Jit2007

AcoustaNomad: A Context-aware Blogging and Learning Platform in Mobile IPv6 Networks 271 AcoustaNomad: A Context-aware Blogging and Learning Platform in Mobile IPv6 Networks Ching-Hui Chen+, Wu-Hsiao Hsu*, Rung-Huei Liang+, Sheng-Cheng Yeh+ Department of Computer and Communication Engineering, Ming-Chuan University * Department of Computer Science and Information Engineering, Ming-Chuan University Taiwan, R.O.C. wuhsiao@mcu.edu.tw + Abstract Recently, the development of mobile applications is more and more popular. The mobile applications will not only take advantage of contextual information, such as location-awareness, to offer greater services to a mobile host (MH) but maintain exiting transport-layer connections as the MH moves from one location to another. This paper exhausts our most recent work: the AcoustaNomad project. AcoustaNomad not only uses the mobile IPv6 to maintain the existing connections even if the MH changes locations and addresses, but utilizes location-aware technique to detect what kind of services the new location provides. In addition, AcoustaNomad provides two mature mobile applications: mobile learning and audio blogging. This paper proposes the architecture of AcoustaNomad and experimental results that demonstrate the ability of AcoustaNomad to enable location-aware services and applications. Keywords: IPv6, mobile learning, audio blog, locationbased service, context-aware. 1 Introduction 1.1 Background Wireless and mobile networks become popular research issues in recent years; as a consequence, mobile wireless networks have evolved to integrate with IP-based infrastructure for multimedia applications. It is becoming clear that mobile host (MH) will be expected to have access to popular applications offered in the traditional wired networks. As a result, the development of mobile applications plays an important role in IP-based wireless networks. Since a MH can roam across different wireless networks, knowledge about locations is the necessity for the MH to detect what kind of services the current location provides. Therefore, the development of mobile applications should take advantage of contextual information, such as location, to offer greater services to the MH. In addition, the mobile IPv6 [1] should also be used to allow transparent routing of IP datagrams in order to maintain exiting transport-layer connections as the MH moves from one location to another. More specifically, mobile IPv6 allows an IPv6 capable MH to be mobile — to arbitrarily change its location on an IPv6 network—and still maintain existing transport layer connections survivability. Mobile technologies also provide an opportunity for a change in education, moving from occasional use of a computer in a lab towards more embedded use in the classroom. Educators and technical developers are in exploiting the capabilities and characteristics of mobile technologies to enable new and engaging forms of learning. Mobile technologies not only have great impact on learning, but also create challenges for the media industry. Digital music technologies, audio compression techniques, increasing data transfer speeds, and the recent success of digital mobile players like the iPod, have forced the music industry to develop and alter the conception of the music by putting it into a mobile space. In Location33 [2], users equipped with a GPS enabled PDA or mobile phone, walk around downtown Culver City, California and explore a new type of musical album by mixing together songs and stories based on their movement. In this paper, we present our recent work: the AcoustaNomad project which is supported by IPv6 R&D division of national information and communications initiative (NICI) Taiwan. AcoustaNomad integrates mobile learning, mobile audio blogging, location-aware technology and mobile IPv6 technologies for any MHs. When a MH roams across multiple wireless networks, its IPv6 address may change in order to maintain connectivity. AcoustaNomad not only uses the mobile IPv6 to maintain the existing connections even if the MH changes locations and addresses, but utilizes location-aware technology to detect what kind of services the new location provides. At the current implementation, AcoustaNomad provides two mature mobile applications: mobile learning and mobile audio blogging. When a MH roams across multiple wireless networks, it will trigger the m-blogging service 272 Journal of Internet Technology Volume 8 (2007) No.3 and allow the MH to upload its audio file that recorded the feelings or stories about a specific location to a remote server. Any MH can download the audio files when it roams to the same location. The system will also activate location specific learning contents in which language learners will be able to move around the physical space and retrieve location-specific content to act out the authentic conversations. 