Lawful Interception For Mobile IP: A Mobile Agent based ...caia.swin.edu.au/talks/CAIA-TALK-040728B.pdf · • Generic LI model • LI in mobile telephony – GSM • Applying telephony
Post on 10-Oct-2020
1 Views
Preview:
Transcript
Lawful Interception ForMobile IP: A Mobile Agent
based Approach
- Andres Rojas
2
Outline
• Generic LI model• LI in mobile telephony – GSM• Applying telephony model to MobileIP• A Mobile Agent based approach• Q?
3
Generic LI model
Network
Legal Document
Operation & MaintenanceCenter
Lawful InterceptionFunction
DeliveryNetwork
Law EnforcementAgency (LEA)
Law EnforcementAgency (LEA)
Activation,Deactivation
Call Content,meta-inf
1
23
4
5
4
LI in Australia
• Governed by Telecommunications(Interception) Act.
• AFP, State Police, Royal Commissions• 2002-2003
– ~3700 active interceptions– Avg. ~40 day intercept period
5
Outline
• Generic LI model• LI in mobile telephony – GSM• Applying telephony model to MobileIP• A Mobile Agent based approach• Q?
6
LI in GSM – InterceptionActivation
Operation &MaintenanceCenter
Legal Document
ControlNetwork
LI System
MSC-A
BSC/BTS
BSC/BTS
BS
BS
BS
BS
MSC-B
BSC/BTS
BSC/BTS
BS
BS
BS
BS
1
23
Activate MSC-A
Activate MSC-B
7
Outline
• Generic LI model• LI in mobile telephony – GSM• Applying telephony model to MobileIP• A Mobile Agent based approach• Q?
8
Comparison of Components
500-2000*1-80No. nodes forsmall citycoverage
8 – 32 port ?2500 – 2 millionsubscribers
Capacity
Router, SwitchRow of Racks -Room
SizeMobile IPGSM (MSC)Characteristic
* Dependent on coverage area of each node
9
Simulation – topology
Central InterceptionController
Leaf nodesBackbone
10
Simulation – results
80.11133111
103.88172812
60.11100010
43.827299
30.785128
20.623437
12.982166
7.511255
3.85644
1.62273
0.4882
Activation time(s)
Total leaf nodesValue of n
11
Simulation – movementexample
-Node and it’s-coverage area
- target- target’s movement
12
Summary of Inefficiencies
• Bandwidth wasted in coverage areas ofnodes where a target does not move to
• Host resources (CPU, IO operations) wasteddue to sniffing effort
• Time to fully activate
13
Outline
• Generic LI model• LI in mobile telephony – GSM• Applying telephony model to MobileIP• A Mobile Agent based approach• Q?
14
Mobile Agents - Definition
• “a self contained piece of software that moveswithin the network with it’s code, state and theability to decide when to move to a new location”
• Movement– Reactive (L2, AAA)– Predictive (Liu & Maguire)
15
MA based approach – operation
Operation &MaintenanceCenter
Legal Document
Network
Central LI
1
MA2
Net-A
AP
LIR
MAPMA
MN
Net-B
AP
LIR
MAPMA
MN
3
4
5
MA –Mobile Agent
MAP – MA Platform
LIR – LI Resource
16
LI focus (1)
• Specific LI application of a more genericcase.
• Other opportunities to explore MIP & MAmarriage
17
LI focus (2) – future work
• Potential for missing target’s traffic– Accuracy of prediction (predictive)– Speed of movement (reactive)
• Security !– MA is a virus– Potential for misuse could be disastrous
• How much CPU/Bandwidth is saved?
18
Outline
• Generic LI model• LI in mobile telephony – GSM• Applying telephony model to MobileIP• A Mobile Agent based approach• Q?
19
Q time.
top related