Ketabi, S., Buckley, M., Pazhooheshy, P., Farahvash, F. and Ganjali, Y. 2022. Correlation-Aware Flow Consolidation for Load Balancing and Beyond. SIGMETRICS Perform. Eval. Rev. 49, 4 (Jun. 2022), 105–110. DOI:https://doi.org/10.1145/3543146.3543170.
@article{ketabi_correlationaware_2022a,
title = {Correlation-Aware Flow Consolidation for Load Balancing and Beyond},
author = {Ketabi, Shiva and Buckley, Matthew and Pazhooheshy, Parsa and Farahvash, Faraz and Ganjali, Yashar},
year = {2022},
month = jun,
journal = {SIGMETRICS Perform. Eval. Rev.},
volume = {49},
number = {4},
pages = {105--110},
issn = {0163-5999},
doi = {10.1145/3543146.3543170},
url = {https://dl.acm.org/doi/10.1145/3543146.3543170},
urldate = {2024-08-10}
}
Existing load balancing solutions rely on direct or indirect measurement of rates (or congestion) averaged over short periods of time. Sudden fluctuations in flow rates can lead to significant undershooting/ overshooting of target link loads. In this paper, we make the case for taking variations and correlations of flows into account in load balancing. We propose correlation-aware flow consolidation, i.e. aggregating inversely correlated (or uncorrelated) flows into superflows and using them as building blocks for load balancing. Superflows are smoother than individual flows, and thus are easier to estimate with a higher confidence, and can reduce overshooting/ undershooting of link capacities. We present heuristic methods combined with predictive models to consolidate flows and show they can lead to significant reductions in rate standard deviations compared to correlation-agnostic solutions (up to 33% and 12% improvements at the 50th and 99th percentiles respectively for 20 superflows based on real traffic traces).
Abbasi Zadeh, S., Zandi, F., Beiruti, M.A. and Ganjali, Y. 2022. Load Migration in Distributed Softwarized Network Controllers. International Journal of Network Management. 32, 6 (2022), e2214. DOI:https://doi.org/10.1002/nem.2214.
@article{abbasizadeh_load_2022,
title = {Load Migration in Distributed Softwarized Network Controllers},
author = {Abbasi Zadeh, Sepehr and Zandi, Farid and Beiruti, Mohammad Amin and Ganjali, Yashar},
year = {2022},
journal = {International Journal of Network Management},
volume = {32},
number = {6},
pages = {e2214},
issn = {1099-1190},
doi = {10.1002/nem.2214},
url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/nem.2214},
urldate = {2022-11-13},
langid = {english}
}
Distributed control solutions were introduced to address controller reliability and scalability issues in software-defined networking (SDN). The dynamic nature of network traffic can lead to load imbalance among controller instances. A highly loaded controller instance can be slow in responding to datapath queries and can slow down the entire control platform, as state synchronization and consensus among controller instances are performed in a cooperative manner. In this paper, we present Efficient, Resilient, Consistent (ERC), a novel protocol for migrating the load of a given switch from a controller instance to a different instance. Our protocol has three distinguishing properties compared with prior works in this area: (1) It is resilient to failures during migration, (2) it maintains consistency among all controller instances, and nevertheless, (3) it is more efficient than existing load migration protocols. Compared with state-of-the-art, ERC reduces the migration time by 23–50% depending on network load. The implicit assumed use case in the design of previous load migration algorithms (including ERC) has been the load balancing scenario. However, as this is not the only possible case, by maintaining the desirable properties of ERC, we introduce four variants of our protocol that can add to the versatility of the load migration handling. This is achieved by considering variations of role exchange between controller instances, which gives us an advantage over the fixed master–slave exchange that vanilla ERC or previous work support. We perform an extensive set of experiments to examine the impact of variable network parameters on the performance metrics of interest and to show the effectiveness of the ERC family of protocols in load migration.
Dong, H., Munir, A., Tout, H. and Ganjali, Y. 2021. Next-Generation Data Center Network Enabled by Machine Learning: Review, Challenges, and Opportunities. IEEE Access. 9, (2021), 136459–136475. DOI:https://doi.org/10.1109/ACCESS.2021.3117763.
@article{dong_nextgeneration_2021,
title = {Next-Generation Data Center Network Enabled by Machine Learning: Review, Challenges, and Opportunities},
shorttitle = {Next-Generation Data Center Network Enabled by Machine Learning},
author = {Dong, Haiwei and Munir, Ali and Tout, Hanine and Ganjali, Yashar},
year = {2021},
journal = {IEEE Access},
volume = {9},
pages = {136459--136475},
issn = {2169-3536},
doi = {10.1109/ACCESS.2021.3117763}
}
Data center network (DCN) is the backbone of many emerging applications from smart connected homes to smart traffic control and is continuously evolving to meet the diverse and ever-increasing computing requirements of these applications. The data centers often have tens of thousands of components such as servers and switches/routers that work together to achieve a common objective and serve these applications. Managing such large data centers is a tedious process and demands automation, intelligent control and decision making within the data center. Recently both the industry and academia have focused on bringing intelligence to the control, automation, and management of DCNs. Despite the variety of works that surveyed ML for networking, to the best of our knowledge, none has focused on DCN, which makes this survey original. Readers in the academic and industrial communities will all benefit from a comprehensive discussion of the ML solutions applied in DCN to address critical essential problems, including workload forecasting, traffic flow control, traffic classification and scheduling, topology management, network state prediction, root cause analysis, and network security. Furthermore, this article outlines the challenges and concludes with the future research venues in adopting ML for automatic, intelligent and autonomous DCNs.
Shirali-Shahreza, S. and Ganjali, Y. 2018. Delayed Installation and Expedited Eviction: An Alternative Approach to Reduce Flow Table Occupancy in SDN Switches. IEEE/ACM Trans. Netw. 26, 4 (Aug. 2018), 1547–1561. DOI:https://doi.org/10.1109/TNET.2018.2841397.
@article{shirali_shahreza_delayed_2018,
title = {Delayed Installation and Expedited Eviction: An Alternative Approach to Reduce Flow Table Occupancy in SDN Switches},
shorttitle = {Delayed Installation and Expedited Eviction},
author = {{Shirali-Shahreza}, Sajad and Ganjali, Yashar},
year = {2018},
month = aug,
journal = {IEEE/ACM Trans. Netw.},
volume = {26},
number = {4},
pages = {1547--1561},
issn = {1063-6692},
doi = {10.1109/TNET.2018.2841397},
url = {https://doi.org/10.1109/TNET.2018.2841397},
urldate = {2018-11-21}
}
Limited flow table size in switches is a major concern for SDN applications. The common approach to overcome this problem is to identify elephant flows and solely focus on them. However, there is no gold standard to assess the effectiveness of such greedy solutions. In this paper, we formally define this problem by choosing a cost function hit ratio and an objective function to optimize the average table occupancy and present the optimum solution i.e., theoretical gold standard for it. We model the problem as a knapsack problem, analyze how its solution minimizes the table occupancy, and the similarities to and differences from the default idle timeout mechanism used in OpenFlow. We also present a new approach to minimize flow table occupancy based on the insight gained from the knapsack model analysis. Our solution expedites rule evictions by forecasting the TCP flow termination from RST/FIN packets and delays rule installation by incubating non-TCP flows. It reduces average flow table occupancy between 16%–62% in various networks with less than 1.5% reduction in hit ratio. Using three real-world packet traces, we compare the performance of our solution with the theoretically optimum solution, the static idle timeout approach used in current OpenFlow systems, and heavy hitter detection approaches that are commonly used to solve this problem. We provide in-depth analysis of when and where our approach outperforms other solutions, while discussing why it might be better to use rate-based heavy hitter detection in some scenarios.
Shirali-Shahreza, S. and Ganjali, Y. 2018. Protecting Home User Devices with an SDN-Based Firewall. IEEE Transactions on Consumer Electronics. 64, 1 (Feb. 2018), 92–100. DOI:https://doi.org/10.1109/TCE.2018.2811261.
@article{shirali_shahreza_protecting_2018,
title = {Protecting Home User Devices with an SDN-Based Firewall},
author = {{Shirali-Shahreza}, S. and Ganjali, Y.},
year = {2018},
month = feb,
journal = {IEEE Transactions on Consumer Electronics},
volume = {64},
number = {1},
pages = {92--100},
issn = {0098-3063},
doi = {10.1109/TCE.2018.2811261}
}
Internet-connected consumer electronics marketed as smart devices (also known as Internet-of-Things devices) usually lack essential security protection mechanisms. This puts user privacy and security in great danger. One of the essential steps to compromise vulnerable devices is locating them through horizontal port scans. In this paper, we focus on the problem of detecting horizontal port scans in home networks. We propose a software-defined networking (SDN)-based firewall platform that is capable of detecting horizontal port scans. Current SDN implementations (e.g., OpenFlow) do not provide access to packet-level information, which is essential for network security applications, due to performance limitations. Our platform uses FleXight, our proposed new information channel between SDN controller and data path elements to access packet-level information. FleXight uses per-flow sampling and dynamical sampling rate adjustments to provide the necessary information to the controller while keeping the overhead very low. We evaluate our solution on a large real-world packet trace from an ISP and show that our system can identify all attackers and 99% of susceptible victims with only 0.75% network overhead. We also present a detailed usability analysis of our system.
Shirali-Shahreza, S. and Ganjali, Y. 2015. ReWiFlow: Restricted Wildcard OpenFlow Rules. SIGCOMM Comput. Commun. Rev. 45, 5 (Sep. 2015), 29–35. DOI:https://doi.org/10.1145/2831347.2831352.
The ability to manage individual flows is a major benefit of Software-Defined Networking. The overheads of this fine-grained control, e.g. initial flow setup delay, can overcome the benefits, for example when we have many time-sensitive short flows. Coarse-grained control of groups of flows, on the other hand, can be very complex: each packet may match multiple rules, which requires conflict resolution. In this paper, we present ReWiFlow, a restricted class of OpenFlow wildcard rules (the fundamental way to control groups of flows in OpenFlow), which allows managing groups of flows with flexibility and without loss of performance. We demonstrate how ReWiFlow can be used to implement applications such as dynamic proactive routing. We also present a generalization of ReWiFlow, called Multi-ReWiFlow, and show how it can be used to efficiently represent access control rules collected from Stanford’s backbone network.
Hassas Yeganeh, S., Tootoonchian, A. and Ganjali, Y. 2013. On Scalability of Software-Defined Networking. IEEE Communications Magazine. 51, 2 (Feb. 2013), 136–141. DOI:https://doi.org/10.1109/MCOM.2013.6461198.
@article{hassasyeganeh_scalability_2013,
title = {On Scalability of Software-Defined Networking},
author = {Hassas Yeganeh, Soheil and Tootoonchian, Amin and Ganjali, Yashar},
year = {2013},
month = feb,
journal = {IEEE Communications Magazine},
volume = {51},
number = {2},
pages = {136--141},
issn = {0163-6804},
doi = {10.1109/MCOM.2013.6461198}
}
In this article, we deconstruct scalability concerns in software-defined networking and argue that they are not unique to SDN. We explore the often voiced concerns in different settings, discuss scalability trade-offs in the SDN design space, and present some recent research on SDN scalability. Moreover, we enumerate overlooked yet important opportunities and challenges in scalability beyond the commonly used performance metrics.
Blumenthal, D.J. et al. 2011. Integrated Photonics for Low-Power Packet Networking. IEEE Journal of Selected Topics in Quantum Electronics. 17, 2 (Mar. 2011), 458–471. DOI:https://doi.org/10.1109/JSTQE.2010.2077673.
@article{blumenthal_integrated_2011,
title = {Integrated Photonics for Low-Power Packet Networking},
author = {Blumenthal, Daniel J. and Barton, John and Beheshti, Neda and Bowers, John E. and Burmeister, Emily and Coldren, Larry A. and Dummer, Matt and Epps, Garry and Fang, Alexander and Ganjali, Yashar and Garcia, John and Koch, Brian and Lal, Vikrant and Lively, Erica and Mack, John and Ma{\v s}anovi{\'c}, Milan and McKeown, Nick and Nguyen, Kim and Nicholes, Steven C. and Park, Hyundai and Stamenic, Biljana and {Tauke-Pedretti}, Anna and Poulsen, Henrik and Sysak, Matt},
year = {2011},
month = mar,
journal = {IEEE Journal of Selected Topics in Quantum Electronics},
volume = {17},
number = {2},
pages = {458--471},
issn = {1558-4542},
doi = {10.1109/JSTQE.2010.2077673},
url = {https://ieeexplore.ieee.org/document/5685653},
urldate = {2024-08-13}
}
Communications interconnects and networks will continue to play a large role in contributing to the global carbon footprint, especially in data center and cloud-computing applications exponential growth in capacity. Key to maximizing the benefits of photonics technology is highly functional, lower power, and large-scale photonics integration. In this paper, we report on the latest advances in the photonic integration technologies used for asynchronous optical packet switching using an example photonic integrated switched optical router, the label switched optical router architecture. We report measurements of the power consumed by the photonic circuits in performing their intended function, the electronics required to bias the photonics, processing electronics, and required cooling technology. Data is presented to show that there is room (potentially greater than 10 \texttimes) for improvement in the router packet-forwarding plane. The purpose of this exercise is not to provide a comparison of all-optical versus electronic routers, rather to present a data point on actual measurements of the power contributions for various photonic integration technologies of an all-optical packet router that has been demonstrated and conclude, where the technology can move to reduce power consumption for high-capacity packet routing systems.
Beheshti, N., Burmeister, E., Ganjali, Y., Bowers, J.E., Blumenthal, D.J. and McKeown, N. 2010. Optical Packet Buffers for Backbone Internet Routers. IEEE/ACM Transactions on Networking. 18, 5 (2010), 1599–1609. DOI:https://doi.org/10.1109/TNET.2010.2048924.
