Name: Itamar Cohen
Type: User
Company: Politecnico di Torino
Bio: I am a lecturer at Ariel University, Israel.
My research interests include distributed caching, scheduling, and decision-making under uncertainty.
Location: Italy
Blog: http://www.bgu.ac.il/~itamarq/
Itamar Cohen's Projects
This project provides tools to simulate several access strategies for distributed caching. The simulator considers a user who is equipped by several caches, and receives from them periodical updates about the cached content. These updates are not totally accurate. The user has to select which caches to access, to obtain the requested datum in the lowest price and maximum certainty possible.
This is a simulator for access strategies for distributed caching. The simulator considers a user who is equipped by several caches, and receives from them periodical updates about the cached content. The problem and algorithms implemented here are detailed in the paper: I. Cohen, G. Einziger, R. Friedman, and G. Scalosub, “Access Strategies for Network Caching”, IEEE/ACM Transactions on Networking, 29(2), pp. 609-622, 2021.
This is a Java/Python implementation of an algorithm for parallel VM placement with high throughput and provalbe success guarantees, as described in the paper: I. Cohen, G. Einziger, M. Goldstein, Y. Sa’ar, G. Scalosub, and E. Waisbard. Parallel VM Deployment with Provable Guarantees, IFIP Networking, 2021, pp. 1-9.
This fork adds to Caffeine's simulator an indicator. An indicator is a compact, lightweight database that allows the user predict whether a requested datum is found in the cache. If the answer is negative, the user can access only a remote server instead of querying the cache for that item, thus saving the overhead of unnecessary cache accesses. This fork further implements and tests the algorithm CAB, which dynamically scales the indicator, and the frequency of sending it to the user.
This project implements a distributed asynchronous protocol for the deployment and migration of Micro-Services (MSs) in the edge-cloud continuum. The protocol is implemented using the Omnet++ simulator, and is based on real-world antenna location, and on SUMO (Simulator of Urban MObility).
Source files for building OMNeT++ related docker images
This project provides tools to simulate access strategies and cache-content advertisement schemes for distributed caching.
This project simulates deployment and migration of Service Function Chains (SFC) in data-centers.
This project studies the usage of sketches and estimators for network telemetry and neural networks. The project includes Python implementation of various sketch and esimation schemes, such as Morris Counters, CEDAR, SEAD, and Count Min Sketch. As the research is still in preliminary stage, the code is not fully debugged yet.