Remix: On-demand Live Randomization

Remix: On-demand Live Randomization
Yue Chen
, Zhi Wang, David Whalley, Long Lu*
Florida State University, Stony Brook University*
Background
•
Buffer Overflow -> Code Injection Attack
Background
•
Buffer Overflow -> Code Injection Attack
–
Defense: Data Execution Prevention (DEP)
•
Write XOR Execute: a block (page) of memory cannot
be marked as both writable and executable at the same
time.
Background
•
Buffer Overflow -> Code Injection Attack
–
Defense: Data Execution Prevention (DEP)
•
Write XOR Execute: a block (page) of memory cannot
be marked as both writable and executable at the same
time.
•
Return-oriented Programming (ROP) Attack
–
Discover gadgets from 
existing
 
code, and chain
them by 
ret
 instructions
–
Can be Turing complete
Code Reuse Attack
Existing Code
Chained Gadgets
ROP Defense Strategy
•
ROP is one example of a broad class of attacks that
require attackers to know or predict the locations of
binary features.
ROP Defense Strategy
•
ROP is one example of a broad class of attacks that
require attackers to know or predict the locations of
binary features.
Defense Goal
Frustrate such attacks by randomizing feature
space, or remove features
ASLR
A
ddress 
S
pace 
L
ayout 
R
andomization
Executable
Library1
Library1
Executable
First-time
Second-time
ASLR - Problem
Library1
Executable
Pointer Leak
ASLR - Problem
Library1
Executable
Pointer Leak
De-randomized
ASLR - Problem
Brute force 
attacks are still possible (When the
entropy is small. E.g., 32-bit systems.)
Randomized only 
once
.
Goal
•
Live
 randomization 
during runtime
•
Finer-grained
 randomization unit
•
Low
 performance overhead
•
Highly 
composable
–
Can and should be combined with other defenses
(traditional ASLR, function randomization, etc.)
Remix
Live basic block (BB) (re-)randomization
within
 functions
A
 
b
a
s
i
c
 
b
l
o
c
k
 
i
s
 
a
 
s
t
r
a
i
g
h
t
-
l
i
n
e
 
c
o
d
e
 
w
i
t
h
o
u
t
j
u
m
p
s
 
i
n
 
o
r
 
o
u
t
 
o
f
 
t
h
e
 
m
i
d
d
l
e
 
o
f
 
t
h
e
 
b
l
o
c
k
.
Remix
Live basic block (BB) (re-)randomization
within
 functions
Advantages:
Advantages:
•
No function pointer migration
•
Good spatial locality
A
 
b
a
s
i
c
 
b
l
o
c
k
 
i
s
 
a
 
s
t
r
a
i
g
h
t
-
l
i
n
e
 
c
o
d
e
 
w
i
t
h
o
u
t
j
u
m
p
s
 
i
n
 
o
r
 
o
u
t
 
o
f
 
t
h
e
 
m
i
d
d
l
e
 
o
f
 
t
h
e
 
b
l
o
c
k
.
Remix
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Space
 
for
Alignment
Other Functions
…
Function A
Remix
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Space
 
for
Alignment
Other Functions
…
Function A
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Other Functions
…
Function A
After
0.46 seconds
Remix
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Space
 
for
Alignment
Other Functions
…
Function A
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Other Functions
…
Function A
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Other Functions
…
Function A
After
0.46 seconds
After
2.13 seconds
Challenges
•
Chain randomized basic blocks together
–
Need extra space to chain basic blocks
–
Need to update instructions
•
Stale pointer migration
Extra Space
Case 1: Extra Jmp
Basic Block 1
Basic Block 2
Basic Block 3
Jmp to BB1
(1) At the beginning of a function
Extra Space
Case 1: Extra Jmp
Basic Block 1
Basic Block 2
Basic Block 3
Jmp to BB1
(1) At the beginning of a function
(2) At the end of a basic block that does 
not end with an instruction like 
jmp
/
ret
mov
 
