uppercase letters, lowercase letters, numerals, and symbols.
With MS-CHAP v1 in Windows Server 2003, users are prompted to change their password before it expires. In addition, you can configure MS-CHAP v1 on the profile of a remote access policy by using the Allow user to change password after it has expired check box to allow or prevent users from changing a password after it has expired.
Note
-
If the Allow user to change password after it has expired check box is disabled, and a user’s password expires, the user is denied access and cannot change the password to regain network access.
MS-CHAP v1 is enabled as follows :
-
As an authentication protocol on the remote access server. MS-CHAP v 1 is not enabled by default on the Routing and Remote Access service. For information about default settings on other NASs, see your NAS documentation.
-
On the appropriate remote access policy. MS-CHAP v 1 is not enabled by default.
-
On a remote access client.
By default, MS-CHAP v1 for Windows Server 2003 does not support LAN Manager authentication. This is a change from Windows 2000.
If you want to enable the use of LAN Manager authentication with MS-CHAP v1 for older operating systems, such as Microsoft Windows NT 3.5x and Microsoft Windows 95, you must change the value of Allow LM Authentication in the Windows registry.
MS-CHAP v2
Microsoft CHAP v2 (MS-CHAP v2) provides mutual authentication, stronger initial data encryption keys, and different encryption keys for sending and receiving. For VPN connections, Windows Server 2003 offers MS-CHAP v 2 before offering MS-CHAP v1. Windows clients that have been updated with the most recent service pack accept MS-CHAP v2 when it is offered.
MS-CHAP v2 is a one-way encrypted password, mutual authentication process that works as follows:
-
The remote access server sends a challenge to the remote access client that consists of a session identifier and an arbitrary challenge string.
-
The remote access client sends a response that contains the following:
-
The user name.
-
An arbitrary peer challenge string.
-
A one-way encryption of the received challenge string, the peer challenge string, the session identifier, and the user’s password.
-
The remote access server checks the response from the client and sends back a response containing:
-
An indication of the success or failure of the connection attempt.
-
An authenticated response based on the sent challenge string, the peer challenge string, the encrypted response of the client, and the encrypted version of the user’s password.
-
The remote access client verifies the authentication response and, if correct, uses the connection. If the authentication response is not correct, the remote access client terminates the connection.
-
If a user authenticates by using MS-CHAP v2 and attempts to use an expired password, MS-CHAP prompts the user to change the password while connecting to the server. Other authentication protocols do not support this feature effectively locking out the user who used the expired password. New to Windows Server 2003 is the ability to prevent a user from changing their expired password using MS-CHAP v2 by clearing the Allow user to change password after it has expired check box, under the Microsoft Encrypted Authentication version 2 (MS-CHAP v2) check box on the Authentication tab on the profile of a remote access policy.
MS-CHAP v2 is enabled as follows:
-
As an authentication protocol on the remote access server. MS-CHAP v2 is enabled by default on the Routing and Remote Access service. For information about default settings on other NASs, see your NAS documentation.
-
On the appropriate remote access policy. MS-CHAP v2 is enabled by default.
-
On the remote access client.
For information about default settings on other NASs, see your NAS documentation.
Note
-
Note Windows 95 and Microsoft Windows 98 support MS-CHAP v2 only for VPN connections. Windows 95 and Windows 98 do not support MS-CHAP v2 for dial-up connections.
EAP
Extensible Authentication Protocol (EAP) is an extension to PPP that works with dial-up, PPTP, L2TP, and 802.1X authentication clients. EAP allows the addition of new authentication methods known as EAP types. Both the access client and the remote access server must support the same EAP type for successful authentication to occur.
Windows Server 2003 includes an EAP infrastructure and two EAP types, EAP-MD5 and EAP-TLS. The IAS proxy implementation in Windows Server 2003 can forward EAP messages to a RADIUS server.
When you configure EAP as an authentication method in a remote access policy on an IAS server, NASs configured with RADIUS authentication can pass EAP messages from access clients to the IAS server or proxy using the RADIUS protocol. The EAP messages sent between the access client and access server are encapsulated and formatted as RADIUS messages between the remote access server and the RADIUS server.
One advantage of EAP is that EAP types do not need to be installed at each NAS, only at the RADIUS server. In a typical use of EAP, a NAS is configured to use EAP and to use RADIUS as its authentication provider. When a connection is made, the network access client negotiates the use of EAP with the NAS. When the client sends an EAP message to the NAS, the NAS encapsulates the EAP message as a RADIUS message and sends it to its configured RADIUS server. The RADIUS server processes the message and sends a RADIUS-encapsulated EAP message back to the NAS. The NAS then forwards the EAP message to the network access client. In this configuration, the NAS is only a pass-through device. All processing of EAP messages occurs at the network access client and the RADIUS server.
