雲端化網路視訊電話設計與實作

全文

(1)國立高雄大學資訊工程學系碩士論文. 雲端化網路視訊電話設計與實作 Design and Implementation of Cloud Computing for Video/Voice over IP. 研究生：林子堯指導教授：張保榮教授. 中華民國一百零一年六月.

(2) 雲端化網路視訊電話設計與實作指導教授：張保榮教授國立高雄大學資訊工程學系. 研究生：林子堯國立高雄大學資訊工程學系碩士班. 1 摘要. 本研究的目的是實現在雲端計算系統上設計一個即時性網路視訊電話 Video/Voice over IP (VVoIP)的應用程式。PC 或 Notebook 經由 web 介面方式且使用以 Authentication、Authorization、Accounting (AAA)管理機制登入雲端計算系統，透過雲端計算系統的 VVoIP 程式與另一位上線使用者做網路視訊電話方式的連線通話。Android 行動裝置的使用者則經由 Android Market 下載及安裝我們發佈雲端的 VVoIP App 後，在 Adobe AIR App 的環境下進行雲端網路視訊通話。它比 P2P 方式網路視訊電話更具有幾項優點，首先使用者通話前不需事先得知對方 IP 位址，其次它將可以實現跨平台的網路視訊電話，並且 VVoIP 應用程式採用 Real Time Messaging Protocol(RTMP)通訊協定來達成及時性的傳輸服務。. 關鍵字：雲端計算、網路視訊電話、Real Time Messaging Protocol (RTMP)。. i.

(3) Design and Implementation of Cloud Computing for Video/Voice over IP Advisor: Dr. CHANG, Bao-Rong Department of Computer Science and Information Engineering National University of Kaohsiung. Student: LIN, Zih-Yao Department of Computer Science and Information Engineering National University of Kaohsiung. 2 ABSTRACT. The objective of this study is to realize a real-time Video/Voice over IP (VVoIP) application implemented by a cloud computing server. PC and notebook users can access the remote server through the web, and log into the cloud computing system via the AAA (Authentication, Authorization, Accounting) management system. This way, any user can carry out an on-line call with another user through VVoIP application running on the cloud computing server. Mobile Android users can download the cloud VVoIP App from Market/Play Store, and conduct the VVoIP call under the Adobe AIR App environment. This technique outperforms P2P method for many reasons. First, the user need not know the other person’s IP before making the VVoIP call. Also, it can achieve cross-platform VVoIP, and the application uses Real Time Messaging Protocol (RTMP) to achieve immediate transmission of services.. Keywords: Cloud Computing 、 Video/Voice over IP(VVoIP) 、 Real Time Messaging Protocol (RTMP). ii.

(4) 3 誌謝本論文能夠順利的完成，首先最要感謝的是我的指導教授張保榮老師，在撰寫碩士論文的這段期間，張老師不厭其煩地給予我論文寫作上的指導，也很有耐心地包容我的缺失。在這兩年的研究所生活中，我學的到不傴傴是學術上的知識，也讓我學會了許多待人處事之道。感謝張保榮老師在這兩年中，一直很有耐心且細心的給予教導，從老師對我與實驗室的關心和付出上，讓我覺得實驗室的成員就如同老師的小孩一樣，除了研究與課業上的關心之外，老師也非常關心我們的生活狀況，都會適時的給予支持與鼓勵，如今我既將畢業了，心中充滿著無限的感激。感謝張保榮老師，不傴讓我學到做研究的態度，也讓我學會許多研究之外的事。接下來要感謝的是與我共度兩年研究生涯的豊鈞同學，我們一起討論課業、分享知識，彼此討論程式難題、一起為畢業論文奮戰、一起吃飯、玩樂，有好多好多難忘且快樂的時光，你是我在研究上共同奮鬥的重要夥伴，也是我在生活中共患難的朋友。感謝熙忠學長，在我碩一的時候，熙忠都相當慷慨地與我分享你的寶貴經驗，有你的帶領，讓我很快的就能夠適應研究生的生活。感謝啟銘，在我碩二的時候，實驗室因為有你的支援，我才能夠專心的致力於研究上。感謝陪我度過研究生活的各位同學，這一路上有你們的陪伴，讓我的研究生活增添了許多色彩。最後，感謝我最敬愛的家人。在我的求學過程中，不管是在精神上還是實質上，有您們的支持與包容，使得我能夠無後顧之憂的專心於學習上，謝謝您們 24 年來不辭辛勞地栽培，希望我能成為您們的驕傲。. iii.

(5) 4 Directory 1 摘要............................................................................................................................. i 2 ABSTRACT ................................................................................................................ii 3 誌謝...........................................................................................................................iii 4 Directory .................................................................................................................... iv 5 List of Figures ............................................................................................................ vi 6 List of Tables ............................................................................................................viii Chapter 1. Introduction .................................................................................................. 1 Chapter 2. Literature Review ......................................................................................... 6 2.1 Windows Live Messenger ................................................................................ 8 2.2 Yahoo! Messenger .......................................................................................... 10 2.3 Skype.............................................................................................................. 12 2.4 LINE .............................................................................................................. 14 2.5 Viber ............................................................................................................... 16 2.6 Tango Video Calls .......................................................................................... 17 2.7 Function comparison for different products................................................... 18 Chapter 3. Research Method ........................................................................................ 20 3.1 Deploy the cloud host－Proxmox Virtual Environment ................................ 20 3.2 Build the cloud VVoIP environment .............................................................. 21 3.2.1 Create virtual machine and install Ubuntu 10.04 OS ......................... 21 3.2.2 Establish application servers in virtual machine ................................. 23 3.2.3 Video/voice server－Adobe Flash Media Server................................ 25 3.3 Communication architecture in server as well as client ................................. 26 3.4 Phone operation procedure within cloud VVoIP ............................................ 28 3.5 Set-up of development environment and deploy VVoIP application ............. 31 Chapter 4. Experimental Results and Discussion ........................................................ 32 4.1 Experimental environment ............................................................................. 32 4.2 Experimental results....................................................................................... 34 4.2.1 Phone test between PCs ...................................................................... 35 4.2.2 Phone test between PC and mobile device ......................................... 37 4.2.3 Phone test between mobile devices..................................................... 39 4.2.4 Off-site quality test.............................................................................. 41 4.2.5 Simultaneous connection stress test .................................................... 43 4.2.6 Power consumption estimate at Android mobile phone ..................... 46 4.2.7 Power consumption estimate at Personal Computer ........................... 49 iv.

(6) 4.2.8 System performance analysis .............................................................. 52 Chapter 5. Conclusion .................................................................................................. 59 References .................................................................................................................... 61. v.

