Connectivity Providers Database

From: owpurbo@nyquist.uwaterloo.ca To: amcgee@netcom.com Date: Tue, 20 Jul 93 14:12:59 -0400 THE BUILDING OF INFORMATION INFRA-STRUCTURE TO SUSTAIN THE CURRENT GROWTH IN INDONESIA Onno W. Purbo Inter University Center on Microelectronics Institute of Technology Bandung Bandung 40135, INDONESIA FAX: (62)-22-438-338 E-mail: indogtw!paume!onno@uunet.uu.net ABSTRACT A five years experience in building a wireless WAN in Indonesia will be described. Some of the basic ideas are actually coming from CIDA's funded CoSy project to Indonesia in 1987-1988. Ways to strengthen the strategic elements, namely, man-power, knowledge and hardware/software resources, is explained. An alternative physical layer using a low cost packet radio technology with possibilities to interconnect the high-speed LANs into a WAN is presented. This approach enables the Indonesian universities and government research institutions to reverse engineering and to build the TCP/IP WAN and packet radio technology which in turn supports us to build a self-sustainable and self- financing wireless WAN required to transfer of information, knowledge and science needed to sustain a bottom-up technological and social development in Indonesia. INTRODUCTION Although, CIDA's funded CoSy project to assist the Indonesian scholars have been completed in 1988, the spirit to use computer medium to inseminate the development in Indonesia has been carried by several Indonesian CoSy participants (of which the author happens to be one of them) to implement a more practical system with the current situation in Indonesia. In this paper, we report on our five years effort to build a wireless wide area computer network (WAN) infra-structure needed to sustain the current growth in Indonesia. Electronic conferences / mailing lists are used as a strategic medium to support the transfer of knowledge needed to built the required man-power for building the WAN. This development is partly motivated by the exposure of some of the Indonesian staff members to the CoSy at University of Guelph under CIDA funded project back in 1987. Electronics conference, such as, CoSy at University of Guelph, has many advantages, such as: 1. It provides a medium for fast exchange of information among the users of the electronics conferences. This in turn enables a quick transfer of knowledge and technology to support bottom-up development. 2. The users can participate into on-going discussions at any time. CoSy at University of Guelph (back in 1987) as an electronic conferencing system has two major disadvantages: 1. It is very centralized. All participants have to logon to the system to participate in the electronic conferences. 2. It relies heavily on phone or public packet switching network (PSN). Wide Area computer Network (WAN) is a crucial infra- structure to support a solid flow of information as well as distribution of knowledge and science. The existence of wireless WAN in Indonesia, which in turn enables a solid flow of information, permits us to move forward from grass-root level by elevating the capabilities of our man-power to support a bottom- up sustainable technological and social development in Indonesia. There are currently many WAN's in operation in Indonesia. Many commercial sites, such as, computer vendors, oil companies and some NGO's, have their own WAN and connected to international network such as InterNet or PeaceNet. Dedicated and leased lines are normally used to establish these networks. It is ashamed that universities and government research institutions, from which the man-power required to sustain the development in Indonesia are prepared, have very little support from the government in establishing our own WAN knowingly that establishing a solid flow of information is crucial to prepare a knowledgable man-power to 1 sustain the country's continuing development. Aside from this un- supportive environment, several committed staff members from several universities and government research institutions manage to built our own wireless WAN using radio (low cost packet radio technology) as our primary physical link - with minimal support from the government and external resources. Why wireless WAN? Wireless WAN can be established using less expensive equipments and very decentralized in nature. Thus, it can be initiated without having to rely on a single authority / company to provide the equipments / services. This situation is very favourable for the end-users as well as the supporting industries. Furthermore, in the long run, a wireless WAN leads to a lower overhead cost than leased dedicated lines. Moreover, the software (including the source code) and the hardware are readily available in public domain. This in turn enables us to reverse engineering the software and hardware which leads to a low-cost investment in building our own wireless WAN infra-structure without having to depend financially and technologically on external sources (Purbo, 1992). In this paper, we will report on our five years experience in building wireless WAN among the universities and government research institutions in Indonesia. Strategic elements in the development of wireless WAN will be discussed. The TCP/IP computer network architecture used will be briefly reviewed. Current connectivities of our universities and research institution will be shown. Furthermore, our current strategies to move forward will be briefly described. We hope this report will give a valuable input for other developing countries in building their own WAN infra-structure. KEY ELEMENTS IN BUILDING WAN IN INDONESIAN There are basically three strategic elements in building the computer network in Indonesia, namely: 1. Knowledgable & committed man-power. 2. Access and the development of knowledge in computer network. 