EAP-MD5
MD5 with CHAP (EAP-MD5) is a required EAP type that uses the same challenge-handshake protocol as PPP-based CHAP, but the challenges and responses are sent as EAP messages. A typical use for EAP-MD5 is to authenticate the credentials of remote access clients by using user name and password security systems. You can use EAP-MD5 to test EAP interoperability.
EAP-TLS
Extensible Authentication Protocol-Transport Layer Security (EAP-TLS) is an EAP type that is used in certificate-based security environments. To use smart cards for remote access authentication, you can deploy a PKI with Certificate Services in Windows Server 2003 or with a non-Microsoft certification authority, and then configure the EAP-TLS authentication method in remote access policies on your IAS servers. The EAP-TLS exchange of messages provides mutual authentication, negotiation of the encryption method, and secured private key exchange between the network access client and the IAS server. EAP-TLS provides the strongest authentication and key exchange method.
Use of the TLS authentication type with EAP (EAP-TLS) reduces latency for authentication requests because, during the first authentication attempt, TLS caches certificate properties on the access client and on the IAS server. For subsequent authentication attempts, TLS uses the cached certificate properties instead of reading the certificate from the certificate store on the IAS server or client computer. In addition, the load on the IAS server that is typically caused by processing certificates is greatly reduced, improving server performance. Improved performance and response time are especially useful for wireless networks because the wireless client frequently authenticates to the network when moved from one AP to another.
EAP-TLS also supports computer authentication by using the computer account and computer certificate (also called machine certificate). For computer authentication with EAP-TLS, you must install a computer certificate on the client computer. This certificate is used to authenticate the client computer so that the computer can obtain network connectivity and Group Policy updates prior to user login. For user authentication with EAP-TLS, after a network connection is made and the user logs in, a user certificate on the client computer is necessary.
The computer certificate is installed on the IAS server computer so that during EAP-TLS authentication, the IAS server has a certificate to send to the client computer for mutual authentication, regardless of whether the client computer authenticates with a computer certificate or a user certificate. The computer and user-certificates that are submitted by the access client and the IAS server during EAP-TLS authentication, must conform to specified requirements. For more information about these requirements, see “Using Certificate-Based Authentication Methods” later in this section.
Deploying EAP
When you configure a remote access policy with EAP, if the EAP types displayed in the Select EAP Providers dialog box are selected as (Server-Configured), then the configuration of the EAP type is common to the server and applies to all remote access policies that are configured on the server. Otherwise, the EAP type configuration is specific to the current remote access policy, in which case the EAP method must be configured for each policy to which you want it to apply.
EAP-TLS is not supported if the IAS server is not a member of the domain.
Note
-
When deploying EAP-TLS, if you have issued a certificate to your IAS server that has a blank Subject, the certificate is not available to authenticate your IAS server. To change this, you can use Certificate Templates to create a new certificate for enrollment on your IAS server. In the Certificate Properties dialog box, on the Subject Name tab, in Subject name format, select a value other than None.
EAP-based authentication is enabled when:
-
EAP is enabled as an authentication protocol on the NAS.
-
EAP is enabled on the IAS server and configure the EAP type on the appropriate remote access policy.
-
EAP is enabled and configure EAP on a network access client.
In addition to the EAP types defined and supported in Windows Server 2003, you can include new EAP authentication methods by using the EAP Software Development Kit.
Note
-
When using EAP-TLS on computers running Microsoft Windows XP (prior to Service Pack 1), you must have user certificates stored on the computer for user authentication (instead of using smart cards). For computers running Windows Server 2003, Windows XP (Service Pack 1 or later), or Windows 2000, you can use either user certificates stored on the computer or a smart card for user authentication.
Protected Extensible Authentication Protocol
Protected Extensible Authentication Protocol (PEAP) is a new member of the family of EAP authentication methods. PEAP uses TLS to create an encrypted channel between an authenticating PEAP client, such as a wireless client computer, and a PEAP authenticator, such as an IAS server. PEAP does not specify an authentication type. Instead, it provides additional security for other EAP authentication protocols, such as MS-CHAP v2, which can operate within the TLS encrypted channel provided by PEAP. PEAP is used as an authentication method for 802.1X connections, such as 802.11 wireless connections and authenticating switches. PEAP is not supported for VPN or other remote access clients.
To use PEAP in your environment with Windows Server 2003, IAS requires little infrastructure change. If your wireless APs support 802.1X and EAP, then you can use PEAP without any additional changes to your wireless APs. In addition, PEAP is available for Microsoft Windows XP, Service Pack 1 and later; Windows 2000, Service Pack 4; and Pocket PC 2002.
To enhance both the EAP protocols and network security, PEAP provides the following:
-
Protection for the EAP method negotiation that occurs between client and server through a TLS channel. This helps prevent an attacker from injecting packets between the client and the NAS to cause the negotiation of a less secure EAP method. The encrypted TLS channel also helps prevent denial of service attacks against the IAS server.
