A novel technique for data transmission that uses vibration motors in all cell phones as transmitters and accelerometers as receivers. By carefully regulating the...
A novel technique for data transmission that uses vibration motors in all cell phones as transmitters and accelerometers as receivers. By carefully regulating the vibrations at the transmitter and sensing them through vibration sensors, two mobile devices can communicate.
This is useful for security-sensitive applications such as mobile payments because vibrations are inherently more secure than RF broadcast in NFC or Bluetooth. The ubiquity of vibration motors in every cell phone, even in developing regions, presents an immediate market for vibratory communication
Dr. Nam Sung Kim has developed a protocol to vastly improve energy savings in low to medium load server environments.
The protocol works entirely on the Network Interface Card and is easy to implement. For low and medium load environments, the protocol can save approximately 50% of the energy used by the management of "sleep" and "wake" stages. It can be applied to servers both of small and large scale.
Network communication can be a significant bottleneck for computing systems and data centers, particularly with the rise of artificial intelligence and the ensuing...
Network communication can be a significant bottleneck for computing systems and data centers, particularly with the rise of artificial intelligence and the ensuing distributed training applications. For reinforcement learning (RL) training in particular, network communication accounts for up to 83% of execution time. Existing techniques for handling gradient aggregation suffer from latency tradeoffs (e.g., when executed at a central server) or scalability issues (e.g., for circular aggregation methods). A new approach is needed that better addresses network communication latency while leveraging existing data center features, such as programmable switches that allow for some leve of computational capacity.
Professor Jian Huanghas developed iSwitch, a technology that reduces latency bottlenecks in data center network communication by shifting aggregation processes to programmable switches. iSwitch is particularly relevant to AI applications such as reinforcement learning (RL)-based training, where frequent gradient aggregation typically requires a large number of network hops. In demonstrations, the technology achieved a system-level speedup of more than 3.5x for both synchronous and asynchronous RL distributed training and also improved scalability.
Application
Networking datacenters
Benefits
iSwitch results in a system-level speedup of more than 3.5x for both synchronous and asynchronous RL distributed training and also improves scalability.
As of 2025, ransomware attacks occur every 11 seconds, with victims including individuals, schools, governments, hospitals, police departments, businesses, and more. These...
As of 2025, ransomware attacks occur every 11 seconds, with victims including individuals, schools, governments, hospitals, police departments, businesses, and more. These attacks can be expensive in terms of both operations downtime and ransoms actually paid, with many victims finding themselves subject to repeated attacks or continued withholding of systems/data even upon payment of ransom. Existing solutions are expensive, struggle to adapt to the rapid pace of malware innovation, and often rely on manual processes and systems (e.g., backups to physical hard drives) that may be time-consuming and unreliable.
Dr. Jian Huang at the University of Illinois Urbana-Champaign has developed a suite of solid-state drive (SSD) technologies (which could be developed into a single product or commercialized separately) that offer hardware and firmware-level solutions to protect machines from ransomware attacks and mitigate potential damage/loss. The TimeSSD firmware and TimeKit toolkit use intrinsic flash properties to retain the history of past storage states for up to several weeks without dedicated data backups. The technology can significantly enhance the security of flash-based storage systems while incurring negligible performance overhead for the majority of common applications. TimeSSD and TimeKits functionalities can provide flexibility and facilitate more efficient, secure system functions such as protecting against malware or encrypted ransomware attacks that corrupt files in storage, recovering user files, retrieving update logs, and providing an evidence chain for storage forensics. Rollback to a previous consistent state may be achieved with minimal software involvement. Dr. Huang's ransomware-aware solid-state drive (RSSD), meanwhile, hardens systems against ransomware attacks by transferring data to a cloud server through NVMe over Ethernet. RSSD faces no performance impacts and even allows for post-attack analysis to carefully identify the cause and methodology of any ransomware attacks faced by the RSSD. RSSD enhances security support from typical ransomware attacks which assume conventional SSDs perform like HDDs by preemptively protecting against 3 novel types of ransomware attacks designed by Dr. Huang. Furthermore, RSSD ensures that stale data is not discarded after a limited amount of time as current SSDs behave, instead opting to allocate the stale data to the cloud storage in a time-dependent order. This cloud storage effectively removes the limitation of storage on SSDs by providing a space to store an infinite amount of data. The NVMe over Ethernet is the novel design that serves to protect from ransomware attacks the transfer of data from the local SSD to the cloud storage.
Benefits
Inexpensive
Easily integration into conventional solid-state hard drive technologies
Allows for rapid forensics following malware attack
Inherent resiliance to malware compromises (unlike existing antivirus software which itself is susceptible to attack/corruption by bad actors
Can extend life and performance of SSD by more evenly distributing write processes