3. Ability to build and assembled the required hardware / software. Among these elements, the existence of a group of knowledgable and committed people (man-power) is the most crucial component to assure the ability of the Indonesian people in building their own network without having to rely heavily on external resources. In this section, we will focus on our experience in building of the knowledgable and committed man-power (historical perspective). Experience learned from CIDA's funded CoSy project in 1987- 1988 indicates that computer medium can be served as a useful medium in transferring technology as well as in motivating people to learn from others and, in turn, increase their self- motivation. Five years, after the completion of CIDA's funded CoSy 2 project in Indonesia, an exposure to such rare opportunity in 1987 has prompted some of the Indonesian scholars aboard who were studying towards their graduate degree to start several Indonesian mailing lists using their computer account at schools. The first known Indonesian electronic mailing list is located at University of California Berkeley in the US with address <indonesians@janus.berkeley.edu> established around 1987-1988. It is now still in operation and is mainly used for general discussions. Subsequently, many Indonesian mailing lists including a newsgroup soc.culture.indonesia in USENET have been established in InterNet some of which have attracted participants from various Canadian NGOs and institutions such as McGill Indonesia Project, EMDI and WUSC. Among these Indonesian electronic discussion groups, there are two major mailing lists established with purpose solely to support the technological and social development in Indonesia, namely: 1. PAU-Mikro-net <pau-mikro@ee.umanitoba.ca> (established in 1990 with total member of around 200) 2. the Indonesia Development Study <ids@suvm.bitnet> (established in 1991 with total member of around 350) The Indonesia Development Study (IDS) is currently dealing with various general issues in Indonesia's development in a more global sense. The member of IDS are coming from many backgrounds with majority in social and economics sciences. Several people from international organizations including EMDI and McGill Indonesia Project join the IDS mailing list. Different from IDS, PAU-Mikro-net is specifically working on a more narrow subject in building the computer network in Indonesia as well as various microelectronics and computer science aspects. Consequently, the majority of the members in PAU-Mikro-net have engineering background. Similar to IDS, several staff members from Western Universities, such as University of Wisconsin, and several international NGO's, are also joining the PAU-Mikro-net. PAU-Mikro-net is more unique in nature as compared to other Indonesian mailing lists in InterNet. PAU-Mikro-net interchanges ideas as well as performing real tasks directly with participants in Indonesia, namely, at UI (Jakarta); ITB (Bandung); STT-Telkom (Bandung); LAPAN (Jakarta); INKOM-LIPI (Bandung); Ministry of Environment (Jakarta); Ministry of Foreign Affair (Jakarta) and BPPT (Jakarta), via computer network. The Paguyuban TCP/IP is currently active in discussing various issues regarding the development of computer network in Indonesia in PAU-Mikro-net. The international E-mail connection between Indonesia and the InterNet is currently provided by the Center for Computer Science at UI in Jakarta. In terms of helping building the information infra-structure in Indonesia, the discussions in PAU-Mikro-net serve several purposes, such as: 1. Transfer of knowledge by means of consultations for any 3 problems encountered in implementing the technology in Indonesia. 2. Feed-back and report of the current network expansion in Indonesia. 3. Finding an alternative international routes and funding for Indonesia. This is mainly motivated by our expensive PSN. 4. Keeping the spirit of our colleagues in Indonesia in building the computer network. This is very crucial as nearly all of us are currently doing this as a volunteer apart from our daily jobs. Most of us agree to keep this voluntarily basis of our institution to reduce the bureaucracy (which usually slow the progress). These purposes have been served very nicely at PAU-Mikro-Net mainly because most of the members abroad have access to InterNet which enable them to find the required information. Furthermore, prior to the implementation of certain methods / concepts precaution can be provided by monitoring newsgroup / other InterNet mailing lists. Thus, in concept, this is actually a transfer of technology process. Within a five years experience, it has been proven that written medium (electronic or a more conventional paper medium) is essential to be able to transfer the knowledge needed to build the committed man-power. In 1987-1989, prior to the establishment of Indonesian electronic mailing lists in the InterNet, most of the attempts to transfer of knowledge and to increase self- motivation were done using written paper medium, such as, newspaper, magazine. In 1990-1991, as above mentioned, we started using electronic computer medium to perform these tasks as well as real tasks to help and to coordinate the building of computer network in Indonesia. This effort in turn leads to the establishment of an informal consortium of committed staff members of the Indonesian universities and government research institutions known as "Paguyuban TCP/IP". Paguyuban TCP/IP is an informal consortium of about 30-50 committed Indonesian staff members (both in Indonesia and abroad) with a single objective to build the computer network in Indonesia. The informal nature of this consortium enables us to surpass the bureaucracy commonly found in many Indonesian institutions / government offices and, thus, enables us to accelerate the building of the network. Before we proceed with a more detail description of the current state of the wireless WAN in Indonesia. A brief review of the architecture of TCP/IP computer network and its attempts to adopt the technology in Indonesia will be presented in the next section. THE ADOPTION OF TCP/IP WAN IN INDONESIA In this section, a brief overview of the TCP/IP computer network and our attempts to adopt the technology will be reported. First, the overview of the TCP/IP WAN technology will 4 be described. Subsequently, we will describe our attempts to adopt