Best Paper! – L7 Packet Switch: Packet Switch Applying Regular Expression to Packet Payload by Shogo Ando and Akihiro Nakao
Abstract – Software-Defined Networking (SDN) is widely studied in research community. Although OpenFlow switches are considered one of the promising SDN technologies, even the latest version of OpenFlow (1.3) only handles 40 tuples including Layer 2 to Layer 4 (L2–L4) headers. However, application developers and end users may not often find ”flow” based traffic control. In this paper, we propose L7 Packet Switch, which is a technique for controlling the packets based on the payload of the Layer 7 (L7) instead of the L2–L4 header information. L7 Packet Switch maintains TCP flow information and performs regular expression matching to L7 payload. If a packet in a flow matches a regular expression, the packets belonging to the same flow are controlled by the same rule, i.e., actions such as deflecting a flow to specific output ports. L7 Packet Switch also provides Southbound API so that the regular expressions to be applied and the corresponding actions can be specified on the fly. We implement L7 Packet Switch on FLARE node which is a programmable node using Click. We also evaluate the performance and show the possibility of operation in a network environment of 10 Gbps.
Best Paper! – Performance evaluation of Flow-Aware Multi-Topology Adaptive Routing by Robert Wojcik , Jerzy Domzał, Zbigniew Dulinski, Piotr Gawłowicz and Dawid Kowalczyk
Abstract – We evaluate a new approach to routing packets in IP networks: Flow-Aware Multi-Topology Adaptive Routing (FAMTAR). FAMTAR combines flow-aware traffic management and an adaptive routing mechanism. FAMTAR allows to create and use additional paths when such demand occurs. The evaluation demonstrates that it is possible to significantly increase the amount of traffic sent in a network in a given time.
Analyzing the Drawbacks of Node-Based Delays in Tor by Kale Timothy Girry, Satoshi Ohzahata, Celimuge Wu and Toshihiko Kato
Abstract—Tor Network is a low-latency anonymity and censorship resistance network, which relaying traffics through onion routers before being forwarded to the destination. The distribution of trust offered by Tor creates resilience in the face of compromise and censorship. However, there are performance and usability issues in the Tor network. In this paper, we analyze the delay contributions from the host TCP stack of the Tor relay node. We measured and analyzed the transfer data passing through the Tor nodes to evaluate the main causes of delays. The experimental analysis show that increasing delays on nodes are affected by the heavy TCP packets queuing, kernel memory exhaustion and TCP advertised window not able to accept more data. We present our experimental analysis and measurement results as a prerequisite for addressing the Tor’s end-to-end increasing latency and poor usability.
A Method for Estimating the Average Packet Error Rates of Multi-carrier Systems With Interference by Zaid Hijaz and Victor S. Frost
Abstract— We propose a method for estimating the average packet error rates (PER) of a coded orthogonal frequency division multiplexed (OFDM) system when the signal is subject to interference with a shaped spectral density. Thus the interference has a different impact on each sub-carrier. Typically, the PER is a function of the received signal-to-noise-ratio (SNR) and the packet bits are distributed evenly amongst a fixed number of sub-carriers where each sub-carrier has the same SNR. The OFDM system under consideration here utilizes convolutional coding for error correction and Mn-ary square quadrature amplitude modulation (M-QAM) for signaling. The proposed method assumes that each sub-carrier is affected by a different interference power. The novelty in the analysis lies in equating the average interference power affecting each sub-carrier to an equivalent additive white Gaussian noise (AWGN) power. We compare PER prediction using this approach to simulation for a number of modulation and coding schemes (MCS) to validate that the proposed approximation can be used to calculate the average PER.
Improving Image Quality and Interactivity in Free-Viewpoint Video Transmission by Hiroya Sannomiya, Norishige Fukushima and Yutaka Ishibashi
Abstract —Free-viewpoint video, which enables us to change viewpoint of the video freely, attracts rising attention as visual media. Transmitting such media over an IP network, the network delay deteriorates qualities of the media. In this paper, we propose three types of control for the depth and image transmission method in free-viewpoint video transmission. One is buffering control, another is prediction control, and the other is a combination use of both types of control. The buffering control buffers the viewpoint information, and the prediction control predicts the information. We investigate the effects of the three types of control by Quality of Experience (QoE) assessment. Assessment results show that the optimum buffering time and optimum prediction time exist in the buffering control and the prediction control, respectively. Also, the buffering control improves the image quality, but the interactivity is degraded. The image quality becomes higher by the prediction control. Furthermore, there exist the optimum values of the buffering time and prediction time in the combination use, and the use further improves QoE.
