S
Shivendra S. Panwar
Researcher at New York University
Publications - 332
Citations - 9246
Shivendra S. Panwar is an academic researcher from New York University. The author has contributed to research in topics: Wireless network & Network packet. The author has an hindex of 46, co-authored 322 publications receiving 8753 citations. Previous affiliations of Shivendra S. Panwar include Princeton University & Fujitsu.
Papers
More filters
Journal ArticleDOI
Collision resolution algorithms for a time-constrained multiaccess channel
TL;DR: Collision resolution algorithms (CRAs) for the ternary feedback multiple access channel with time constraints are considered and a nonnested CRA is described and its performance is compared with a nested CRA for values of K >.
Journal ArticleDOI
Efficient buffer sharing in shared memory ATM systems with space priority traffic
R. Roy,Shivendra S. Panwar +1 more
TL;DR: This letter studies the problem of the optimal design of buffer management policies within the class of pushout and expelling policies for a shared memory asynchronous transfer mode (ATM) switch or demultiplexer fed by traffic containing two different space priorities.
Proceedings ArticleDOI
On the performance of ATM-UBR with early selective packet discard
K. Cheon,Shivendra S. Panwar +1 more
TL;DR: It is observed that the ESPD scheme improves the effective throughput over the EPD by up to 16% with the network model, and is more effective in alleviating the TCP's unfairness among connections which have different roundtrip times.
Journal ArticleDOI
Deep Neural Network Approximated Dynamic Programming for Combinatorial Optimization
TL;DR: A general framework for combining deep neural networks (DNNs) with dynamic programming to solve combinatorial optimization problems and results show that NDP can achieve considerable computation time reduction on hard problems with reasonable performance loss.
Proceedings ArticleDOI
Realtime Scheduling and Power Allocation Using Deep Neural Networks
TL;DR: Simulation results show that compared with Geometric Programming based power allocation and exhaustive search based scheduling, the proposed method achieves over five orders of magnitude speed-up with less than nine percent performance loss, making real time usage practical.