Honguk Woo

GreenBag: Energy-Efficient Bandwidth Aggregation for Real-Time Streaming in Heterogeneous Mobile Wireless Networks

Modern mobile devices are equipped with multiple network interfaces, including 3G/LTE and WiFi. Bandwidth aggregation over LTE and WiFi links offers an attractive opportunity of supporting bandwidth-intensive services, such as high-quality video streaming, on mobile devices. However, achieving effective bandwidth aggregation in mobile environments raises several challenges related to deployment, link heterogeneity, network fluctuation, and energy consumption. We present GreenBag, an energy-efficient bandwidth aggregation middleware that supports real-time data-streaming services over asymmetric wireless links, requiring no modifications to the existing Internet infrastructure and servers. GreenBag employs several techniques, including medium load balancing, efficient segment management, and energy-aware mode control, to resolve such challenges. We implement a prototype of GreenBag on Android-based mobile devices which hosts, to the best knowledge of the authors, the first LTE-enabled bandwidth aggregation prototype for energy-efficient real-time video streaming. Our experiment results in both emulated and real-world environments show that GreenBag not only achieves good bandwidth aggregation to provide QoS in bandwidth-scarce environments but also efficiently saves energy on mobile devices. Moreover, energy-aware GreenBag can minimize video interruption while consuming 14-25% less energy than the non-energy-aware counterpart in real-world experiments.

Enhanced cooperative communication MAC for mobile wireless networks

In this paper, we propose a new MAC (medium access control) protocol called enhanced cooperative communication MAC (ECCMAC) based on IEEE 802.11. The major objective of ECCMAC is to maximize the benefits of cooperative communication. We first propose a scheme for selecting and maintaining the best relay node. Second, since both cooperative communication and network coding rely on the selection of a relay node, we consider exploiting a network coding technique for additional throughput improvement. Third, to accommodate asymmetric link rates between a sender and a relay node, we employ ECCMAC to measure forward and reverse link rates, whereas prior works have simply assumed symmetric rates. ECCMAC is evaluated in this paper through theoretical analysis, extensive simulation, and simulation with measured data, and the results show that ECCMAC effectively improves wireless network performance.

A Virtualized, Programmable Content Delivery Network

In this paper, we present an open platform of content delivery networks (CDNs), namely vCDN, by which a wide range of delivery strategies can be dynamically deployed via a centralized controller and a pool of geographically dispersed cloud resources. A control application can be written for representing a specific strategy w.r.t required scale, responsiveness, and security in content delivery, and then translated into a virtualized delivery network, a form of an overlay network on cloud storages and virtual machines. In doing so, we explore the unique combination of named data networking (NDN), software defined networking (SDN), and distributed monitoring so as to achieve to a certain extent programmability in content delivery. Different from traditional CDNs, the strategy of content delivery in vCDN is not statically fixed but can be tailored by each service provider via the deployment of its own control application. Throughout a few applications running on our test bed at 7 geographical locations with multiple cloud resources, we demonstrate the feasibility of the programmable CDN and the delivery performance enhancement by inter-cloud network route managements.

Named data networking for infrastructure wireless networks

Named Data Networking(NDN) has emerged to support content-oriented services that are currently prevalent in the Internet. It successfully minimizes redundant transmission on wired links by adopting name-based access and in-network caching. In this paper, we discuss NDN in the IOT world, and propose its enhancement for efficient usage in IOT.

NDN-Based Pub/Sub System for Scalable IoT Cloud

The Internet of Things (IoT) has recently become mainstream, different from the incomplete realization of ubiquitous computing, sensor networks, and others in past that commonly shared the notion of visionary hyper-connectivity, i.e., everything is connected. Among many, one significant reason for this achievement of IoT is the nowadays availability of cloud computing which can cost effectively deal with the sheer number of globally distributed devices and data generated from those devices. In this paper, we address the issue of IoT scalability in cloud environments by proposing the unique integration scheme between pub/sub systems and Named Data Networking (NDN). While the pub/sub architecture has been applied in modern IoT cloud platforms, its implementation and deployment practice have not been fully studied for large scale IoT applications. Our approach concentrates on how to leverage the essential scalability of NDN for building pub/sub systems, thereby achieving scalable IoT cloud services.