…
add
 
…
mov
 
…
mov
 
…
add
 
…
jle
 
…
Extra Space
Case 2: Larger Displacement
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Jump to BB3
Before Remix
Extra Space
Case 2: Larger Displacement
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Jump to BB3
Jump to BB3
Before Remix
After Remix
Extra Space
Case 2: Larger Displacement
One-byte displacement:
jmp +0x10
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Jump to BB3
Jump to BB3
Four-byte displacement:
jmp +0x00001000
Before Remix
After Remix
Extra Space
Solution
With
 Source Code:
Modify the compiler to:
1.  
Insert an extra 5-byte NOP 
instruction after each basic block
2.  
Only generate instructions with
4-byte displacement
Enough Space Guaranteed!
Extra Space
Solution
With
 Source Code:
Modify the compiler to:
1.  
Insert an extra 5-byte NOP 
instruction after each basic block
2.  
Only generate instructions with
4-byte displacement
Enough Space Guaranteed!
Without
 Source Code:
Leverage existing NOP paddings:
•
Function alignment
•
Loop alignment
Can insert 
random bytes
 into NOP
space after randomization, making
attack even harder.
Instruction Update
Q:  Which instructions need updating ?
A:  
Control-flow
 related ones, to adjust displacement
Instruction Update
Two-step update 
(e.g., unconditional direct jmp)
:
Basic Block 1
Basic Block 2
Basic Block 3
Original
After BB reordering
Step one:
Jumps to original
address of BB 3
Step two:
Jumps to current
address of BB 3
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 1
Basic Block 2
Basic Block 3
jmp
Instruction Update
•
Direct call
: step-one update
•
Indirect call
: no update needed
•
Direct jump
: step-one and step-two update
•
Indirect jump
: discussed later
•
PC-relative addressing
: step-one update
Indirect Jump
•
Jump to functions – unchanged
–
PLT/GOT
–
Tail/Sibling Call
•
Jump to basic blocks – see next
Basic Block Pointer Conversion
•
Why?
-
Migrate stale 
pointers to basic blocks
, to ensure
correctness
Basic Block Pointer Conversion
•
Why?
-
Migrate stale 
pointers to basic blocks
, to ensure
correctness
•
Where?
-
Return address
-
Jump table (switch/case)
-
Saved context (e.g., setjmp/longjmp)
-
Kernel exception table
…
…
Call Foo
Return Site
…
…
main
Foo
Illustration - Return Address
…
Call Foo
Return Site
…
…
main
Foo
Illustration - Return Address
…
Call Foo
Return Site
…
…
main
Foo
Illustration - Return Address
Optimization
To Speed up the randomization procedure:
•
 
Pre-store
 the required information
–
 Basic block information (e.g., locations)
–
 Code/data that need updating
Optimization
To speed up execution:
•
 Probabilistic loop bundling
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Loop
Bundle
Optimization
To speed up execution:
•
 Probabilistic loop bundling
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Loop
Basic Block 1
Bundled BB
Basic Block 4
Basic Block 5
Bundle
Optimization
To speed up execution:
•
 Probabilistic loop bundling
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Loop
Basic Block 1
Bundled BB
Basic Block 4
Basic Block 5
Basic Block 1
Bundled BB
Basic Block 4
Basic Block 5
Randomize
Bundle
Optimization
To speed up execution:
•
 Probabilistic loop bundling
Basic Block 1
Basic Block 2
Basic Block 3
Basic Block 4
Basic Block 5
Loop
Basic Block 1
Bundled BB
Basic Block 4
Basic Block 5
The bundling layout is 
different
 from time to time. – 
Unpredictable
!
Basic Block 1
Bundled BB
Basic Block 4
Basic Block 5
Randomize
Implementation
•
Can work on source code using a slightly
modified LLVM, or work directly on binaries.
•
C
a
n
 
