Kernel-level attacks: KernelSnitch highlights how unprivileged attackers can exploit variable-size kernel data structures like hash tables and trees to leak sensitive information via side-channel timing differences.
Wearable authentication: Researchers introduced a smartwatch continuous authentication system utilizing low-frequency multi-channel PPG signals, showing practicality through significant power consumption reductions while retaining high accuracy.
Blockchain efficiency: A new generic sharding blockchain consensus pattern, Kronos, was proposed to achieve scaling and optimized overhead, showcasing substantial improvements in throughput and latency.
KernelSnitch: Side Channel-Attacks on Kernel Data Structures#
The authors introduce KernelSnitch, a novel software side-channel attack targeting variable-size kernel data structures like hash tables and trees.
By exploiting the timing differences caused by varying occupancy levels of these data structures, an unprivileged attacker can leak sensitive information.
The paper demonstrates a website fingerprinting attack with an F1 score of 89.5% and a covert channel achieving a capacity of 580 kbit/s at an error rate of 2.8%.
Know Me by My Pulse: Toward Practical Continuous Authentication on Wearable Devices via Wrist-Worn PPG#
This work presents a continuous authentication system for smartwatches using low-frequency (25 Hz) multi-channel Photoplethysmography (PPG) signals to reduce power consumption.
The system employs a Bi-LSTM model with an attention mechanism to extract features from short 4-second windows of PPG data.
The real-world implementation achieved an 88.11% accuracy, a 0.48% False Acceptance Rate (FAR), and an 11.77% False Rejection Rate (FRR) while reducing sensor power consumption by 53% compared to a 512 Hz setup.
Kronos: A Secure and Generic Sharding Blockchain Consensus with Optimized Overhead#
The paper proposes Kronos, a secure sharding blockchain consensus pattern designed to optimize overhead for cross-shard transactions.
Kronos utilizes a shared buffer managed by shard members to handle transactions, requiring no Byzantine fault tolerance (BFT) protocol execution for efficient rejection in happy paths.
Experimental evaluations over up to 1000 AWS EC2 nodes demonstrated that Kronos can scale to thousands of nodes, achieving a peak throughput of 320.2 ktx/sec with a latency of 2.0 sec.