(7) 5 List of Figures Figure 1. Windows Live Messenger’s Login interface on PC platform (in Chinese) ............................................................................................................. 8 Figure 2. Windows Live Messenger’s Web Interface Access (in Chinese) ............... 9 Figure 3. Windows Live Messenger’s Login interface on Smart Phone platform (in Chinese) ........................................................................................................ 9 Figure 4. Yahoo! Messenger’s Login interface on PC platform (in Chinese).......... 11 Figure 5. Yahoo! Email combined with Yahoo! Messenger Web interface version (in Chinese) ...................................................................................................... 11 Figure 6. Yahoo! Messenger’s Login interface on Smart Phone platform (in Chinese) ........................................................................................................... 12 Figure 7. Skype’s Login interface on PC platform (in Chinese).............................. 13 Figure 8. Skype’s Login interface on Smart Phone platform (in Chinese) .............. 14 Figure 9. LINE’s Login interface on Smart Phone platform (in Chinese) ............... 15 Figure 10. LINE’s Login interface on PC platform (in Chinese) ............................ 15 Figure 11. Web interface of the LINE for Chrome browser only (in Chinese) ........ 16 Figure 12. Viber’s login interface on smart phone platform .................................... 16 Figure 13. Tango’s Login interface on Smart Phone platform (in Chinese) ............ 17 Figure 14. Tango’s Login interface on PC platform (in Chinese) ............................ 17 Figure 15. Proxmox Virtual Environment cloud system cluster architecture .......... 21 Figure 16. Install Ubuntu 10.04 OS in the virtual machine ..................................... 22 Figure 17. Operating system Ubuntu10.04 in virtual machine ................................ 23 Figure 18. Connection between client and cloud computer ..................................... 24 Figure 19. Diagram for communication architecture in server ................................ 26 Figure 20. Diagram for communication architecture in server as well as client ..... 27 Figure 21. Mobile, tablet, notebook, PDA, and PC connected to cloud computing 27 Figure 22. Flowchart of cloud VVoIP initialization as well as telephone idle ......... 29 Figure 23. Flowchart of cloud VVoIP telephone idle in detail................................. 30 Figure 24. Web login screen of VVoIP on Chrome browser on the PC ................... 34 Figure 25. VVoIP login screen in Android smart phone .......................................... 35 Figure 26. PC browser screen: no calls diagram ...................................................... 36 Figure 27. Video call display in PC for user01 ........................................................ 36 Figure 28. Video call display in PC for user02 ........................................................ 37 Figure 29. Video call display in PC for user01 ........................................................ 38 Figure 30. Video call display in Android device #1 for user03 ............................... 38 vi.

(8) Figure 31. Video call display in Android device #1 for user03 ............................... 39 Figure 32. Video call display in Android device #2 for user02 ............................... 40 Figure 33. A snapshot of video call display in Android device #1 .......................... 40 Figure 34. A snapshot of video call display in Android device #2 .......................... 41 Figure 35. Phone test using off-site Wi-Fi ............................................................... 42 Figure 36. Phone test using on-site Wi-Fi ................................................................ 42 Figure 37. The first group of the multi-user map ..................................................... 43 Figure 38. The second group of the multi-user map ................................................ 43 Figure 39. The third group of the multi-user map.................................................... 44 Figure 40. The fourth group of the multi-user map ................................................. 44 Figure 41. The fifth group of the multi-user map .................................................... 45 Figure 42. In the beginning of VVoIP test operation ............................................... 47 Figure 43. Almost at the end of VVoIP test operation ............................................. 48 Figure 44. CPU (i5-750) utilization rate of 100% of current measurement ............ 49 Figure 45. CPU (i5-750) utilization rate for the idle phone running in a PC .......... 50 Figure 46. Core voltage for the idle phone running in a PC .................................... 50 Figure 47. Login interface of the flash media server management tool .................. 52 Figure 48. Control interface of the flash media server management tool ................ 53 Figure 49. Active connections － control interface ............................................... 53 Figure 50. Bandwidth － control interface ............................................................ 54 Figure 51. CPU and Memory Usage － control interface ..................................... 54 Figure 52. Line chart for CPU, memory, and bandwidth in PC-to-PC .................... 57 Figure 53. Line chart for CPU, memory, and bandwidth in PC-to-Phone ............... 57 Figure 54. Line chart for CPU, memory, and bandwidth in Phone-to-Phone .......... 58 Figure 55. Composite index line chart for PC-to-PC, PC-to-Phone and Phone-to-Phone ................................................................................................ 58. vii.

(9) 6 List of Tables Table 1. Program features comparison table ............................................................ 18 Table 2. Communication structure of the program .................................................. 19 Table 3. Source code ................................................................................................ 19 Table 4. Virtual machine environment ..................................................................... 21 Table 5. Software packages installed ....................................................................... 24 Table 6. Development environment and package .................................................... 31 Table 7. Node environment ...................................................................................... 32 Table 8. Virtual Machine’s hardware configuration ................................................. 32 Table 9. The test environment .................................................................................. 33 Table 10. Sony Ericsson XPERIA Ray specifications ............................................. 33 Table 11. Mobile phone power consumption estimate............................................. 46 Table 12. Measured data in average for CPU .......................................................... 51 Table 13. FMS Data ................................................................................................. 55 Table 13. Wi-Fi specification ................................................................................... 60. viii.

(10) Chapter 1. Introduction Present cloud-computing and services are dominated by large-size public cloud and enterprise-owned private cloud. As a matter of fact, the small and medium enterprises (SMEs), schools, and social groups need private small-cloud computing and services (PSCCS) [8]. With this service, the department’s information costs can be drastically reduced as well as it can quickly increase the competitiveness of its information system due to the following reasons: centralized monitoring, quick management, dynamic optimization, and efficient backup. Technically, unexpected situations with service-type servers, such as websites, databases, AP servers, and file servers, bring much trouble for enterprises. Once a service comes to a halt, it can cause faulty data, stalled production lines, and interrupted operation procedures, leading to multiple losses. However, a physical host has problems of promptness of service transfer to another host, restarting the service, and inability to update data in real time. The hardware, information, and data will be obstacles for enterprises to overcome. To solve above stated issues, cloud virtual server systems (Hypervisor), such as VMware ESX/ESXi Server [30], Microsoft Hyper-V R2 [17], and Proxmox Virtual Environment [25], have integrated virtualization, virtual machines, and virtual services. With this method, users need only to connect to the system using low cost thin clients (a low-end PC or PDA) to complete general tasks [4], reducing IT purchasing fees considerably. As thin clients are easier to setup, chances of malfunctions, heat crashes, and computer viruses will also decrease, thus saving money on costs and electricity. It is worth mentioning that, in the previous research on this case [6] [7] [8] [9] 1.

(11) [10] [11] [13], the embedded device was completed. Through high-speed Wi-Fi or Ethernet connection, it is now able to connect to the cloud virtual server system. It uses the standard J2ME [21] environment, implemented on the JamVM [20] virtual machine, and adopts the GNU Classpath [16] as the Java Class libraries. There already exist service systems of ―public clouds‖ such as Google App Engine [15], Amazon Web Services [5], and Microsoft Azure [31], ―private clouds‖ like IBM Blue Cloud [18], ―open clouds‖ like Open Nebula [23], Eucalyptus [14], Apache Hadoop [32], and NCDM Sector/Sphere [26], where IBM Blue Cloud [18] is also a ―hybrid cloud‖. The government sector is pushing forth the large-scale cloud server system using the large cluster computer networks as the testing platform. There are also plans to set up regional mid-size cloud servers in a couple of large-scale universities in order to provide cloud services to various regions. It is worth noting that there have not been plans to develop small-cloud services (private small-cloud server systems) for SMEs, schools, and social groups. Therefore, it is a great opportunity is to invest and develop in small-cloud services in order to bridge with the nation’s large, medium and small cloud services network, especially suitable for a number of meddle size universities. Developing suitable private small-cloud server systems can provide a niche for SMEs, schools, and social groups, and act as a cloud service option. With thin clients and embedded platforms, private small-cloud computing servers are able to give real time feedbacks. This short-range cloud controller service reacts quickly to user requests and inquiries, whereas remote cloud services on internet cannot guarantee an immediate response. To resolve the issue of limited bandwidth, if the task can be completed by short-range private small-cloud computing system, then they will not request services from remote cloud server host. This reduces the usage rate of large bandwidths, as well as saves on the costs of data transportation. While 2.