w
o
r
k
 
o
n
 
L
i
n
u
x
 
u
s
e
r
-
s
p
a
c
e
 
a
p
p
l
i
c
a
t
i
o
n
s
,
 
a
n
d
F
r
e
e
B
S
D
 
k
e
r
n
e
l
 
m
o
d
u
l
e
s
.
Evaluation - Security
•
Finer-grained randomization:
–
Adds about four bits of entropy to each instruction.
•
Live randomization during runtime:
–
Destroy discovered gadgets immediately after each
re-randomization.
Want more entropy? – Insert more NOP space!
Evaluation - Performance
SPEC CPU 2006 Performance Overhead (randomized once)
Evaluation - Performance
SPEC CPU 2006 Size Increase
Evaluation - Performance
Apache Web Server Performance Overhead (by ApacheBench)
Randomization Time Interval
Randomization time interval can be random !
Performance depends on hardware speed.
Evaluation - Performance
ReiserFS Performance Overhead (by IOZone)
Randomization Time Interval
Randomization time interval can be random !
Performance depends on hardware speed.
Q&A

								

								
								

										

												
Slide Note

												

										

								


								
						

							
						


							

								

									
										
Embed
									
									
										
Share
									

									
									
									

								

								
							


							

							 
 							 
 							 
						  


							
							

								
This content delves into defense mechanisms against code injection attacks, specifically focusing on techniques like Data Execution Prevention (DEP), Write XOR Execute, Return-oriented Programming (ROP), and ASLR. It explores how these strategies thwart attackers' attempts by randomizing feature space or removing features, ultimately aiming to protect systems from vulnerabilities and unauthorized access.


								
    
																		
  •   Code Injection
    • 
																	
  •    Defense Strategies
    • 
																	
  •    Data Execution Prevention
    • 
																	
  •    ROP
    • 
																	
  •    ASLR
    • 
																

								

								

									

										fikyogu
										

											 fikyogu Follow
											

										

									

									
Uploaded on Feb 23, 2025 | 0 Views

								

								


								

									
Download Presentation

									
								

								
Please find below an Image/Link to download the presentation.

								
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.


									
									

									


										 
										 
										 
										 

									

									

									

								
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.

								

									 
									
Download Presentation
								

								
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.


								

									
 E N D 

								


							
Presentation Transcript

							

							
  1. Remix: On-demand Live Randomization Yue Chen, Zhi Wang, David Whalley, Long Lu* Florida State University, Stony Brook University*
    2. 
            
            
            
  2. Background     Buffer Overflow -> Code Injection Attack
    3. 
            
            
            
  3. Background     Buffer Overflow -> Code Injection Attack     Defense: Data Execution Prevention (DEP)     Write XOR Execute: a block (page) of memory cannot  be marked as both writable and executable at the same  time.
    4. 
            
            
            
  4. Background     Buffer Overflow -> Code Injection Attack     Defense: Data Execution Prevention (DEP)     Write XOR Execute: a block (page) of memory cannot  be marked as both writable and executable at the same  time.     Return-oriented Programming (ROP) Attack     Discover gadgets from existing code, and chain  them by ret instructions     Can be Turing complete
    5. 
            
            
            
  5. Code Reuse Attack Existing Code Chained Gadgets
    6. 
            
            
            
  6. ROP Defense Strategy     ROP is one example of a broad class of attacks that  require attackers to know or predict the locations of  binary features.
    7. 
            
            
            
  7. ROP Defense Strategy     ROP is one example of a broad class of attacks that  require attackers to know or predict the locations of  binary features. Defense Goal Frustrate such attacks by randomizing feature  space, or remove features
    8. 
            
            
            
  8. ASLR Address Space Layout Randomization Library1 Library1 Executable Executable First-time Second-time
    9. 
            
            
            
  9. ASLR - Problem Pointer Leak Library1 Executable
    10. 
            
            
            
  10. ASLR - Problem Pointer Leak Library1 De-randomized Executable
    11. 
            
            
            
  11. ASLR - Problem Brute force attacks are still possible (When the entropy is small. E.g., 32-bit systems.) Randomized only once.
    12. 
            
            
            
  12. Goal     Live randomization during runtime     Finer-grained randomization unit     Low performance overhead     Highly composable     Can and should be combined with other defenses  (traditional ASLR, function randomization, etc.)
    13. 
            
            
            
  13. Remix Live basic block (BB) (re-)randomization  within functions A basic block is a straight-line code without  jumps in or out of the middle of the block.
    14. 
            