@article{beheshti_optical_2010,
title = {Optical Packet Buffers for Backbone Internet Routers},
author = {Beheshti, Neda and Burmeister, Emily and Ganjali, Yashar and Bowers, John E. and Blumenthal, Daniel J. and McKeown, Nick},
year = {2010},
journal = {IEEE/ACM Transactions on Networking},
volume = {18},
number = {5},
pages = {1599--1609},
issn = {1063-6692, 1558-2566},
doi = {10.1109/TNET.2010.2048924},
url = {http://ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=547108},
urldate = {2012-01-05}
}
If optical routers are to become reality, we will need several new optical technologies, one of which is to build sufficiently large optical buffers. Building optical buffers for routers is daunting: Today’s electronic routers often hold millions of packets, which is well beyond the capabilities of optical technology. In this paper, we argue that two new results offer a solution. First, we show that the size of buffers in backbone routers can be made very small-just about 20 packets per linecard-at the expense of a small loss in throughput. Second, we show that integrated delay line optical buffers can store a few dozen packets on a photonic chip. With the combination of these two results, we conclude that future Internet routers could use optical buffers.
Markopoulou, A., Iannaccone, G., Bhattacharyya, S., Chuah, C.-N., Ganjali, Y. and Diot, C. 2008. Characterization of Failures in an Operational IP Backbone Network. Networking, IEEE/ACM Transactions on. 16, 4 (2008), 749–762. DOI:https://doi.org/10.1109/TNET.2007.902727.
@article{markopoulou_characterization_2008,
title = {Characterization of Failures in an Operational IP Backbone Network},
author = {Markopoulou, Athina and Iannaccone, Gianluca and Bhattacharyya, Supratik and Chuah, Chen-Nee and Ganjali, Yashar and Diot, Christophe},
year = {2008},
journal = {Networking, IEEE/ACM Transactions on},
volume = {16},
number = {4},
pages = {749--762},
issn = {1063-6692},
doi = {10.1109/TNET.2007.902727}
}
As the Internet evolves into a ubiquitous communication infrastructure and supports increasingly important services, its dependability in the presence of various failures becomes critical. In this paper, we analyze IS-IS routing updates from the Sprint IP backbone network to characterize failures that affect IP connectivity. Failures are first classified based on patterns observed at the IP-layer; in some cases, it is possible to further infer their probable causes, such as maintenance activities, router-related and optical layer problems. Key temporal and spatial characteristics of each class are analyzed and, when appropriate, parameterized using well-known distributions. Our results indicate that 20% of all failures happen during a period of scheduled maintenance activities. Of the unplanned failures, almost 30% are shared by multiple links and are most likely due to router-related and optical equipment-related problems, respectively, while 70% affect a single link at a time. Our classification of failures reveals the nature and extent of failures in the Sprint IP backbone. Furthermore, our characterization of the different classes provides a probabilistic failure model, which can be used to generate realistic failure scenarios, as input to various network design and traffic engineering problems.
Ganjali, Y. and McKeown, N. 2006. Update on Buffer Sizing in Internet Routers. SIGCOMM Comput. Commun. Rev. 36, 5 (Oct. 2006), 67–70. DOI:https://doi.org/10.1145/1163593.1163605.
@article{ganjali_update_2006,
title = {Update on Buffer Sizing in Internet Routers},
author = {Ganjali, Yashar and McKeown, Nick},
year = {2006},
month = oct,
journal = {SIGCOMM Comput. Commun. Rev.},
volume = {36},
number = {5},
pages = {67--70},
issn = {0146-4833},
doi = {10.1145/1163593.1163605},
url = {https://dl.acm.org/doi/10.1145/1163593.1163605},
urldate = {2024-08-13}
}
In the past two years, several papers have proposed rules that suggest two to five orders of magnitude reduction in Internet core router buffers. Others present scenarios where buffer sizes need to be significantly increased. So why the different rules? In this paper we briefly compare the different results and proposals, and summarize some recent preliminary experiments to validate the proposals. We’ll see that different results apply to different parts of the network, and depend on several assumptions. For example, we believe that buffers can be safely reduced by an order of magnitude in the routers in service provider backbone networks; but it would be premature to reduce them in routers closer to the edge.
Ganjali, Y., Keshavarzian, A. and Shah, D. 2005. Cell Switching versus Packet Switching in Input-Queued Switches. IEEE/ACM Trans. Netw. 13, 4 (Aug. 2005), 782–789. DOI:https://doi.org/10.1109/TNET.2005.852884.
@article{ganjali_cell_2005,
title = {Cell Switching versus Packet Switching in Input-Queued Switches},
author = {Ganjali, Yashar and Keshavarzian, Abtin and Shah, Devavrat},
year = {2005},
month = aug,
journal = {IEEE/ACM Trans. Netw.},
volume = {13},
number = {4},
pages = {782--789},
issn = {1063-6692},
doi = {10.1109/TNET.2005.852884},
url = {https://dl.acm.org/doi/10.1109/TNET.2005.852884},
urldate = {2024-08-13}
}
Input Queued (IQ) switches have been well studied in the past two decades by researchers. The main problem concerning IQ switches is scheduling the switching fabric in order to transfer packets from input ports to output ports. Scheduling is relatively easier when all packets are of the same size. However, in practice, packets are of variable length. In the current implementation of switches, variable length packets are segmented into fixed length packets–also knowns as cells–for the purpose of scheduling. However, such cell-based switching comes with some significant disadvantages: (a) loss of bandwidth due to the existence of incomplete cells; and (b) additional overhead of segmentation of packets and re-assembly of cells. This is a strong motivation to study packet-based scheduling, i.e., scheduling the transfer of packets without segmenting them.The problem of packet scheduling was first considered by Marsan et al. They showed that under any admissible Bernoulli IID (independent and identically distributed) arrival traffic, a simple modification of the Maximum Weight Matching (MWM) algorithm achieves 100% throughput. In this paper, we first show that no work-conserving (i.e., maximal) packet-based algorithm is stable for arbitrary admissible arrival processes. Thus, the results of Marsan et al. are strongly dependent on the arrival distribution. Next, we propose a new class of "waiting" algorithms. We show that the "waiting"-MWM algorithm is stable for any admissible traffic using the fluid limit technique.We would like to note that the algorithms presented in this paper are distribution independent or universal. The algorithms and proof methods of this paper may be useful in the context of other scheduling problems.
Enachescu, M., Ganjali, Y., Goel, A., McKeown, N. and Roughgarden, T. 2005. Part III: Routers with Very Small Buffers. SIGCOMM Comput. Commun. Rev. 35, 3 (Jul. 2005), 83–90. DOI:https://doi.org/10.1145/1070873.1070886.
@article{enachescu_part_2005,
title = {Part III: Routers with Very Small Buffers},
shorttitle = {Part III},
author = {Enachescu, Mihaela and Ganjali, Yashar and Goel, Ashish and McKeown, Nick and Roughgarden, Tim},
year = {2005},
month = jul,
journal = {SIGCOMM Comput. Commun. Rev.},
volume = {35},
number = {3},
pages = {83--90},
issn = {0146-4833},
doi = {10.1145/1070873.1070886},
url = {https://dl.acm.org/doi/10.1145/1070873.1070886},
urldate = {2024-08-13}
}
Internet routers require buffers to hold packets during times of congestion. The buffers need to be fast, and so ideally they should be small enough to use fast memory technologies such as SRAM or all-optical buffering. Unfortunately, a widely used rule-of-thumb says we need a bandwidth-delay product of buffering at each router so as not to lose link utilization. This can be prohibitively large. In a recent paper, Appenzeller et al. challenged this rule-of-thumb and showed that for a backbone network, the buffer size can be divided by pN without sacrificing throughput, where N is the number of ows sharing the bottleneck. In this paper, we explore how buffers in the backbone can be significantly reduced even more, to as little as a few dozen packets, if we are willing to sacrifice a small amount of link capacity. We argue that if the TCP sources are not overly bursty, then fewer than twenty packet buffers are sufficient for high throughput. Specifically, we argue that O(log W) buffers are sufficient, where W is the window size of each ow. We support our claim with analysis and a variety of simulations. The change we need to make to TCP is minimal–each sender just needs to pace packet injections from its window. Moreover, there is some evidence that such small buffers are sufficient even if we don’t modify the TCP sources so long as the access network is much slower than the backbone, which is true today and likely to remain true in the future. We conclude that buffers can be made small enough for all-optical routers with small integrated optical buffers.
@article{biedl_balanced_2005,
title = {Balanced Vertex-Orderings of Graphs},
author = {Biedl, Therese and Chan, Timothy and Ganjali, Yashar and Hajiaghayi, Mohammad Taghi and Wood, David R.},
year = {2005},
month = apr,
journal = {Discrete Applied Mathematics},
volume = {148},
number = {1},
pages = {27--48},
issn = {0166-218X},
doi = {10.1016/j.dam.2004.12.001},
url = {https://www.sciencedirect.com/science/article/pii/S0166218X04003828},
urldate = {2024-08-13}
}
In this paper we consider the problem of determining a balanced ordering of the vertices of a graph; that is, the neighbors of each vertex v are as evenly distributed to the left and right of v as possible. This problem, which has applications in graph drawing for example, is shown to be NP-hard, and remains NP-hard for bipartite simple graphs with maximum degree six. We then describe and analyze a number of methods for determining a balanced vertex-ordering, obtaining optimal orderings for directed acyclic graphs, trees, and graphs with maximum degree three. For undirected graphs, we obtain a 13/8-approximation algorithm. Finally we consider the problem of determining a balanced vertex-ordering of a bipartite graph with a fixed ordering of one bipartition. When only the imbalances of the fixed vertices count, this problem is shown to be NP-hard. On the other hand, we describe an optimal linear time algorithm when the final imbalances of all vertices count. We obtain a linear time algorithm to compute an optimal vertex-ordering of a bipartite graph with one bipartition of constant size.
Ganjali, Y. and Hajiaghayi, M.T. 2004. Characterization of Networks Supporting Multi-Dimensional Linear Interval Routing Schemes. Theoretical Computer Science. 326, 1 (Oct. 2004), 103–116. DOI:https://doi.org/10.1016/j.tcs.2004.06.013.
@article{ganjali_characterization_2004,
title = {Characterization of Networks Supporting Multi-Dimensional Linear Interval Routing Schemes},
author = {Ganjali, Yashar and Hajiaghayi, MohammadTaghi},
year = {2004},
month = oct,
journal = {Theoretical Computer Science},
volume = {326},
number = {1},
pages = {103--116},
issn = {0304-3975},
doi = {10.1016/j.tcs.2004.06.013},
url = {https://www.sciencedirect.com/science/article/pii/S0304397504003986},
urldate = {2024-08-13}
}
An Interval Routing Scheme (IRS) is a well-known, space efficient routing strategy for routing messages in a distributed network. In this scheme, each node of the network is assigned an integer label and each link at each node is labeled with an interval. The interval assigned to a link e at a node v indicates the set of destination addresses of the messages which should be forwarded through e at v. A Multi-dimensional Interval Routing Scheme (MIRS) is a generalization of IRS in which each node is assigned a multi-dimensional label (which is a list of d integers for the d-dimensional case). The labels assigned to the links of the network are also multi-dimensional (a list of d 1-dimensional intervals). The class of networks supporting linear IRS (in which the intervals are not cyclic) is already known for the one-dimensional case (13th Annu. ACM Symp. Principles of Distributed Computing (PODC), ACM Press, New York, August 1994, pp. 216–224). In this paper, we generalize this result and completely characterize the class of networks supporting linear MIRS (or MLIRS) for a given number of dimensions d. We show that by increasing d, the class of networks supporting MLIRS is strictly expanded. We also give a characterization of the class of networks supporting strict MLIRS (which is an MLIRS in which the intervals assigned to the links incident to a node v, does not contain the label of v).
Ganjali, Y. 2003. Optimum Multi-Dimensional Interval Routing Schemes on Networks with Dynamic Cost Links. Computing and Informatics. 22, 1 (2003), 1–18.
@article{ganjali_optimum_2003,
title = {Optimum Multi-Dimensional Interval Routing Schemes on Networks with Dynamic Cost Links},
author = {Ganjali, Yashar},
year = {2003},
journal = {Computing and Informatics},
volume = {22},
number = {1},
pages = {1--18}
}
Hajiaghayi, M.T. and Ganjali, Y. 2002. A Note on Consecutive Ones Submatrix Problem. Information Processing Letters. 83, 3 (Aug. 2002), 163–166.
@article{hajiaghayi_note_2002,
title = {A Note on Consecutive Ones Submatrix Problem},
author = {Hajiaghayi, MohammadTaghi and Ganjali, Yashar},
year = {2002},
month = aug,
journal = {Information Processing Letters},
volume = {83},
number = {3},
pages = {163--166},
url = {http://www.cs.toronto.edu/~yganjali/test}
}
Ganjali, Y., Ghebleh, M., Hajiabolhasan, H., Mirzazadeh, M. and Sadjad, B.S. 2002. Uniquely 2-List Colorable Graphs. Discrete Applied Mathematics. 119, 3 (Jul. 2002), 217–225.
@article{ganjali_uniquely_2002,
title = {Uniquely 2-List Colorable Graphs},
author = {Ganjali, Yashar and Ghebleh, Mohammad and Hajiabolhasan, Hossein and Mirzazadeh, Mahdi and Sadjad, Bashir S.},
year = {2002},
month = jul,
journal = {Discrete Applied Mathematics},
volume = {119},
number = {3},
pages = {217--225}
}
Conference and Workshop Papers
Bahnasy, M., Hossein Mortazavi, S., Munir, A., Shafieirad, H. and Ganjali, Y. 2023. Host-Assisted Transport Layer in Data Centers Using Network-Aware Rate Adjustment. GLOBECOM 2023 - 2023 IEEE Global Communications Conference (Dec. 2023), 4674–4679.