(12) internet should be down for a period of time, private small-cloud computing system will still work through LAN, and will maintain operation of short-range usage. Those having security concerns about placing confidential information (like enrollment status, medical records, and trade secrets) on remote cloud will instead turn to private small-cloud site for executions and storage. This explains the advantages private small-cloud has over remote cloud. In the recent years, the wireless mobile technology has become the most exciting domain in telecommunications and networking. Wireless mobile communications enable us to implement mobile applications associated with real-time video/voice over IP (VVoIP) on various portable computing devices, ranging from PDA and tablet PC to smart phone. For example, vehicle-to-vehicle (V2V) and smart phone communications is currently one of the most popular topics related to mobile wireless communication. Integrating wireless networks into existing protocols pose a new challenge. The main issue is that the TCP/IP-based Internet technologies were designed for wired networks with mostly fixed hosts. Audio streaming adopts Real-Time Transport Protocol (RTP), it relies on UDP, and has no loss recovery, which is unreliable. Instead of UDP, TCP leads to a longer delay time and may possibly encounter the obsolete data received at the peer side. Therefore, mobile users communicating via VVoIP between the nearest base stations have to consider the aspect of wireless paths based on ad hoc network, or the cloud computing client/server scheme. In the recent years, more and more VoIP [28] [29] has been applied to cloud computing system [1]. In order to deal with three crucial problems: (a) head-of-line blocking, (b) handoff interruption, and (c) non-real-time transmission in TCP-based protocols, previous work had adopted PR-SCTP protocol instead of TCP to implement VVoIP application on x86 machines called SCTP-IHU[7]. For the purpose of mobile VVoIP 3.

(13) via P2P connection (called P2P-VVoIP [19] [22]), a mobile version of SCTP-IHU was made by transplanting VVoIP application into ARM-based embedded platforms and smart phones. However, the computation load for real-time video phone call at the thin client side was too large and thus ARM-based embedded platforms cannot maintain the goal of power-saving. Moreover, there are some shortcomings for P2P-VVoIP as described below. The peer side must install VVoIP application on the mobile devices, and both users need to know their respective IP addresses before initiating P2P connection via wired or wireless network. In order to resolve the problems mentioned above, this study intends to establish a real-time VVoIP application running on a cloud computing system (denoted Cloud-VVoIP). For PC users, clients use browsers, where the Web interface is accessible to video/voice streaming through cloud computing server. For Android users, thin clients are thought of as mobile devices, where Android Apps interface is applicable to access video/voice streaming through cloud computing server. At the same time, access control has been designed to realize authentication, authorization, and accounting (AAA [12]) security in cloud computing system. People can enjoy the convenience and high-performance of Internet phone calls. First, the caller does not need the destination IP in order to make calls. As well, based on client-server architecture, users can make VVoIP calls simply by using a low-cost thin client in through the VVoIP application. Therefore we adopt TCP-based RTMP related protocols instead of PR-SCTP. This scheme greatly reduces the computation load and achieves power efficiency at the thin client side. Technically, cloud computing with TCP-based protocol has better performance and higher efficiency than PR-SCTP. Even though there are multiple benefits in cloud computing, which can help many companies break the physical ties between the IT infrastructure and its users, security assurance over IaaS, PaaS, and/or SaaS is still a challenging issue to the 4.

(14) cloud computing system. Regarding intruder detection, the use of access control systems to prevent illegal intrusions outside the cloud is taken into account in this study. From the energy point of view, cloud computing owns four characteristics: large amounts of data, low cost, efficiency, and reliability. This chapter describes the research background, research motivation and purpose. The second chapter is the technical discussion of relevant literature. Chapter 3 explains in detail the method proposed in this paper. Chapter 4 provides experiment snapshots and results. Chapter 5 summarizes and explores the defects of this method, and how future research can be improved and extended for further development.. 5.

(15) Chapter 2. Literature Review Voice over IP (VoIP, or Voice over Internet Protocol) commonly refers to the communication protocols, technologies, methodologies, and transmission techniques involved in the delivery of voice communications and multimedia sessions over Internet Protocol (IP) networks, such as the Internet. Other terms commonly associated with VoIP are IP telephony, Internet telephony, voice over broadband (VoBB), broadband telephony, and broadband phone. Internet telephony refers to communications services —voice, fax, SMS, and voice-messaging applications— that are transported via the Internet, rather than the public switched telephone network (PSTN). The steps involved in originating a VoIP telephone call are signaling and media channel setup, digitization of the analog voice signal, encoding, encapsulation, and transmission as Internet Protocol (IP) packets over a packet-switched network. On the receiving side, similar steps (usually in the reverse order) such as reception of the IP packets, decoding, and digital-to-analog conversion are utilized to reproduce the original voice stream. Even though IP Telephony and VoIP are terms that are used interchangeably, they are actually different; IP telephony is a general term describing digital telephony systems that use IP protocols for voice communication, while VoIP is actually a subset of IP Telephony. VoIP is a technology used by IP telephony as a means of transporting phone calls. Early providers of Voice over IP services offered business models (and technical solutions) that mirrored the architecture of the legacy telephone network. Second generation providers, such as Skype have built closed networks for private user bases, offering the benefit of free calls and convenience, while denying their users the ability to call out to other networks. This has severely limited the ability of users to 6.

(16) mix-and-match third-party hardware and software. Third generation providers, such as Google Talk have adopted the concept of Federated VoIP - which is a complete departure from the architecture of the legacy networks. These solutions typically allow arbitrary and dynamic interconnection between any two domains on the Internet whenever a user wishes to place a call. VoIP systems employ session control protocols to control the set-up and tear-down of calls, as well as audio codec which encodes speech, allowing transmission of digital audio via audio stream over an IP network. The choice of codec varies between different implementations of VoIP, depending on application requirements and network bandwidth; some implementations rely on narrowband and compressed speech, while others support high fidelity stereo codecs. Some popular codecs include u-law and a-law versions of G.711, G.722, which are high-fidelity codecs marketed as HD Voice by Polycom, a popular open source voice codec known as iLBC, a codec that only uses 8 Kbit/s each way called G.729, and many others. VoIP is available on many smartphones and Internet devices so that users of portable devices that are not phones, may place calls or send SMS messages over 3G or Wi-Fi. Through the years, many applications have been developed and are widely used for PCs and smart phones. This chapter will introduce several common VoIP and VVoIP applications.. 7.