1.2 Design Elements of AcoustaNomad Necessary elements of AcoustaNomad are listed as follows: A. Time The blog is another way of storytelling and the time sequence is embedded. The most important property of the sequential data is that it updates promptly which makes the m-blogging keep the magical attraction ceaselessly, and always with unknown things to be discovered. B. Locations The development of mobile devices makes the mobile blogging possible in which people can blog everywhere while the sense of place is removed. The location element is amplified in m-blogging and m-learning with the audio content of unique culture from the location of the content creator; and therefore, the locative sense is kept. C. Sound The popularity of the consumer electronics and the digital recorders forms the background of using the vocal media. In m-blogging, audio creates the immersive feelings and voice augments the virtual story to promote the story conversation and social interaction. On the other hand, m-learning also uses audio as material to facilitate learning while one is moving and engaged by visual cues. D. Narration Narration in m-blogging is the unique style in storytelling which makes the story plots more dramatic than the traditional blogs and forms a new “storytelling language” which is collaborative and contributed from all of the participants. E. Interaction The interaction property is implied in m-blogging by active listening. People listen to or create audio contents to response to other bloggers. Therefore, the interaction reveals the social relationships of the story roles and the participants. Furthermore, the m-blogging is a form of the postmodern literature by actually listening to audio contents in arbitrary order. The remainder of this paper is organized as follows. Section 2 introduces the previous works related to mobile context-aware applications, mobile learning and audio blogging. The architecture of AcoustaNomad is explained in section 3. In section 4, we describe our current implementation status and present some results. Some conclusions are given in section 5. 2 Related Work Many research projects have developed the mobile context-aware applications in IP networks. Some related works are presented in the following subsections. 2.1 The Context-aware IPv6 Applications The article [3] developed a novel wireless access point protocol designed to support the development of next generation mobile context-aware applications in their local environments. Once deployed, the architecture will allow ordinary citizens secure, accountable and convenient access to a set of tailored applications including location, multimedia and context based services, and public Internet. The developed architecture utilizes packet marking and network level packet filtering techniques within a modified mobile IPv6 protocol stack to perform access control over a range of wireless network technologies. The project offers a number of significant advantages, such as support for fine-grained access control policies, soft-state based authorization, support for continuous media or streaming applications while roaming and scalability, over existing systems. The article [4] discussed the issues related the application of wireless communication and positioning technologies to the development of handheld electronic tour guides. The goal is to allow tourists to navigate both the physical space and related information space when visiting a museum or a city. The system detects the position of the tourist, and provides personalized and contextual information. 2.2 The Location-aware Services Location-aware services have received great attention in wireless networks. The Global Positioning System (GPS) is based on a well-known principle called the triangulation technique, which can work accurately outdoors but fails indoor due to the blocking of buildings between satellites and the GPS receiver [5]. Therefore, several approaches have been adopted for estimating the location of an MH inside buildings [6]. Among them, the existing environment of wireless LANs offers an inexpensive solution to the indoor location-aware problem. The first locating system based on such an approach is called RADAR [7]. It operates by measuring and recording the signal strength (SS) of a number of sampled locations during the off-line phase, and infers the location of a MH based on the recorded SS database information in the AcoustaNomad: A Context-aware Blogging and Learning Platform in Mobile IPv6 Networks 273 real-time phase. RADAR is able to estimate a user’s location to within a few meters of the MH’s actual position. This suggests that a large class of location-aware services can be built over wireless LANs. 2.3 Immersive Audio The article [8] developed a model for authoring media content linked to physical space, which showed the concept of a virtual layer overlapped upon the physical space. Similar concept was found in Soundscape Composition [9] which provides pre-recorded audio content at specific locations. In the earlier works, these kinds of contents lack the relation of location and media, and thus they are against Salamensky’s theory [10]. As a result, the recent works try to link multimedia with location using location-specific technology. SoundWalk [11] provides a tour experience with vocal and background sounds, however, the listening mode is linear. In Location33 [2], the received multimedia contents are independent to specific individuals which breaks up the traditional linear music listening mode, and uses the individual’s walking path to create a new song, but it is still against Salamensky’s theory. The way that people are both receivers and creators solves the problem as Urban Tapestries [12] enables people to leave their path in the city and form a complex network. However, these works do not provide the opportunity for people to interact with each other as suggested by Salamensky’s theory. Patholog [13] uses weblogs to create interactions among users based upon the suggestions by Hourigan [14] that a blog could create more direct dialogue between readers and writers. Nevertheless, although