@inproceedings{bahnasy_hostassisted_2023,
title = {Host-Assisted Transport Layer in Data Centers Using Network-Aware Rate Adjustment},
booktitle = {GLOBECOM 2023 - 2023 IEEE Global Communications Conference},
author = {Bahnasy, Mahmoud and Hossein Mortazavi, Seyed and Munir, Ali and Shafieirad, Hossein and Ganjali, Yashar},
year = {2023},
month = dec,
pages = {4674--4679},
issn = {2576-6813},
doi = {10.1109/GLOBECOM54140.2023.10437035},
url = {https://ieeexplore.ieee.org/document/10437035},
urldate = {2024-04-23}
}
Next generation applications for datacenters, such as Distributed Machine Learning (DML) and Big Data, have complex communication patterns that demand a scalable, stateless and application-aware optimal transport protocol to maximize network utilization and improve application performance. Recent transport protocols either provide limited benefits due to lack of information sharing between application and network; or implement complex stateful mechanisms to improve the application performance. In this paper, we present Omni- Transport Mechanism (Omni-TM) as a message-based congestion control protocol. Omni-TM allows exchanging message information with the network to negotiate the optimal transmission rate without maintaining a per-flow state at the switches (i.e., stateless). Omni- Tmis designed to reach maximum link capacity in one shot. Our simulation results show that Omni- Tmdemonstrates better traffic control decisions (i.e., close to zero queue length while maintaining high link utilization). Furthermore, Omni- Tmreduces Flow Completion Time (FCT) up to 45 % in a realistic workload compared to DCTCP.
Pazhooheshy, P., Abbasloo, S. and Ganjali, Y. 2023. Harnessing ML For Network Protocol Assessment: A Congestion Control Use Case. Proceedings of the 22nd ACM Workshop on Hot Topics in Networks (New York, NY, USA, Nov. 2023), 213–219.
@inproceedings{pazhooheshy_harnessing_2023a,
title = {Harnessing ML For Network Protocol Assessment: A Congestion Control Use Case},
shorttitle = {Harnessing ML For Network Protocol Assessment},
booktitle = {Proceedings of the 22nd ACM Workshop on Hot Topics in Networks},
author = {Pazhooheshy, Parsa and Abbasloo, Soheil and Ganjali, Yashar},
year = {2023},
month = nov,
series = {HotNets '23},
pages = {213--219},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/3626111.3628182},
url = {https://dl.acm.org/doi/10.1145/3626111.3628182},
urldate = {2024-08-10},
isbn = {9798400704154}
}
In this paper, our primary objective is to showcase that the application of machine learning techniques extends beyond network protocol design. We aim to demonstrate that performance assessment of network protocols, a vital aspect of improving network infrastructures and developing better protocol designs, can be modernized through the utilization of machine learning. As a step towards this goal, we have designed and introduced Mahak, the first tool that harnesses active learning techniques to automate the performance assessment of congestion control schemes. Mahak actively learns to optimize the evaluation process of congestion control schemes so that they can generate their performance maps over a desired space without exhaustively testing them in every scenario. Mahak treats schemes under the test as black boxes. This protocol-agnostic aspect of Mahak enables users to directly assess the performance of the actual implementation of a protocol instead of their over-simplified mathematical models or simplified simulated versions.
Zadeh, S.A., Munir, A., Bahnasy, M.M., Ketabi, S. and Ganjali, Y. 2023. On Augmenting TCP/IP Stack via eBPF. Proceedings of the 1st Workshop on eBPF and Kernel Extensions (New York, NY, USA, Sep. 2023), 15–20.
@inproceedings{zadeh_augmenting_2023,
title = {On Augmenting TCP/IP Stack via eBPF},
booktitle = {Proceedings of the 1st Workshop on eBPF and Kernel Extensions},
author = {Zadeh, Sepehr Abbasi and Munir, Ali and Bahnasy, Mahmoud Mohamed and Ketabi, Shiva and Ganjali, Yashar},
year = {2023},
month = sep,
series = {eBPF '23},
pages = {15--20},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/3609021.3609300},
url = {https://dl.acm.org/doi/10.1145/3609021.3609300},
urldate = {2023-10-10},
isbn = {9798400702938}
}
As the data center networks’ bandwidth-delay product is increasing and the applications are moving to nano services (with many small flows), managing flows in the network is becoming more challenging. Current TCP/IP stack faces fundamental limitations to meet these challenges. First, it lacks the ability to accurately estimate the network state under dynamic network settings. Second, the current stack is not flexible enough to be extended easily. In this work, we propose a framework, Augmenter, that augments (i.e., increases the network visibility of) the TCP/IP stack to address these challenges. Leveraging eBPF, Augmenter gathers the state of ongoing flows and uses this information to manage other flows that are currently active or arriving in the future. We present one specific use case of setting the initial congestion window of flows dynamically based on network conditions. Our initial tests, show that Augmenter can improve the application performance by up to 1.4x compared to the fixed initial window-based solutions. Implementing Augmenter in the TCP/IP stack itself is not trivial. Augmenter employs eBPF to implement its desired functionality as it enables introducing such changes relatively easy. We discuss potential challenges and solutions in designing and implementing Augmenter applications.
Zadeh, S.A., Zandi, F., Buckley, M. and Ganjali, Y. 2023. Meta-Migration: Reducing Switch Migration Tail Latency Through Competition. 2023 IFIP Networking Conference (IFIP Networking) (Jun. 2023), 1–9.
@inproceedings{zadeh_metamigration_2023a,
title = {Meta-Migration: Reducing Switch Migration Tail Latency Through Competition},
shorttitle = {Meta-Migration},
booktitle = {2023 IFIP Networking Conference (IFIP Networking)},
author = {Zadeh, Sepehr Abbasi and Zandi, Farid and Buckley, Matthew and Ganjali, Yashar},
year = {2023},
month = jun,
pages = {1--9},
issn = {1861-2288},
doi = {10.23919/IFIPNetworking57963.2023.10186446},
url = {https://ieeexplore.ieee.org/document/10186446},
urldate = {2024-08-10}
}
Resource management in distributed network control planes plays a vital role in the performance of the data plane and therefore the performance of network applications. Overwhelmed controller instances or underutilized instances could reshape their workloads by exchanging their load, i.e., switches that they control. To safely implement this exchange procedure, switch migration protocols are being used. As the migration procedure pauses processing new flows for a few milliseconds, these protocols are designed to be as fast as possible. Faster protocols add to the agility of the network to rapidly cope with the changing demand. In this paper, we introduce a general framework, called Meta-Migration, which focuses on expediting the existing time-sensitive controller load migration protocols. Based on the observation that these protocols impose low overheads on the involved parties, we modify them in a way that they can run in parallel toward multiple candidate destinations. Unlike the usual Fixed protocols that have to decide their destinations before running the protocol, here we rely on the real-time probes that we obtain from multiple systems and commit to only one of them in the middle of the procedure. Typically, migrations can complete on sub-second timescales, but sudden traffic bursts or system-level glitches can significantly slow down these protocols. We observe that by using Meta-Migration, we can dramatically diminish these negative effects. We show theoretical justifications for why this approach improves the overall performance of the migration, namely, its mean finishing time, and the tail latency of the migration. In addition, by developing a distributed controller simulator over real physical devices, we thoroughly measure the effectiveness of this approach as well as its incurred overheads. Our testbed results show up to a 53% tail reduction in the migration time.
Ketabi, S., Chen, H., Dong, H. and Ganjali, Y. 2023. A Deep Reinforcement Learning Framework for Optimizing Congestion Control in Data Centers. NOMS 2023-2023 IEEE/IFIP Network Operations and Management Symposium (May 2023), 1–7.
@inproceedings{ketabi_deep_2023,
title = {A Deep Reinforcement Learning Framework for Optimizing Congestion Control in Data Centers},
booktitle = {NOMS 2023-2023 IEEE/IFIP Network Operations and Management Symposium},
author = {Ketabi, Shiva and Chen, Hongkai and Dong, Haiwei and Ganjali, Yashar},
year = {2023},
month = may,
pages = {1--7},
issn = {2374-9709},
doi = {10.1109/NOMS56928.2023.10154411},
url = {https://ieeexplore.ieee.org/document/10154411},
urldate = {2024-08-10}
}
Various congestion control protocols have been designed to achieve high performance in different network environments. Modern online learning solutions that delegate the congestion control actions to a machine cannot properly converge in the stringent time scales of data centers. We leverage multi-agent reinforcement learning to design a system for dynamic tuning of congestion control parameters at end-hosts in a data center. The system includes agents at the end-hosts to monitor and report the network and traffic states, and agents to run the reinforcement learning algorithm given the states. Based on the state of the environment, the system generates congestion control parameters that optimize network performance metrics such as throughput and latency. As a case study, we examine BBR, an example of a prominent recently-developed congestion control protocol. Our experiments demonstrate that the proposed system has the potential to mitigate the problems of static parameters.
Zandi, F., Zadeh, S.A., Abbasloo, S., Pazhooheshy, P., Ganjali, Y. and Hu, Z. 2023. Live Stateful Migration of a Virtual Sub-Network. NOMS 2023-2023 IEEE/IFIP Network Operations and Management Symposium (May 2023), 1–9.
@inproceedings{zandi_live_2023,
title = {Live Stateful Migration of a Virtual Sub-Network},
booktitle = {NOMS 2023-2023 IEEE/IFIP Network Operations and Management Symposium},
author = {Zandi, Farid and Zadeh, Sepehr Abbasi and Abbasloo, Soheil and Pazhooheshy, Parsa and Ganjali, Yashar and Hu, Zhenhua},
year = {2023},
month = may,
pages = {1--9},
issn = {2374-9709},
doi = {10.1109/NOMS56928.2023.10154281},
url = {https://ieeexplore.ieee.org/document/10154281},
urldate = {2024-08-10}
}
Traffic processing on cloud-scale bandwidths has given rise to a new type of network structure, comprising a large number of highly-structured virtual entities working in close harmony. This structure, which we call a virtual sub-network, might be in need of migration, for reasons of load-balancing, maintenance, and disaster prevention. In this paper, we argue that the common migration schemes are not adequate for the complexity of this task. Therefore, we present Qanat, a migration system specifically optimized for the live migration of a virtual sub-network in its entirety to a different physical location. We show how Qanat employs widely-used techniques, such as traffic prioritization, buffering, and network tunnels, to overcome the main issues of live migration. In the paper, we categorize the main challenges of the migration task, provide an analytical study of Qanat’s algorithms, and measure its performance metrics through large-scale simulations. We conclude that Qanat can efficiently and transparently migrate virtual sub-networks and can provide a useful tool for system administrators.
Munir, A., Mortazavi, S.H., Bahnasy, M.M., Baniamerian, A., Wang, S., Guan, S. and Ganjali, Y. 2022. SmartTags: Bridging Applications and Network for Proactive Performance Management. Proceedings of the ACM SIGCOMM Workshop on Network-Application Integration (New York, NY, USA, Aug. 2022), 46–52.
@inproceedings{munir_smarttags_2022,
title = {SmartTags: Bridging Applications and Network for Proactive Performance Management},
shorttitle = {SmartTags},
booktitle = {Proceedings of the ACM SIGCOMM Workshop on Network-Application Integration},
author = {Munir, Ali and Mortazavi, Seyed Hossein and Bahnasy, Mahmoud Mohamed and Baniamerian, Amir and Wang, Shimiao and Guan, Shichao and Ganjali, Yashar},
year = {2022},
month = aug,
series = {NAI '22},
pages = {46--52},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/3538401.3546601},
url = {https://doi.org/10.1145/3538401.3546601},
urldate = {2022-08-28},
isbn = {978-1-4503-9395-9}
}
Sudden changes in the applications and events in the network are often related. Many of the datacenter applications go through sudden state changes (such as a query-response in MemCached application) that may result in an event in the network (such as utilization, packet drops, etc.). Existing works do not fully leverage the relationship between application state changes and network events and as a result provide limited performance improvements. An ideal application and network management system should be able to automatically identify the sources of sudden changes in the application, host or network and relate these changes to network events to enable proactive network management. In this work, we propose SmartTags, a system that automatically learns which of the application state changes (Tags) are related to the network events, and uses this information for proactive network management. At a high level, smartTags is orthogonal to current NAI approaches. It provides a systemic way for application developers and network designers to automatically learn the relationship between application behavior and network events. Through very simple small scale real testbed based experiments, we demonstrate that smartTags can improve the training time of a distributed machine learning application by 27% while minimizing loss to zero. Similarly, it can improve the query completion time of MemCached by 32% while achieving near zero loss. We envision much more gains in large scale distributed systems.
Mortazavi, S.H., Munir, A., Bahnasy, M.M., Dong, H., Wang, S. and Ganjali, Y. 2022. EarlyBird: Automating Application Signalling for Network Application Integration in Datacenters. Proceedings of the ACM SIGCOMM Workshop on Network-Application Integration (New York, NY, USA, Aug. 2022), 40–45.
@inproceedings{mortazavi_earlybird_2022,
title = {EarlyBird: Automating Application Signalling for Network Application Integration in Datacenters},
shorttitle = {EarlyBird},
booktitle = {Proceedings of the ACM SIGCOMM Workshop on Network-Application Integration},
author = {Mortazavi, Seyed Hossein and Munir, Ali and Bahnasy, Mahmoud Mohamed and Dong, Haiwei and Wang, Shimiao and Ganjali, Yashar},
year = {2022},
month = aug,
series = {NAI '22},
pages = {40--45},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/3538401.3546599},
url = {https://doi.org/10.1145/3538401.3546599},
urldate = {2022-08-28},
isbn = {978-1-4503-9395-9}
}
Many recent studies in datacenter networking have proposed the idea of using information from applications for optimizing and resource planning. These Application-Aware Networks generally assume that applications can provide an accurate view about their requirements from the network and their traffic characteristics in real time. However, relying on the applications and developers to convey the traffic information is not realistic. We believe that automating the process of information extraction from applications is a crucial step towards realizing the idea of Network-Application Integration (NAI). In this paper, we investigate whether we can automatically identify places in the application code that, when executed, lead to predictable changes in the host’s network output. By augmenting the application code at these execution places, we can generate explicit signals that can be used to predict local network events (such as changes in rate, bursts, etc.). This creates a mechanism for automatic adjustment of the network based on application signals. We propose a combination of simple heuristics and learning methods to minimize the burden of the application developer for NAI. To the best of our knowledge, this is the first study that attempts to automatically realize NAI on the application. Our experimental evaluation shows a high accuracy (over 87%) of our prototype in predicting local network events such as micro-bursts.
