There are a number of basic ways an automated network access control (NAC) system can identify unauthorized users and keep them from accessing your network. However, there are pros and cons to using these different NAC methods. This article will discuss both the basic network access control principles and the different trade-offs each brings to the table, as well as explore some additional NAC considerations. Geared toward the Internet service provider, hotel operator, library, or other public portal operator who provides Internet service and wishes to control access, this discussion will give you some insight into what method might be best for your network.
The NAC Strategies
MAC addresses are unique to every computer connected to the network, and thus many NAC systems use them to grant or deny access. Since MAC addresses are unique, NAC systems can use them to identify an individual customer and grant them access.
While they can be effective, there are limitations to using MAC addresses for network access. For example, if a customer switches to a new computer in the system, it will not recognize them, as their MAC address will have changed. As a result, for mobile customer bases, MAC address authentication by itself is not viable.
Furthermore, on larger networks with centralized authentication, MAC addresses do not propagate beyond one network hop, hence MAC address authentication can only be done on smaller networks (no hops across routers). A work-around for this limit would be to use a distributed set of authentication points local to each segment. This would involve multiple NAC devices, which would automatically raise complexity with regard to synchronization. Your entire authentication database would need to be replicated on each NAC.
Finally, a common question when it comes to MAC addresses is whether or not they can be spoofed. In short, yes, they can, but it does require some sophistication and it is unlikely a normal user with the ability to do so would go through all the trouble to avoid paying an access charge. That is not to say it won’t happen, but rather that the risk of losing revenue is not worth the cost of combating the determined isolated user.
I mention this because some vendors will sell you features to combat spoofing and most likely it is not worth the incremental cost. If your authentication is set up by MAC address, the spoofer would have to also have the MAC address of a paying user in order to get in. Since there is no real pattern to MAC addresses, guessing another customer’s MAC address would be nearly impossible without inside knowledge.
IP addresses allow a bit more flexibility than MAC addresses because IP addresses can span across a network segment separated by a router to a central location. Again, while this strategy can be effective, IP address authentication has the same issue as MAC addressing, as it does not allow a customer to switch computers, thus requiring that the customer use the same computer each time they log in. In theory, a customer could change the IP address should they switch computers, but this would be way too much of an administrative headache to explain when operating a consumer-based network.
In addition, IP addresses are easy to spoof and relatively easy to guess should a user be trying to steal another user’s identity. But, should two users log on with the same IP address at the same time, the ruse can quickly be tracked down. So, while plausible, it is a risky thing to do.
User ID Combined with MAC Address or IP Address
This methodology solves the portability issue found when using MAC addresses and IP addresses by themselves. With this strategy, the user authenticates their session with a user ID and password and the NAC module records their IP or MAC address for the duration of the session.
For a mobile consumer base, this is really the only practical way to enforce network access control. However, there is a caveat with this method. The NAC controller must expire a user session when there is a lack of activity. You can’t expect users to always log out from their network connection, so the session server (NAC) must take an educated guess as to when they are done. The ramification is that they must log back in again. This usually isn’t a major problem, but can simply be a hassle for users.
The good news is the inactivity timer can be extended to hours or even days, and should a customer login in on a different computer while current on a previous session, the NAC can sense this and terminate the old session automatically.
The authentication method currently used with the NetEqualizer is based on IP address and user ID/password, since it was designed for ISPs serving a transient customer base.
Other Important Considerations
NAC and Billing Systems
Many NAC solutions also integrate billing services. Overlooking the potential complexity and ballooning costs with a billing system has the potential to cut into efficiency and profits for both customer and vendor. Our philosophy is that a flat rate and simple billing are best.
To name a few examples, different customers may want time of day billing; billing by day, hour, month, or year; automated refunds; billing by speed of connections; billing by type of property (geographic location); or tax codes. It can obviously go from a simple idea to a complicated one in a hurry. While there’s nothing wrong with these requests, history has shown that costs can increase exponentially when maintaining a system and trying to meet these varied demands, once you get beyond simple flat rate.
Another thing to look out for with billing is integration with a credit card processor. Back-end integration for credit card processing takes some time and energy to validate. For example, the most common credit card authentication system in the US, Authorize.net, does not work unless you also have a US bank account. You may be tempted to shop your credit card billing processor based on fees, but if you plan on doing automated integration with a NAC system, it is best to make sure the CC authorization company provides automated tools to integrate with the computer system and your consulting firm accounts for this integration work.
You cannot purchase and install a NAC system without some network analysis. Most NAC systems will re-direct unauthorized users to a Web page that allows them to sign up for the service. Although this seems relatively straight forward, there are some basic network features that need to be in place in order for this redirection to work correctly. The details involved go beyond the scope of this article, but you should expect to have a competent network administrator or consultant on hand in order to set this up correctly. To be safe, plan for eight to 40 hours of consulting time for troubleshooting and set-up above and beyond the cost of the equipment.
Network Access for Organizational Control
Thus far we have focused on the basic ways to restrict basic access to the Internet for a public provider. However, in a private or institutional environment where security and access to information are paramount, the NAC mission can change substantially. For example, in the Wikipedia article on network access control, a much broader mission is outlined than what a simple service provider would require. The article reads:
“Network Access Control aims to do exactly what the name implies—control access to a network with policies, including pre-admission endpoint security policy checks and post-admission controls over where users and devices can go on a network and what they can do.”
This paragraph was obviously written by a contributor that views NAC as a broad control technique reaching deep into a private network. Interestingly, there is an ongoing dispute on Wikipedia stating that this definition goes beyond the simpler idea of just granting access.
The rift on Wikipedia can be summarized as an argument over whether a NAC should be a simple gatekeeper for access to a network, with users having free rein to wander once in, or whether the NAC has responsibilities to protect various resources within the network once access is attained. Both camps are obviously correct, but it depends on the customer and type of business as to what type of NAC is required.
Therefore, in closing, the overarching message that emerges from this discussion is simply that implementing network access control requires an evaluation not only of the network setup, but also how the network will be used. Strategies that may work perfectly in certain circumstances can leave network administrators and users frustrated in other situations. However, with the right amount of foresight, network access control technologies can be implemented to facilitate the success of your network and the satisfaction of users rather than serving as an ongoing frustrating limitation.