Kotaro Kataoka

A rapidly deployable disaster communications system for developing countries

We address the problem of designing a communication system for deploying under disaster situations where the pre-disaster communication infrastructure has been uprooted. The disaster communication system is required to aid in rescue effort by providing text, voice, and video communication links between the field worker and the operation center. It is also required to provide some communication services to the affected citizens so that the overwhelming demands made on the communication infrastructure during the times of disaster are mitigated to some extent. The other highlights of the design are its low-cost nature by exploiting the economies of scale and the fact that basic mobile phones with text-only features are widely available in developing countries.

liteFlow: Lightweight and distributed flow monitoring platform for SDN

Traditional network monitoring involving packet capturing or flow sampling has many challenges such as scalability, accuracy and availability of processing resource when networks become large-scale, high-speed and heterogeneous. SDN is a promising approach to address these challenges, but each SDN switch has its own capacity limitation, such as its cache memory called TCAM, and thus it needs coordination of resources with other network nodes to monitor the network in a scalable manner. liteFlow introduces an intelligent framework for authority switch selection, which is in charge of monitoring all flows between an end-hosts pair, while the other path switches simply forward packets without any monitoring process. The proposed system distributes the load of monitoring flows among SDN switches, and makes the scalability and accuracy of network monitoring manageable. This paper also implements a few metrics for flow-based network monitoring as example applications of liteFlow. The proof of concept implementation has been done using SDN testbed partially involving campus network of IIT Hyderabad.

Auto-Configuration of SDN Switches in SDN/Non-SDN Hybrid Network

This paper proposes an auto-configuration mechanism for a newly attached SDN (Software-defined Networking) switch and intermediate switches in an SDN/non-SDN hybrid network. Automation of initial configuration of SDN switches brings the benefit of reducing the installation costs and maintaining the operational consistency in an enterprise network. However, the heterogeneity of a hybrid network introduces difficulty in achieving such benefits. The proposed system achieves the following features: 1) detecting a new SDN switch in a hybrid network, 2) providing the switch with appropriate configuration parameters in an SDN/non-SDN hybrid network and 3) ensuring seamless service across the SDN and non-SDN segments. The proposed solution also introduces DHCP-SDN as an extension of Dynamic Host Configuration Protocol (DHCP) to support SDN parameters. The proposed system enables a new SDN switch to automatically start operating even though the intermediate switches are not configured with the SDN-specific VLAN setup.

LightChain: A lightweight optimisation of VNF placement for service chaining in NFV

In network function virtualization (NFV), the placement of the virtual network function (VNF) significantly affects the load on switches and the efficiency of bandwidth utilization. Suboptimal placement of VNFs in service chains increases the flow rules in the switches and the ping-pong traffic among the VNFs. This paper presents LightChain, an efficient polynomial run time approach for optimizing VNF placement and service chaining in NFV. LightChain forms the directed acyclic graph of the given service chains and applies topological sorting to the graph. The significance of the proposed approach is as follows: 1) The minimum hop-count in each service chain, 2) Reduction of flow rules on SDN switches, and 3) Elimination of ping-pong traffic. Evaluation results show that LightChain can determine the placement locations of the VNFs in a reasonably short time. Our comparison results with other approach show that LightChain installs a lower number of flow rules in the switches.

Segmented proactive flow rule injection for service chaining using SDN

Service chaining is one of the challenges in network function virtualization (NFV), where a sequence of middleboxes apply their service to particular traffic. The challenge of service chaining is that a middlebox may alter the content of packet headers; thus, the context of service chaining for particular traffic gets lost. This paper proposes Segmented Proactive Flow Rule Injection (SPFRI) using Software Defined Networking (SDN). SPFRI maintains the consistency of service chain even though one or more middleboxes may alter the packets at both outbound and inbound directions of a flow. The network configuration of middleboxes must be appropriately set to make them perform in the service chain. The significant benefits of the proposed SPFRI are as fallows: 1) no requirement of modifying to the middlebox operating system or software, 2) longer and more chains and 3) friendliness to non-SDN switches. This paper details the segmentation of the service chain and calculates the required number of flow rules. We also discuss the performance of SPFRI and a realistic approach to practical implementation.

Performance evaluation of wireless ad-hoc network for post-disaster recovery using Linux Live USB nodes

Wireless ad-hoc networks are unplanned, quickly deployable, infrastructureless, and realistic for networking in emergency or restricted situations. They have actively been studied and proposed for post-disaster recovery. Our approach uses Linux Live USB nodes, which can be any available laptop computer at disaster-affected site, to boot the guest OS using USB flash drive with a variety of customized software packages demanded in post-disaster situations. Using an additional Wi-Fi USB adapter, Linux Live USB nodes help to easily form wireless ad-hoc network and serve user devices through Wi-Fi hotspot. While a lot of studies have been done over simulation, it is crucial to understand how wireless ad-hoc networks behave and their performance in the real field. In this paper, we examine different scenarios of wireless ad-hoc network deployment inside a building: 1) the assignment of frequency (Single Channel or Multi Channel), and 2) choice of mesh and routing protocol (OSLR or batman-adv). According to the results of field trials, our primary option of deployment will be Multi Channel batman-adv to cover multiple rooms inside the building expected to be evacuation shelter.