users will be able to read weblogs anywhere, they won’t be able to associate the location with the media. In AcoustaNomad, the relation between location and people from their cognition echoes Shotter’s theory [15]; the relation between multimedia and people from their individual cultural background echoes Bakhtin’s theory [16]. Moreover, the inherent property of interaction within blogging and the game property are embedded in AcoustaNomad. Therefore, AcoustaNomad realizes the theory of Salamensky to form an interactive and immersive audio world. experience, but also learned to help each other, to offer suggestions and ideas, and to try things out through practice. TANGO project [18] proposed vocabulary learning in a ubiquitous computing environment. Each student held a PDA that attached a RFID tag reader/writer. At the beginning English language class, students looked for objects in the room according to the questions read aloud by the PDA. A label that has written the name of the object was stuck on the corresponding object with a RFID tag. After the experiment, students found the system interested and would like to keep using it. 3 The Architecture of AcoustaNomad As shown in Figure 1, the architecture of AcoustaNomad is divided into four components: m- Learning, m-Blogging, locating engine and mobile IPv6 technologies. More specifically, the components of m-Learning and m-Blogging are applications that must be based on the components of locating engine and mobile IPv6 technologies. When a MH changes its location, it uses the locating engine to detect and decide what kinds of services should be provided in the specific location. At the same time, the locating engine employs mobile IPv6 technology to maintain the existing connection. Afterward, the MH uses the services of m-Learning and m-Blogging via Sockets API. The functions provided by these components are described in following subsections. 3.1 Mobile IPv6 The main function of mobile IPv6 is to maintain an existing connection. That is, when an MH changes its location, it might also change its IPv6 address. The key benefit of mobile IPv6 is that even though the MH changes 2.4 Mobile Learning Mobile technologies enable learners to interact simultaneously with both the physical world and the digital information. The Savannah study [17] explored the use of mobile devices to enable an interactive learning experience. Each student wore headphones for auditory experience and carried a PDA which could be tracked using GPS. The learners not only gained the knowledge of learning through e AcoustaNomad a Figure 1 Th rchitecture 274 Journal of Internet Technology Volume 8 (2007) No.3 locations and addresses, the existing connections through which the MH is communicating are always maintained. In AcoustaNomad, the KAME [19] mobile IPv6 stack had been used since it works well and is regarded reference code. Specifically, an IP layer handover is triggered when the default router is no longer bi-directionally reachable. In this case, the MH must discover a new default router. To discover the new default router, the MH will send out “router solicitation”. After receiving the reply with the new router’s prefix, the MH will then obtain a care of address (CoA) with the received prefix and its own MAC address. 3.2 Locating Engine A locating technique over wireless LANs can adopt the network-based or client-based deployment. For a network-based algorithm, a group of access points (APs) collect the RF signal measurements from a MH and send them to the central server for locating estimation. In AcoustaNomad, we propose a client-based locating engine as shown in Figure 2. Each MH’s wireless adapter gathers real-time SS from multiple APs to match the fingerprinting of the signal strength database, which was established during the off-line phase. In the proposed locating algorithm, a MH’s position is obtained by using the following Euclidean Distance method: ∑(SS − SS ) n Euclidean Distance = n i i i 2 3.3 m-Learning The purpose of m-learning feature is to provide location-specific content to help the language learners to engage in the real world activities. Based on the characteristics of specific locations, the learning materials were designed to meet the needs of learners and were evaluated by an expert language teacher. Each dialogue was displayed in three formats; English, Chinese character and pin-yin, accompanied with correct pronunciations. When a MH enters the zone of a specific location, the selection menu will be popped up on the screen of the MH as shown in Figures 3 and 4. By clicking on the selected categories, specific dialogues will be presented and the MH can click on the play icon to listen to the pronunciations. (1) =1 where SS , SS , n and i denote SS stored in the database during the off-line phase, the real-time accessed SS by the MH, the amount of AP and the AP , respectively. Figure 3 Selection Menu for m-Learning Figure 2 The framework of location engine Figure 4 Learning Content n i i i AcoustaNomad: A Context-aware Blogging and Learning Platform in Mobile IPv6 Networks 275 3.4 m-Blogging afterward. Most of users agreed that m-blogging is more attractive than the traditional story and drew them to listen to a story related to a specific location. They were also satisfied with the system and the user interface. Ten students from different countries were used as the subjects to evaluate the m-learning feature. They are 4 males and 6 females with average 2 years Chinese language learning experience. They were told and encouraged to go through all the learning materials and tried out the system functions. Each student took about 20 minutes for the learning session and took a survey after the experiment. The result showed that most of them were pleased with the system and the interface, found this kind of learning method useful, and would like to keep using it. However, they also expressed that pronunciations provided by the system were too fast for the beginning learners, and hoped the system could offer more learning tools. 