• No results found

Mobile Phone Network Security

N/A
N/A
Protected

Academic year: 2021

Share "Mobile Phone Network Security"

Copied!
57
0
0

Loading.... (view fulltext now)

Full text

(1)

Mobile Phone Network Security

Internet Security [1] VU

Adrian Dabrowski, Markus Kammerstetter,

Georg Merzdovnik, Stefan Riegler and Aljosha Judmayer [email protected]

(2)

Mobile phone networks

– 1G • Not standardized – 2G (GSM) • From 1980es • Introduced SIM • Started as CS, now PS – 3G (UMTS) • From 1990es – 4G (LTE) • From 2000nd Planes – User Plane • Voice, Data, SMS – Signaling Plane • Call setup, ... – Management Plane • Network organization

(3)

Structure of a 2G/3G System

(4)

Legend

User Equipment (UE) → Radio Access Network (RAN) → Core Network (CN)

Universal Terrestrial Radio Access Network (UTRAN) GSM/EDGE Radio Access Network (GERAN)

GPRS Support Node (SGSN)

Gateway GPRS Support Node (GGSN) Mobile Switching Centre (MSC)

(5)

Radio Layer

Physical Channels != Logical Channels Broadcast Channels

– Carry “Beacon” Information

– Paging & signaling to idle devices – Unencrypted

Dedicated Channels

– Communication to a specific UE(ME). – Often encrypted

(6)

GSM Encryption

A5/0

– No encryption, banned from many networks (i.e. Tmobile Austria)

A5/1

– Standard today A5/2

– Export Version, broken 199 A5/3 + 4

(7)
(8)

IMEI – IMSI - TMSI

International Mobile Equipment Identifier – The phone

International Mobile Subscriber Identifier – The SIM card (i.e., the user)

Temporary Mobile Subscriber Identifier – A temporary UserID / SessionID

– (should) Prevent tracking since signaling plane is unencrypted

(9)

Attacks: TMSI deanonymization

– TMSI deanonymization • Record Paging Traffic

• Call known Number, hangup before full channel setup (=before starts ringing)

• Use set of recorded pages to filter candidates – Aka sieve

(10)

Attacks: Internet interconnectivity

– GPRS tunneling over Internet

• Publicly reachable tunnel endpoints have been found via port scanning at several Telcos

– DNS infrastructure,

(11)

Attacks: SIM Cloning

– COMP128 weakness

• Key derivation algorithm

• Secret key recovery by analyzing thousands of responses

• SIM card cloning

• Used via programmable multi-SIMs and development SIM cards

(12)

Attacks: Decryption

– GSM Cipher

• Rainbow tables available (~2TB)

• Decode session key (eavesdropping) • In seconds...

(13)

Attacks: SS7

– SS7 Attacks

• “Signaling System 7”

– Signaling Backbone within and between many Telcos

– Telcos fully trust each other • e.g. Anytime Interrogation

– Find cell ids (=location) of any phone • Share Session key in case of roaming, etc

(14)

Attacks: IMSI Catchers aka

Stingray

Used for

– Tracking users

– Eavesdropping calls, data, texts – Man-in-the-Middle

– Attack phone using operator system messages (e.g. Management Interface, reprogram APN, HTTP-Proxy, SMS/WAP-Server...)

– Attack SIM (c.f. SIM card rooting, otherwise filtered by most mobile carriers), Attack Baseband

– Geotargeting ads (e.g. SMS)

(15)
(16)

A Mobile Network

with a Mobile Station

(17)
(18)
(19)

A wild IMSI Catcher appears –

(20)

A Real Network

(21)

Cell tower density

(22)

“IMSI Catchers”

Identification only

Retrive

IMSI / IMEI / TMSI

Reject Location Update Tracking

Traffic Man-in-the-Middle

Hold in Cell

Actively intercept traffic

– Relay to real network – Active or passive

decryption

Hold but intercept passively

Imprison in cell, so phone is not lost to a neighbor cell UMTS downgrade Blocking UMTS transmission Spoofing System messages

(23)

“IMSI Catchers”

(24)

IC: Car Installation

(25)

IC: Car Installation

(26)

Car Installation

(27)

IC: Car Installation

(28)

IC: Car Installation

(29)

Body IMSI Catcher

(30)

Only for Law Enforcement?

Known Producers

– Rohde & Schwarz – Gamma Group – Ability – IAI Elta – Septier – Meganet – NeoSoft – Proximus – Cyttek – …

DIY – USRP based

– Kirstin Paget • DEFCON 19 • US$1,500 – D. Werhle • Master's Thesis • Freiburg – B. Postl • Master's Thesis • Vienna

(31)

How to catch an

IMSI Catcher?