(17) 2.1 Windows Live Messenger Windows Live Messenger (formerly named MSN Messenger) is an instant messaging client created by Microsoft that is currently designed to work with Windows XP (up to Wave 3), Windows Vista, Windows 7, Windows Mobile, Windows Phone, Windows CE, Xbox 360, Blackberry OS, iOS, Java ME, S60 on Symbian OS 9.x, and Zune HD. The client has been part of Microsoft's Windows Live set of online services since 2005. It connects to Microsoft's .NET Messenger Service. The client was first released as MSN Messenger on July 22, 1999, and as Windows Live Messenger on December 13, 2005. In June 2009, Microsoft reported the service attracted over 330 million active users each month. Windows Live Messenger is one of the most widely used instant messaging software in the world, in addition to basic text-message, the current version also supports video conferencing, voice chat, multiplayer sessions, and game extensions, among many others.. Figure 1. Windows Live Messenger’s Login interface on PC platform (in Chinese). 8.

(18) Figure 2. Windows Live Messenger’s Web Interface Access (in Chinese). Figure 3. Windows Live Messenger’s Login interface on Smart Phone platform (in Chinese). 9.

(19) 2.2 Yahoo! Messenger Yahoo!. Messenger. (sometimes. abbreviated. YIM. or. YM). is. an. advertisement-supported instant messaging client and associated protocol provided by Yahoo!. Yahoo! Messenger is provided free of charge and can be downloaded and used with a generic "Yahoo! ID" which also allows access to other Yahoo! services, such as Yahoo! Mail, where users can be automatically notified when they receive new email. Yahoo! also offers PC-PC, PC-Phone and Phone-to-PC service, file transfers, webcam hosting, text messaging service, and chat rooms in various categories. Yahoo! Messenger was originally launched under the name Yahoo! Pager on March 9, 1998. In addition to instant messaging features similar to those offered by ICQ, it also offers (on Microsoft Windows) features such as: IMVironments (customizing the look of Instant Message windows, some of which include authorized themes of famous cartoons such as Garfield or Dilbert), address-book integration and Custom Status Messages. It was also the first major IM client to feature BUZZing and music-status. Another recently added feature is customized avatars. Yahoo! Messenger also allows users to access Yahoo! Chat Rooms. Another Smart Phone Messenger Apps plus video voice plug Apps can also call the Video Phone.. 10.

(20) Figure 4. Yahoo! Messenger’s Login interface on PC platform (in Chinese). Figure 5. Yahoo! Email combined with Yahoo! Messenger Web interface version (in Chinese). 11.

(21) Figure 6. Yahoo! Messenger’s Login interface on Smart Phone platform (in Chinese). 2.3 Skype Skype is a proprietary Voice over Internet Protocol service and software application originally created by Niklas Zennstrom and Janus Friis in 2003, and owned by Microsoft since 2011. Skype’s VoIP[29] has been used in cloud computing system in 2010. The service allows users to communicate with peers by voice, video, and instant messaging over the Internet. Phone calls to recipients may be placed on the traditional telephone networks. Calls to other users within the Skype service are free of charge, while calls to landline telephones and mobile phones are charged via a debit-based user account system. Skype has also become popular for its additional features, including file transfer, and video conferencing. Its competitors include SIP and H.323-based services, such as Empathy, Linphone, Ekiga as well as the Google Talk service.. 12.

(22) Skype has 663 million registered users as of September 2011. The network is operated by Microsoft, which has its Skype division headquarters in Luxembourg. Most of the development team and 44% of the employees of the division are situated in Tallinn and Tartu, Estonia. Unlike most VoIP services, Skype is a hybrid peer-to-peer and client–server system. It makes use of background processing on computers running Skype software. Skype's original proposed name (Sky Peer-to-Peer) reflects this fact. Some network administrators have banned Skype on corporate, government, home, and education networks, citing reasons such as inappropriate usage of resources, excessive bandwidth usage, and security concerns. In addition, it is one of the few Video Phone Apps in the Android Market.. Figure 7. Skype’s Login interface on PC platform (in Chinese). 13.

(23) Figure 8. Skype’s Login interface on Smart Phone platform (in Chinese). 2.4 LINE LINE is a smart phone application released in June, 2011. LINE users can cross-border and cross-telecommunicate with other network operators. Users can use free VoIP, or send text messages through the wireless network or UMTS to other users. The web version for the Tablet PC, and the computer version released in March 2012 is able to send instant graphical messages. However, VoIP still requires the use of smart phones as well; it does not yet support VVoIP. LINE is said to be one of the highest-profile smart phone applications in Taiwan’s market right now, having more than 40 million users worldwide.. 14.

(24) Figure 9. LINE’s Login interface on Smart Phone platform (in Chinese). Figure 10. LINE’s Login interface on PC platform (in Chinese). 15.

(25) Figure 11. Web interface of the LINE for Chrome browser only (in Chinese). 2.5 Viber Viber is of Internet phone software installed on smart phones. Its operating system is iOS, Android or Windows Phone. Free calls can be made and SMSs sent between its users through a wireless network or UMTS. The difference between Viber and other VoIP providers is that it is free, and the user does not need to register or pay, as long as both sides have installed the software.. Figure 12. Viber’s login interface on smart phone platform 16.

(26) 2.6 Tango Video Calls Tango Video Calls is a smart phone based VVoIP. Users can make free phone and video calls through 3G, 4G and Wi-Fi to any smart phone, tablet device, or PC that have also installed Tango. You can make a Tango call with voice control, and whenever you wish to share something, you can easily switch to video calling. It is one of the few VVoIP applications on Android phones.. Figure 13. Tango’s Login interface on Smart Phone platform (in Chinese). Figure 14. Tango’s Login interface on PC platform (in Chinese) 17.

(27) 2.7 Function comparison for different products This section will compare some differences between Cloud-VVoIP and other VVoIP. First the VVoIP version is compared, as well as the number of functions under the various versions listed in Table 1. For example, Windows Live Messenger can use video/voice over IP on the PC version, but not on the web and the Android version. ―Yahoo! Messenger‖, ―Skype‖, and ―Tango Video Calls‖ can use video/voice over IP on the PC and the Android version, but not on the web version, because these applications didn’t develop the web version. The Cloud-VVoIP mentioned in this study can use the video/voice over IP whether it is the PC version, web version or Android version. Table 1. Program features comparison table PC version. Web version. Android version. Windows Live Messenger. VVoIP. ×. ×. Yahoo! Messenger. VVoIP. ×. VVoIP. Skype. VVoIP. ×. VVoIP. Line. ×. ×. VoIP. Viber. ×. ×. VoIP. Tango Video Calls. VVoIP. ×. VVoIP. Cloud-VVoIP. VVoIP. VVoIP. VVoIP. 18.