Accepting More General Telephone Calls in Emergency Situations by Limiting General Call Holding Time under Trunk Reservation Control by Kazuki Tanabe, Sumiko Miyata and Katsunori Yamaoka
Abstract —We investigated the effect of limiting the holding time of general telephone calls on their acceptance rate during an emergency situation. Simulation using a queuing system model of a telephone exchange showed that limiting the holding time of general calls increased the number of accepted general calls even when their acceptance was threshold-controlled while still enabling the required number of emergency calls to be accepted. An investigation of the relationship between the trafﬁc intensity of accepted general calls and the general call-blocking rate showed that an upper limit on the trafﬁc intensity can be set by using threshold control. Limiting the holding time enables more general callers to share the acceptable trafﬁc intensity, which reduces the general call-blocking rate. Calculation of the call-blocking rate for various numbers of lines in a telephone exchange showed that the effect of a general call holding time limitation on the general call-blocking rate is equivalent to that obtained by adding a certain number of lines.
Channel-Aware Transmission of Feedback Messages for WiMAX ARQ by Tsung-Kang Hung, Feng-Zhi Zhuang and Hsu-Feng Hsiao
Abstract —In the 4G communications standards, including Worldwide Interoperability for Microwave Access (WiMAX) and Long Term Evolution (LTE), Automatic Repeat reQuest (ARQ) is one of the important tools to coop with error-prone transmission environments. The primary function of ARQ is to provide reliable communication between a transmitter and its receivers through the retransmission of lost ARQ blocks. The retransmission at the sender can be triggered by either of the two events. The first one is for the sender to retransmit an ARQ block whose retry timer expires. The second one is when the transmitter receives the feedback message reports, which indicate block reception situation from the receiver, the flagged lost ARQ blocks will then be sent again. There are four types of feedback messages defined in the WiMAX standard. How and when the receiver reports feedback messages will have impact on the transmission efficiency. If the receiver reports feedback messages more frequently, the lost ARQ blocks can be retransmitted more promptly at the cost of higher bandwidth consumption of the feedback channel. By contrast, if the receiver waits for a longer time before sending a feedback message, feedback messages can be formed more efficiently in terms of bandwidth usage with longer transmission time incurred. In this paper, a performance evaluation metric is defined to consider the influence of both transmission latency and the required bandwidth to transmit feedback messages. A feedback message transmission scheme is further proposed to determine the interval and the types of feedback messages dynamically, based on the condition of channel conditions. The proposed method is aimed to find a balance between the transmission delay and the bandwidth usage efficiently, and the improvement of the proposed method is confirmed by the experiment results.
Link-failure assessment in redundant ICS networks supported by the Interconnected-asset Ontology by Henk Birkholz and Ingo Sieverdingbeck
Abstract – Ethernet-based networks for Industrial Control Systems (ICS) rely on redundant links to mitigate failures, to increase availability, but also masking failures and complicating assessment of redundancy remaining. We use the Interconnected-asset Ontology to represent topologies from automated snapshots without the need for changes, infrastructure or fundamental operation of the ICS network. The presented approach and prototype provides assessments seamlessly integrated in a state-of-the-art industrial control system.
A MATLAB Platform for Characterizing MIMO-OFDM Communications with Software Defined Radios by Ryan Measel, Donald J. Bucci, Christopher S. Lester, Kevin Wanuga, Richard Primerano, Kapil R. Dandekar and Moshe Kam
Abstract — A new MATLAB-based, wireless measurement platform using an existing software-defined radio architecture is presented. It augments IEEE 802.11g MIMO-OFDM physical layer schemes with new designs such as Maximal Ratio Combining, Alamouti coding, and Spatial Multiplexing. The platform provides a series of metrics, including channel capacity, Error Vector Magnitude (EVM), and Post-Processing Signal-to-Noise Ratio (PP-SNR) to characterize link and network performance. The software implementation and test protocol of the platform are presented with a validation study demonstrating its application.
Utility Maximization for Chunk-based OFDMA Systems with Multiple BER Requirements by Tianzhou He, Xin Wang and Wei Ni
Abstract — In wireless orthogonal frequency division multiple access (OFDMA) standards, subcarriers are grouped into chunks and a chunk of subcarriers is made as the minimum unit for subcarrier allocation. We investigate the chunk-based resource allocation for OFDMA downlink, where data streams contain packets with diverse bit-error-rate (BER) requirements. Supposing that adaptive transmissions are based on a number of discrete modulation and coding modes, we propose the optimal resource allocation scheme that maximizes a utility function of average user rates. With proper formulation, the relevant optimization problem is cast as an mixed-integer program, and it is shown that the optimal strategy for this problem can be obtained through Lagrange dual-based gradient iterations with fast convergence and low computational complexity per iteration. Furthermore, a novel on-line algorithm is developed to approach the optimal strategy without knowledge of intended wireless channels a priori. Numerical results show that the proposed optimal chunk allocation schemes with different α-fair utility functions can nicely trade off total network throughput for fairness among users