Baniamerian, A., Munir, A., Sobhani, A., Bahnasy, M.M., Thomas, R., Yan, S., Chu, X. and Ganjali, Y. 2022. NCE: An ECN Dual Mechanism to Mitigate Micro-Bursts. Proceedings of the ACM SIGCOMM Workshop on Network-Application Integration (New York, NY, USA, Aug. 2022), 7–12.
@inproceedings{baniamerian_nce_2022,
title = {NCE: An ECN Dual Mechanism to Mitigate Micro-Bursts},
shorttitle = {NCE},
booktitle = {Proceedings of the ACM SIGCOMM Workshop on Network-Application Integration},
author = {Baniamerian, Amir and Munir, Ali and Sobhani, Ashkan and Bahnasy, Mahmoud Mohamed and Thomas, Ron and Yan, Si and Chu, Xingjun and Ganjali, Yashar},
year = {2022},
month = aug,
series = {NAI '22},
pages = {7--12},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/3538401.3546597},
url = {https://doi.org/10.1145/3538401.3546597},
urldate = {2022-08-28},
isbn = {978-1-4503-9395-9}
}
Due to unique characteristics of datacenter applications, such as high volume and dynamic traffic patterns of short flows, micro-bursts significantly impact the application performance. Moreover, the reactive nature of congestion control (CC) algorithms makes CC ineffective against micro-bursts. This paper introduces the Notification of Change Events (NCE) mechanism. Compared to the Explicit Congestion Notification (ECN) schemes, where switches send information to hosts, NCE uses a dual approach that sends information from "transport layer" of sources to the network. This design provides a proactive in-network solution that mitigates micro-bursts in real-time with the help of the information from the application. Simulation results demonstrate NCE can improve the flow completion time of short flows up to 62% with negligible impacts on long flow throughput.
Buckley, M., Abbasi-Zadeh, S., Beiruti, M.A., Abbasloo, S. and Ganjali, Y. 2022. Switch Migration Scheduling in Distributed SDN Controllers. 2022 IEEE 8th International Conference on Network Softwarization (NetSoft) (Jun. 2022), 348–356.
@inproceedings{buckley_switch_2022,
title = {Switch Migration Scheduling in Distributed SDN Controllers},
booktitle = {2022 IEEE 8th International Conference on Network Softwarization (NetSoft)},
author = {Buckley, Matthew and {Abbasi-Zadeh}, Sepehr and Beiruti, Mohammad Amin and Abbasloo, Soheil and Ganjali, Yashar},
year = {2022},
month = jun,
pages = {348--356},
issn = {2693-9789},
doi = {10.1109/NetSoft54395.2022.9844120}
}
Due to the dynamic nature of traffic, networks must rapidly adapt to changing conditions. This is especially true in the context of the control plane which must ensure continuous and seamless operation. Switch migration, the process of changing the controller associated with a switch, is an important tool in facilitating this goal. In this work, we study the problem of minimizing the overall time to migrate a set of switches. We examine the problem subject to constraints on controller resources and QoS groups. We show that the problem is NP-hard and provide heuristic algorithms for solving large instances in practice. Through extensive experiments, we demonstrate that the heuristics achieve performance close to optimal while reducing the running time by several orders of magnitude.
Ramezan, G., Abdelnasser, A. and Ganjali, Y. 2022. KnowMe: A Module to Improve the Efficiency of Resource Allocation in Data Center Networks. 2022 12th International Conference on Cloud Computing, Data Science Engineering (Confluence) (Jan. 2022), 176–183.
@inproceedings{ramezan_knowme_2022,
title = {KnowMe: A Module to Improve the Efficiency of Resource Allocation in Data Center Networks},
shorttitle = {KnowMe},
booktitle = {2022 12th International Conference on Cloud Computing, Data Science Engineering (Confluence)},
author = {Ramezan, Gholamreza and Abdelnasser, Amr and Ganjali, Yashar},
year = {2022},
month = jan,
pages = {176--183},
doi = {10.1109/Confluence52989.2022.9734200}
}
In recent years, a new paradigm called network-application integration (NAI) has been introduced which enables applications to express their requirements to the network. By applying NAI in data center networks (DCNs), the DCN controller can better serve applications. We refer to this as data center network-application integration (DC-NAI). However, a critical issue in DC-NAI is how applications use the assigned resources. Selfish applications may under-utilize or over-utilize the assigned resources which can negatively impact the existing applications. To overcome this challenge, in this paper, we propose a DCN controller add-on module called KnowMe that takes the applications’ behavior into consideration and improves the efficiency of the resource allocation. The KnowMe module is composed of two components: the resource utilization prediction and the resource assignment learning. The first component predicts the future applications’ resource usage by looking at their previous resource usage behavior. The second component makes decisions about the applications’ resource requests considering the current network status. Simulation results confirm the efficiency of our proposed module and its ability to mitigate the selfish behavior of applications. The results show that under network resource limitation, the selfish applications in KnowMe-enabled DCNs experience 40% higher resource request rejection rate comparing to the well-behaved applications.
Mortazavi, S.H., Shafieirad, H., Bahnasy, M., Munir, A., Cheng, Y., Das, A. and Ganjali, Y. 2021. Accord: Application-Driven Networking in the Datacenter. Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing (New York, NY, USA, Dec. 2021), 1–10.
@inproceedings{mortazavi_accord_2021,
title = {Accord: Application-Driven Networking in the Datacenter},
shorttitle = {Accord},
booktitle = {Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing},
author = {Mortazavi, Seyed Hossein and Shafieirad, Hossein and Bahnasy, Mahmoud and Munir, Ali and Cheng, Yuanhui and Das, Anudeep and Ganjali, Yashar},
year = {2021},
month = dec,
series = {UCC '21},
pages = {1--10},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/3468737.3494102},
url = {https://doi.org/10.1145/3468737.3494102},
urldate = {2021-12-18},
isbn = {978-1-4503-8564-0}
}
Resource optimization algorithms in the cloud are ever more data-driven and decision-making has become reliant on more and more data flowing from different cloud components. Applications and the network control layer on the other hand mainly operate in isolation without direct communication. Recently, increased integration between the network and application has been advocated to benefit both the application and the network but the information exchange has mostly been limited to flow level information. We argue that in the realm of datacenter networks, sharing additional information such as the function processing times and deployment data for planning jobs and tasks can result in major optimization benefits for the network. In this study we present Accord as a Network Application Integration solution to achieve a holistic network-application management solution. We propose a protocol as an API between the network and application then we build a system that uses the processing and networking data from the application to perform network scheduling and routing optimizations. We demonstrate that for a sample distributed learning application, an Accord enhanced solution that uses the application processing information can yield up to 27.8% reduction in Job Completion Time (JCT). In addition, we show how Accord can yield better results for routing decisions through a reinforcement learning algorithm that outperforms first shortest path first by %13.
Ketabi, S. and Ganjali, Y. 2020. Hierarchical Congestion Control (HCC): Cooperation of Uncorrelated Flows for Better Fairness and Throughput. NOMS 2020 - 2020 IEEE/IFIP Network Operations and Management Symposium (Apr. 2020), 1–5.
@inproceedings{ketabi_hierarchical_2020,
title = {Hierarchical Congestion Control (HCC): Cooperation of Uncorrelated Flows for Better Fairness and Throughput},
shorttitle = {Hierarchical Congestion Control (HCC)},
booktitle = {NOMS 2020 - 2020 IEEE/IFIP Network Operations and Management Symposium},
author = {Ketabi, S. and Ganjali, Y.},
year = {2020},
month = apr,
pages = {1--5},
issn = {2374-9709},
doi = {10.1109/NOMS47738.2020.9110367}
}
Congestion control protocols face several challenges for achieving max-min fairness and high throughput. First, each flow has a limited view of the network state. In the absence of a centralized congestion control entity, coordination is left (directly or indirectly) to individual flow sources. Second, most flows are very volatile by nature: flow rates/demands change significantly from one instant to another.In this paper, we present a hierarchical congestion control scheme to tackle these challenges. We aggregate flows with low correlation in a hierarchical manner, and recursively compute and allocate rates to these flows. Our experiments show that correlation-aware aggregation results in significant improvements in terms of fairness (around 15% increase) and throughput (around 100% increase) compared to schemes that neglect the correlation metric. Also, we evaluate our rate allocation scheme and show that it can achieve near optimal rates for 95th percentile of flows in most time intervals.
Beiruti, M.A. and Ganjali, Y. 2020. Load Migration in Distributed SDN Controllers. NOMS 2020 - 2020 IEEE/IFIP Network Operations and Management Symposium (Apr. 2020), 1–9.
@inproceedings{beiruti_load_2020,
title = {Load Migration in Distributed SDN Controllers},
booktitle = {NOMS 2020 - 2020 IEEE/IFIP Network Operations and Management Symposium},
author = {Beiruti, Mohammad Amin and Ganjali, Yashar},
year = {2020},
month = apr,
pages = {1--9},
issn = {2374-9709},
doi = {10.1109/NOMS47738.2020.9110292}
}
Distributed control solutions were introduced to address controller reliability and scalability issues in Software-Defined Networking (SDN). The dynamic nature of network traffic can lead to load imbalance amongst controller instances. A highly loaded controller instance can be slower in responding to datapath queries, and can slow down the entire control platform as state synchronization, and consensus amongst controller instances are performed cooperatively. In this paper, we present a new and efficient load migration protocol, called ERC, for shifting input load associated with overloaded controller instances towards lightly loaded instances. Our protocol has three distinguishing properties compared to prior works on this area: it is extremely efficient, resilient to failures during migration, and ensures consistency among all controller instances. ERC can be used for a wide range of network applications including load balancing, power saving, and resource optimization. It is also significantly more efficient than existing load migration protocols with 25-55% reduction in migration time, and 10-20% reduction in required migration buffer size.
Beheshti, N., Lapukhov, P. and Ganjali, Y. 2019. Buffer Sizing Experiments at Facebook. Proceedings of the 2019 Workshop on Buffer Sizing (Palo Alto, CA, USA, Dec. 2019), 1–6.
@inproceedings{beheshti_buffer_2019,
title = {Buffer Sizing Experiments at Facebook},
booktitle = {Proceedings of the 2019 Workshop on Buffer Sizing},
author = {Beheshti, Neda and Lapukhov, Petr and Ganjali, Yashar},
year = {2019},
month = dec,
series = {BS '19},
pages = {1--6},
publisher = {Association for Computing Machinery},
address = {Palo Alto, CA, USA},
doi = {10.1145/3375235.3375244},
url = {https://doi.org/10.1145/3375235.3375244},
urldate = {2020-05-28},
isbn = {978-1-4503-7745-4}
}
In this paper, we present the results of several rounds of buffer sizing experiments at Facebook. We describe how we change buffer sizes in data center and backbone routers and study the impact of buffer sizes on several performance metrics, including flow completion time, latency, link utilization, and packet drop rate. Our observations suggest reducing buffers from millions of packets to a few thousand, and even a few hundred packets, does not lead to a general breakdown in network performance. In some cases we observed mostly tolerable degradation in some metrics (e.g., packet drop rates) and significant enhancements in others (e.g., latency).
Ketabi, S. and Ganjali, Y. 2019. Perfect Is the Enemy of Good: Lloyd-Max Quantization for Rate Allocation in Congestion Control Plane. 2019 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS) (Sep. 2019), 1–4.
@inproceedings{ketabi_perfect_2019,
title = {Perfect Is the Enemy of Good: Lloyd-Max Quantization for Rate Allocation in Congestion Control Plane},
shorttitle = {Perfect Is the Enemy of Good},
booktitle = {2019 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)},
author = {Ketabi, Shiva and Ganjali, Yashar},
year = {2019},
month = sep,
pages = {1--4},
doi = {10.1109/ANCS.2019.8901887}
}
Decoupling congestion control plane from datapath expedites the development of new congestion control solutions and creates opportunities for explicit rate allocation schemes. However, dealing with large numbers of flows remains a major challenge. Max-min fairness - the gold standard for rate allocation - has a running complexity proportional to the number of flows, which might be prohibitive in large-scale networks. To accelerate explicit rate allocation, we suggest using rate quantization, i.e. mapping the continuous range of flow rates to a small number of bins. We use Lloyd-max, a quantization method that generates bins according to the distribution of flow rates, to dynamically adjust the quantization bins over time. Our experimental evaluation shows that the distortion caused by this quantization scheme is small, while reducing the max-min rate allocation running time by 60 - 90%.
Ghorbani, S., Yang, Z., Godfrey, P.B., Ganjali, Y. and Firoozshahian, A. 2017. DRILL: Micro Load Balancing for Low-Latency Data Center Networks. Proceedings of the Conference of the ACM Special Interest Group on Data Communication (SIGCOMM) (Los Angeles, CA, USA, 2017), 225–238.
@inproceedings{ghorbani_drill_2017,
title = {DRILL: Micro Load Balancing for Low-Latency Data Center Networks},
shorttitle = {DRILL},
booktitle = {Proceedings of the Conference of the ACM Special Interest Group on Data Communication (SIGCOMM)},
author = {Ghorbani, Soudeh and Yang, Zibin and Godfrey, P. Brighten and Ganjali, Yashar and Firoozshahian, Amin},
year = {2017},
series = {SIGCOMM '17},
pages = {225--238},
publisher = {ACM},
address = {Los Angeles, CA, USA},
doi = {10.1145/3098822.3098839},
url = {http://doi.acm.org/10.1145/3098822.3098839},
urldate = {2018-04-25},
isbn = {978-1-4503-4653-5}
}
The trend towards simple datacenter network fabric strips most network functionality, including load balancing, out of the network core and pushes it to the edge. This slows reaction to microbursts, the main culprit of packet loss in datacenters. We investigate the opposite direction: could slightly smarter fabric significantly improve load balancing? This paper presents DRILL, a datacenter fabric for Clos networks which performs micro load balancing to distribute load as evenly as possible on microsecond timescales. DRILL employs per-packet decisions at each switch based on local queue occupancies and randomized algorithms to distribute load. Our design addresses the resulting key challenges of packet reordering and topological asymmetry. In simulations with a detailed switch hardware model and realistic workloads, DRILL outperforms recent edge-based load balancers, particularly under heavy load. Under 80% load, for example, it achieves 1.3-1.4x lower mean flow completion time than recent proposals, primarily due to shorter upstream queues. To test hardware feasibility, we implement DRILL in Verilog and estimate its area overhead to be less than 1%. Finally, we analyze DRILL’s stability and throughput-efficiency.
