Wireless Security: WPA Vs WAPI, should we be
worried?
Chris Rose
Technology Research Institute of Florida, Inc.
Abstract. The IEEE 802.11 standard includes an optional encryption
capability, the Wired Equivalent Privacy (WEP) however, WEP is vulnerable to
attack and can easily be broken, therefore WEP was recently temporarily
supplemented with Wi-Fi Protected Access (WPA) which offers improvements
over WEP while a newer, more secure protocol 802.11i is developed. However,
China has now introduced a new security protocol called WLAN Authentication
and Privacy Infrastructure (WAPI) which is not a part of the IEEE 802.11
standard and is not interoperable with 802.11 which means that there will be
two incompatible wireless security standards which will be inconvenient and
more expensive for users. More importantly, because China is a large
developing market, WAPI is an attempt by China to exert itself internationally
as a major player in technological innovation by creating their own standard,
even though this has the possibility to cause immense economic harm to
established networking companies and disruption in the worldwide
technological infrastructure.
Introduction
Wireless networking has brought convenience to both the home and office
since it is now possible to abandon the wires which formally connected all the
computers in a network together. The problem with wireless is that unlike wired
networks where data travels over physical cables or fiber, the wireless LAN
broadcasts signals in the air and these can be intercepted by anyone within range.
Although most wireless LANs implement an encryption scheme, the authentication
protocols are also broadcast through the air and these can be intercepted by an
eavesdropper who can then generate a key and decrypt the data.
WEP – a Brief History
The IEEE 802.11 standard includes an optional encryption capability called
Wired Equivalent Policy (WEP), which uses the RSA RC4 security algorithm in the
media access controller (MAC). With WEP all the passwords are stored in both the
access points and on each computer on the network with the intention that WEP will
Rose C. (2004).
Wireless Security: WPA Vs WAPI, should we be worried?.
In Proceedings of the 3rd International Workshop on Wireless Information Systems, pages 18-23
DOI: 10.5220/0002682800180023
Copyright
c
SciTePress
encrypt all the transmissions between the access point and the devices on the network.
This encryption does not implement a high level of security in a public network, since
it would also have to publish the password [8].
WEP was designed with the following factors in mind:
Reasonably strong - It must meet customers’ needs.
Self-synchronizing - Stations quite frequently go in and out of coverage.
Computationally efficient - The WEP algorithm may be implemented in
hardware or software. If it is efficient, it allows low-MIPS devices to still
implement it in software.
Exportable - It can be exported outside the US and imported to other
countries.
Optional - It is an option not required in an 802.11-compliant system [12]
WEP specifies the use of 40 bit keys and this type of key was chosen because at
the time the protocol was drafted, the United States Government had restrictions on
the export of technology containing cryptography. However, this key length is short
enough to make brute force attacks easily accomplished with just about any amount of
computing power available today. However, many equipment manufacturers extend
this protocol and offer a stronger encryption in the form of a 128-bit key, which is
really a 104-bit key with a 24-bit initialization vector (IV). By using this extension,
brute force attacks are nearly impossible with today’s computing platforms but there
are other methods that do not require an attack on the key, so even the 128-bit key is
not totally secure.
Various vulnerabilities have been demonstrated with WEP and among these are
those documented by Borisov, Goldberg and Wagner (2001) [2] who demonstrated
four types of attacks that are easily perpetrated on wireless networks by anyone using
readily-available, off-the-shelf equipment. These are:
1. Passive attacks to decrypt traffic based on statistical analysis.
2. Active attack to inject new traffic from unauthorized mobile stations, based
on known plaintext.
3. Active attacks to decrypt traffic, based on tricking the access point.
4. Dictionary-building attack that, after analysis of about a day's worth of
traffic, allows real-time automated decryption of all traffic.
WPA – the Temporary Solution
Because WEP has been found to be vulnerable to attacks, various
alternatives have been investigated. The Wireless Ethernet Compatibility Alliance
(WECA) is the industry group that verifies and certifies 802.11x wireless products for
interoperability. On January 21, 2002, WECA (also called the Wi-Fi Alliance)
completed a draft of IEEE's 802.11i spec to improve security on 802.11 wireless
networks and this is now being circulated within the engineering community for
editing and eventual approval. This draft is for a security algorithm called Temporal
Key Integrity Protocol (TKIP) which was developed with the help of some of the
same encryption experts that exposed the vulnerabilities in WEP. Although TKIP, like
WEP, is based on RC4 encryption it is implemented in a different way that addresses
some those vulnerabilities, including the ability to generate a new set of encryption
19
keys for every 10,000 packets. Most current Wi-Fi certified products should be
upgradeable to TKIP and those that cannot be upgraded will still interoperate with
products that use TKIP but will only use WEP for security [1].