            
            
  14. Remix Live basic block (BB) (re-)randomization  within functions Advantages:         No function pointer migration Good spatial locality A basic block is a straight-line code without  jumps in or out of the middle of the block.
    15. 
            
            
            
  15. Remix Function A Basic Block 1 Basic Block 2 Basic Block 3 Basic Block 4 Basic Block 5 Space for  Alignment Other Functions    
    16. 
            
            
            
  16. Remix Function A Function A Basic Block 1 Basic Block 3 Basic Block 2 Basic Block 5 Basic Block 3 After 0.46 seconds Basic Block 4 Basic Block 2 Basic Block 5 Basic Block 4 Space for  Alignment Basic Block 1 Other Functions     Other Functions    
    17. 
            
            
            
  17. Remix Function A Function A Function A Basic Block 1 Basic Block 2 Basic Block 3 Basic Block 2 Basic Block 5 Basic Block 5 Basic Block 3 After 2.13 seconds After 0.46 seconds Basic Block 3 Basic Block 4 Basic Block 2 Basic Block 5 Basic Block 1 Basic Block 4 Space for  Alignment Basic Block 4 Basic Block 1 Other Functions     Other Functions     Other Functions    
    18. 
            
            
            
  18. Challenges     Chain randomized basic blocks together     Need extra space to chain basic blocks     Need to update instructions     Stale pointer migration
    19. 
            
            
            
  19. Extra Space Case 1: Extra Jmp (1) At the beginning of a function Jmp to BB1 Basic Block 3 Basic Block 2 Basic Block 1
    20. 
            
            
            
  20. Extra Space Case 1: Extra Jmp (1) At the beginning of a function (2) At the end of a basic block that does  not end with an instruction like jmp/ret Jmp to BB1 mov add mov             Basic Block 3 Basic Block 2 mov add jle             Basic Block 1
    21. 
            
            
            
  21. Extra Space Case 2: Larger Displacement Before Remix Basic Block 1 Jump to BB3 Basic Block 2 Basic Block 3 Basic Block 4 Basic Block 5
    22. 
            
            
            
  22. Extra Space Case 2: Larger Displacement Before Remix After Remix Basic Block 1 Basic Block 1 Jump to BB3 Jump to BB3 Basic Block 2 Basic Block 2 Basic Block 3 Basic Block 4 Basic Block 4 Basic Block 5 Basic Block 5 Basic Block 3
    23. 
            
            
            
  23. Extra Space Case 2: Larger Displacement Before Remix After Remix Basic Block 1 Basic Block 1 Jump to BB3 Jump to BB3 Basic Block 2 Basic Block 2 One-byte displacement: jmp +0x10  Basic Block 3 Basic Block 4 Basic Block 4 Four-byte displacement: jmp +0x00001000 Basic Block 5 Basic Block 5 Basic Block 3
    24. 
            
            
            
  24. Extra Space Solution With Source Code: Modify the compiler to: 1.  Insert an extra 5-byte NOP  instruction after each basic block 2.  Only generate instructions with 4-byte displacement     Enough Space Guaranteed!
    25. 
            
            
            
  25. Extra Space Solution Without Source Code: With Source Code: Leverage existing NOP paddings: Modify the compiler to: 1.  Insert an extra 5-byte NOP  instruction after each basic block 2.  Only generate instructions with 4-byte displacement         Function alignment Loop alignment Can insert random bytes into NOP space after randomization, making attack even harder.     Enough Space Guaranteed!
    26. 
            
            
            
  26. Instruction Update Q:  Which instructions need updating ? A:  Control-flow related ones, to adjust displacement
    27. 
            
            
            
  27. Instruction Update Two-step update (e.g., unconditional direct jmp): Step one: Jumps to original address of BB 3 Step two: Jumps to current address of BB 3 Original After BB reordering Basic Block 1 Basic Block 3 Basic Block 3 Basic Block 3 jmp Basic Block 2 Basic Block 2 Basic Block 2 Basic Block 2 Basic Block 3 Basic Block 1 Basic Block 1 Basic Block 1
    28. 
            