-
Support for the fragmentation and reassembly of messages, allowing the use of EAP types that do not provide this.
-
Wireless clients that can authenticate the IAS or RADIUS server. Because the server also authenticates the client, mutual authentication occurs.
-
Support for computer authentication using the client computer account and computer certificate.
-
Protection against the deployment of an unauthorized wireless access point (AP) or authenticating switch when the EAP client authenticates the certificate provided by the IAS server. In addition, the TLS master secret created by the PEAP authenticator and client is not shared with the access server. Because of this, the access server cannot decrypt the messages protected by PEAP.
-
PEAP fast reconnect reduces the latency for authentication requests and responses between an access client and the IAS server when the access client moves from one AP to another. The IAS server caches a portion of the access client credentials and uses these cached credentials to quickly authorize the connection after the client is associated with a new AP.
PEAP authentication process
There are two stages in the PEAP authentication process. The first stage establishes a secure channel between the EAP client and the IAS server. The second stage provides EAP authentication between the EAP client and the IAS server. The following example describes the two stage PEAP authentication process for wireless clients:
TLS encrypted channel
The wireless client associates with a wireless access point. An IEEE 802.11-based association provides an Open System or Shared Key Authentication before a secure association is created between the client and AP. If Shared Key Authentication is configured for use at the AP, then 802.1X is not used. If the AP is configured for Open System Authentication, 802.1X is used. After the IEEE 802.11-based association is successfully established between the client and AP, the TLS session is negotiated with the RADIUS server. The key that is derived during this negotiation is used to encrypt all subsequent communication, and the TLS encrypted channel is established.
EAP-authenticated communication
An EAP negotiation in which the wireless client authenticates to the IAS server occurs through the TLS channel. The IAS server authenticates the user or client computer with the method that is determined by the EAP type and selected for use within PEAP (either TLS or MS-CHAP v2). The AP only forwards messages between wireless client and RADIUS server. The AP (or person monitoring the AP) cannot decrypt these messages because the AP is not the TLS endpoint.MS-CHAP v2
802.11 wireless deployments using PEAP
You can choose either of these two EAP types to use with PEAP: MS-CHAP v2 or TLS.
MS-CHAP v2 uses credentials (user name and password) for user authentication, and a certificate in the server computer certificate store for server authentication.
TLS uses either certificates installed in the client computer certificate store or a smart card to authenticate a user or client computer, and a certificate in the server computer certificate store for server authentication.
PEAP with EAP-MS-CHAP v2
PEAP with EAP-MS-CHAP v2 (PEAP-MS-CHAP v2) is easier to deploy than EAP-TLS because user authentication is accomplished with password-based credentials (user name and password) instead of certificates or smart cards. Only the IAS server is required to have a certificate. Additionally, the server certificate can be issued by a public CA that is trusted by the client computer, meaning that the public CA certificate already exists in the Trusted Root Certification Authority folder in the client computer certificate store. In this case, the server certificate is not downloaded and added to the client trusted root certificate store, and the user is not prompted to make a decision about whether to trust the server.
PEAP-MS-CHAP v2 provides improved security over MS-CHAP v2 by using mutual authentication, by preventing an unauthorized server from negotiating the least secure authentication method, and by providing key generation with TLS. PEAP-MS-CHAP v2 requires that the client trust the certificates provided by the server.
PEAP with EAP-TLS
Public Key certificates provide a much stronger authentication method than those that use password-based credentials. PEAP with EAP-TLS (PEAP-TLS) uses certificates for server authentication and either certificates or smart cards for user and client computer authentication. To use PEAP-TLS, you must deploy a PKI.
PEAP-TLS also supports computer authentication by using the computer account and computer certificate. For computer authentication with PEAP-TLS, you must install a computer certificate, on the client computer. A computer certificate installed on the client computer is used to authenticate the client computer so that the computer can obtain network connectivity and Group Policy updates prior to user login. For user authentication with PEAP-TLS after a network connection is made and the user logs in, a user certificate on the client computer is necessary.
The computer certificate is installed on the IAS server computer so that during PEAP-TLS authentication, the IAS server has a certificate to send to the client computer for mutual authentication, regardless of whether the client computer authenticates with a computer certificate or a user certificate. The computer and user certificates submitted by the access client and the IAS server during PEAP-TLS authentication must conform to the requirements specified in the section "Using Certificate-Based Authentication Methods."
TLS caching of certificate properties
When using the PEAP-TLS and EAP-TLS authentication methods with certificates, TLS uses cached certificate properties instead of reading the certificate from the certificate store on the IAS server.
If a certificate is deleted, its TLS credential cache entry is removed. If a certificate is renewed, its TLS credential cache entry is removed and TLS caches properties for the renewed certificate. If a certificate expires, TLS continues using outdated cached certificate information until the cache expires or is refreshed.
If you change or replace a certificate, you can refresh the TLS cache by restarting the server computer.
