September 2018

 In my doctoral thesis, I introduced how to realize NFV service chains at the emerging and utterly challenging link speeds at 100 Gbps using commodity hardware, while obliterating latency.


May 2018

My latest research is featured in the ACM Technews, PHYS.ORG, ECN, and KTH.


April 2018

In our NSDI 2018 paper "Metron: NFV Service Chains at the True Speed of the Underlying Hardware", we focus on how to realize high performance NFV service chains at the true speed of the underlying hardware. We solve this challenging problem by exploiting the synergy between available network resources (i.e., programmable switches and network cards) and commodity servers, while eliminating inter-core communication among the service chain components. We demonstrate, via 40-Gbps and 100-Gbps experiments, that our approach achieves: (i) 2.75-6.5x better efficiency, (ii) up to 4.7x lower latency, and (iii) up to 7.8x higher throughput than the state of the art.


February 2017

In our SCC journal article, we built an NFV profiler that tracks the movement of packets across the system’s memory hierarchy by collecting key hardware and OS-level performance counters. Leveraging the profiler’s data, our Service Chain Coordinator’s (SCC) runtime accelerates user-space NFV service chains by carefully combining scheduling with I/O multiplexing. Specifically, with SCC service chains are granted enough time for processing an entire batch of packets before yielding a CPU, thereby greatly reducing context switching. SCC results in three-fold latency reduction due to cache and main memory utilization improvements. More importantly, SCC reduces the latency variance of NFV service chains by up to 40x compared to standard FastClick service chains by making the average case for an NFV chain to perform as well as the best case.


November 2016

In our SNF journal article, we present a framework that synthesizes (S) network function (NF) service chains by eliminating redundant I/O and repeated elements, while consolidating stateful cross layer packet operations across the chain. SNF uses graph composition and set theory to determine traffic classes handled by a service chain composed of multiple elements. It then synthesizes each traffic class using a minimal set of new elements that apply single-read-single-write and early-discard operations. SNF realizes stateful and long service chains at 40 Gbps throughput with low latency on only one machine with 8 CPU cores. This performance is up to 8.5x greater than the state of the art FastClick NFV framework.