Wireless LANs - University of St. Thomas

Download Report

Transcript Wireless LANs - University of St. Thomas

CISC 210 - Class Today
•
•
•
•
•
Wireless LANs - recap
Link Encryption – book style
Link encryption – LAN style
WEP
WPA
March 2005
R. Smith - University of St Thomas - Minnesota
1
802 Protocol in general
• Traditionally evolved from Ethernet
– Unreliable (unACKed, unchecksummed)
– Broadcast between nearby stations
– As fast and cheap as possible
• Ethernet: CSMA/CD
– Detect ‘free’ channel; detect collisions
– Exponential backoff
• Wireless (802.11): CSMA/CA
– All stations can’t always hear each other; CD isn’t practical
– Wireless is noisier than Ethernet; more dropped packets
– Impractical to completely ignore reliability
March 2005
R. Smith - University of St Thomas - Minnesota
2
802.11 Protocol
• Virtual Carrier Sense
– Send “RTS” to ask for permission to send
• Gives source, destination, and duration of “real”
transmission
– If no other traffic, recipient sends back CTS
– Then sender sends the actual data
– Recipient sends an ACK
• Collisions most likely during RTS
– They’re very short messages, reduce collision risk
– Other stations see the RTS/CTS, wait to transmit till done
• Packets are smaller on 802.11 than Ethernet
– Big packets are more likely to be corrupted by noise
March 2005
R. Smith - University of St Thomas - Minnesota
3
Hooking Up
• Base Stations
– May serve as ‘Access Point’ (AP) - Provide a link to a
‘backbone’ – i.e. Internet access
• “Service Sets”
– “Basic Service Set” (BSS) – Environment where everyone is
within range of a single base station
– “Extended Service Set” (ESS) – where two or more base
stations are connected via a common backbone to provide
more coverage (I do this at home)
• Service Set ID (“SSID”)
– That magic text string that pops up from a base station and
identifies the service set you’re in (default ‘linksys’ on many)
March 2005
R. Smith - University of St Thomas - Minnesota
4
Link Encryption Objectives
• Confidentiality on isolated set of computers
– Computers only talk to one another
– They do not talk to other computers
• No communication with outsiders
– Avoid both intentional and accidental data disclosure
• Hide traffic as much as possible
– Don’t disclose traffic patterns; don’t disclose data
• Safety and familiarity paramount
– Shouldn’t interfere with computer or network operation
– Should always work with minimum of fuss
– Extra cost is acceptable
March 2005
R. Smith - University of St Thomas - Minnesota
5
Link Level encryption: properties/features
• Red/Black separation
– Everything that goes out is encrypted
– Everything inside is cleartext
• Good algorithm; good keys
– A problem with older wireless (we’ll see later)
– Good keys = over 100 bits
– Good algorithm = AES, maybe triple DES (slower)
• Protect against replay & rewrite attacks
–
–
–
–
Duplicates must be detectable: packet serial numbers, etc.
Cryptographic checksum that outsiders can’t forge
Good stream cipher or block mode
Random data to confound “known plaintext” attacks
March 2005
R. Smith - University of St Thomas - Minnesota
6
Routing and LAN Encryption
• Point to Point Encryption
– Kind of a dead horse today
• Everyone uses multipoint LANs, like Ethernet
• Is everything encrypted that goes out?
– What about MAC addresses?
– If we need an address it must be in plaintext
• A wireless router
– All data on the wireless is encrypted
• Including IP addresses
– We strip off the wireless encryption when it leaves the wireless
LAN
March 2005
R. Smith - University of St Thomas - Minnesota
7
Wireless Crypto
• WEP, WPA – what do they encrypt?
• What objectives do we achieve?
• Link Encryption Objectives
–
–
–
–
Confidentiality on isolated set of computers
No communication with outsiders
Hide traffic as much as possible
Safety and familiarity paramount
March 2005
R. Smith - University of St Thomas - Minnesota
8
WEP versions
• “Wired Equivalent Privacy”
– Describes the hope, not the achievement
• Shared key encryption protocol
– 64-bit keys (original WEP)
– 128-bit keys (WEP 2)
– Uses RC-4 stream cipher (hard to use safely)
• Poorly constructed encryption
– 64-bit keys broken in 40-bit time
– 128-bit keys broken in 64-bit time
