A New DCQCN Rate Increase Algorithm with Adaptive Byte Counterby Daisuke Sugahara (Kansai University, Japan); Osamu Shiraki and Eiji Yoshida (Fujitsu Laboratories Ltd., Japan); Miki Yamamoto (Kansai University, Japan)
Abstract – Datacenter networks are constructed of high speed networks, such as 40Gbps Ethernet, and quite severe commu- nication quality, low latency and high throughput, is required from applications. RDMA (Remote Direct Memory Access) is now a hot topic in datacenter networking because it can operate high speed data transmission without TCP/IP protocol overhead. DCQCN (Datacenter Quantized Congestion Notification) has been proposed as congestion control for RDMA over Ethernet. In this paper, we reveal with our preliminary evaluation for DCQCN that 1) byte counter behavior causes fairness issues, 2) parameter setting of small timer for time counter and large byte counter makes time counter a dominant factor in rate increase algorithm, and 3) time counter behavior which is independent of transmission rate resolves fairness issues. And we also show that DCQCN has a technical problem of delaying transition timing to Hyper-Active Increase phase due to large byte counter setting, which causes slow convergence to the desirable transmission rate. We propose a new setting for byte counter in DCQCN so that its interval of counter increment is independent of transmission rate. This algorithm, named Adaptive Byte Counter DCQCN (ABC- DCQCN), enables shorter interval of byte counter increment and can resolve both fairness issue and fast convergence of transmission rate.
Real-World Implementation of Function Chaining in Named Data Networking for IoT Environments by Yohei Kumamoto (Waseda University); Hiroki Yoshii and Hidenori Nakazato (Waseda University, Japan)
Abstract – In this paper, we discuss how to implement function chaining in Named Data Networking (NDN), an incarnation of information centric networking technology, for real-world IoT environments. We explain our new architecture, called NDN-FC, for function chaining over NDN, and how to extend existing NDN software to support function chaining. The key features discussed in this paper are Interest and Data packet structure, forwarding methods, and segmentation and reassembly methods of a content. Even in IoT environments, it is possible that most content, such as image and video, does not fit into a single Data packet. Segmentation and reassembly of a content is therefore crucial. The feasibility of our proposed protocol for segmentation and reassembly is displayed through a prototype implementation. In order to support lightweight operation of functions, the implementation is extended to use Docker container technology to run functions. The performance of Docker implementation and virtual machine implementation are compared.
Performance Analysis of Periodic Cellular-IoT Communication with Immediate Release of Radio Resources by Shuya Abe (Osaka University, Japan); Go Hasegawa (Tohoku University, Japan); Masayuki Murata (Osaka University, Japan)
Abstract – Mobile cellular networks are now serving all kinds of Internet of Things (IoT) communications. Since current contention-based random access and radio resource allocation are optimized for traditional human communications, massive IoT communications cannot be efficiently accommodated. For this reason, standardization activities for connecting IoT devices, such as Cellular-IoT (C-IoT), have emerged. However, there have been few studies devoted to the evaluation of the performance of the C-IoT communications with periodic data transmissions, despite their being the common characteristics of many IoT communications.
Herein, we evaluate the capacity of mobile cellular networks in accommodating periodic C-IoT communications, focusing on differences in performance between LTE and Narrowband- IoT (NB-IoT) networks. To achieve this, we conduct end-to-end performance analyses of both control and data planes, including the random access procedure, radio resource allocation, and bearer establishment in EPC network. Moreover, we determined the effect of immediate release of radio resources considered in 3GPP. Numerical evaluation results show that NB-IoT can accommodate more IoT devices than LTE, although this results in significant latency in data transmission. Furthermore, we find that the number of IoT devices that can be accommodated increases up to 20.7 times with immediate release of radio resources.
A Hierarchical, Scalable Approach for Availability Analysis of Software Defined Networks by Swapna S. Gokhale (University of Connecticut, USA); Veena B. Mendiratta (NOKIA Bell Labs, USA); Lalita J Jagadeesan (Nokia Bell Labs, USA)
Abstract – Software-Defined Networking (SDN) separates the network control and data planes in communication networks, thereby enabling the dynamic reconfiguration of the data plane at run-time through the control plane software. SDNs can be large, comprising of tens of controllers and thousands of switches, where combinatorial models for availability analysis can lead to state space explosion. Further complicating matters, the logically centralized SDN control plane is realized in practice in a distributed fashion to provide horizontal scale-out and redundancy to avoid a single point of failure. This distribution necessitates the use of consensus mechanisms to ensure consistency on key data in the context of failures, introducing further considerations into SDN models. To this end, we present the first hierarchical analytical model of SDN availability, under control and data plane failures, that takes consistency and recovery into account, and that is scalable. Our experiments using this model demonstrate the interplay between characteristics of the distributed control plane and data plane in overall network availability.