Yeganeh, S.H. and Ganjali, Y. 2016. Beehive: Simple Distributed Programming in Software-Defined Networks. Proceedings of the Symposium on SDN Research (New York, NY, USA, Mar. 2016), 1–12.
@inproceedings{yeganeh_beehive_2016,
title = {Beehive: Simple Distributed Programming in Software-Defined Networks},
shorttitle = {Beehive},
booktitle = {Proceedings of the Symposium on SDN Research},
author = {Yeganeh, Soheil Hassas and Ganjali, Yashar},
year = {2016},
month = mar,
series = {SOSR '16},
pages = {1--12},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/2890955.2890958},
url = {https://dl.acm.org/doi/10.1145/2890955.2890958},
urldate = {2024-08-14},
isbn = {978-1-4503-4211-7}
}
In this paper, we present the design and implementation of Beehive, a distributed control platform with a simple programming model. In Beehive, control applications are centralized asynchronous message handlers that optionally store their state in dictionaries. Beehive’s control platform automatically infers the keys required to process a message, and guarantees that each key is only handled by one light-weight thread of execution (i.e., bee) among all controllers (i.e., hives) in the platform. With that, Beehive transforms a centralized application into a distributed system, while preserving the application’s intended behavior. Beehive replicates the dictionaries of control applications consistently through mini-quorums (i.e., colonies), instruments applications at runtime, and dynamically changes the placement of control applications (i.e., live migrates bees) to optimize the control plane. Our implementation of Beehive is open source, high-throughput and capable of fast failovers. We have implemented an SDN controller on top of Beehive that can handle 200K of OpenFlow messages per machine, while persisting and replicating the state of control applications. We also demonstrate that, not only can Beehive tolerate faults, but also it is capable of optimizing control applications after a failure or a change in the workload.
Ghorbani, S., Godfrey, B., Ganjali, Y. and Firoozshahian, A. 2015. Micro Load Balancing in Data Centers with DRILL. Proceedings of the 14th ACM Workshop on Hot Topics in Networks (New York, NY, USA, Nov. 2015), 1–7.
@inproceedings{ghorbani_micro_2015,
title = {Micro Load Balancing in Data Centers with DRILL},
booktitle = {Proceedings of the 14th ACM Workshop on Hot Topics in Networks},
author = {Ghorbani, Soudeh and Godfrey, Brighten and Ganjali, Yashar and Firoozshahian, Amin},
year = {2015},
month = nov,
series = {HotNets-XIV},
pages = {1--7},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/2834050.2834107},
url = {https://dl.acm.org/doi/10.1145/2834050.2834107},
urldate = {2024-08-14},
isbn = {978-1-4503-4047-2}
}
The trend towards simple data center network fabric strips most network functionality, including load balancing capabilities, out of the network core and pushes them to the edge. We investigate a different direction of incorporating minimal load balancing intelligence into the network fabric and show that this slightly smarter fabric significantly enhances performance. We provide a very simple in-network load balancing scheduling algorithm called DRILL which is purely local to each switch. DRILL leverages local load sensing and randomization concepts to distribute load among multiple paths. Through simulation, we show that this simple approach outperforms CONGA, a recent global edge-based load balancing scheme for data centers. We also formally prove the switch-level stability and throughput-efficiency of DRILL’s scheduling algorithm.
Eftekhar, M., Koudas, N. and Ganjali, Y. 2015. Reaching a Desired Set of Users via Different Paths: An Online Advertising Technique on Micro-Blogging Platforms. (2015).
@inproceedings{eftekhar_reaching_2015,
title = {Reaching a Desired Set of Users via Different Paths: An Online Advertising Technique on Micro-Blogging Platforms},
shorttitle = {Reaching a Desired Set of Users via Different Paths},
author = {Eftekhar, Milad and Koudas, Nick and Ganjali, Yashar},
year = {2015},
publisher = {OpenProceedings.org},
doi = {10.5441/002/EDBT.2015.17},
url = {https://openproceedings.org/2015/conf/edbt/paper-145.pdf},
urldate = {2024-08-14},
langid = {english}
}
Social media and micro-blogging platforms have been successful for communication and information exchange enjoying vast number of user participation. Given their millions of users, it is natural that there is a lot of interest for marketing and advertising on these platforms as attested by the introduced advertising platforms on Twitter and Facebook. In this paper, inspired by micro-blogging advertising platforms, we introduce two problems to aid ad and marketing campaigns. The first problem identifies topics (called analogous topics) that have approximately the same audience in a micro-blogging platform as a given query topic. The main idea is that by bidding on an analogous topic instead of the original query topic, we reach approximately the same audience while spending less of our budget. Then, we present algorithms to identify expert users on a given query topic and categorize these experts to finely understand their diversified expertise. This is imperative for word of mouth marketing where individuals have to be targeted precisely.
Yeganeh, S.H. and Ganjali, Y. 2014. Beehive: Towards a Simple Abstraction for Scalable Software-Defined Networking. Proceedings of the 13th ACM Workshop on Hot Topics in Networks (New York, NY, USA, Oct. 2014), 1–7.
@inproceedings{yeganeh_beehive_2014,
title = {Beehive: Towards a Simple Abstraction for Scalable Software-Defined Networking},
shorttitle = {Beehive},
booktitle = {Proceedings of the 13th ACM Workshop on Hot Topics in Networks},
author = {Yeganeh, Soheil Hassas and Ganjali, Yashar},
year = {2014},
month = oct,
series = {HotNets-XIII},
pages = {1--7},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/2670518.2673864},
url = {https://dl.acm.org/doi/10.1145/2670518.2673864},
urldate = {2024-08-14},
isbn = {978-1-4503-3256-9}
}
Simplicity is a prominent advantage of Software-Defined Networking (SDN), and is often exemplified by implementing a complicated control logic as a simple control application on a centralized controller. In practice, however, SDN controllers turn into distributed systems due to performance and reliability limitations, and the supposedly simple control applications transform into complex logics that demand significant effort to design and optimize.In this paper, we present Beehive, a distributed control platform aiming at simplifying this process. Our proposal is built around a programming abstraction which is almost identical to a centralized controller yet enables the platform to automatically infer how applications maintain their state and depend on one another. Using this abstraction, the platform automatically generates the distributed version of each control application, while preserving its behavior. With runtime instrumentation, the platform dynamically migrates applications among controllers aiming to optimize the control plane as a whole. Beehive also provides feedback to identify design bottlenecks in control applications, helping developers enhance the performance of the control plane. Our prototype shows that Beehive significantly simplifies the process of realizing distributed control applications.
Shirali-Shahreza, S., Penn, G., Ganjali, Y. and Balakrishnan, R. 2014. Do You Want to Talk to Your Router? Probably Not! Proceedings of Designing Speech and Language Interactions Workshop at 2014 SIGCHI Conference on Human Factors in Computing (CHI 2014) (Toronto, Apr. 2014), 4.
@inproceedings{shirali_shahreza_you_2014,
title = {Do You Want to Talk to Your Router? Probably Not!},
booktitle = {Proceedings of Designing Speech and Language Interactions Workshop at 2014 SIGCHI Conference on Human Factors in Computing (CHI 2014)},
author = {{Shirali-Shahreza}, Sajad and Penn, Gerald and Ganjali, Yashar and Balakrishnan, Ravin},
year = {2014},
month = apr,
pages = {4},
address = {Toronto},
url = {publications/test.pdf}
}
Hassas Yeganeh, S. and Ganjali, Y. 2014. Turning the Tortoise to the Hare: An Alternative Perspective on Event Handling in SDN. Proceedings of the International Workshop on Software-Defined Ecosystems (BigSystem) (Vancouver, Canada, 2014).
@inproceedings{hassasyeganeh_turning_2014,
title = {Turning the Tortoise to the Hare: An Alternative Perspective on Event Handling in SDN},
booktitle = {Proceedings of the International Workshop on Software-Defined Ecosystems (BigSystem)},
author = {Hassas Yeganeh, Soheil and Ganjali, Yashar},
year = {2014},
address = {Vancouver, Canada}
}
Shirali-Shahreza, S. and Ganjali, Y. 2013. Efficient Implementation of Security Applications in OpenFlow Controller with FleXam. Proceedings of the 21st Annual Symposium on High-Performance Interconnects (HOTI) (Aug. 2013), 49–54.
@inproceedings{shirali_shahreza_efficient_2013,
title = {Efficient Implementation of Security Applications in OpenFlow Controller with FleXam},
booktitle = {Proceedings of the 21st Annual Symposium on High-Performance Interconnects (HOTI)},
author = {{Shirali-Shahreza}, S. and Ganjali, Y.},
year = {2013},
month = aug,
pages = {49--54},
doi = {10.1109/HOTI.2013.17}
}
Current OpenFlow specifications provide limited access to packet-level information such as packet content, making it very inefficient, if not impossible, to deploy security and monitoring applications as controller applications. In this paper, we propose FleXam, a flexible sampling extension for OpenFlow designed to provide access to packet level information at the controller. Simplicity of FleXam makes it possible to implement it easily in OpenFlow switches and operate at line rate without requiring any additional memory. At the same time, its flexibility allows implementation of various monitoring and security applications in the controller, while maintaining balance between overhead and collected information details. FleXam realizes the advantages of both proactive and reactive routing schemes by providing a tunable trade-off between the visibility of individual flows, and the controller load. As an example, we demonstrate how FleXam can be used to implement a port scan detection application with an extremely low overhead.
Shirali-Shahreza, S. and Ganjali, Y. 2013. Empowering Software Defined Network Controller with Packet-Level Information. In Proceedings of the 1st IEEE Workshop on Traffic Identification and Classification for Advanced Network Services and Scenarios (TRICANS) (Jun. 2013), 1335–1339.
@inproceedings{shirali_shahreza_empowering_2013,
title = {Empowering Software Defined Network Controller with Packet-Level Information},
booktitle = {In Proceedings of the 1st IEEE Workshop on Traffic Identification and Classification for Advanced Network Services and Scenarios (TRICANS)},
author = {{Shirali-Shahreza}, S. and Ganjali, Y.},
year = {2013},
month = jun,
pages = {1335--1339},
doi = {10.1109/ICCW.2013.6649444}
}
Packet level information, such as packet content and inter-arrival time, are necessary for some network monitoring and control applications. However, current Software Defined Networks (SDN) such as OpenFlow provide limited access to packet-level information in the controller. In this paper, we propose an extension that enables the controller to access packet-level information through per-flow sampling. Our extension is flexible and powerful, yet it can be implemented entirely in the data plane at line rate. We present a set of possible applications that can take advantage of this new packet-level information, including examples that are extremely difficult, if not impossible in current SDN.
Shirali-Shahreza, S., Penn, G., Balakrishnan, R. and Ganjali, Y. 2013. SeeSay and HearSay CAPTCHA for Mobile Interaction. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (Apr. 2013), 2147–2156.
@inproceedings{shirali_shahreza_seesay_2013,
title = {SeeSay and HearSay CAPTCHA for Mobile Interaction},
booktitle = {Proceedings of the SIGCHI Conference on Human Factors in Computing Systems},
author = {{Shirali-Shahreza}, Sajad and Penn, Gerald and Balakrishnan, Ravin and Ganjali, Yashar},
year = {2013},
month = apr,
pages = {2147--2156},
publisher = {ACM},
url = {http://dl.acm.org/citation.cfm?id=2481295},
urldate = {2014-05-10}
}
Shirali-Shahreza, S., Hassas Yeganeh, S. and Ganjali, Y. 2013. Smarter Home Networks with Simple Gateways. Proceedings of the 32nd IEEE International Conference on Computer Communications (Turin, Italy, Apr. 2013).
@inproceedings{shirali_shahreza_smarter_2013,
title = {Smarter Home Networks with Simple Gateways},
booktitle = {Proceedings of the 32nd IEEE International Conference on Computer Communications},
author = {{Shirali-Shahreza}, Sajad and Hassas Yeganeh, Soheil and Ganjali, Yashar},
year = {2013},
month = apr,
address = {Turin, Italy}
}
Eftekhar, M., Koudas, N. and Ganjali, Y. 2013. Bursty Subgraphs in Social Networks. Proceedings of the Sixth ACM International Conference on Web Search and Data Mining (New York, NY, USA, Feb. 2013), 213–222.
@inproceedings{eftekhar_bursty_2013,
title = {Bursty Subgraphs in Social Networks},
booktitle = {Proceedings of the Sixth ACM International Conference on Web Search and Data Mining},
author = {Eftekhar, Milad and Koudas, Nick and Ganjali, Yashar},
year = {2013},
month = feb,
series = {WSDM '13},
pages = {213--222},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/2433396.2433423},
url = {https://dl.acm.org/doi/10.1145/2433396.2433423},
urldate = {2024-08-10},
isbn = {978-1-4503-1869-3}
}
Data available through social media and content sharing platforms present opportunities for analysis and mining. In the context of social networks, it is interesting to formalize and locate bursts of activities amongst users, related to a particular event and to report sets of socially connected users participating in such bursts. Such collections present new opportunities for understanding social events, and render new ways of online marketing.In this paper, we model social information using two conceptualized graph models. The first one (the action graph) provides a detailed model of all activities of all users while the second one (the holistic graph) provides an aggregate view on each user in the social media. We also propose two models to define the notion of "burst". The first model (intrinsic burst model) takes the intrinsic characteristics of each user into account to recognize the bursty behaviors; while the second model (social burst model) considers neighbors’ influences when identifying bursts. We provide two linear algorithms to detect bursts based on the proposed models. These algorithms have been extensively evaluated on a month of full Twitter dataset certifying the practicality of our approach. A detailed qualitative study of our techniques is also presented.