            
            
  28. Instruction Update     Direct call: step-one update     Indirect call: no update needed     Direct jump: step-one and step-two update     Indirect jump: discussed later     PC-relative addressing: step-one update
    29. 
            
            
            
  29. Indirect Jump     Jump to functions     unchanged     PLT/GOT     Tail/Sibling Call     Jump to basic blocks     see next
    30. 
            
            
            
  30. Basic Block Pointer Conversion     Why? - Migrate stale pointers to basic blocks, to ensure  correctness
    31. 
            
            
            
  31. Basic Block Pointer Conversion     Why? - Migrate stale pointers to basic blocks, to ensure  correctness     Where? - Return address - Jump table (switch/case) - Saved context (e.g., setjmp/longjmp) - Kernel exception table    
    32. 
            
            
            
  32. Illustration - Return Address main Foo         Call Foo Return Site    
    33. 
            
            
            
  33. Illustration - Return Address main Foo         Call Foo Return Site    
    34. 
            
            
            
  34. Illustration - Return Address main Foo         Call Foo Return Site    
    35. 
            
            
            
  35. Optimization To Speed up the randomization procedure:     Pre-store the required information     Basic block information (e.g., locations)     Code/data that need updating
    36. 
            
            
            
  36. Optimization To speed up execution:     Probabilistic loop bundling Basic Block 1 Basic Block 2 Loop Basic Block 3 Basic Block 4 Basic Block 5
    37. 
            
            
            
  37. Optimization To speed up execution:     Probabilistic loop bundling Basic Block 1 Basic Block 1 Basic Block 2 Loop Bundled BB Bundle Basic Block 3 Basic Block 4 Basic Block 4 Basic Block 5 Basic Block 5
    38. 
            
            
            
  38. Optimization To speed up execution:     Probabilistic loop bundling Basic Block 1 Basic Block 1 Bundled BB Basic Block 2 Loop Bundled BB Bundle Randomize Basic Block 3 Basic Block 1 Basic Block 4 Basic Block 4 Basic Block 5 Basic Block 5 Basic Block 5 Basic Block 4
    39. 
            
            
            
  39. Optimization To speed up execution:     Probabilistic loop bundling Basic Block 1 Basic Block 1 Bundled BB Basic Block 2 Loop Bundled BB Bundle Randomize Basic Block 3 Basic Block 1 Basic Block 4 Basic Block 4 Basic Block 5 Basic Block 5 Basic Block 5 Basic Block 4 The bundling layout is different from time to time.     Unpredictable!
    40. 
            
            
            
  40. Implementation     Can work on source code using a slightly  modified LLVM, or work directly on binaries.     Can work on Linux user-space applications, and  FreeBSD kernel modules.
    41. 
            
            
            
  41. Evaluation - Security     Finer-grained randomization:     Adds about four bits of entropy to each instruction.     Live randomization during runtime:     Destroy discovered gadgets immediately after each  re-randomization. Software Apache nginx lighttpd Average Basic Block Number  per Function 15.3 18.8 14.4 Average NOP Space (bytes)  per Function 19.3 26.2 22.1 Want more entropy?     Insert more NOP space!
    42. 
            
            
            
  42. Evaluation - Performance SPEC CPU 2006 Performance Overhead (randomized once)
    43. 
            
            
            
  43. Evaluation - Performance SPEC CPU 2006 Size Increase
    44. 
            
            
            
  44. Evaluation - Performance Randomization Time Interval Apache Web Server Performance Overhead (by ApacheBench) Performance depends on hardware speed. Randomization time interval can be random !
    45. 
            
            
            
  45. Evaluation - Performance Randomization Time Interval ReiserFS Performance Overhead (by IOZone) Performance depends on hardware speed. Randomization time interval can be random !
    46. 
            