However, not all researchers are satisfied that TKIP will be able to do what it
is designed to do. Some researchers believe the real problem is the fundamental way
in which Wi-Fi works. Although rapid rekeying of WEP keys, which will be
implemented in TKIP, will make it more difficult to crack, the complete design is not
good security. If you are relying on a confidentiality mechanism, in general that is
considered bad design and TKIP does not eliminate the fundamental flaw in Wi-Fi
security. "If anybody breaks TKIP, they not only break the confidentiality but they
also break the access control and authentication so one break breaks everything. That
is not good design. Each security mechanism should stand on its own" [9].
However, TKIP is just considered an interim solution for WEP that mandates
that each client uses different keys and frequently changes them before a hacker has
time to decipher them. TKIP uses a 128 bit key along with the user’s MAC address
plus an initialization vector. In May 2002, the U.S. Secretary of Commerce chose the
Advanced Encryption Standard (AES) which is up to 256 bits and is even stronger
than DES, as an authorized government standard and it will eventually supersede the
Data Encryption Standard (DES) in new deployments. However, companies with
deployed wireless LANs can upgrade their networks to TKIP security using firmware.
AES requires hardware acceleration using a co processor to off load the encryption
and decryption or it would slow the throughput down to an unacceptable level. It also
requires new Wi-Fi cards in the client devices and new access points. AES was
available in the first quarter of 2003 [9].
The Wi-Fi Alliance has certified Wi-Fi Protected Access (WPA) and
although it is not an official IEEE standard, it has been adopted by the Wi-Fi Alliance
as a temporary measure as an enhancement to the security of wireless networks until
work on 802.11i has been completed [6]. WPA is used with TKIP and the Michael
message integrity check (MIC) algorithm, to provide enhanced security for wireless
networks
The Michael security algorithm calculates an 8-byte Message Integrity Code
(MIC) and this uses a calculation features which is available on existing wireless
devices. The calculated MIC is placed between the data of the IEEE 802.11 frame and
the 4-byte ICV. The MIC field is encrypted together with the frame data and the 4-
byte Integrity Check Value (ICV). In addition, Michael provides protection against
replay attacks and it does this by including a new frame counter in the IEEE 802.11
frame.
WAPI
The 802.11i security standard went out to be voted on for confirmation in
late 2003 (to date no results are available) but in a surprising twist, the Chinese
government has announced their own specification. The Standardization
Administration of China (SAC) announced in May 2003, that China would adopt a
WLAN standard, called GB15629.11-2003. The WLAN equipment that is sold in
China is required to comply with this standard from Dec. 1, 2003 but a transition
20
period has been granted and this extends the compliance deadline for a few WLAN
products until June 1, 2004 [6].
The recently announced Chinese WLAN standard is very similar in many
ways to the IEEE's 802.11 wireless networking standard but the one important
difference is that it uses a security protocol, called WLAN Authentication and Privacy
Infrastructure (WAPI). WAPI is not part of the 802.11 standard, since the IEEE
standard relies on Wired Equivalent Privacy (WEP). However, if there are two
different WLAN standards in existence, one for China and one for the rest of the
world, this could cause the market for wireless networking equipment to splinter in
two [6].
WAPI uses a block cipher for encryption and an authentication mechanism
that appears to be similar to the IEEE 802.1x standard, which is part of the upcoming
IEEE 802.11i security standard. Diagrams shown during technical discussions with
Chinese officials indicated something like a RADIUS (Remote Authentication Dial-In
User Service) server being used for authentication and in addition, they seemed to
show a central RADIUS system for authenticating all users on all WLANs in China
[6]. This is possibly a method of monitoring all wireless devices being used in China.
In addition, this new Chinese specification will possibly cost more and be
harder to integrate into portable devices like PDAs and phones since it will probably
require more memory. This might also affect the entire notebook market since it will
have an effect on other products that utilize wireless networking. To make matters
worse, if a foreign vendor wishes to produce a product that complies with this new
Chinese standard then they have to sign coproduction agreements with a limited
number of specifically designated Chinese companies, which China justifies by citing
national security concerns. However, the foreign company would have no control
over what goes inside these security products, which obviously raises liability
questions, and of course the Chinese company might simply delay production for the
foreign company giving alternative Chinese manufacturers an unfair advantage. Even
more worrying is that these Chinese firms might demand full disclosure of the foreign
technology, claiming it is impossible to implement WAPI without this information
[6].
Since China is the most important developing economy in the world and it
accounts for a large portion of manufactured goods, including high-tech products,
China will in the next few years, become a major power in engineering and
technology. One method of establishing a country for a position of leadership in
technology is to set national standards that are incompatible with what the rest of the
world is doing. What in effect this does is that it forces everyone else to deal with
technologies that are otherwise unnecessary and this can be used to protect domestic
industries [7].