Ghobadi, M. and Ganjali, Y. 2013. TCP Pacing in Data Center Networks. Proceedings of the 21st Annual Symposium on High-Performance Interconnects (HOTI) (2013), 25–32.
@inproceedings{ghobadi_tcp_2013,
title = {TCP Pacing in Data Center Networks},
booktitle = {Proceedings of the 21st Annual Symposium on High-Performance Interconnects (HOTI)},
author = {Ghobadi, Monia and Ganjali, Y.},
year = {2013},
pages = {25--32},
doi = {10.1109/HOTI.2013.18}
}
This paper studies the effectiveness of TCP pacing in a data center setting. TCP senders inject bursts of packets into the network at the beginning of each round-trip time. These bursts stress the network queues which may cause loss, reduction in throughput and increased latency. Such undesirable effects become more pronounced in data center environments where traffic is bursty in nature and buffer sizes are small. TCP pacing is believed to reduce the burstiness of TCP traffic and to mitigate the impact of small buffering in routers. Unfortunately, current research literature has not always agreed on the overall benefits of pacing. In this paper, we present a model for the effectiveness of pacing. Our model demonstrates that for a given buffer size, as the number of concurrent flows are increased beyond a Point of Inflection (PoI), non-paced TCP outperforms paced TCP. We present a lower bound for the PoI and argue that increasing the number of concurrent flows beyond the PoI, increases inter-flow burstiness of paced packets and diminishes the effectiveness of pacing.
Eftekhar, M., Ganjali, Y. and Koudas, N. 2013. Information Cascade at Group Scale. Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (New York, NY, USA, 2013), 401–409.
@inproceedings{eftekhar_information_2013,
title = {Information Cascade at Group Scale},
booktitle = {Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining},
author = {Eftekhar, Milad and Ganjali, Yashar and Koudas, Nick},
year = {2013},
series = {KDD '13},
pages = {401--409},
publisher = {ACM},
address = {New York, NY, USA},
doi = {10.1145/2487575.2487683},
url = {http://doi.acm.org/10.1145/2487575.2487683},
urldate = {2018-11-21},
isbn = {978-1-4503-2174-7}
}
Identifying the k most influential individuals in a social network is a well-studied problem. The objective is to detect k individuals in a (social) network who will influence the maximum number of people, if they are independently convinced of adopting a new strategy (product, idea, etc). There are cases in real life, however, where we aim to instigate groups instead of individuals to trigger network diffusion. Such cases abound, e.g., billboards, TV commercials and newspaper ads are utilized extensively to boost the popularity and raise awareness. In this paper, we generalize the "influential nodes" problem. Namely we are interested to locate the most "influential groups" in a network. As the first paper to address this problem: we (1) propose a fine-grained model of information diffusion for the group-based problem, (2) show that the process is submodular and present an algorithm to determine the influential groups under this model (with a precise approximation bound), (3) propose a coarse-grained model that inspects the network at group level (not individuals) significantly speeding up calculations for large networks, (4) show that the diffusion function we design here is submodular in general case, and propose an approximation algorithm for this coarse-grained model, and finally by conducting experiments on real datasets, (5) demonstrate that seeding members of selected groups to be the first adopters can broaden diffusion (when compared to the influential individuals case). Moreover, we can identify these influential groups much faster (up to 12 million times speedup), delivering a practical solution to this problem.
Ghobadi, M., Salmon, G., Ganjali, Y., Labrecque, M. and Steffan, J.G. 2012. Caliper: Precise and Responsive Traffic Generator. 2012 IEEE 20th Annual Symposium on High-Performance Interconnects (HOTI) (Aug. 2012), 25–32.
@inproceedings{ghobadi_caliper_2012,
title = {Caliper: Precise and Responsive Traffic Generator},
shorttitle = {Caliper},
booktitle = {2012 IEEE 20th Annual Symposium on High-Performance Interconnects (HOTI)},
author = {Ghobadi, M. and Salmon, G. and Ganjali, Y. and Labrecque, M. and Steffan, J.G.},
year = {2012},
month = aug,
pages = {25--32},
doi = {10.1109/HOTI.2012.16}
}
This paper presents Caliper, a highly-accurate packet injection tool that generates precise and responsive traffic. Caliper takes live packets generated on a host computer and transmits them onto a gigabit Ethernet network with precise inter-transmission times. Existing software traffic generators rely on generic Network Interface Cards which, as we demonstrate, do not provide high-precision timing guarantees. Hence, performing valid and convincing experiments becomes difficult or impossible in the context of time-sensitive network experiments. Our evaluations show that Caliper is able to reproduce packet inter-transmission times from a given arbitrary distribution while capturing the closed-loop feedback of TCP sources. Specifically, we demonstrate that Caliper provides three orders of magnitude better precision compared to commodity NIC: with requested traffic rates up to the line rate, Caliper incurs an error of 8 ns or less in packet transmission times. Furthermore, we explore Caliper’s ability to integrate with existing network simulators to project simulated traffic characteristics into a real network environment. Caliper is freely available online.
Hassas Yeganeh, S., Eftekhar, M., Ganjali, Y., Keralapura, R. and Nucci, A. 2012. CUTE: Traffic Classification Using TErms. 2012 21st International Conference on Computer Communications and Networks (ICCCN) (Aug. 2012), 1–9.
@inproceedings{hassasyeganeh_cute_2012,
title = {CUTE: Traffic Classification Using TErms},
shorttitle = {CUTE},
booktitle = {2012 21st International Conference on Computer Communications and Networks (ICCCN)},
author = {Hassas Yeganeh, Soheil and Eftekhar, Milad and Ganjali, Yashar and Keralapura, Ram and Nucci, Antonio},
year = {2012},
month = aug,
pages = {1--9},
doi = {10.1109/ICCCN.2012.6289207}
}
Among different traffic classification approaches, Deep Packet Inspection (DPI) methods are considered as the most accurate. These methods, however, have two drawbacks: (i) they are not efficient since they use complex regular expressions as protocol signatures, and (ii) they require manual intervention to generate and maintain signatures, partly due to the signature complexity. In this paper, we present CUTE, an automatic traffic classification method, which relies on sets of weighted terms as protocol signatures. The key idea behind CUTE is an observation that, given appropriate weights, the occurrence of a specific term is more important than the relative location of terms in a flow. This observation is based on experimental evaluations as well as theoretical analysis, and leads to several key advantages over previous classification techniques: (i) CUTE is extremely faster than other classification schemes since matching flows with weighed terms is significantly faster than matching regular expressions; (ii) CUTE can classify network traffic using only the first few bytes of the flows in most cases; and (iii) Unlike most existing classification techniques, CUTE can be used to classify partial (or even slightly modified) flows. Even though CUTE replaces complex regular expressions with a set of simple terms, using theoretical analysis and experimental evaluations (based on two large packet traces from tier-one ISPs), we show that its accuracy is as good as or better than existing complex classification schemes, i.e. CUTE achieves precision and recall rates of more than 90%. Additionally, CUTE can successfully classify more than half of flows that other DPI methods fail to classify.
Ghorbani, S. and Ganjali, Y. 2012. Would You Be My Friend? Privacy in Online Social Networks. Proceedings of the IEEE International Conference on Information Society (i-Society 2012) (London, UK, Jun. 2012).
@inproceedings{ghorbani_would_2012,
title = {Would You Be My Friend? Privacy in Online Social Networks},
booktitle = {Proceedings of the IEEE International Conference on Information Society (i-Society 2012)},
author = {Ghorbani, Soudeh and Ganjali, Yashar},
year = {2012},
month = jun,
address = {London, UK}
}
Ghorbani, S. and Ganjali, Y. 2012. Will You Be My Friend? Privacy Implications of Accepting Friendships in Online Social Networks. 2012 International Conference on Information Society (i-Society) (Jun. 2012), 340–345.
@inproceedings{ghorbani_will_2012,
title = {Will You Be My Friend? Privacy Implications of Accepting Friendships in Online Social Networks},
shorttitle = {Will You Be My Friend?},
booktitle = {2012 International Conference on Information Society (i-Society)},
author = {Ghorbani, S. and Ganjali, Y.},
year = {2012},
month = jun,
pages = {340--345}
}
Online social networks (OSNs) have become extremely popular in recent years. Users actively interact in these networks and share large amounts of personal information. This has led to emergence of a treasure trove of data for many entities, from marketers and spammers to employers and intelligence agencies, which has become a serious privacy concern. Previous works have addressed many aspects about privacy in OSNs such as characterizing potential privacy leakage [14], possible ways for inferring sensitive private information [9], [18], and appropriateness of default privacy settings [11]. In contrast, we focus on the entity who plays the main role in guarding privacy: the user. By sending out friend requests to unknown users in one of the largest OSNs, we provide evidence that a considerable portion of OSN users are willing to let a stranger, possibly an adversary, into their social network, thus granting her access to the users’ personal information and to some extent to those of their friends. We study several factors that might foster such behavior, and measure the amount of information that will consequently become accessible. We find that for more than 95% of the users who accept our friend requests, we gained access to personal information that would not otherwise be accessible. We also show that the majority of the users who accept the requests have indeed changed their default privacy settings to restrict access to some parts of their personal information to their friends while making them publicly inaccessible.
Tootoonchian, A., Gorbunov, S., Casado, M., Sherwood, R. and Ganjali, Y. 2012. On Controller Performance in Software-Defined Networks. Proceedings of the 2nd USENIX Workshop on Hot Topics in Management of Internet, Cloud, and Enterprise Networks and Services (HotICE) (San Jose, CA, Apr. 2012).
@inproceedings{tootoonchian_controller_2012,
title = {On Controller Performance in Software-Defined Networks},
booktitle = {Proceedings of the 2nd USENIX Workshop on Hot Topics in Management of Internet, Cloud, and Enterprise Networks and Services (HotICE)},
author = {Tootoonchian, Amin and Gorbunov, Sergey and Casado, Martin and Sherwood, Robert and Ganjali, Yashar},
year = {2012},
month = apr,
address = {San Jose, CA}
}
Hassas Yeganeh, S. and Ganjali, Y. 2012. Kandoo: A Framework for Efficient and Scalable Offloading of Control Applications. Proceedings of the First Workshop on Hot Topics in Software Defined Networks (Helsinki, Finland, 2012), 19–24.
@inproceedings{hassasyeganeh_kandoo_2012,
title = {Kandoo: A Framework for Efficient and Scalable Offloading of Control Applications},
shorttitle = {Kandoo},
booktitle = {Proceedings of the First Workshop on Hot Topics in Software Defined Networks},
author = {Hassas Yeganeh, Soheil and Ganjali, Yashar},
year = {2012},
pages = {19--24},
publisher = {ACM},
address = {Helsinki, Finland},
doi = {10.1145/2342441.2342446},
url = {http://doi.acm.org/10.1145/2342441.2342446},
urldate = {2012-10-31},
isbn = {978-1-4503-1477-0}
}
Limiting the overhead of frequent events on the control plane is essential for realizing a scalable Software-Defined Network. One way of limiting this overhead is to process frequent events in the data plane. This requires modifying switches and comes at the cost of visibility in the control plane. Taking an alternative route, we propose Kandoo, a framework for preserving scalability without changing switches. Kandoo has two layers of controllers: (i) the bottom layer is a group of controllers with no interconnection, and no knowledge of the network-wide state, and (ii) the top layer is a logically centralized controller that maintains the network-wide state. Controllers at the bottom layer run only local control applications (i.e., applications that can function using the state of a single switch) near datapaths. These controllers handle most of the frequent events and effectively shield the top layer. Kandoo’s design enables network operators to replicate local controllers on demand and relieve the load on the top layer, which is the only potential bottleneck in terms of scalability. Our evaluations show that a network controlled by Kandoo has an order of magnitude lower control channel consumption compared to normal OpenFlow networks.
Ghobadi, M., Yeganeh, S.H. and Ganjali, Y. 2012. Rethinking End-to-End Congestion Control in Software-Defined Networks. Proceedings of the 11th ACM Workshop on Hot Topics in Networks (Redmond, WA, USA, 2012), 61–66.
@inproceedings{ghobadi_rethinking_2012,
title = {Rethinking End-to-End Congestion Control in Software-Defined Networks},
booktitle = {Proceedings of the 11th ACM Workshop on Hot Topics in Networks},
author = {Ghobadi, Monia and Yeganeh, Soheil Hassas and Ganjali, Yashar},
year = {2012},
pages = {61--66},
publisher = {ACM},
address = {Redmond, WA, USA},
doi = {10.1145/2390231.2390242},
url = {http://doi.acm.org/10.1145/2390231.2390242},
urldate = {2013-03-30},
isbn = {978-1-4503-1776-4}
}
TCP is designed to operate in a wide range of networks. Without any knowledge of the underlying network and traffic characteristics, TCP is doomed to continuously increase and decrease its congestion window size to embrace changes in network or traffic. In light of emerging popularity of centrally controlled Software-Defined Networks (SDNs), one might wonder whether we can take advantage of the global network view available at the controller to make faster and more accurate congestion control decisions. In this paper, we identify the need and the underlying requirements for a congestion control adaptation mechanism. To this end, we propose OpenTCP as a TCP adaptation framework that works in SDNs. OpenTCP allows network operators to define rules for tuning TCP as a function of network and traffic conditions. We also present a preliminary implementation of OpenTCP in a ~4000 node data center.
Kaffash Bokharaei, H., Sahraei, A., Ganjali, Y., Keralapura, R. and Nucci, A. 2011. You Can {SPIT} but You Can’t Hide: Spammer Identification in Telephony Networks. 30th IEEE International Conference on Computer Communications. Proceedings (Shanghahi, China, 2011).