(28) Table 2 compares the communication architecture. ―Yahoo! Messenger‖, ―Skype‖, and ―Tango Video Calls‖ are P2P mode, but Cloud-VVoIP in this study is client/server mode. Table 2. Communication structure of the program Communication structure Yahoo! Messenger. Peer-to-Peer Mode. Skype. Peer-to-Peer Mode. Tango Video Calls. Peer-to-Peer Mode. Cloud-VVoIP. Client/Server Mode. Table 3 compares the software source code. ―Yahoo Messenger‖, ―Skype‖, and ―Tango Video Calls‖ are business software (close source code). but Cloud-VVoIP of in study is free software (open source code). Table 3. Source code free or business software Yahoo! Messenger. business software(close source code). Skype. business software(close source code). Tango Video Calls. business software(close source code). Cloud-VVoIP. free software(open source code). 19.

(29) Chapter 3. Research Method This research implements Video/Voice over IP and deploys it onto the cloud system. The following briefly describes the deployment of cloud systems, setting up of the Server-side environment, Adobe Flash Media Server Video voice server, Client-side and Server-side communication architecture, cloud VVoIP master phone architecture, building the development environment, and distribution of the VVoIP Program.. 3.1 Deploy the cloud host－Proxmox Virtual Environment Proxmox Virtual Environment has all the packages needed to deploy cloud architecture. Its advantages are: easy installation, promptness, correctness, ease of use, ease of maintenance, ease of upgrade, and free download. In the future, we have expected it will become private cloud’s mainstream cloud controller. The Proxmox Virtual. Environment. download. site. is. provided. here. http://pve.proxmox.com/wiki/Downloads Proxmox Virtual Environment cloud system cluster architecture is shown in Fig. 15.. 20.

(30) Figure 15. Proxmox Virtual Environment cloud system cluster architecture. 3.2 Build the cloud VVoIP environment In Proxmox Virtual Environment web management page, select 1 Node to create Cloud-VVoIP server environment from the cluster as referred to in section 3.2.1.. 3.2.1 Create virtual machine and install Ubuntu 10.04 OS Create a virtual machine; the virtual machine’s environment is listed in Table 4. Table 4. Virtual machine environment Nature. Equipment. Virtual Type. Fully virtualized (KVM). Computer CPU cores. Quad-Core(4-core). Memory size. 2 Gigabyte. Disk space. 32 Gigabyte. Operating System. Ubuntu10.04. 21.

(31) Then install the operating system Ubuntu10.04 in the virtual machine as shown in Figs 16 and 17.. Figure 16. Install Ubuntu 10.04 OS in the virtual machine. 22.

(32) Figure 17. Operating system Ubuntu10.04 in virtual machine. 3.2.2 Establish application servers in virtual machine In the Ubuntu OS under the Virtual Machine environment, install Apache, MySQL, PHP, Flash Media Server, zend-framework and other packages, and then install phpMyAdmin to manage the database. Set proper network settings so that the Internet can connect to the Virtual Machine as shown in Fig. 18. Establishing application servers in virtual machine is related to PHP for Flash FMS [2].. 23.

(33) Figure 18. Connection between client and cloud computer Table 5 lists all software packages needed in Ubuntu 10.04. Table 5. Software packages installed Nature. Packages name. Web Server. apache2. MySQL. mysql-server-5.1. PHP Server. php5. MySQL management tool. phpmyadmin. Video/voice server. Adobe Flash Media Development Server 4.5. Other. zend-framework , 24.

(34) 3.2.3 Video/voice server－Adobe Flash Media Server Flash Media Server (FMS) can be thought of as the distribution center of media, such as video and audio files. When users want to access these video voice data, the video and audio player (SWF file) will use the Real Time Messaging Protocol (RTMP) to call the server, find the data stream ID, then load and play the video voice data. The benefit of this process is it will start playing immediately when we press the play button. When data is passed to the active Adobe Flash Player in the browser, video or audio data will immediately show up. No need to wait for the contents to load. At the same time, no streaming data will be downloaded to the browser’s cache, which will ensure the security of the Video/Voice over IP, and prevent bootlegging. RTMP is a dedicated and specific protocol, using the Transmission Control Protocol (TCP) to transfer data packets between Flash Player and Flash Media Server (FMS). For a variety of transmission devices, the real advantage of this process is that RTMP is designed to pass video (FLV, MP4, MOV) and audio (MP3, AAC, and Nellymoser) to SWF-embedded web pages or mobile phones, or even desktop Adobe AIR applications. The RTMP operation is as follows: media content is delivered to the SWF file and is not stored on the Web server, but rather the server folder where Flash Media Server is running. Install the development kit of Flash Media Development Server. In fact, there is not much difference between it and Flash Media Interactive Server. The only difference is the number of simultaneous connections (Flash Media Development Server can only allow 10 Client connections to the FMS). Adobe Flash Media Development Server 4.5 download website: https://www.adobe.com/cfusion/tdrc/index.cfm?loc=zh_tw&product=flashmediaserve r. 25.

(35) 3.3 Communication architecture in server as well as client After Ubuntu10.04 has installed apache2, PHP, MySQL, and adobe flash media server, its internal management architecture will look like Fig. 19.. Figure 19. Diagram for communication architecture in server. 26.

(36) The Client opens the Server’s VVoIP website through a browser, then the Client will use the Cloud-VVoIP Services in web interface; its architecture is illustrated in Fig. 20.. Figure 20. Diagram for communication architecture in server as well as client The PC, NB, and a variety of mobile devices connected through wired or wireless network to our cloud computing, its architecture is shown in Fig. 21.. Figure 21. Mobile, tablet, notebook, PDA, and PC connected to cloud computing 27.

(37) 3.4 Phone operation procedure within cloud VVoIP This section will describe the VVoIP’s program process, as well as the phone's idle process. After the user logs in through the login page, he or she will enter the main program page. Having entered the main program page, the program will check the variable ―logined‖. If it is false, it will return to login page, or else, it will define some variables and function.. The program will detect the microphone and camera device if no microphone or camera is detected, the program will halt and prompt the user to install these devices. If the microphone and camera is present, the program will begin to connect to flash media server and enter the idle process. The idle process will access the database (MySQL) every few seconds to check for incoming calls. database (MySQL) every few seconds, program take to see if have an incoming call. If there is a call, then the program will handle the incoming call event. A simplified flowchart is shown in Fig 22.. 28.

(38) Figure 22. Flowchart of cloud VVoIP initialization as well as telephone idle. Fig. 22 is merely the simplified process. Fig. 23 contains a more detailed explanation on the idle process.. 29.

(39) 30. Figure 23. Flowchart of cloud VVoIP telephone idle in detail.

(40) 3.5 Set-up of development environment and deploy VVoIP application Install the PHP language code editor (Notepad++) in the Windows 7 operating system. Then install the Flash development tool (Adobe Flash Professional CS5.5). The Flash development interface will set the default player to Flash Player 10.1. Use this tool to develop Flash program. Table 6. Development environment and package. Development Environment Operating System Programming Languages PHP Code Editor Flash program development tool. Package or Tool Windows 7 Flash Action Script 3.0、PHP Notepad++ Adobe Flash Professional CS5.5. This study uses Flash Action Script 3.0 language to program the VVoIP (Video/Voice over IP) application. After PHP code has uploaded to virtual machine web server, VVoIP program will start to communicate with MySQL via PHP. Once VVoIP application has be released as .swf file, and .html file is produced in Adobe Flash Professional CS5.5, two files are uploaded to the web server and then the VVoIP webpage can be browsed. VVoIP webpage displays are shown in Figs 24 and 25. For Android users, Android Apps ( .Apk files) are uploaded to the Web Server or the Android Market for user download Android users need only to install Adobe AIR Apps to use VVoIP Apps, and conduct video calls with other online users.. 31.

