Timing-based Side-Channel Attacks and Mitigations in Multicore Processors
Digital Document
Document
Persons |
Persons
Creator (cre): Ali, Usman
Major Advisor (mja): Khan, Omer
Co-Major Advisor (cma): Chandy, John A.
Associate Advisor (asa): Wang, Lei
|
||||||||
---|---|---|---|---|---|---|---|---|---|
Title |
Title
Title
Timing-based Side-Channel Attacks and Mitigations in Multicore Processors
|
||||||||
Origin Information |
Origin Information
|
||||||||
Parent Item |
Parent Item
|
||||||||
Resource Type |
Resource Type
|
||||||||
Digital Origin |
Digital Origin
born digital
|
||||||||
Description |
Description
Multicore processors are extensively deployed in safety-critical environments due
to efficient utilization of shared hardware resources (i.e., caches, network-on-chip, memory controllers). The sharing of underlying hardware brings massive performance benefits but opens up security challenges. Concurrently executing applications on multicore processors compete for shared hardware and induce contentions, consequently creating timing variations in execution time. Timing variations allow an adversary application to leak security-critical information using timing-based side-channel attacks. This dissertation proposes novel timing-based side-channel attacks and mitigations that target network-on-chip hardware and combinations of network-on-chip, cache, and memory controller hardware in multicore processors. The network-on-chip has low timing variations and has challenges of external noise. It uses information theory-based techniques to overcome the challenges of low-timing variations and external noise. It demonstrates practical covert communication and information leakage attacks in a noisy environment to demonstrate the practicality of attacks. The proposed attacks are implemented on real Tilera TileGX-72 multicore processors, indicating their effectiveness. It proposes novel multi-level mitigation schemes to mitigate these attacks and evaluates performance implications using real-world graphs and machine learning workloads. The mitigation schemes are implemented on an industry-standard multicore RISC-V simulator. |
||||||||
Language |
Language
|
||||||||
Genre |
Genre
|
||||||||
Organizations |
Organizations
Degree granting institution (dgg): University of Connecticut
|
||||||||
Held By | |||||||||
Rights Statement |
Rights Statement
|
||||||||
Use and Reproduction |
Use and Reproduction
These Materials are provided for educational and research purposes only.
|
||||||||
Note |
Note
|
||||||||
Degree Name |
Degree Name
Doctor of Philosophy
|
||||||||
Degree Level |
Degree Level
Ph.D.
|
||||||||
Degree Discipline |
Degree Discipline
Electrical Engineering
|
||||||||
Local Identifier |
Local Identifier
S_45079476
|