@inproceedings{kaffashbokharaei_you_2011,
title = {You Can \{SPIT\} but You Can't Hide: Spammer Identification in Telephony Networks},
booktitle = {30th IEEE International Conference on Computer Communications. Proceedings},
author = {Kaffash Bokharaei, Hossein and Sahraei, Alireza and Ganjali, Yashar and Keralapura, Ram and Nucci, Antonio},
year = {2011},
address = {Shanghahi, China}
}
Kaffash Bokharaei, H., Ganjali, Y., Keralapura, R. and Nucci, A. 2011. Telephony Network Characterization for Spammer Identification. IFIP Performance 2011 (Amsterdam, Netherlands, 2011).
@inproceedings{kaffashbokharaei_telephony_2011,
title = {Telephony Network Characterization for Spammer Identification},
booktitle = {IFIP Performance 2011},
author = {Kaffash Bokharaei, Hossein and Ganjali, Yashar and Keralapura, Ram and Nucci, Antonio},
year = {2011},
address = {Amsterdam, Netherlands}
}
Shirali-Shahreza, S., Ganjali, Y. and Balakrishnan, R. 2011. Verifying Human Users in Speech-Based Interactions. Proceedings of the 13th Annual Conference of the International Speech Communication Association (Florence, Italy, 2011).
@inproceedings{shirali_shahreza_verifying_2011,
title = {Verifying Human Users in Speech-Based Interactions},
booktitle = {Proceedings of the 13th Annual Conference of the International Speech Communication Association},
author = {{Shirali-Shahreza}, Sajad and Ganjali, Yashar and Balakrishnan, Ravin},
year = {2011},
address = {Florence, Italy}
}
Gill, P., Ganjali, Y., Wong, B. and Lie, D. 2010. Dude, Where Is That IP? Circumventing Measurement-Based IP Geolocation. Proceedings of the 19th USENIX Security Symposium (Washington, DC, USA, Aug. 2010).
@inproceedings{gill_dude_2010,
title = {Dude, Where Is That IP? Circumventing Measurement-Based IP Geolocation},
booktitle = {Proceedings of the 19th USENIX Security Symposium},
author = {Gill, Phillipa and Ganjali, Yashar and Wong, Bernard and Lie, David},
year = {2010},
month = aug,
address = {Washington, DC, USA}
}
Tabatabaei, H. and Ganjali, Y. 2010. Preserving Pacing in Real Networks - An Experimental Study Using NetFPGA. Proceedings of the 2nd North American NetFPGA Developers Workshop (Stanford, CA, USA, Aug. 2010).
@inproceedings{tabatabaei_preserving_2010,
title = {Preserving Pacing in Real Networks - An Experimental Study Using NetFPGA},
booktitle = {Proceedings of the 2nd North American NetFPGA Developers Workshop},
author = {Tabatabaei, Hamed and Ganjali, Yashar},
year = {2010},
month = aug,
address = {Stanford, CA, USA}
}
Tootoonchian, A. and Ganjali, Y. 2010. HyperFlow: A Distributed Control Plane for OpenFlow. Internet Network Management Workshop/Workshop on Research on Enterprise Networking (San Jose, CA, USA, Apr. 2010).
@inproceedings{tootoonchian_hyperflow_2010,
title = {HyperFlow: A Distributed Control Plane for OpenFlow},
booktitle = {Internet Network Management Workshop/Workshop on Research on Enterprise Networking},
author = {Tootoonchian, Amin and Ganjali, Yashar},
year = {2010},
month = apr,
address = {San Jose, CA, USA},
url = {http://www.cs.toronto.edu/~yganjali/}
}
Tootoonchian, A., Ghobadi, M. and Ganjali, Y. 2010. Traffic Matrix Estimation Using OpenFlow. Passive and Active Measurement Conference (Zurich, Switzerland, Apr. 2010).
@inproceedings{tootoonchian_traffic_2010,
title = {Traffic Matrix Estimation Using OpenFlow},
booktitle = {Passive and Active Measurement Conference},
author = {Tootoonchian, Amin and Ghobadi, Monia and Ganjali, Yashar},
year = {2010},
month = apr,
address = {Zurich, Switzerland}
}
Rastegarfar, H., Ghobadi, M. and Ganjali, Y. 2009-11-30/2009-12-04. Emulation of Optical PIFO Buffers. IEEE Global Telecommunications Conference, 2009. GLOBECOM 2009 (2009-11-30/2009-12-04), 1–6.
@inproceedings{rastegarfar_emulation_2009,
title = {Emulation of Optical PIFO Buffers},
booktitle = {IEEE Global Telecommunications Conference, 2009. GLOBECOM 2009},
author = {Rastegarfar, H. and Ghobadi, M. and Ganjali, Y.},
year = {2009-11-30/2009-12-04},
pages = {1--6},
publisher = {IEEE},
doi = {10.1109/GLOCOM.2009.5426009},
isbn = {978-1-4244-4148-8},
langid = {english}
}
With recent advances in optical technology, we are closer to building all-optical routers than ever before. A major problem in this area, however, is the lack of all-optical memories similar to what we have in electronics. To overcome this problem, recently, there have been several proposals that show how we can emulate First-In First-Out (FIFO) queues using a combination of fiber delay lines and switches. Unfortunately, FIFO queues cannot be used for implementing many link scheduling policies including weighted fair queuing, weighted round-robin, or strict priority, which are essential components of any modern router today. In this paper, we introduce an architecture based on fiber delay lines and optical switches that can be used for emulating Push-In First-Out (PIFO) queues. In a PIFO queue, an incoming packet can be pushed anywhere in the queue, and therefore it can be used for the implementation of various link scheduling policies. We describe a scheduling algorithm for this architecture and show that with a small speedup, we can build a PIFO queue of size N - 1 using only O(log2 N) 3 ÃÂ 3 optical switches. The resulting system has a minimum reliability of 99.5%, and even for the small portion of departure requests that cannot be fulfilled immediately, the requested packet is ready to depart within approximately five time slots from the request time.
Tootoonchian, A., Saroiu, S., Ganjali, Y. and Wolman, A. 2009. Lockr: Better Privacy for Social Networks. Proceedings of the 5th ACM International Conference on Emerging Networking EXperiments and Technologies (CoNEXT) (Rome, Italy, Dec. 2009).
@inproceedings{tootoonchian_lockr_2009,
title = {Lockr: Better Privacy for Social Networks},
booktitle = {Proceedings of the 5th ACM International Conference on Emerging Networking EXperiments and Technologies (CoNEXT)},
author = {Tootoonchian, Amin and Saroiu, Stefan and Ganjali, Yashar and Wolman, Alec},
year = {2009},
month = dec,
address = {Rome, Italy}
}
Salmon, G., Labrecque, M., Ghobadi, M., Ganjali, Y. and Steffan, G. 2009. NetFPGA-Based Precise Traffic Generation. Proceedings of NetFPGA Developers Workshop (Stanford, Califronia, USA, Aug. 2009).
@inproceedings{salmon_netfpgabased_2009,
title = {NetFPGA-Based Precise Traffic Generation},
booktitle = {Proceedings of NetFPGA Developers Workshop},
author = {Salmon, Geoffrey and Labrecque, Martin and Ghobadi, Monia and Ganjali, Yashar and Steffan, Greg},
year = {2009},
month = aug,
address = {Stanford, Califronia, USA},
url = {http://www.cs.toronto.edu/~yganjali/}
}
Submitted to NetFPGA Developers Workshop 2009.
Salmon, G., Labrecque, M., Ghobadi, M., Ganjali, Y. and Steffan, G. 2009. NetThreads: Programming NetFPGA with Threaded Software. Proceedings of NetFPGA Developers Workshop (Stanford, California, USA, Aug. 2009).
@inproceedings{salmon_netthreads_2009,
title = {NetThreads: Programming NetFPGA with Threaded Software},
booktitle = {Proceedings of NetFPGA Developers Workshop},
author = {Salmon, Geoffrey and Labrecque, Martin and Ghobadi, Monia and Ganjali, Yashar and Steffan, Greg},
year = {2009},
month = aug,
address = {Stanford, California, USA},
url = {http://www.cs.toronto.edu/~yganjali/}
}
Havary-Nassab, V., Koulakezian, A. and Ganjali, Y. 2009. Denial of Service Attacks in Networks with Tiny Buffers. INFOCOM Workshops 2009, IEEE (2009), 1–6.
@inproceedings{havary-nassab_denial_2009,
title = {Denial of Service Attacks in Networks with Tiny Buffers},
booktitle = {INFOCOM Workshops 2009, IEEE},
author = {{Havary-Nassab}, V. and Koulakezian, A. and Ganjali, Y.},
year = {2009},
pages = {1--6},
url = {publications/test.pdf}
}
Recently, several papers have studied the possibility of shrinking buffer sizes in Internet core routers to just a few dozen packets under certain constraints. If proven right, these results can open doors to building all-optical routers, since a major bottleneck in building such routers is the lack of large optical memories. However, reducing buffer sizes might pose new security risks: it is much easier to fill up tiny buffers, and thus organizing Denial of Service (DoS) attacks seems easier in a network with tiny buffers. To the best of our knowledge, such risks have not been studied before; all the focus has been on performance issues such as throughput, drop rate, and flow completion times. In this paper, we study DoS attacks in the context of networks with tiny buffers. We show that even though it is easier to fill up tiny buffers, synchronizing flows is more difficult. Therefore to reduce the network throughput, the attacker needs to utilize attacks with high packet injection rates. Since such attacks are easily detected, we conclude that DoS attacks are in fact more difficult in networks with tiny buffers.
Beheshti, N., Ganjali, Y., Ghobadi, M., McKeown, N. and Salmon, G. 2008. Experimental Study of Router Buffer Sizing. Proceedings of Internet Measurement Conference (IMC) (Vouliagmeni, Greece, Oct. 2008), 197–210 (Best paper).
@inproceedings{beheshti_experimental_2008,
title = {Experimental Study of Router Buffer Sizing},
booktitle = {Proceedings of Internet Measurement Conference (IMC)},
author = {Beheshti, Neda and Ganjali, Yashar and Ghobadi, Monia and McKeown, Nick and Salmon, Geoff},
year = {2008},
month = oct,
pages = {197-210 (Best paper)},
address = {Vouliagmeni, Greece}
}
Tootoonchian, A., Gollu, K., Saroiu, S., Ganjali, Y. and Wolman, A. 2008. Lockr: Social Access Control for Web 2.0. Proceedings of ACM SIGCOMM Workshop on Online Social Networks (WOSN) (Seattle, Washington, USA, Aug. 2008).
@inproceedings{tootoonchian_lockr_2008,
title = {Lockr: Social Access Control for Web 2.0},
booktitle = {Proceedings of ACM SIGCOMM Workshop on Online Social Networks (WOSN)},
author = {Tootoonchian, Amin and Gollu, Kiran and Saroiu, Stefan and Ganjali, Yashar and Wolman, Alec},
year = {2008},
month = aug,
address = {Seattle, Washington, USA},
url = {http://www.cs.toronto.edu/~yganjali/}
}
Beheshti, N., Ganjali, Y., Goel, A. and McKeown, N. 2008. Obtaining High Throughput in Networks with Tiny Buffers. Proceedings of 16th International Workshop on Quality of Service (IWQoS) (Enschede, Netherlands, Jun. 2008).
@inproceedings{beheshti_obtaining_2008,
title = {Obtaining High Throughput in Networks with Tiny Buffers},
booktitle = {Proceedings of 16th International Workshop on Quality of Service (IWQoS)},
author = {Beheshti, Neda and Ganjali, Yashar and Goel, Ashish and McKeown, Nick},
year = {2008},
month = jun,
address = {Enschede, Netherlands},
url = {http://www.cs.toronto.edu/~yganjali/}
}
Beheshti, N., Ganjali, Y., Ghobadi, M., Naous, J., McKeown, N. and Salmon, G. 2008. Performing Time-Sensitive Network Experiments. Proceedings of the 4th ACM/IEEE Symposium on Architecture for Networking and Communications Systems (ANCS) (San Jose, CA, 2008), 127–128.
@inproceedings{beheshti_performing_2008,
title = {Performing Time-Sensitive Network Experiments},
booktitle = {Proceedings of the 4th ACM/IEEE Symposium on Architecture for Networking and Communications Systems (ANCS)},
author = {Beheshti, Neda and Ganjali, Yashar and Ghobadi, Monia and Naous, Jad and McKeown, Nick and Salmon, Geoff},
year = {2008},
pages = {127--128},
address = {San Jose, CA},
url = {http://www.cs.toronto.edu/~yganjali/test.pdf}
}
Beheshti, N., Naus, J., Ganjali, Y. and McKeown, N. 2007. Experimenting with Buffer Sizes in Routers. Proceedings of the 3rd ACM/IEEE Symposium on Architecture for Networking and Communications Systems (ANCS) (Orlando, Florida, USA, Dec. 2007), 41–42.
@inproceedings{beheshti_experimenting_2007,
title = {Experimenting with Buffer Sizes in Routers},
booktitle = {Proceedings of the 3rd ACM/IEEE Symposium on Architecture for Networking and Communications Systems (ANCS)},
author = {Beheshti, Neda and Naus, Jad and Ganjali, Yashar and McKeown, Nick},
year = {2007},
month = dec,
pages = {41--42},
address = {Orlando, Florida, USA}
}
Beheshti, N. and Ganjali, Y. 2007. Packet Scheduling in Optical FIFO Buffers. Proceedings of IEEE INFOCOM High-Speed Networks Workshop (Anchorage, Alaska, USA, May 2007).
@inproceedings{beheshti_packet_2007,
title = {Packet Scheduling in Optical FIFO Buffers},
booktitle = {Proceedings of IEEE INFOCOM High-Speed Networks Workshop},
author = {Beheshti, Neda and Ganjali, Yashar},
year = {2007},
month = may,
address = {Anchorage, Alaska, USA},
url = {http://www.cs.toronto.edu/~yganjali/}
}
Wang, M. and Ganjali, Y. 2007. The Effects of Fairness in Buffer Sizing. NETWORKING 2007. Ad Hoc and Sensor Networks, Wireless Networks, Next Generation Internet (Atlanta, Georgia, USA, 2007), 867–878.