Leontief-Based Data Cleaning Workload Distribution Strategy for EH-MWSN by Concepcion Sanchez Aleman, Niki Pissinou and Sheila Alemany (Florida International University, USA)
Abstract – The use of energy-harvesting technologies in mobile wireless sensor networks (MWSN) delivers a promising opportunity to mitigate the limitations that irreplaceable energy sources impose over conventional MWSN. We propose Leontief-Data Cleaning Distribution Strategy (Leontief-DCD), an economic model-based method designed to distribute the data cleaning workload in energy harvesting MWSN powered by predictable energy sources, such as solar energy. Leontief-DCD creates interdependencies among sensor nodes to predict the required cooperation from each node in the data cleaning process. Different from existing task allocation methods, the interdependencies in Leontief-DCD allows for to plan a workload distribution that benefits the network as a whole, rather than only individual sensors, which consequently benefit the overall system performance. Our results show that when employing our method to distribute data cleaning workload in highly dirty, real-world datasets in scenarios with high and low energy, our method increased the number of data samples engaged in data cleaning processes by up to 25.57%, the count of active sensor nodes by up to 44.01%, and the network overall well-being by up to 55.42% compared to data cleaning performed by each node individually.
Connection-Oriented BLE Traffic Servicing Characteristics on Android Devices by Joshua Siva and Christian Poellabauer (University of Notre Dame, USA)
Abstract – We seek to better understand Bluetooth Low Energy (BLE) connection-oriented communication on Android devices. We vary packet size, connection interval, and method of communication, and measure the time for communication operations to complete and packet arrival times on a variety of Android devices. The collected observations help to characterize the traffic and service characteristics of BLE and Android devices, respectively. Although BLE is touted as an ideal protocol for many Internet of Things (IoT) applications, the observations suggest that any IoT operations that require latency guarantees might be unable to meet such guarantees if communication includes Android smartphones.
Determining the Indoor Location of an Emergency Caller in a Multi-story Building by Luke Logan (Illinois Institute of Technology, USA); Carol Davids (Illinois Institute of Technology & School of Applied Technology, USA); Cary Davids (IIT, USA)
Abstract – A method to determine the indoor location of a caller who uses a smartphone to call for emergency assistance in a multi-story building is described. The system uses the signals broadcast by an array of BlueTooth Low Energy iBeacons deployed in the building as input to a trilateration calculation. The method described could be applied to RF emitters other than BlueTooth, including WiFi access points. The results of experiments conducted in a live classroom building in which the iBeacons are deployed demonstrate 100% accurate floor identification and mean distance accuracies within 5 meters. The solution’s multistory accuracy and its low cost in terms of money, data collection effort, and simplicity of the calculations demonstrate this as a reasonable solution to the problem of locating emergency callers in multistory buildings.
Firewall Configuration and Path Analysis for SmartGrid Networks by Nastassja Gaudet, Abhijeet Sahu and Ana E Goulart (Texas A&M University, USA); Edmond Rogers (IT TECHNICAL ASSOCIATE, USA); Katherine Davis (Texas A&M University, USA)
Abstract –Firewalls are needed in electrical utility companies to protect their substations, control centers and their communication with the balancing authorities. Firewalls also allow the creation of demilitarized zones (DMZ’s) in the utilities, where information about the utility’s operation can be accessed by the corporate network and outside contractors. As an additional step in creating a cyber topology for a synthetic power system, in this paper we model an electrical utility and the main data flows in and out of its control center. This allows the creation of use cases and firewall rules for each case. The path of selected use cases are analyzed in terms of open ports and risk level.
Data Processing and Model Selection for Machine Learning-based Network Intrusion Detection by Abhijeet Sahu, Zeyu Mao, Katherine Davis and Ana E Goulart (Texas A&M University, USA)
Abstract – Signature-based Intrusion Detection Systems (IDSes) such as Snort, BRO or Suricata depend on specific patterns and byte sequences in network traffic to detect intrusions; hence, they cannot prevent intrusions for unknown zero-day attacks. Various anomaly-based IDSes that have been proposed based on machine learning (ML) techniques incur high false positives. To overcome this, we explore different types of data processing, i.e. data balancing, feature correlation, normalization, and feature reduction, and whether they are necessary for datasets with different feature dimensions: Coburg Intrusion Detection Data Sets (CIDDS) with five features and Knowledge Discovery and Data Mining (KDD) with 41 features. Further, we perform model selection by comparing the performance of various linear and non-linear classifiers. Generally, our results show that non- linear classifiers outperformed linear ones and that using data balancing and normalization improves the overall accuracy for most classifiers.