(41) Chapter 4. Experimental Results and Discussion In this chapter, we will do some tests to show the effectiveness of this study, such as PC-to-PC call quality testing, PC-to-Phone call quality testing, and Phone-to-Phone call quality testing, off-site call test, simultaneous calling stress test.. 4.1 Experimental environment The equipment for the experimental cloud server environment is listed in Table 7. Table 7. Node environment Nature. Equipment. Computer CPU cores. Quad-Core(4-core). CPU model. Intel(R) Core(TM) i5CPU 750 @2.67GHz @2.66 GHz. Memory size. 4 gigabyte. Disk space. 500 gigabyte. Operating System. Proxmox Virtual Environment(Linux). Virtual Machine’s hardware configuration is listed in Table 8. Table 8. Virtual Machine’s hardware configuration Nature. Equipment. Computer CPU cores. Quad-Core(4-core). Memory size. 2 gigabyte. Disk space. 32 gigabyte. In this test, the Android device is Sony Ericsson XPERIA Ray, with a PC browser Chrome 19.0.. 32.

(42) Table 9. The test environment Test Environment. Equipment or Program. PC browser. Chrome 19.0. Android device. Sony Ericsson XPERIA Ray. Sony Ericsson XPERIA Ray specifications are listed in Table 10. Table 10. Sony Ericsson XPERIA Ray specifications Android Device Nature. Equipment. Operating System Processor video camera network. Android 2.3.4 Qualcomm Snapdragon MSM8255 1GHz Front VGA video camera support Supports Wi-Fi 802.11b/g/n , 3.5G network. 33.

(43) 4.2 Experimental results In this study, Internet video calls can be placed via the VVoIP program across different operating systems, such as Firefox and Windows IE. In addition, the VVoIP program can be installed on the Android mobile devices. Internet video calls can be made between Android mobile devices. Calls can also be made between different platforms, such as Andriod devices and PC. Cloud-VVoIP’s web login page is shown in Fig. 24.. Figure 24. Web login screen of VVoIP on Chrome browser on the PC. 34.

(44) The Android smart phone login screen after starting Cloud-VVoIP app is shown in Fig. 25.. Figure 25. VVoIP login screen in Android smart phone. 4.2.1 Phone test between PCs VVoIP program supports internet video calling between PCs provided that the PCs have Webcams and microphones. The PC needs to install a newer version of the Adobe Flash Player10.1. The Cloud-VVoIP screen after user login is shown in Fig. 26.. 35.

(45) Figure 26. PC browser screen: no calls diagram The cloud-VVoIP screens when two users make an Internet video call through the browser are shown in Figs 27 and 28.. Figure 27. Video call display in PC for user01. 36.

(46) Figure 28. Video call display in PC for user02. 4.2.2 Phone test between PC and mobile device VVoIP program supports Internet video calling between the PC and the Android Mobile device, provided each have a Webcam and a microphone. The PC needs to install a newer version of the Adobe Flash Player10.1, while the Android Mobile device needs to install Adobe AIR Apps. The cloud-VVoIP screens when two users make an Internet video call between PC and mobile device are shown in Figs 29 and 30.. 37.

(47) Figure 29. Video call display in PC for user01. Figure 30. Video call display in Android device #1 for user03. 38.

(48) 4.2.3 Phone test between mobile devices VVoIP program supports Internet video calling between Android Mobile devices provided that the Android Mobile devices have Webcams and microphones. And the Android Mobile devices need to install Adobe AIR Apps. The cloud-VVoIP screens when two users make an Internet video call between mobile devices are shown in Figs 31 and 32.. Figure 31. Video call display in Android device #1 for user03. 39.

(49) Figure 32. Video call display in Android device #2 for user02 The cloud-VVoIP actual screen shots when two users make an Internet video call between mobile devices are shown in Figs 33 and 34.. Figure 33. A snapshot of video call display in Android device #1 40.

(50) Figure 34. A snapshot of video call display in Android device #2. 4.2.4 Off-site quality test Since the program was developed using laboratory Wi-Fi, off-site call tests need to be conducted, such as home Wi-Fi connection or using ADSL connection with another TAnet user. Off-site quality test screens are shown in Figs 35 and 36.. 41.

(51) Figure 35. Phone test using off-site Wi-Fi. Figure 36. Phone test using on-site Wi-Fi. 42.

(52) 4.2.5 Simultaneous connection stress test Due to FMS development version, we can only have ten clients connect to Flash Media Development Server at any one time. Fig. 37 shows the first group of simultaneous connection stress test.. Figure 37. The first group of the multi-user map Fig. 38 shows the second group of simultaneous connection stress test.. Figure 38. The second group of the multi-user map 43.

(53) Fig. 39 shows the third group of simultaneous connection stress test.. Figure 39. The third group of the multi-user map Fig. 40 shows the fourth group of simultaneous connection stress test.. Figure 40. The fourth group of the multi-user map. 44.

(54) Fig. 41 shows the fifth group of simultaneous connection stress test.. Figure 41. The fifth group of the multi-user map. 45.

(55) 4.2.6 Power consumption estimate at Android mobile phone According to the electricity specification indicated on the casing of mobile phone battery, for example Sony Ericsson Xperia Ray, battery capacity has marked 1460mAh with operating rate voltage 3.7V; in other words it can theoretically deliver the operational rate power about 5.402Wh, as listed in Table 11, when it works continuously and exhaustively. As a result, the measured highest rating of power consumption 5.6721Wh at Android mobile phone tested by software ZDbox [33] shows a little bit higher than theoretical one as indicated below in Table 11. Table 11. Mobile phone power consumption estimate Theoretical. Measured. power consumption. power consumption. Battery capacity. 1460mAh. 1460mAh. Voltage. 3.7 V. 3.85V~3.92V. Watt hour. 1.46A*3.7V=5.402Wh. 1.46A*3.885V= 5.6721Wh. (Battery enclosure marked 5.5Wh). 46.

(56) As shown in Fig. 42, the display at the up-left and up-right corners have indicated the marked time and battery power usage, respectively, in the beginning of VVoIP test operation.. Figure 42. In the beginning of VVoIP test operation As shown in Fig. 43, the display at the up-left and up-right corners have indicated the marked time and battery power usage, respectively, almost at the end of VVoIP test operation.. 47.