@inproceedings{wang_effects_2007,
title = {The Effects of Fairness in Buffer Sizing},
booktitle = {NETWORKING 2007. Ad Hoc and Sensor Networks, Wireless Networks, Next Generation Internet},
author = {Wang, Mei and Ganjali, Yashar},
year = {2007},
pages = {867--878},
address = {Atlanta, Georgia, USA},
url = {http://dx.doi.org/10.1007/978-3-540-72606-7_74},
urldate = {2008-05-14}
}
Buffer sizing in Internet routers is a fundamental problem that has major consequences in the design, implementation, and economy of the routers, as well as on the performance observed by the end users. Recently, there have been some seemingly contradictory results on buffer sizing. On the one hand, Appenzeller et al. show that as a direct consequence of desynchronization of flows in the core of the Internet, buffer sizes in core routers can be significantly reduced without any major degradation in network performance. On the other hand, Raina and Wischik show that such reduction in buffer sizing comes at the cost of synchronization and thus instability in the network. This work unifies these results by studying the effects of fairness in buffer sizing. We show that the main difference arises from the implicit assumption of fairness in packet dropping in the latter result. We demonstrate that desynchronization among flows observed by Appenzeller et al. is caused by unfair packet dropping when a combination of TCP-Reno and the drop-tail queue management is used. We also show that bringing fairness in packet dropping will introduce synchronization among flows, and will make the system unstable as predicted by Raina and Wischik. Our analysis suggests that there is an intrinsic trade-off between fairness in packet drops and desynchronization among TCP-Reno flows when routers use the drop-tail queue management. Achieving fairness, desynchronization, small buffer size, and 100% link utilization at the same time is desirable and feasible yet challenging. The studies in this paper provide insights for further explorations in reaching this goal.
Lin, M. and Ganjali, Y. 2006. Power-Efficient Rate Scheduling in Wireless Links Using Computational Geometric Algorithms. Proceedings of the International Wireless Communications and Mobile Computing Conference (IWCMC) (Vancouver, Canada, Jul. 2006).
@inproceedings{lin_powerefficient_2006,
title = {Power-Efficient Rate Scheduling in Wireless Links Using Computational Geometric Algorithms},
booktitle = {Proceedings of the International Wireless Communications and Mobile Computing Conference (IWCMC)},
author = {Lin, Mingjie and Ganjali, Yashar},
year = {2006},
month = jul,
address = {Vancouver, Canada}
}
Enachescu, M., Ganjali, Y., Goel, A., McKeown, N. and Roughgarden, T. 2006. Routers with Very Small Buffers. Proceedings of the IEEE INFOCOM’06 (Barcelona, Spain, Apr. 2006).
@inproceedings{enachescu_routers_2006,
title = {Routers with Very Small Buffers},
booktitle = {Proceedings of the IEEE INFOCOM'06},
author = {Enachescu, Mihaela and Ganjali, Yashar and Goel, Ashish and McKeown, Nick and Roughgarden, Tim},
year = {2006},
month = apr,
address = {Barcelona, Spain}
}
Also available as technical report TR05-HPNG-060606, High Performance Networking Group, Stanford University.
Beheshti, N., Ganjali, Y., Rajaduray, R., Blumenthal, D. and McKeown, N. 2006. Buffer Sizing in All-Optical Packet Switches. 2006 Optical Fiber Communication Conference and the National Fiber Optic Engineers Conference (Anaheim, CA, USA, 2006), 3 pp.
@inproceedings{beheshti_buffer_2006,
title = {Buffer Sizing in All-Optical Packet Switches},
booktitle = {2006 Optical Fiber Communication Conference and the National Fiber Optic Engineers Conference},
author = {Beheshti, N. and Ganjali, Y. and Rajaduray, R. and Blumenthal, D. and McKeown, N.},
year = {2006},
pages = {3 pp.},
publisher = {IEEE},
address = {Anaheim, CA, USA},
doi = {10.1109/OFC.2006.215714},
url = {http://ieeexplore.ieee.org/document/1636745/},
urldate = {2022-12-18},
isbn = {978-1-55752-803-2}
}
Dukkipati, N., Ganjali, Y. and Zhang-Shen, R. 2005. Typical versus Worst Case Design in Networking. Proceedings of the Fourth ACM Workshop on Hot Topics in Networks (HotNets-IV) (College Park, Maryland, Nov. 2005).
@inproceedings{dukkipati_typical_2005,
title = {Typical versus Worst Case Design in Networking},
booktitle = {Proceedings of the Fourth ACM Workshop on Hot Topics in Networks (HotNets-IV)},
author = {Dukkipati, Nandita and Ganjali, Yashar and {Zhang-Shen}, Rui},
year = {2005},
month = nov,
address = {College Park, Maryland},
url = {http://www.cs.toronto.edu/~yganjali/}
}
Ganjali, Y. and McKeown, N. 2005. Routing in a Highly Dynamic Topology. Sensor and Ad Hoc Communications and Networks, 2005. IEEE SECON 2005. 2005 Second Annual IEEE Communications Society Conference On (2005), 164–175.
@inproceedings{ganjali_routing_2005,
title = {Routing in a Highly Dynamic Topology},
booktitle = {Sensor and Ad Hoc Communications and Networks, 2005. IEEE SECON 2005. 2005 Second Annual IEEE Communications Society Conference On},
author = {Ganjali, Y. and McKeown, N.},
year = {2005},
pages = {164--175}
}
Ganjali, Y. and Keshavarzian, A. 2004. Load Balancing in Ad Hoc Networks: Single-Path Routing vs. Multi-Path Routing. INFOCOM 2004. Twenty-Third AnnualJoint Conference of the IEEE Computer and Communications Societies (2004), 1120–1125 vol.2.
@inproceedings{ganjali_load_2004,
title = {Load Balancing in Ad Hoc Networks: Single-Path Routing vs. Multi-Path Routing},
shorttitle = {Load Balancing in Ad Hoc Networks},
booktitle = {INFOCOM 2004. Twenty-Third AnnualJoint Conference of the IEEE Computer and Communications Societies},
author = {Ganjali, Y. and Keshavarzian, A.},
year = {2004},
volume = {2},
pages = {1120-1125 vol.2},
doi = {10.1109/INFCOM.2004.1356998}
}
Multi-path routing has been studied thoroughly in the context of wired networks. Ii has been shown that using multiple paths to route messages between any source-destination pair of nodes (instead of using a single path) balances the load more evenly throughout the network. The common belief is that the same is true for ad hoc networks, i.e., multi-path routing balances the load significantly better than single-path routing. We show that this is not necessarily the case. We introduce a new model for evaluating the load balance under multi-path routing, when the paths chosen are the first K shortest paths (for a pre-specified K). Using this model, we show that unless we use a very large number of paths (which is very costly and therefore infeasible) the load distribution is almost the same as single shortest path routing. This is in contrary to the previous existing results which assume that multi-path routing distributes the load uniformly.
Ganjali, Y., Keshavarzian, A. and Shah, D. 2003. Input Queued Switches: Cell Switching vs. Packet Switching. Proceedings of the IEEE INFOCOM’03 (San Francisco, CA, USA, Apr. 2003), Best paper runner up.
@inproceedings{ganjali_input_2003,
title = {Input Queued Switches: Cell Switching vs. Packet Switching},
booktitle = {Proceedings of the IEEE INFOCOM'03},
author = {Ganjali, Yashar and Keshavarzian, Abtin and Shah, Devavrat},
year = {2003},
month = apr,
pages = {Best paper runner up},
address = {San Francisco, CA, USA},
url = {http://www.google.com}
}
Selected as one of the 10 best papers
Ganjali, Y. 2001. Characterization of Networks Supporting Multi-Dimensional Linear Interval Routing Schemes. Proceedings of the 8th International Colloquium on Structural Information and Communication Complexity (SIROCCO’01) (Barcelona, Spain, Jun. 2001), 163—178.
@inproceedings{ganjali_characterization_2001,
title = {Characterization of Networks Supporting Multi-Dimensional Linear Interval Routing Schemes},
booktitle = {Proceedings of the 8th International Colloquium on Structural Information and Communication Complexity (SIROCCO'01)},
author = {Ganjali, Yashar},
year = {2001},
month = jun,
pages = {163---178},
address = {Barcelona, Spain}
}
King, A.J., Begen, M., Cojocaru, M., Fowler, E., Ganjali, Y., Lai, J., Lee, T., Navasca, C. and Ryan, D. 2001. Web Hosting Service Level Agreements. Proceedings of the 5th Pacific Institute for Mathematical Sciences 2001 Industrial Problem Solving Workshop (University of Washington, Seattle, 2001).
@inproceedings{king_web_2001,
title = {Web Hosting Service Level Agreements},
booktitle = {Proceedings of the 5th Pacific Institute for Mathematical Sciences 2001 Industrial Problem Solving Workshop},
author = {King, Alan J. and Begen, Mehmet and Cojocaru, Monica and Fowler, Ellen and Ganjali, Yashar and Lai, Judy and Lee, Tijin and Navasca, Carmeliza and Ryan, Daniel},
year = {2001},
address = {University of Washington, Seattle}
}
Also as IBM Research Report RC22301 (W0201-59) January 8, 2002 Mathematics
Ganjali, Y. and Ghodsi, M. 1999. An Approximate Algorithm for Shortest Paths on Weighted Triangulated Irregular Networks. Proceedings of 4th CSI Computer Conference (CSICC’98) (Tehran, Iran, Jan. 1999).
@inproceedings{ganjali_approximate_1999,
title = {An Approximate Algorithm for Shortest Paths on Weighted Triangulated Irregular Networks},
booktitle = {Proceedings of 4th CSI Computer Conference (CSICC'98)},
author = {Ganjali, Yashar and Ghodsi, Mohammad},
year = {1999},
month = jan,
address = {Tehran, Iran},
url = {http://www.cs.toronto.edu/~yganjali}
}
In Persian
Patents
Munir, A., Ketabi, S., Bahnasy, M.M., ZADEH, S.A.B.B.A.S.I. and Ganjali, Y. 2024. Methods and Apparatus for Improved Congestion Signaling. May 2024.
@patent{munir_methods_2024,
title = {Methods and Apparatus for Improved Congestion Signaling},
author = {Munir, Ali and Ketabi, Shiva and Bahnasy, Mahmoud Mohamed and ZADEH, Sepehr ABBASI and Ganjali, Yashar},
year = {2024},
month = may,
url = {https://patents.google.com/patent/US20240163220A1/en},
urldate = {2024-08-14}
}
Ganjali, Y., Hu, Z., ZADEH, S.A.B.B.A.S.I., Abbasloo, S., Pazhooheshy, P. and Zandi, F. 2024. Systems and Methods to Migrate a Virtual Sub-Network. Feb. 2024.
@patent{ganjali_systems_2024,
title = {Systems and Methods to Migrate a Virtual Sub-Network},
author = {Ganjali, Yashar and Hu, Zhenhua and ZADEH, Sepehr ABBASI and Abbasloo, Soheil and Pazhooheshy, Parsa and Zandi, Farid},
year = {2024},
month = feb,
url = {https://patents.google.com/patent/US11895013B1/en},
urldate = {2024-08-14}
}
Bahnasy, M.M., Munir, A., Thomas, R., Si, Y.A.N. and Ganjali, Y. 2024. Hardware Accelerated Temporal Congestion Signals. Jan. 2024.
@patent{bahnasy_hardware_2024,
title = {Hardware Accelerated Temporal Congestion Signals},
author = {Bahnasy, Mahmoud Mohamed and Munir, Ali and Thomas, Ron and Si, Y. A. N. and Ganjali, Yashar},
year = {2024},
month = jan,
url = {https://patents.google.com/patent/US11863451B2/en},
urldate = {2024-08-14}
}
Zheng, X., Chu, X. and Ganjali, Y. 2023. Method and Apparatus for Managing a Packet Received at a Switch. Dec. 2023.
@patent{zheng_method_2023,
title = {Method and Apparatus for Managing a Packet Received at a Switch},
author = {Zheng, Xiaolong and Chu, Xingjun and Ganjali, Yashar},
year = {2023},
month = dec,
url = {https://patents.google.com/patent/US20230412518A1/en},
urldate = {2024-08-14}
}
Bahnasy, M.M., Ketabi, S., ZADEH, S.A.B.B.A.S.I., Ganjali, Y. and Li, F. 2023. Methods, Systems and Devices for Network Management Using Control Packets. Dec. 2023.
@patent{bahnasy_methods_2023,
title = {Methods, Systems and Devices for Network Management Using Control Packets},
author = {Bahnasy, Mahmoud Mohamed and Ketabi, Shiva and ZADEH, Sepehr ABBASI and Ganjali, Yashar and Li, Fenglin},
year = {2023},
month = dec,
url = {https://patents.google.com/patent/US11848868B2/en},
urldate = {2024-08-14}
}
Baniamerian, A., Chu, X., Ganjali, Y., Munir, A. and Sobhani, A. 2023. System and Method for a Scalable Source Notification Mechanism for In-Network Events. Oct. 2023.
@patent{baniamerian_system_2023,
title = {System and Method for a Scalable Source Notification Mechanism for In-Network Events},
author = {Baniamerian, Amir and Chu, Xingjun and Ganjali, Yashar and Munir, Ali and Sobhani, Ashkan},
year = {2023},
month = oct,
url = {https://patents.google.com/patent/US20230344768A1/en},
urldate = {2024-08-14}
}
Munir, A., Baniamerian, A., Bahnasy, M.M., Mortazavi, S.H. and Ganjali, Y. 2023. Methods and Systems for Predicting Sudden Changes in Datacenter Networks. Oct. 2023.
@patent{munir_methods_2023,
title = {Methods and Systems for Predicting Sudden Changes in Datacenter Networks},
author = {Munir, Ali and Baniamerian, Amir and Bahnasy, Mahmoud Mohamed and Mortazavi, Seyed Hossein and Ganjali, Yashar},
year = {2023},
month = oct,
url = {https://patents.google.com/patent/US20230336470A1/en},
urldate = {2024-08-14}
}