SFO: SubFlow Optimizer for MPTCP in SDN

Multipath TCP (MPTCP) brings promising features of throughput improvement and load balancing by utilizing the multiple available paths. However, adding a subflow without awareness of the network state will result in wasting network resources and degrading MPTCP performance. Finding and operating the proper number of subflows are necessary to achieve the best performance of MPTCP. However, an MPTCP host does not have knowledge about the number of available paths in the network, and it may create the suboptimal number of subflows. This paper proposes SubFlow Optimizer (SFO) for MPTCP in SDN that estimates the Optimum Number of Subflows (ONS) and informs it to the MPTCP hosts. The estimated ONS reflects the number of optimal paths found out of the set of feasible paths between the server and client. Selecting such an optimal path for a new subflow improves its throughput whereas detecting ONS makes efficient use of the overall network. SFO also enables the quick negotiation of ONS between MPTCP hosts during its session establishment that reduces the time required for ONS negotiation down to one Round-Trip Time (RTT). Therefore, MPTCP hosts can immediately achieve the best performance without incremental try and catch of subflow creation. The results of performance evaluation show that SFO brings the significant throughput improvement to the overall MPTCP session compared with the cases of improper estimation of the available number of subflows.

FastSplit: Fast and Dynamic IP Mobility Management in SDN

IP Mobility has many unresolved issues including triangle routing, session survivability, long delay and packet loss during hand-off. SDN-based IP Mobility promises to mitigate them by taking the advantage of network-wide visibility and centralized control. However, SDN-based approach still has some problems namely 1) large signaling overhead of controller and 2) high hand-off latency as the number of nodes and size of the network become large. This paper proposes a simple approach to efficiently handle IP mobility in SDN called “FastSplit”. FastSplit reduces the signaling overhead to reestablish the new path after movement by reusing the existing path between correspondent node (CN) and mobile node (MN). Our solution is two fold: 1) it finds out the redirection/split point in polynomial time by restricting the search area along the previous path between MN and CN, and 2)using Sub-Optimality Factor (SOF) for determining the cost-effective split point in order to manage the trade-off between the optimality of reestablished path and the reduction of signaling overhead for handling host mobility. We present an algorithm to find out a split point between MN and CN so that the existing path configuration can be reused. The evaluation results showed that FastSplit contributes to achieving significantly less signaling cost by reducing the number of SDN switches to reconfigure, by up to 50% as well as shortening the hand-off duration by 17% in comparison to OpenFlow based Mobile IP. FastSplit is also proven that such performance can be achieved without compromising the overall end-to-end throughput between CN and MN at the new location.

Scaling a broadcast domain of Ethernet: Extensible transparent filter using SDN

Scalability of Ethernet is a known issue where communication is disturbed by broadcast traffic caused by a large number of nodes in a single broadcast domain. This paper proposes Extensible Transparent Filter (ETF) for Ethernet using SDN, that suppresses broadcast traffic in a broadcast domain by selecting an appropriate outgoing port of the switch through which the target host of broadcast packet is reachable. ETF maintains both consistent functionality and backward compatibility of the existing networking protocols such as Address Resolution Protocol, Dynamic Host Configuration Protocol, and other possible protocols that work with the broadcast of a packet. ETF was developed using Trema and Floodlight, which are open source SDN controller supporting OpenFlow. The number of flooding was significantly reduced in a broadcast domain where user clients connecting to the same VLAN across multiple production SDN switches successfully performed DHCP and ARP respectively. This paper also discusses the practical deployment and operation of SDN and ETF as a part of campus network in Indian Institute of Technology Hyderabad. The proposed approach improves scalability of a broadcast domain of Ethernet as an enhancement to the existing enterprise networks using SDN.

A stand-alone quickly-deployable communication system for effective post-disaster management

We address the problem of designing a communication system for deploying under disaster situations where the pre-disaster communication infrastructure has been uprooted. The disaster communication system is required to aid in rescue effort by providing text, voice, and video communication links between the field worker and the operation center. It is also required to provide some communication services to the affected citizens so that the overwhelming demands made on the communication infrastructure during the times of disaster are mitigated to some extent. The other highlights of the design are (i) its low-cost nature achieved by exploiting the economies of scale, (ii) constant communication between the field worker and manager, (iii) the use of basic mobile phones with text-only features for gathering victim data and populating the victim database in real-time, and (iv) web-based applications for managing victim data.