            
            
  46. Q&A
    47. 
            
            
            
						

						

					

					
					
					

						
Related

						

						


							
							
							
							

						

							

									

      

        
          

            

              Demand Estimation and Demand Forecasting
            

            

              
Demand Estimation and Demand Forecasting

              
roxy

            

          

        
      


     

      

        
          

            

              Discussion of Randomized Experiments and Experimental Design Challenges
            

            

              
Discussion of Randomized Experiments and Experimental Design Challenges

              
rowen

            

          

        
      


     

      

        
          

            

              Principles of Traffic Demand Analysis and Highway Demand Forecasting
            

            

              
Principles of Traffic Demand Analysis and Highway Demand Forecasting

              
lolarose

            

          

        
      


     

      

        
          

            

              Demand in Economics: Chapter 4 Overview
            

            

              
Demand in Economics: Chapter 4 Overview

              
marcello

            

          

        
      


     

      

        
          

            

              The Law of Demand in Economics
            

            

              
The Law of Demand in Economics

              
selkie

            

          

        
      


     

      

        
          

            

              Randomization and the American Put: A Comprehensive Overview
            

            

              
Randomization and the American Put: A Comprehensive Overview

              
nasir

            

          

        
      


     

      

        
          

            

              The Hidden Benefit of Cool Roofs in Reducing Peak Energy Demand
            

            

              
The Hidden Benefit of Cool Roofs in Reducing Peak Energy Demand

              
effinger_j

            

          

        
      


     

      

        
          

            

              Create a Bigo Live Clone for Android & iOS Live Streaming Apps
            

            

              
Create a Bigo Live Clone for Android & iOS Live Streaming Apps

              
ridh

            

          

        
      


     

      

        
          

            

              Demand in Macroeconomics: Key Concepts and Laws
            

            

              
Demand in Macroeconomics: Key Concepts and Laws

              
emanuella

            

          

        
      


     

      

        
          

            

              Supply and Demand in a Competitive Market
            

            

              
Supply and Demand in a Competitive Market

              
raon566

            

          

        
      


     

      

        
          

            

              Privacy Issues in IEEE 802.11 Networks: Tracking and MAC Randomization
            

            

              
Privacy Issues in IEEE 802.11 Networks: Tracking and MAC Randomization

              
kard154

            

          

        
      


     

      

        
          

            

              Randomization and Initial Dispensing in ASPIRE Training
            

            

              
Randomization and Initial Dispensing in ASPIRE Training

              
addily

            

          

        
      


     

      

        
          

            

              Supply and Demand in Economics
            

            

              
Supply and Demand in Economics

              
chr_jea

            

          

        
      


     

      

        
          

            

              Supply and Demand in Economics
            

            

              
Supply and Demand in Economics

              
ale_cr

            

          

        
      


     

      

        
          

            

              Demand and Supply in Market Economics
            

            

              
Demand and Supply in Market Economics

              
caes713

            

          

        
      


     

      

        
          

            

              The Law of Demand in Economics
            

            

              
The Law of Demand in Economics

              
rabalais_j

            

          

        
      


     

      

        
          

            

              CREST Demand Model v2.0 Eoghan McKenna & Murray Thomson
            

            

              
CREST Demand Model v2.0 Eoghan McKenna & Murray Thomson

              
jon_def

            

          

        
      


     

      

        
          

            

              Elasticity of Demand and Factors Affecting It
            

            

              
Elasticity of Demand and Factors Affecting It

              
hoeft_g

            

          

        
      


     

      

        
          

            

              Degrees of Elasticity of Demand
            

            

              
Degrees of Elasticity of Demand

              
rayl_1

            

          

        
      


     

      

        
          

            

              Supply and Demand in Competitive Markets
            

            

              
Supply and Demand in Competitive Markets

              
ejat

            

          

        
      


     

      

        
          

            

              Emerging Trends in Live Streaming: Challenges and Opportunities
            

            

              
Emerging Trends in Live Streaming: Challenges and Opportunities

              
cian_41

            

          

        
      


     

      

        
          

            

              MTN-028 Study Product Considerations Overview
            

            

              
MTN-028 Study Product Considerations Overview

              
tigheco

            

          

        
      


     							

						

					

					

				


				

			

		
		

			

				
				

					
More Related Content

					

						

														

					

				

				
			

		












 





  
  
  



giItT1WQy@!-/#giItT1WQy@!-/#giItT1WQy@!-/#giItT1WQy@!-/#giItT1WQy@!-/#giItT1WQy@!-/#