4 The System Implementation 4.1 The Network Topology and Experimental Testbed The m-Blogging service aims at supporting locative audio interaction in urban living situation. Any MH can easily keep personal audio blogs in specific locations via our m-blogging authoring system. As shown in Figure 5, a MH can press recording button to start his/her voice recording and upload the audio file immediately right at the location. Our tools also support previous and next selection buttons to facilitate audio selection at the corresponding location. In addition, a playback button is made to support both audio previews before upload and audio listening after download. The interface is designed according to a metaphor of portable mp3 player and recorder. Furthermore, this mp3-like system is context-aware enabled by our locating engine. Different blogs are kept by people in different places. The implementation of the proposed architecture is currently completed. Both MH and server run FreeBSD 5.4-SNAP operation system which includes KAME IPv6 reference implementation. The KAME project was a joint effort of six companies in Japan to provide a free stack of IPv6, IPsec, and Mobile IPv6 for FreeBSD variants. Therefore, it provides a lot of useful features of IPv6 such as Address autoconfiguration, any cast addresses, mandatory multicast addresses, mobile IPv6, and IPv6-toIPv4 transition mechanisms. Eleven graduate students, 10 male and 1 female, participated in m-blogging experiment and took a questionnaire Figure 5 M-Blogging User Interface The experimental testbed of AcoustaNomad is located on the 6th floor of our department building. The layout of the 6th floor is shown in Figure 7. This floor was divided into eight areas to represent eight Taipei Metro Rapid Transit stations. As shown in Figure 6, we place four 11 Mbps IEEE 802.11b compliant APs in the 6th floor. Each AP is connected to a CISCO 2811 router which is in turn connected to other CISCO 2811 routers in order to form a network topology. The operation system, called IOS in CISCO router 2811, is both mobile IPv6 and IPv6 capable router. Since there are eight networks in the network topology, each network must have a different IPv6 address. A dynamic routing protocol, such as RIPv6, is used between CISCO 2811 routers in order to route packets to different networks. Figure 6 The proposed network topology 276 Journal of Internet Technology Volume 8 (2007) No.3 4.2 The Implementation Results and Discussion As shown in Figure 7, when a MH stays in a simulated station, the location engine will estimate the MH’s location and trigger the interactive menu for the specific location. The yellow and red icon indicates m-Blogging and m-Learning services, respectively. When the MH continues to roam to next simulated station, there is a short period of time in which the MH is unreachable because both the layer 2 and layer 3 handovers occurred. The short period time can vary between 5 and 10 ms in our experimental testbed. At the same time, the location engine will estimate the MH’s new location, use the mobile IPv6 to maintain the existing connection and trigger the new interactive menu for the new location. The existing connection will be broken temporarily due to the layer 2 handovers, the provided m-Blogging service will be delayed only for a short period of time. Hence, the MH will see simply the yellow icon appeared on the MH and can click on the icon to activate the menu shown in Figure 5. Figure 8 indicates that the location only has m-Learning service. Therefore, the MH can click on the red icon to activate the menu as shown in Figure 3 for a specific location. Figure 9 shows another analysis result, the cumulative distribution function (CDF) of the error distance for the locating scheme, of our implementation. It is obvious that increasing the total number of APs from 1 to 2 or more can provide a significant locating accuracy. However, the performance improvement is saturated when the total number of APs is increased from 3 to 4. It is worth noting from the experimental results that more than 80 percent of the estimated locations have error distance less than 5 meters. As a result, many services can be predicted in advance accurately in a specific location when a MH roams to the specific location. 