Deploying PEAP
PEAP is used as an authentication method for 802.1X-based connections and is not supported for VPN or other remote access clients. When you deploy both PEAP and EAP unprotected by PEAP on the same network, do not use the same authentication type (MS-CHAP v2 or TLS) with and without PEAP.
For example, if you deploy PEAP with EAP-TLS (PEAP-TLS), do not also deploy EAP-TLS without PEAP. Deploying authentication methods with the same type — one with and the other without external protection of PEAP or IPSec — creates a security issue.
If you need to deploy both PEAP and EAP on the same network, use different authentication types for each deployment.
An example of a deployment of both authentication methods that does not introduce a security issue is to use PEAP-MS-CHAP v2 for secure password authentication for wireless clients while using EAP-TLS as the authentication method for remote access clients using VPN connections. For this configuration, two remote access policies are configured at the IAS server, one for wireless clients and one for VPN clients, with the appropriate authentication method and type configured for each policy.
PEAP fast reconnect
PEAP fast reconnect enables wireless clients to move between wireless APs on the same network without being reauthenticated each time they associate with a new AP.
Wireless APs are configured as RADIUS clients to RADIUS servers. If a wireless client roams between APs that are configured as clients of the same RADIUS server, the client is not required to be authenticated with each new association.
PEAP fast reconnect reduces the response time for authentication between client and authenticator because the authentication request is forwarded from the new server to the original server. Because both the PEAP client and authenticator use previously cached TLS connection properties (the collection of which is named the TLS handle), the authenticator can quickly determine that the client connection is a reconnect.
If the original PEAP authenticator is unavailable, full authentication must occur between the client and the new authenticator. The TLS handle of the new PEAP authenticator is cached by the client. The client can cache TLS handles for multiple PEAP authenticators. For smart cards or PEAP-MS-CHAP v2 authentication, the user is asked to supply the personal identification number (PIN) or credentials, respectively.
The following tables, PEAP-MS-CHAP v2 and “PEAP-TLS” describe authentication behavior with PEAP-MS-CHAP v2 and PEAP-TLS when wireless clients roam from one wireless AP to a new AP. PEAP authentication behavior differs when wireless APs are configured either as clients to the same RADIUS server or as clients to different RADIUS servers.
PEAP-MS-CHAP v2
|
New AP Is RADIUS Client to Same RADIUS Server
|
New AP Is RADIUS Client to a Different RADIUS Server
|
|
The user is not prompted for credentials each time the client computer associates with a new AP.
|
The user is prompted for credentials on this initial association. By default, the domain credentials of the currently logged in user are sent automatically. The next time the client computer associates with an AP that is a client to this server, user credentials are not required.
|
|
The RADIUS server is not required to provide a certificate.
|
The RADIUS server provides a certificate on this initial association so that the wireless client can authenticate to the RADIUS server. The next time the client computer associates with an AP that is a client to this server, the server is not required to be reauthenticated.
|
PEAP-TLS
|
New AP is RADIUS Client to Same RADIUS Server
|
New AP Is RADIUS Client to a Different RADIUS Server
|
|
The wireless client and RADIUS server are not required to exchange certificates.
|
The wireless client and RADIUS server exchange certificates on this initial association. The next time the client computer associates with an AP that is a RADIUS client to this RADIUS server, certificates are not exchanged.
|
|
The user is not prompted for a smart card PIN each time the wireless client computer associates with a new AP.
|
The user is prompted for a smart card PIN on this initial association. The next time the client computer associates with an AP that is a RADIUS client to this RADIUS server, the user is not prompted for the PIN. When the TLS cache expires, the user is prompted to reenter the PIN.
|
PEAP fast reconnect requires the following:
-
Both the PEAP client (802.11 wireless client) and PEAP authenticator (RADIUS server) must have fast reconnect enabled.
-
All APs to which the PEAP client roams must be configured as RADIUS clients to a RADIUS server (the PEAP authenticator) for which PEAP is configured as the authentication method for wireless connections.
-
All APs to which the PEAP client associates must be configured to prefer the same RADIUS server (PEAP authenticator) in order to avoid being prompted for credentials from every RADIUS server. If the AP cannot be configured to prefer a RADIUS server, you can configure an IAS RADIUS proxy with a preferred RADIUS server.
PEAP-based authentication is enabled when:
-
You install a certificate on the IAS server that meets the minimum server certificate requirements.
-
You configure PEAP as the authentication method for a remote access policy on your IAS server.
-
You enable and configure PEAP on a remote access client.
Certificates are used for network access authentication because they provide strong security for authenticating users and computers and eliminate the need for less secure password-based authentication methods.
Two authentication methods use certificates: EAP-TLS and PEAP. Both methods always use certificates for server authentication, and both the remote access client and the IAS server are TLS endpoints during the authentication process, providing strong security.
