The next step in the evolution of Wireless networking is 8011.N that should provide an improvement over G networks in speed and range. This specification is undergoing final ratification at the IEEE (Institute of Electrical and Electronics Engineers). In anticipation of the new spec, networking manufacturers such as Belkin, Linksys, and Netgear have come up with some new devices that are considered “draft pre-N”. There is no guarantee that these devices will be upgradeable to the new standard. Users may have to buy hardware complying with the N standard. On the other hand, the draft 1.0 status of these devices should in a perfect world mean that draft-n hardware could be upgradeable to the final specs via firmware updates.

After testing these devices however, PC Magazine’s Oliver Kaven discovered that the pre-N hardware’s performance was very disappointing considering the expectations surrounding the 8011.N specification. In many cases G network devices outperformed the pre-N devices. In addition, the pre-N hardware required numerous firmware upgrades to even make them functional. Another issue that he encountered is that cross vendor compatibility was a major issue. If you absolutely require the added capabilities that N wireless networks provide, buy equipment from the same manufacturers.

At its inception, WEP(“wired equivalent privacy”) had been maligned by the press for being a flawed wireless encryption solution. This was not helped by the fact that there was a less than impressive, well publicized demonstration of the then new standard. However, this has not stopped many users from using WEP as the "password" to their wireless network.

The reality of the state of WEP is that using existing technology, patient “war drivers”(wireless network crackers) can defeat its encryption. Data that is transmitted wirelessly is done so in embedded frames within packets. Under the WEP standard, each frame is encrypted, or hidden by a string of pseudorandom numbers called a keystream.

The problem is that the pseudorandom numbers are generated with a 24-bit random number seed called an initialization vector (IV). This number seed is transmitted with every encrypted frame unencrypted. Therefore, a cracker can look at two encrypted frames and know whether they were encrypted with the same IV. To make cracking the code more difficult, each frame is transmitted with a different IV. The problem is that there are only 16,777,216 different IV values. If you use a Wireless B network, these values can be exhausted after six hours since it transmits data at 11 megabytes per second, or even in less time if you are using a Wireless G network.

You would think that six hours would be a long time to crack a key, but WEP has a complication, “weak IVs”. By a mathematical quirk, 2% of those 16,777,216 IV values turn traitor and leak a little information about the data they encrypt.
As a result, password cracker utilities like AirSnort can be used to find enough “weak” IV values until it has enough to crack the WEP key.

So what is the alternative? The best solution to this problem is to use the more current WPA standard. WPA doubles the size of the initialization vector from 24 to 48 bits. Under WPA, a network can pass 281 trillion packets before the pool of unique IVs is exhausted. This would take about a thousand years. Furthermore, WPA changes keys often enough so that it is impossible for a cracker to gather enough “traitor packets” to reverse-engineer a key.