What is happening is that the world’s most populous nation, with the world’s
fastest growing economy, which will shortly be the second most important market for
computer products (after the USA) is flexing their muscles. This is not the first time
that China has done this as they have their own version of Linux called RedFlag, their
own CPU architecture called Dragon and just recently implemented their own
proprietary DVD standard. Perhaps this is also related to the Chinese government’s
effort to develop products that are not subject to the payment of patent royalties which
21
they are now obligated to pay ever since they joined the World Trade Organization.
Skeptics might consider WAPI to be just another effort at this.
Standards and Network Externalities
The concept of a network externality emerges from the fact that some goods
and services are more valuable when more people consume or utilize that good or
service and yet these goods and services have little or no value if they are used
without a network [4]. Examples of network externalities include telephones and fax
machines as people who own these products would form a network so as to be able to
exchange information and provide a way for people to communicate with each other.
The more people who have telephones the more valuable this network becomes and
the same applies to wireless network equipment. In markets where network
externalities are powerful and people are able to choose from different standards, the
advantages of conforming to a popular platform must be weighed against the
advantages of horizontally differentiating output since conforming to a common
standard exploits the added value associated with network externalities [3].
If a specific technology penetrates a significant portion of the market for any
product, it becomes extremely difficult to dislodge simply because people do not wish
to switch to a new piece of hardware or software, even if the new product is proven to
be superior. This is because they will lose both the time they invested in the old
product and the ability to share data with others who would also have the old product
[11]. This is exactly what has happened with the 802.11 standard, in fact, what has
occurred is that the wireless network 802.11 standard has effectively caused a lock-in.
“A lock-in occurs when the self reinforcing effects increase demand for a specific
product or service and make it difficult for innovative alternatives to be considered”
[5]. Since a de facto lock-in has occurred with 802.11 throughout the rest of the
world, it will be extremely difficult for China’s WAPI to succeed outside of China,
but the problem will exist for manufacturers who wish to sell goods to China or for
visitors to China who wish to use their 802.11 equipment.
Conclusion
Although WAPI has been portrayed as the Chinese solution to the problem
of wireless security it appears to have many of the same problems as WEP. Many
believe that WAPI will be insecure and cause problems for manufacturers who will
now have to conform to two standards, one for China and another for the rest of the
world [10].
It is also left to be seen if the rest of the world will accept this Chinese
wireless standard. For an encryption standard to be thought of as trustworthy, it would
first have to be written by a well-respected member of the cryptography community
and be also subjected to many years of peer review, and only if it is able to survive
without any major weakness being discovered will it be accepted in the encryption
22
community. It is very doubtful that China would open up their standard for peer
review by the rest of the world.
In addition, since the 802.11 standard has enjoyed worldwide acceptance,
network externalities are very strong which has caused a lock-in, whereby people are
going to be resistant to move to another wireless standard. Therefore, while WAPI
will enjoy government mandated success in China and cause severe disruptions
among foreign wireless technology companies that wish to sell their products in the
Chinese market, it will not penetrate the market outside of China. Unfortunately, it
will also cause immense inconvenience for foreign visitors in China.
References
1. Arar, Y.: Beefing up 802.11b Security PCWorld (2002, February 4)
2. Borisov, N. Goldberg, I., Wagner, D.: Intercepting Mobile Communications: The Insecurity
of 802.11. Proceedings of the 7th International Conference on Mobile Computing and
Networking, Rome, Italy (July, 2001)
3. Economides, N., Flyer, F.: Compatibility and Market Structure for Network Goods. New
York University. (November, 1997)
4. Katz, M., Shapiro, C.: Network Externalities, Competition and Compatibility. American
Economic Review (June 1985)
5. Krechmer, K.: The Fundamental Nature of Standards: Economics Perspective
Communications Standards Review (2000)
6. Lemon, S., Lawson, S.: Clouds hang low over Chinese WLAN standard IDG News Service.
(December 19, 2003). Retrieved from the World Wide Web on 1/14/04 from
http://wireless.itworld.com/4276/031219chinawlan/page_1.html
7. Mathias, C.: Chinese food for thought. EETimes. (January 5, 2004) Retrieved from the
World Wide Web on 1/14/04 from
http://www.eetimes.com/comm/c/rd/OEG20040105S0043
8. Oraskari, J.: Bluetooth versus WLAN IEEE 802.11x. Department of Computer Science and
Engineering, Helsinki University of Technology (2001)
9. Schwartz, E.: Researchers crack new wireless security spec. InfoWorld (February 14, 2002)
10. TIA: China Promotes New Wireless Encryption Standard. Pulse Online. (December 2003)
Retrieved from the World Wide Web on 1/14/04 from
http://pulse.tiaonline.org/article.cfm?ID=1911
11. Varian, H.: The Information Economy: How much will two bits be worth in the digital
marketplace? Scientific American (September 1995)
12. Weatherspoon, S.: Overview of IEEE 802.11b Security. Intel Technology
Journal (2000,
Q2)
23