Depending on the authentication type configured with the authentication method, certificates might also be used for user authentication and client computer authentication. Certificates are stored in one of two places: either in the computer or user certificate store, or embedded in a computer chip on a smart card. If you deploy EAP-TLS, user authentication can be performed by using either a smart card or a certificate in the Current User certificate store on the client computer. If you deploy PEAP-TLS, the same is true. If you deploy PEAP-MS-CHAP v2, certificates are not used on the client computer; the user supplies password-based credentials when logging on to the network.
You can configure remote access policies in IAS to use EAP-TLS to perform certificate-based authentication for many types of network access, including VPN and wireless connections. In addition, you can use PEAP as the authentication method for wireless or wired 802.1X connections.
To deploy certificates to users and computers, you must first do the following:
-
Design a PKI.
-
Deploy a CA by using Certificate Services or by using a third party certification authority.
-
Design and publish one or more certificates using Certificate Templates, which are installed with Certificate Services.
-
Select one or more distribution methods (also called certificate enrollment methods) to install the certificates on computers and distribute them to users.
If you deploy PEAP-MS-CHAP v2 for wireless clients, you can use a non-Microsoft CA to obtain a server certificate.
Because user authentication is accomplished with password-based credentials with PEAP-MS-CHAP v2, you do not need to deploy a PKI or a CA on your network, however the server certificate you obtain from a third party CA must conform to server certificate requirements.
All certificates that are used for network access authentication must meet the requirements for X.509 certificates and work for connections that use Secure Sockets Layer-Transport Layer Security (SSL/TLS). After this minimum requirement is met, both server and client certificates must meet the additional requirements listed in the “Server certificate requirements” section.
Server certificate requirements
You can configure clients to validate server certificates by using the Validate server certificate option. Using PEAP-MS-CHAP v2, PEAP-TLS, or EAP-TLS as the authentication method, the client accepts the authentication attempt of the server when the certificate meets the following requirements:
-
The Subject name contains a value. If you issue a certificate to your IAS server that has a blank Subject, the certificate is not available to authenticate your IAS server.
-
The computer certificate on the server chains to a trusted root CA and does not fail any of the checks that are performed by CryptoAPI.
-
The IAS or VPN server computer certificate is configured with the Server Authentication purpose in Enhanced Key Usage (EKU) extensions (the object identifier for Server Authentication is 1.3.6.1.5.5.7.3.1).
-
The server certificate is configured with a required cryptographic service provider (CSP) value of Microsoft RSA SChannel Cryptographic Provider.
-
The Subject Alternative Name (SubjectAltName) extension, if used, must contain the DNS name of the server. With PEAP and EAP-TLS, servers display a list of all installed certificates in the computer’s certificate store, with the following exceptions:
-
Certificates that do not contain the Server Authentication purpose in EKU extensions are not displayed.
-
Certificates that do not contain a Subject name are not displayed.
-
Servers do not display registry-based and smart card-logon certificates.
Note
-
You can designate which certificate is used by the IAS or VPN server in the remote access policy profile. When you configure EAP and PEAP authentication methods that require a certificate for server authentication and you do not select a specific certificate in the Smart Card or other Certificate Properties dialog box, the IAS or VPN server automatically selects a certificate from the computer certificate store. If the server obtains a newer certificate, it uses the new certificate. This might cause the IAS or VPN server to use a certificate that is not correctly configured for authentication, causing authentication to fail as the result. To prevent this, always manually select a server certificate when configuring PEAP and EAP authentication methods that require one.
Client certificate requirements
With EAP-TLS or PEAP-TLS, the server accepts the client authentication attempt when the certificate meets the following requirements:
-
The client certificate is issued by an enterprise CA or mapped to a user or computer account in Active Directory.
-
The user or computer certificate on the client chains to a trusted root CA, includes the Client Authentication purpose in EKU extensions (the object identifier for Client Authentication is 1.3.6.1.5.5.7.3.2), and fails neither the checks that are performed by CryptoAPI and specified in the remote access policy nor the Certificate object identifier checks that are specified in IAS remote access policy.
-
The 802.1X client does not use registry-based certificates that are either smart card-logon or password-protected certificates.
-
For user certificates, the Subject Alternative Name (SubjectAltName) extension in the certificate contains the user principal name (UPN).
-
For computer certificates, the Subject Alternative Name (SubjectAltName) extension in the certificate must contain the fully qualified domain name (FQDN) of the client, which is also called the DNS name. With PEAP-TLS and EAP-TLS, clients display a list of all installed certificates in the Certificates snap-in with the following exceptions:
-
Wireless clients do not display registry-based and smart card-logon certificates.
-
Wireless clients and VPN clients do not display password-protected certificates.
-
Certificates that do not contain the Client Authentication purpose in EKU extensions are not displayed.