Global-Scale Event Dissemination on Mobile Social Channeling Platform

We aim to develop a new pub/sub service through our XMPP-based platform for mobile social channeling applications. This pub/sub service is required to support the dissemination of publications from the social channels to billions of mobile clients with general interest as well as those with specific interest. Due to the wider reachability of publications, this pub/sub service imposes higher scalability requirement than the conventional ones. Meeting the scalability requirement becomes even more difficult when the pub/sub clients are geographically distributed and exhibit highly dynamic behavior. Given the dependency on the XMPP-based social channels as a substrate for publication dissemination, we first employed the state-of-the-art implementation of the pub/sub extension to XMPP servers. However, this implementation falls short in handling the aforementioned global scale and dynamic nature of our new pub/sub service. Therefore, we address the shortcomings by devising a new XMPP-based pub/sub system called SES (Scalable Eventing System). SES facilitates pub/sub communication on inter-regional clusters of pub/sub servers. SES improves scalability through a novel load-balancing technique that adapts consistent hashing and virtual node placement strategies. Also, the procedure of publication notification is enhanced for higher throughput and lower latency. We demonstrate the advancements through empirical analysis conducted on Amazon Elastic Compute Cloud (EC2).

Buffering in proxy mobile IPv6: implementation and analysis

Proxy mobile IPv6 (PMIPv6) is a network-based mobility management protocol that improves performance in terms of handover latency, signaling cost, and packet loss compared to host-based mobility management protocols. However, still packet loss occurs during the handover of the mobile node (MN). Several attempts have been made to improve the reliability of PMIPv6 service by proposing schemes in which packets are buffered in network entities during the handover of the MN to prevent packet loss, and performance improvement has been demonstrated via simulations. So far, there have been no implementations of buffering functions in the literature. This paper addresses design of buffering function and its implementation to prevent packet loss, and demonstrates the results. We have implemented a PMIPv6 testbed based on open source resources. We discuss the functional and performance enhancements, comparing PMIPv6 with the buffering implemented and standard PMIPv6. We also propose an improved buffering function where the packet forwarding rate of the buffer is adjusted. The results through the testbed show that the buffering function in PMIPv6 effectively prevents packet loss during the handover of the MN. We have found out that we can manage the amount of packets in the buffer without further increment by adjusting the packet forwarding rate of the buffer as well.

A large scale presence network for pervasive social computing

In this paper, we present a scalable presence network architecture for emerging pervasive social applications in which the interaction of not only people but surrounding physical objects seamlessly occurs. The architecture combines the server-based messaging and the peer-to-peer communication schemes, thereby rendering itself deployable at the large scale beyond social networks so as to incorporate a sheer number of embedded devices in a ubiquitous environment.

Architecture of a network performance monitor for application services on multi-clouds

Recently, many new services are being deployed on the cloud due to its flexibility of on-demand allocation and deallocation of computing resources. As the services require more advanced features such as fault-tolerance and high performance, much interest is focused on the multi-clouds. Hence, the needs for effectively managing the quality of services on multi-clouds emerge rapidly. Among the many factors affecting the quality of services, we focus on the network performance in this paper. We identify essential requirements of network performance monitor for multi-clouds and propose an architecture. In particular we address the necessity of supporting external agents and discuss how to integrate with them in a flexible and extensible way. In addition, the issues of timely delivery and off-line analysis of measured results are addressed.

A Revisit to Web Browsing on Wearable Devices.

Wearable devices and smartwatches have become prevalent in recent years, yet consuming web contents on those devices are not common mainly due to their restricted IO capabilities. In this paper, we revisit the web browser model and propose the notion of fast access browsing that incorporates the lightweight, always-on web snippets, namely widget view, into web applications. This allows smartwatch users to rapidly access web contents (i.e., within 200ms) similarly as they interact with notification. To do so, we analyse about 90 smartwatch applications, identify the quick preview pattern, and then define the constrained web specifications for smartwatches. Our implementation, the wearable device toolkit for fast access browsing, is now being tested and deployed on commercialized products and the developer tool for building widget view enabled web applications will be soon available as the smartwatch SDK extensions.