March 2005
R. Smith - University of St Thomas - Minnesota
9
WEP Crypto
• WEP Encryption (diagram)
– Pick an IV (Initialization Vector, Nonce), 24 bits
– Concatenate to ‘root key’ -> k | IV
• This is the packet key (up to 128 bits)
– Calculate CRC32 over the data (the “ICV”)
– Encrypt data and ICV using the packet key
– Transmit IV and encrypted data
March 2005
R. Smith - University of St Thomas - Minnesota
10
WEP Weaknesses
• ICV only protects against random errors
– Possible to modify a packet's contents and CRC without
knowing the crypto key (think of the bit flip example)
• “Related Key” attacks
– Attacker knows part but not all of the key
– Algorithm is vulnerable if
• Knowing info about crypto with one key…
• Yields info about encryption with a “similar” key
– RC-4 is vulnerable.
– Lots of ‘crib’ available (ARP)
• “Chop chop” attack
– Intercept and retransmit a packet
• Change the last byte of data through trial and error
March 2005
R. Smith - University of St Thomas - Minnesota
11
WPA
• WPA – a stopgap to replace WEP ASAP
– 128-bit keys using RC-4
• Used existing hardware
• Better integrity protection – MIC using ‘Michael’
• Still uses ICV function as well
• Larger effective keys
– TKIP
• Similar to WEP, but ‘mixes’ the IV and key
March 2005
R. Smith - University of St Thomas - Minnesota
12
WPA2 –
• Implements 802.11i enhancements
– Use AES instead of RC-4
– Permanent keys to authenticate; temporary for data
– Can use RADIUS authentication server
• Counter Mode with CBC MAC
– Integrates encryption, integrity checking, and key variation
• Key update protocol
– Each packet has a unique key
– Derived from packet serial #, shared secret, MAC addr
March 2005
R. Smith - University of St Thomas - Minnesota
13
Projects
• Find a project and get started!
• The ‘final date due’ for your proposal is LATE
Escrowed Encryption
• Obsolete but interesting technology
– Security implications?
– Political implications?
• All packets include LEAF
– Encrypted with a special shared secret key. Contains
• Device ID
• Session key used to encrypt this message
• Checksum on the LEAF
– There’s a special decryption system
• Has a database of DeviceID/Decryption key
March 2005
R. Smith - University of St Thomas - Minnesota
15
Protocols and Layers
• We use layering for several things
– Organize the software
– Format the packets
• What it really does:
Establish a relationship between software
components on different computers
– Layers communicate with each other at same layer
• IP – IP or TCP – TCP or HTTP – HTTP
– They ‘use’ the lower layers to carry their messages
March 2005
R. Smith - University of St Thomas - Minnesota
16
Protocol Layering Examples
• Network class –
bear with me
• Pizza delivery
example
– How do we order
pizza at a party?
March 2005
R. Smith - University of St Thomas - Minnesota
17
Network Protocol Layering
Usually a ‘funnel’ shape
• Top level = Applications
– Lots of choices: e-mail, web, file exchange,
– Uses ‘socket interface’ to talk to networks
• Mid levels = “The Protocol Stack”
– Transport layer: UDP/TCP
– Internet layer: IP
– Link layer: LAN protocols
• Bottom level = device driver connections
– Hardware-specific software, configuration
– Uses device driver interface to link to the protocol stack
– Uses a cable or antenna to link to the network
March 2005
R. Smith - University of St Thomas - Minnesota
18
Packets follow the layers
• Upper layer data = innermoust
• Lower layer data = outermost
• Innermost data usually travels the network
unchanged
• Outermost data gets swapped with each hop
through a router
March 2005
R. Smith - University of St Thomas - Minnesota
19
Diagramming the Crypto
• Elements
–
–
–
–
Protocol stack elements
Where the crypto goes
What is encrypted
What is plaintext
March 2005
R. Smith - University of St Thomas - Minnesota
20
That’s it
• Questions?
Creative Commons License
This work is licensed under the Creative Commons Attribution-Share Alike 3.0 United
States License. To view a copy of this license, visit
http://creativecommons.org/licenses/by-sa/3.0/us/ or send a letter to Creative
Commons, 171 Second Street, Suite 300, San Francisco, California, 94105, USA.
March 2005
R. Smith - University of St Thomas - Minnesota
21