Certificate enrollment methods and domain membership
The domain membership of computers for which you want to enroll certificates affects the certificate enrollment method that you can choose. Certificates for domain member computers can be enrolled automatically (also known as auto-enrollment), while an administrator must enroll certificates for non-domain member computers using the CA Web enrollment pages or a floppy disk. The certificate enrollment method for non-domain member computers is known as a trust bootstrap process, through which certificates are created and then manually requested or distributed securely by administrators, providing administrators the ability to build common trust between non-domain member computers and the domain.
Domain member certificate enrollment
If your VPN server, IAS server, or client running Windows 2000 or Windows XP Professional is a member of a domain running Windows Server 2003 and Active Directory, you can configure the auto-enrollment of computer and user certificates through Group Policy. After auto-enrollment is configured and enabled, all domain member computers receive computer certificates the next time Group Policy is refreshed whether the refresh is triggered manually by using the gpupdate command or automatically when logging on to the domain.
If your computer is a member of a domain where Active Directory is not installed, you can install computer certificates manually by requesting them through the Certificates snap-in.
Note
-
Computers running Windows 2000 can auto-enroll computer certificates only.
Non-domain member certificate enrollment
Certificate enrollment for computers that are not domain members cannot be done with auto-enrollment. When a computer is joined to a domain, a trust is established that allows auto-enrollment to occur without administrator intervention. When a computer is not joined to a domain, trust is not established and a certificate is not issued. Trust must be established using one of the following methods:
-
An administrator (who is, by definition, trusted) must request a computer or user certificate by using the CA Web enrollment tool.
-
An administrator must save a computer or user certificate to a floppy disk and install it on the non-domain member computer. Or, when the computer is not accessible to the administrator (for example, a home computer connecting to an organization network by means of an L2TP/IPSec VPN connection), a domain user whom the administrator trusts can install the certificate.
-
An administrator can distribute a user certificate on a smart card issued to a user (computer certificates are not distributed on smart cards).
Many network infrastructures contain VPN and IAS servers that are not domain members. For example, a VPN server in a perimeter network might not be a domain member for security purposes. In this case, a computer certificate with the Server Authentication purpose contained in the EKU extensions must be installed on the non-domain member VPN server before it can successfully negotiate L2TP/IPSec-based VPN connections with clients. Note that if the non-domain member VPN server is used as an endpoint for a VPN connection with another VPN server, EKU extensions must contain both the Server Authentication and Client Authentication purposes.
Choosing certificate templates and certificate enrollment methods
Windows Server 2003, Standard Edition, provides different certificate templates from Windows Server 2003, Enterprise Edition; and Windows Server 2003, Datacenter Edition.
If you run an enterprise CA on a computer that runs Windows Server 2003, Standard Edition, you can use the following table, “Certificate Use for Windows Server 2003, Standard Edition,” to determine the following: the best certificate templates; the certificate purposes configured in the EKU extensions of the certificate; and the certificate enrollment method for your requirements.
Certificate Use for Windows Server 2003, Standard Edition
|
Object and Domain Membership
|
Certificate Template
|
Certificate Purposes
|
Preferred Certificate Enrollment Method
|
Alternate Certificate Enrollment Method
|
|
VPN or IAS server, domain member
|
Computer
|
Server Authentication
|
Auto-enrollment
|
Request a certificate with the Certificates snap-in
|
|
VPN server with site-to-site connection, domain member
|
Computer
|
Server Authentication and Client Authentication
|
Auto-enrollment
|
Request a certificate with the Certificates snap-in
|
|
Windows XP Professional client, domain member
|
Computer
|
Client Authentication
|
Auto-enrollment
|
Request a certificate with the Certificates snap-in
|
|
VPN or IAS server, non-domain member
|
Computer
|
Server Authentication
|
CA Web enrollment tool
|
Install from a floppy disk
|
|
VPN server with site-to-site connection, non-domain member
|
Computer
|
Server Authentication and Client Authentication
|
CA Web enrollment tool
|
Install from a floppy disk
|
|
Windows XP Professional client, non-domain member
|
Computer
|
Client Authentication
|
CA Web enrollment tool
|
Install from a floppy disk
|
|
User, domain user
|
User
|
Client Authentication
|
Auto-enrollment
|
Use a smart card or the CA Web enrollment tool
|
If you are running an enterprise CA on a computer running Windows Server 2003, Enterprise Edition; Windows Server 2003, Datacenter Edition; the 64-bit of Windows Server 2003, Enterprise Edition; or the 64-bit version of Windows Server 2003, Datacenter Edition, you can use the following table, “Certificate Use for Other Versions of Windows Server 2003,” to determine the best certificate templates and certificate enrollment method for your requirements.