5 Conclusions We have introduced AcoustaNomad, an implementation work for mobile learning and audio blogging applications in this paper. AcoustaNomad is based on locating engine which uses empirical signal strength measurements as well as mobile IPv6 which maintain the exiting transport-layer connections. In addition, we also present the architecture of AcoustaNomad and the experimental testbed. The analysis results indicate that more than 80 percent of the estimated locations have error distance less than 5 meters. This means that the empirical method is superior and more accurate. As a result, any MH can detect the services easier when it moves to a new location. The results also revealed that subjects of the experiments had great interests in our system; they not only like the friendly user interfaces, but also satisfied with the services that the system could provide. It proves that emerging contextaware technology with the location-specific content is able to offer better opportunities for cultural explorations and social interactions. It also draws people from isolated environments together to our magic AcoustaNomad world. We believe that is our most significant contribution. Figure 7 A simulated station with m-blogging and m-learning service Figure 8 A simulated station with m-learning service Figure 9 The CDF of the error distance for the locating scheme AcoustaNomad: A Context-aware Blogging and Learning Platform in Mobile IPv6 Networks 277 Reference [1] D. Johnson, C. Perkins, J. Arkko, “IP mobility Support for IPv6,” IETF RFC 3775, June 2004 [2] W. Carter, S. L. Liu, “Location33: A Mobile Musical,” In Proceedings of the International Conference on New Interfaces for Musical Expression, Vancouver, BC, Canada, May 26-28, 2005. [3] Adrian Friday, Maomao Wu, Joe Finney, Stefan Schmid and Keith Cheverst, “Network layer Access control for Context-Aware IPv6 applications,” Wireless Networks 9, 299-309, 2003. [4] J. Broadbent and P. Marti, “CyberGuide: A mobile context-aware tour guide,” Wireless network 3, 421433, 1997. [5] Guolin Sun, Jie Chen, Wei Guo, and K.J. Ray Liu, ”Signal Processing Techniques in Network-Aided Positioning,” IEEE Signal Processing Magazine, July 2005, pp.12-23. [6] Chin-Liang Wang, Yih-Shyh Chiou, and ShengCheng Yeh, “A Location Algorithm Based on Radio Propagation Modeling for Indoor Wireless Local Area Networks,” IEEE Vehicular Technology Conference Spring 2005, Stockholm, Sweden, May 2005. [7] P. Bahl and V. N. Padmanabhan, “RADAR: An In-Building RF-BASED User Location and Tracking System,” IEEE INFOCOM 2000, May 2000, pp. 775-784. [8] S. Fisher, “Environment Media: Linking Virtual Environments to the Real World,” In Creative Digital Media: Its Impact on the New Century, Keio University COE international Symposium, Keio University Press, Tokyo, Japan, Dec., 2001. [9] H. Westerkamp, “Soundscape Composition: Linking Inner and Outer Worlds,” Written for Soundscape before 2000 Conference, Amsterdam, Holland, Nov. 19-26, 1999. [10] S. I. Salamensky, “Talk, Talk, Talk: The Cultural Life of Everyday Conversation,” article by H. Bhabha, and S. L. Gilman, “Just Talking: Tete-A-Tete,” Routledge, London and New York, 2001, pp. 3-14. [11] Oversampling, INC. Soundwalk NYC., Jan. 2005. Retrieved from http://www.soundwalk.com [12] G. Lane, “Urban Tapestries: Wireless Networking, Public Authoring and Social Knowledge,” In Forth Wireless World Conference, Jul. 17-18, 2003. [13] W. Carter, S. Fisher, T. Furmanski, K. Macdonald, T. Millican, “Patholog: Creating Location-based Web Logs,” In Proceedings of the Tenth International Conference on Virtual Systems and Multimedia, Ogaki City, Japan, Nov. 17-19, 2004. [14] J. D. Lasica, “When Bloggers Commit Journalism,” Sep., 2002. Retrieved from http://www.ojr.org/ojr/ lasica/1032910520.php. [15] J. Shotter, “Conversational Realities: Constructing Life Through Language,” Sage, London, 1993. [16] M. Crang, “Thinking Space,” Critical Geographies Series, article by Holloway, Julian and Kneale, J. M. Bakhtin, “Dialogics of Space,” Routledge, London and New York, 2002. [17] H. Ogata, Y. Yano,”Context-aware support for computersupported ubiquitous learning,” Proceedings of the The 2nd IEEE International Workshop on Wireless and Mobile Technologies in Education (WMTE’04),2004. [18] K. Facerw, R. Joiner, D. Stanton, J. Reidz, R. Hullz, D. Kirk, ”Savannah: mobile gaming and learning? ” Journal of Computer Assisted Learning 20, 2004, pp. 399-409. [19] The overview of KAME project, http://www.kame. net/project-overview.html Biographies is an associate professor of Dept. of Computer and Communication Engineering at MingChuan University. Her research interests include developing of e-learning content and environment, and implementing mobile devices in learning. She obtained the doctoral degree in instructional media and technology from University of Connecticut. Ching-Hui Chen received the Ph.D. degree in the Department of computer Science and Information Engineering from Tamkang University, Taiwan in 1999. He is currently an assistant professor in department of computer science and information engineering at Ming Chuan University, Taiwan. His research interests are in QoS unicast/multicast routing, communication protocols and wireless networks. Wu-Hsiao Hsu is an assistant professor of Dept. of Computer and Communication Engineering at Ming Chuan University. His research interests include mixed reality, computer graphics, and human computer interface. He received the B.S. and Ph. D. degrees in computer science from National Taiwan University. Rung-Huei Liang 278 Journal of Internet Technology Volume 8 (2007) No.3 received the B.S.E.E. degree from National Taiwan University of Science and Technology in 1991, the M.S.E.E. degree from National Central University, Taiwan, R.O.C. in 1993, and the Ph.D. in electrical engineering from National Central University, Taiwan in 2000. Since 2003, Dr Yeh is an associate professor of the Department of Computer and Communication Engineering at Ming Chuan University, Taiwan. His research interests include computer networks and wireless communications. This author became a Member of IEEE in 2000. Sheng-Cheng Yeh