(32)

Artifact: Frequency

Unused or guard channel

• Only found in Full Scan

Announced neighbor freq., but unused

• Careful not to create interference Detactability – Frequency plans • e.g. radio regulatory • Self created

(33)

Artifact: Cell ID

New CID/LAC needed – To provoke

“Location Update Request”

– Random?

– Use real one not used in that

geographical region

Detectability

– Cell IDs are very stable – Cell Database (local) • Also for frequencies – Correlation with GPS coordinates

(34)

Artifact: Location Update /

Register

Just providing a better signal Is not enough

– Timers, Hysteresis – Unpredictable

radio environment RF Jamming?

– Forcing full scan

Detectability:

– Watching noise levels

(35)

Artifact: UMTS handling

Downgrading to GSM – e.g. Mayer and

Wetzel, 2005 [1] • GSM layer in most deployed UMTS networks – (selectively) Jamming – Downgrade LUR – Others... Detectability:

– Noise and Signal levels – Database of regions where UMTS is available, and GSM usage is unlikely • Cell Database

[1] Mayer and Wetzel, “A man-in-the-middle attack on UMTS”, ACM Workshop on Wireless security, 2005

(36)

Downgrade 4G → 3G → 2G

Pre-authentication traffic is unprotected

- includes GET_IDENTITY (IMSI, IMEI)

Location Updates can be rejected unauthenticated

– Needed for Roaming case

– Reject cause: “You don't have a

subscription for this service”

(37)

Encryption

Older IMSI Catchers:

Downgrade encryption to 'none' (A5/0)

A5/1 and A5/2 can be decrypted with rainbow tables

– In realtime

A5/3 rolled out at the moment

– IC will have to do active MITM again

Detectability: – Cipher Indicator • Feature request in Android, 2009, assigned 2013 – Roaming!

(38)

Artifact: Cell Imprisonment

Networks provides up

to 32 neighbor

frequencies

– MS stores typ. 6+1 – Used for hand

overs, LAR, …

IC will likely provide an

empty (eq.) NL

– To not loose phone to a neighbor cell

Detectability:

(39)

Traffic forwarding

a) relay via other MS – Loose caller ID – No incoming calls b) via SS7 or similar – Caller ID correct – Loose incoming calls

c) recover secret SIM key

– Impersonate to

network with victims identity

Detectability:

(40)

Usage Pattern

Identification Mode – Short living cells MITM Mode

– Longer living cells

Both:

– Unusual locations for cells

(41)

Cell capabilities and parameter

fingerprinting

Cell capabilities & parameters

Organization of logical channels on physical channels

Timeout values

Can be different on each cell, but typically they are the

same over the whole network

– Differ between networks

Detectability:

– Cell and network database

(42)

Network Monitor Mode

(43)
(44)

Two approaches

Mobile IMSI Catcher Catcher – Standard Android API – No need to root phone – No need for a specific chipset (e.g. GoldX) – Easy Interface Stationary IMSI Catcher Catcher – Network of measuring stations – Good locations, larger coverage – Cheap • RaspberryPi based

(45)
(46)
(47)

Two approaches - Features

Mobile IMSI Catcher Catcher – GPS + Neighbor cell listing • Geographical correlation • Cell-IDs – Cell Capabilities – RF and NCL manipulations – Limited to NCL but mobile Stationary IMSI Catcher Catcher – Cell-ID mapping – Frequency usage – Cell lifetime – Cell capabilities, network parameters – Jamming

(48)
(49)
(50)
(51)
(52)
(53)
(54)
(55)

WIP: Network operator has no

global view

Currently in a project with a major Austrian carrier – Finding IMSI Catcher based on operator data Network operator has no global view

– Some transactions are designed decentralized – One transaction can leave trace on many

levels/protocols

– 2G/3G/4G interaction grown historically

– Monitoring solutions have to carry high load

• e.g. > 100K LUR/min

(56)

Work in Progress

Verified with USRP based IMSI Catcher

– Need commercial devices for testing Build dense measurement network – Goal: 20-40 stations Implement 3G + LTE Android-APP

– Fine tune ruleset for everyday situations – Problem zones:

• Tunnels

(57)

(Bachalor's and Master's)

Thesis' opportunities

– Port 2G broadcast sniffing to FPGA (BladeRF) – Implement 3G

• GNU Radio , SDR

– Implement 4G

• GNU Radio, SDR

– Implement client stack – More ideas?

References

Related documents