Certificate Use for Other Versions of Windows Server 2003
|
Object and Domain Membership
|
Certificate Template
|
Certificate Purpose
|
Preferred Certificate Enrollment Method
|
Alternate Certificate Enrollment Method
|
|
VPN or IAS server, domain member
|
RAS and IAS Server
|
Server Authentication
|
Auto-enrollment
|
Request a certificate with the Certificates snap-in
|
|
Windows XP Professional client, domain member
|
Workstation Authentication
|
Client Authentication
|
Auto-enrollment
|
Request a certificate with the Certificates snap-in
|
|
VPN or IAS server, non-domain member
|
RAS and IAS Server
|
Server Authentication
|
CA Web enrollment tool
|
Install from a floppy disk
|
|
Windows XP Professional client, non-domain member
|
Workstation Authentication
|
Client Authentication
|
CA Web enrollment tool
|
Install from a floppy disk
|
Note
-
If your server running IAS is not a domain controller but is a member of a domain with a Windows 2000 mixed functional level, you must add the server to the access control list (ACL) of the RAS and IAS Server certificate template. You must also configure the correct permissions for auto-enrollment. There are different procedures for adding single servers and groups of servers to the ACL.
You can add an individual server to the ACL for the RAS and IAS Server certificate template by doing the following: In Certificate Templates, select the template RAS and IAS Server, and then add the IAS server to the template Security properties. For more information, see "To allow subjects to request a certificate that is based on the template" in Help and Support Center for Windows Server 2003. After you add your IAS server to the ACL, grant Read, Enroll, and Auto-enroll permissions.
You can manage a group of servers by adding the servers to a new global or universal group, and then adding the group to the ACL of the certificate template:
-
In Active Directory Users and Computers, create a new global or universal group for IAS servers. Next, add to the group all computers that are IAS servers, are not domain controllers, and are members of a domain with a Windows 2000 mixed functional level.
-
In Certificate Templates, select the RAS and IAS Server template, and then add the group you created to the template Security properties by completing the steps.
After you add your new group, grant Read, Enroll, and Auto-enroll permissions.
The unauthenticated access method allows remote access users to log on without having their identity verified by IAS. For example, IAS does not verify the user’s name and password against an account in Active Directory when unauthenticated access is enabled. The only user validation performed with the unauthenticated access method is authorization to verify that the user has permission to access the network.
Enabling unauthenticated access presents security risks that you must carefully consider to determine whether to enable this authentication method. Three scenarios that illustrate typical security considerations for unauthenticated access include the following: Guest Access; Dialed Number Identification Service (DNIS) authorization; and Automatic Number Identification/Calling Line Identification (ANI/CLI) authorization.
Guest access for PPP users
Guest access is the ability to log on to a domain without a user name or a password. Both Routing and Remote Access service and IAS must be configured to support unauthenticated access.
When a remote access server receives a connection attempt, it negotiates with the user different authentication types enabled at the server. If the client accepts one of them, it sends the appropriate credentials for the accepted authentication type. It the user refuses authentication, Routing and Remote Access service checks its properties to verify if unauthenticated access is enabled and, if enabled, forwards the Access-Request message to IAS. This Access-Request message does not contain a User-Name attribute or any other credentials.
When IAS receives the message without a User-Name attribute, it processes the message as an attempt by the user to connect as guest. In this case, IAS uses the name of the guest account in a domain as the user identity. IAS then evaluates policies to determine the right profile. If a match is found, and unauthenticated access is enabled in the profile, other authorizations are validated, and an Access-Accept message is returned. The accounting log file logs the user identity and authentication type, which can be used to determine whether the user is logged on with guest access.
You can enable Guest access by doing the following:
-
Enable unauthenticated access on the remote access server.
-
Enable unauthenticated access on the appropriate remote access policy.
-
Enable the Guest account.
-
Set the remote access permission on the Guest account to either Allow access or Control access through Remote Access Policy depending on your remote access policy administrative model.
If you do not want to enable the Guest account, do the following:
-
Create a user account and set the remote access permission to either Allow access or Control access through Remote Access Policy.
-
Set the Default User Identity registry value on the authenticating server (either the remote access server or the IAS server) to the name of the account.
Following is an example of guest access.
-
During PPP negotiation, the dial-in client rejects all of the PPP authentication protocols of the NAS.
-
If the NAS is configured to allowed unauthenticated access, the NAS sends an Access-Request message without the User-Name attribute and without a password. For the Windows Server 2003 or Windows 2000 Routing and Remote Access service, unauthenticated access is enabled from the Authentication tab on the properties of a server in the Routing and Remote Access snap-in.
-
Because the User-Name attribute is not included in the Access-Request message and by default the IAS user identity is using the User-Name attribute, the user identity is set to Guest (or the value of Default User Identity).
-
With the user identity of Guest and an unauthenticated connection attempt, The authentication and authorization process as discussed earlier in the chapter is performed. If the connection attempt matches a policy whose profile settings have unauthenticated access enabled and the Guest account is enabled and has the appropriate remote access permission, IAS sends an Access-Accept message to the NAS.
DNIS authorization
Dialed Number Identification Service (DNIS) authorization is the authorization of a connection attempt based on the number called. This attribute is referred to as Called-Station-ID. DNIS is used by standard telecommunication companies. This service returns the number called to the called party. Based on the Called-Station-ID attribute, IAS can deliver different services to dial-up/remote access users.