(57) Figure 43. Almost at the end of VVoIP test operation After the test of power consumption for Android mobile phone, we found that the phone under idle state dissipates the battery power only out of 2% of its capacity for 70 minutes. Based on that, we intuitively are able to estimate total 100% power consumed exhaustively to be last around 3500 minutes (58.33 hours). Furthermore, in Cloud-VVoIP application, the wattage for an hour consumed in the idle phone can be figured out by using the following formula in Eq. (1) or Eq. (2). C (Wh) / P(Wh / h)  t (h). (1). C (Wh) / t (h)  P(Wh / h). (2). According to Eq. (2), we can calculate the power consumption for an hour in the idle phone as follows. 5.6721(Wh) / 58.33 (hrs) = 0.09724(Wh/h) = 97.24(mWh/h) Typically, the idle phone for Cloud-VVoIP application just consumes 0.09724W for an hour and thus the full rating of battery capacity will be available for 58.33 hours. Therefore, it is assured that Cloud-VVoIP application consumes battery power not too much.. 48.

(58) 4.2.7 Power consumption estimate at Personal Computer This session will discuss the power consumption for personal computer when Cloud-VVoIP is applied to it. According to the report from the web site in Internet so far., CPU (i5-750) utilization rate of 100% of current measurement based on the electricity specification of CPU is roughly 1.66 Amp per second as shown in Fig. 44 (Source: http://www.mobile01.com/topicdetail.php?f=168&t=1643921).. Figure 44. CPU (i5-750) utilization rate of 100% of current measurement In the Cloud-VVoIP application, CPU (i5-750) utilization rate is roughly 7%~13% and its average 10% while the idle phone is running in a PC.. 49.

(59) Figure 45. CPU (i5-750) utilization rate for the idle phone running in a PC It is worth noting that CPU voltage goes up and down ranged from 0.96 V to 1.32 V and its average 1.14 V while Cloud-VVoIP is running in a PC.. Figure 46. Core voltage for the idle phone running in a PC. 50.

(60) The necessary data about CPU electricity as mentioned above, while the idle phone is running in a PC, will be summarized herein in the following Table 12. As a result, in VVoIP application, the power consumption for an hour is about 681.264 watt while the idle phone is running in a PC. Table 12. Measured data in average for CPU Electricity. Measured Data. Voltage. 1.14 Volt. Current. 1.66 Amp per second. Watt hour. 1.66A*10%*1.14V*3600= 681.264 Watt per hour. 51.

(61) 4.2.8 System performance analysis Flash Media Server has its own web management tools. After logging in with the administrator’s account, password, and server IP, the user can enter the management page. The login page of the management tools is shown in Fig. 47.. Figure 47. Login interface of the flash media server management tool. 52.

(62) From the management page, the administrator can view real-time information of the Flash Media Server. For example, the server’s active connections, bandwidth, CPU and memory usage and other information are illustrated in Fig. 48.. Figure 48. Control interface of the flash media server management tool The first chart, Fig. 49, represents the number of client connections to flash media server.. Figure 49. Active connections － control interface. 53.

(63) The second chart, Fig. 50, stands for Flash Media Server's bandwidth usage.. Figure 50. Bandwidth － control interface. The third chart, Fig. 51, means Flash Media Server’s CPU and memory usage.. Figure 51. CPU and Memory Usage － control interface. 54.

(64) When there are ten clients connected to Flash Media Server, we can organize the Flash Media Server workload through the flash media server management tools, as listed in Table 13. Table 13. FMS Data Option. Quantity. Clients number. 10. Bandwidth. 1.6~2.6Mbps. CPU Usage. 1~3 %. Memory Usage. 18 % ( About 375 MegaByte). Through the flash media server management tools, we can collect and organize CPU, memory and bandwidth information into a chart, depending on the number of simultaneous calls, as shown in Fig. 47, 48 and 49. The scheme we proposed herein is to come up with a composite index (CI) through a linear combination form. This has implied the utility rate of hardware resources in Cloud-VVoIP system. We denote three input variables, which are the tasks for CPU utility, memory utility, and bandwidth utility in an on-line phone call over Cloud-VVoIP, denoted as Cutility, Mutility, and Butility, respectively. As Cloud-VVoIP is operating in an on-line phone call via Ethernet or WiFi, we may estimate CPU utility (Cutility), memory utility (Mutility), and bandwidth utility (Butility) according to Eqs (3), (4), and (5), respectively, where Ci represents a sampled CPU utility rate, Mj stands for a sample memory utility rate, and Bk means a sampled bandwidth utility rate. Furthermore the weight of respective variable to form a formula of linear combination is abbreviated to be w1, w2, and w3 in Eq. (6) as follows.. For the. purpose of expediting the evaluation of composite index [24] [27], we have intuitively. 55.

(65) simplified the equal-weighted linear combination of the equation in Eq. (7). Finally, the outcome of this linear combination represents a composite index that accounts for the remaining amount of resources can be utilized further for the operations of tasks to utilize the resources CPU, memory, and/or bandwidth. N. C utility . C. i. i 1. (3). N M. M M utility . j. j 1. (4). M L. B. k. Butility . k 1. L. (5). CI  w1  C utility  w 2  M utility  w 3  Butility s. t . w1  w 2  w 3  1. w1  w 2  w3 , w1  w 2  w3  1. (6). (7). The grade of composite index stands for how much percentage the remaining amount of resources left where higher grade indicate more resources are available in system; otherwise, few resources left. We have employed composite index to go for analyzing the performance of FMS as illustrated in Figs 52, 53, 54 and 55. Adobe flash player as we know works higher efficiently with PC-x86 machine than Adobe AIR in Android ARM-based mobile device. So FMS needs fewer hardware resources supplied when connected to PCs; in contrast, FMS requires more hardware resources in cloud computing server when connected to Android mobile devices.. 56.

(66) Figure 52. Line chart for CPU, memory, and bandwidth in PC-to-PC. Figure 53. Line chart for CPU, memory, and bandwidth in PC-to-Phone. 57.

(67) Figure 54. Line chart for CPU, memory, and bandwidth in Phone-to-Phone. Figure 55. Composite index line chart for PC-to-PC, PC-to-Phone and Phone-to-Phone. 58.

(68) Chapter 5. Conclusion General Video / Voice over IP applications are installed on the PC, Notebook. These applications are rarely able to support a variety of different operating systems, not to mention a completely different device (such as mobile phones and tablet PCs) to do cross-platform calls. This study’s VVoIP application is not installed onto the PC, but rather built on the cloud system. Communication with other online users is done through the web interface. The video streaming service for cloud VVoIP uses the RTMP protocol, in which the two sides will conduct a session through Flash Media Server. Voice / video streaming is done through Flash Media Server between two wireless mobile devices (or PCs). The users need not know in advance the actual IP address of the other users. Users can perform real-time online video calls through the cloud web interface (or cell phone Apps interface). This system has the following advantages: the user's information is stored in the cloud system's database, such as the phone book. Because of the Server-Client architecture, the server can immediately be aware of the connection numbers and each usage condition, and then make the appropriate call flow allocations. The login system uses the Authentication, Authorization, Accounting (AAA) management system, placing account management and user authentication mechanism into practice. In the future, the server can grant various levels of authorization to different identities, so to increase management securities. When the programmer updates VVoIP, users need only to refresh the webpage to obtain its latest version and execute it. VVoIP requires about 5MB of memory and a bandwidth of 260Kbps in Android devices. Currently, Android devices use the 3.5G network. In theory, the ideal. 59.