You can enable DNIS authorization by doing the following tasks:
-
Enable unauthenticated access on the remote access server.
-
Create a remote access policy on the authenticating server (remote access server or IAS server) for DNIS-based authorization with the Called-Station-ID condition set to the phone number.
-
Enable unauthenticated access on the remote access policy for DNIS-based authorization.
ANI authorization
ANI authorization is based on the number the user calls from. This attribute is referred to as Calling Station ID, or Caller ID. Based on the Calling-Station-ID attribute, IAS can deliver different services to dial-up/remote access users.
Using ANI authorization is different from using the Caller ID dial-in property of a user account. ANI authorization is performed when the user does not type any user name or password, and refuses to use any valid authentication method. In this case, IAS receives Calling-Station-ID, and no user name and password. To support ANI authorization, the Active Directory must have user accounts with Caller IDs as user names. This kind of authentication is used with the cellular phone authentication and by ISPs in Germany and Japan.
When using the Caller ID property on a user account, the user types in credentials, such as a user name and password, and uses a valid authentication method to log on. IAS authenticates the user with the credentials, and then compares the Calling-Station-ID attribute in the Access-Request to the Caller ID property of the user account to authorize the connection attempt.
You can enable ANI authorization by doing the following:
-
Enable unauthenticated access on the remote access server.
-
Enable unauthenticated access on the appropriate remote access policy for ANI/CLI-based authentication.
-
Create a user account for each number calling, for which you want to provide ANI/CLI authorization. The name of the user account must match the number that the user is dialing from. For example, If a user is dialing in from 555-0100, create a "5550100" user account.
-
Set the User Identity Attribute registry value to 31 on the authenticating server.
To always use the calling number as the user identity, set the Override User-Name registry value to 1 on the authenticating server.
Note
-
If the Override User-Name is set to 1 and the User Identity Attribute is set to 31, then the authenticating server can perform only ANI/CLI-based authentication. The result is that normal authentication using authentication protocols such as MS CHAP v1, CHAP, and EAP is disabled.
ANI example
The following example explains how ANI/CLI authorization works for a dial-up client with an existing user account called 5550100. The client is dialing in from the phone number 555-0100.
-
During PPP negotiation, the dial-in client rejects all of the PPP authentication protocols of the NAS.
-
If the NAS is configured to allow unauthenticated access, the NAS sends an Access-Request message without the User-Name attribute and without a password. (For the Routing and Remote Access service, you can enable unauthenticated access on the Authentication tab in the properties of a server in the Routing and Remote Access snap-in.)
-
Because the User-Name attribute is not included in the Access-Request message, and the IAS user identity is set to use the Calling-Station-ID attribute, the user identity is set to 5550100.
-
With the user identity of 5550100 and an unauthenticated connection attempt, the authentication and authorization process is performed. IAS sends an Access-Accept message to the NAS if the following two conditions are met:
-
The connection attempt matches a policy with a profile setting of unauthenticated access enabled.
-
The 550100 account has the appropriate remote access permissions.
MAC address authorization
Media Access Control (MAC) address authorization functions in the same way as ANI authorization, but it is used for wireless clients and clients connecting to your network by using an 802.1X authenticating switch.
MAC address authorization is based on the MAC address of the network adapter installed in the user’s client computer. Like ANI authorization, MAC address authorization uses the Calling-Station-ID attribute instead of user name and password or certificate-based credentials to identify the user during the connection attempt.
MAC address authorization is performed when the user does not type in any user name or password, and refuses to use any valid authentication method. In this case, IAS receives Calling-Station-ID, and no user name and password. To support MAC address authorization, the Active Directory must have user accounts with MAC addresses as user names.
MAC address authorization is enabled when you do the following:
-
Enable MAC address authorization on access servers (such as wireless APs).
-
Enable unauthenticated access on the appropriate remote access policy for MAC address-based authentication, and enable PAP.
-
Create a user account for each MAC address for which you want to provide MAC address authorization. The name of the user account must match the MAC address of the network adapter installed in the computer from which the user is connecting. The format of the password assigned to the account is determined by the network access server vendor. Review the network access server documentation to determine the appropriate password.
-
Set the User Identity Attribute registry value to 31 on the authenticating server. This registry value location is: HKLM\SYSTEM\CurrentControlSet\Services\RemoteAccess\Policy
-
The User Identity Attribute registry DWORD decimal value: 31
-
To always use the MAC address as the user identity, set the Override User-Name registry value to 1 on the IAS server. This registry value location is: HKLM\SYSTEM\CurrentControlSet\Services\RemoteAccess\Policy
-
The Override User-Name registry DWORD decimal value: 1
Caution |
|
Incorrectly editing the registry can severely damage your system. Before making changes to the registry, you should back up any valued data on the computer. |