(69) transmission 3.6/7.2/14.4 Mbps, however, the actual network flow is unable to 1~2Mbps. Android devices can also Wi-Fi network. Wi-Fi is a wireless LAN setting in the IEEE 802.11. Table 13 as listed below will compare the bandwidths of several Wi-Fi specifications. It can be seen that whether it is Wi-Fi or 3.5G network, they can fulfill the VVoIP bandwidth demand. Table 14. Wi-Fi specification option \ specification maximum bandwidth. 802.11b. 802.11a. 802.11g. 11Mbps. 54Mbps. 54Mbps. In the future, this research can be improved in several ways： (a) Parallel database processing： We used MySQL in this study. However, it may be dangerously overburdened if the usage was dramatically increased in the future. So we may need switch to Hadoop Database (HBase) to resolve this problem. Hadoop HDFS as its file storage system; HBase also uses Hadoop MapReduce to process the vast amounts of data in HBase. (b) Distributing the workload the Flash Media Server： When a single flash media server is unable to meet the demand of large numbers of users, we need to re-design the VVoIP program in order to correspond to more Flash Media Servers. Then configure multiple Flash Media Servers and their independent IPs to provide more bandwidth. (c) Developing VVoIP for iOS devices： Adobe Flash Professional CS5.5 has published AIR for iOS applications. AIR for iOS applications can be executed on Apple iPhones and iPads. The .FLA file is converted to a native iPhone application. Flash type videos and voices first need to be processed in Flash Media Server 4.5, and then we can stream to iOS devices.. 60.

(70) References [1] [2] [3] [4]. [5] [6]. [7]. [8]. [9]. [10]. [11]. [12] [13]. 王鵬，2010，雲端運算的關鍵技術與應用實例，台北:佳魁資訊. 張亞飛，2009，PHP for Flash FMS 動態網站開發手札，台北:上奇資訊. 廖冠雄, 林聿中, 謝欽旭.，2010，VoIP 網路電話實驗系統，臺灣優奎士. A. Kimball, S. Michels-Slettvet, and C. Bisciglia, ―Cluster Computing for Web-Scale Data Processing, ― ACM SIGCSE Bulletin - SIGCSE 08, vol. 40, no. 1, pp. 116-120, March 2008. Amazon Web Services (AWS), 2012. http://aws.amazon.com/ B. R. Chang, C.-F. Huang, H. F. Tsai, and Z.-Y. Lin, ―Rapid Access Control on Ubuntu Cloud Computing with Facial Recognition and Fingerprint Identification,‖ Journal of Information Hiding and Multimedia Signal Processing, Vol. 3, No. 2, pp. 176-190, April 2012. B. R. Chang, C.-P. Young, H. F. Tsai, and R.-Y. Fang, ―Timed PR-SCTP for Fast Voice/Video over IP in Wired/Wireless Environments,‖ Journal of Information Hiding and Multimedia Signal Processing, Vol. 2, No. 4, pp. 320-331, October 2011. B. R. Chang, H.-F. Tsai, C.-F. Huang, and H.-C. Huang, ―Private Small-Cloud Computing in Connection with WinCE Thin Client,‖ Lecture Notes in Artificial Intelligence, vol. 6422, pp. 172-182, Nov. 10, 2010. B. R. Chang, H. F. Tsai, C.-P. Young, and R.-Y. Fang, ―Embedded System for Inter-Vehicle Heterogeneous-Wireless-based Real-Time Multimedia Streaming and Video/Voice over IP,‖ Proc. of 4th International Conference on Innovative Computing, Information and Control (ICICIC-2009), Paper ID: B06-01, Ambassador Kaohsiung, Taiwan, December 7-9, 2009. B. R. Chang, H.-F. Tsai, Z.-Y. Lin, C.-M. Chen, and Chien-Feng Huang, ―Adaptive Performance for VVoIP Implementation in Cloud Computing Environment,‖ Lecture Notes in Artificial Intelligence, Vol. 7198, Part III, pp. 356-365, March, 2012. B. R. Chang, H. F. Tsai, Z.-Y. Lin, and C.-F. Huang, ―Fast Access Security on Cloud Computing: Ubuntu Enterprise Server & Cloud with Face and Fingerprint Identification,‖ Advances in Intelligent and Soft Computing (Springer), Vol. 144, pp. 451-457, January 2012. C. Metz, ―AAA protocols: authentication, authorization, and accounting for the Internet,‖ Internet Computing (IEEE), vol. 3, no. 6, pp. 75-79, Nov. 1999. C.-P. Young, B. R. Chang, and Z.-L. Qiu, ―Scheduling Optimization for Vector Graphics Acceleration on Multiprocessor Systems,‖ Journal of Information 61.

(71) Hiding and Multimedia Signal Processing, Vol. 3, No. 3, pp. 248-278, July [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]. [25] [26] [27]. [28] [29] [30] [31] [32] [33]. 2012. Eucalyptus, 2012. http://open.eucalyptus.com/ Google App Engine, 2012. http://groups.google.com/group/google-appengin GNU Classpath , GNU Classpath, Essential Libraries for Java, in 2012. http://www.gnu.org/software/classpath/ Hyper-V Server R2, 2012. http://www.microsoft.com/en-us/server-cloud/hyper-v-server/ IBM Cloud Computing, 2012. http://www.ibm.com/ibm/cloud/ I Hear U Project Homepage, 2012. http://ihu.sourceforge.net/ JamVM -- A compact Java Virtual Machine, 2012. http://jamvm.sourceforge.net/ Java 2 Platform, Micro Edition (J2ME), 2012. http://www.java.com/zh_TW/download/faq/whatis_j2me.xml J.-Y. Li, ―Peer-to-Peer System Structures and Data Transmission Strategies: Design and Implementation,‖ Ph.D. Dissertation, Globe Thesis, Feb. 28, 2010. OpenNebula, 2012. http://www.opennebula.org/ Positive Effects on System Performance http://www.auslogics.com/en/downloads/boost-speed/5/auslogics_boostspeedpositive_effects_on_system_performance.pdf Proxmox Server Solution GmbH, 2012. http://www.proxmox.com/ Sector/Sphere, National Center for Data Mining, 2012. http://sector.sourceforge.net/ Testing the Effect of Encryption on Disk Performance http://www.castleforce.co.uk/assets/File/Check_Point/Testing-Encryption-Disk -Performance_from_Castleforce.pdf Voice over IP in a Cloud, IIS, 2009. http://www.17freecall.com/, http://www.loip.com/ VoIP in cloud computing, Skpye, 2010. http://www.chinalabs.com/html/chanyezhuanxing/2010/1214/41378.html VMware Virtualization, 2011. http://www.vmware.com/virtualization/ Windows Azure- A Microsoft Solution to Cloud, 2012. http://tech.cipper.com/index.php/archives/332 Welcome to Apache Hadoop, 2012. http://hadoop.apache.org/ ZDbox, version: 3.5.222, 2012. https://play.google.com/store/apps/details?id=com.zdworks.android.toolbox&f eature=search_result#?t=W251bGwsMSwxLDEsImNvbS56ZHdvcmtzLmFuZ HJvaWQudG9vbGJveCJd. 62.

(72)