the technology and to transfer the technology to our local industries in Indonesia to sustain the current growth of the computer network. A simple network architecture is used in the implementation of TCP/IP-based computer network (see Fig. 1). It consists of five protocol layers. These layers, in terms of its functionality (from the lowest to the highest level) are physical layer, link layer, network layer, transport layer and application layer. There is no need for an end user to understand how these layers interacts to use the computer network. InterNet Protocol (IP) (Postel, 1981) and Transport Control Protocol (TCP) (Postel, 1981) are used in network and transport layer, respectively. This leads to the term TCP/IP-network in referring to this architecture. The physical and link layer protocols used in a computer network may vary depending on the form of the network. In most high-speed Local Area Network (LAN)s, 10Mbps Ethernet or Token Ring physical layer, the IEEE 802 link layer protocol (Stallings, 1987) is normally used. To form a Wide Area Network, commercial Packet Switching Network (PSN) or even ISDN may be used with various link layer protocol such as CCITT X.25 (CCITT). The interconnection of various physical and link layer protocols in various LAN / WAN to form a nation wide or even worldwide computer network is transparent to the users by using InterNet Protocol (IP) in TCP/IP-based WAN. The TCP/IP-based WAN has currently emerged into the largest worldwide computer network known as InterNet. We use mostly PC-based TCP/IP routers and switches to minimize the overhead costs. TCP/IP protocol is implemented in KA9Q Network Operating System (NOS) originally develop in the amateur radio (Karn, 1985), (Karn, 1988). NOS is running on PC with both source and executable code available in public domain for non-commercial purposes. We have managed to reverse engineering the NOS software for our purposes. The use of TCP/IP protocol enables us to merge various high-speed LANs into WAN over radio as well as leased line. We retain the current policy in distributing NOS software that is "public domain" provided with the source code and manual. This is a very important contribution to sustain the growth of the computer network. Packet radio as physical layer is of interest especially due to its nature which allows more flexibility for the end user to build their network without having to depend on external sources as well as to reduce the overhead costs. Let us now review some of the alternative configurations of packet radio technology. These configurations are currently being pursued into mass production lines to support further expansion of our wireless WAN in Indonesia. Aside from the radio and the microcomputer, the typical configuration of packet radio station can be classified into: 1. 1200 bps AFSK modems connected directly to PC's serial 5 port. 2. PC add-on Terminal Node Controller (TNC) with AFSK modems. 3. High-speed 56 Kbps packet radio systems. In the first configuration, a 1200 bps AFSK modem is connected directly to PC's serial port. This configuration is primarily intended to provide an end user with a lowest possible cost hardware necessary to become a part of the packet radio network. Typical cost to built such AFSK modem is in the range of US$20-$40. We are currently working very closely with several NGO's in transferring the knowledge in building this lowest cost solution to the small industries in Indonesia. We hope to see the first shipment of the generation of low cost modem in 1994. For advanced packet radio applications, e.g., dedicated TCP/IP operations, the PC computing power should be freed to perform higher-level networking tasks. This can be done in the second option by leaving lower level protocol operations to a dedicated hardware. This configuration is process to be produced in mass production in Indonesia by Prof. Iskandar Alisyahbana group from ITB. As the network grows to interlink various high-speed LANs into WAN, it is most likely that the long-distance packet switching backbone nodes will experiencing a heavy traffic which might create network congestion. Work is currently underway to investigate the use of a high-speed PC add-on card connected to a 56 Kbps packet radio high-speed modem for high-speed packet radio backbone. It is interesting to note that the commercial packet switching network (PSN) in Indonesia operates at top speed of only 9600 bps. Please note that the packet radio technology is especially used to link regional and long distance networks. Within a building or campus complex, a high-speed LAN would be a better investment. Furthermore, TCP/IP protocol allows us to integrate these two different technologies (LAN and packet radio technology) into a single WAN which transparent to end-users. Since most of the alternative configurations are currently being produced in Indonesia, the expansion of the wireless WAN can be done at low cost without having to depend on external sources. IMPLEMENTATION OF WIRELESS WAN IN INDONESIA In this section, the actual implementation of packet radio technology in our WAN will be presented. The wireless WAN is currently in operation (and keep expanding) to connect: 1. University of Indonesia (UI) in Jakarta. 2. Institute of Technology Bandung (ITB) in Bandung. 3. National Aerospace Institute (LAPAN) in Jakarta and Bogor. 4. Agency for the Assessment and Application of Technology (BPPT) in Jakarta. 5. School of the Telecommunication Agency (STT-Telkom) in Bandung. 6 6. The Indonesian Science Institute (P3INKOM-LIPI) in Bandung. 7. The Ministry of Environment in Jakarta. 8. The Ministry of Foreign Affair in Jakarta. Several other institutions (such as University of Gadjah Mada in Jogyakarta, University of Parahiyangan in Bandung, University of Satya Watjana in Salatiga, University of Brawidjaja in Malang) are currently preparing their man-power and equipments to join the network. All of these activities are initiated by the committed staff members of each institution without any direct intervention from the top government policy-makers. It is to say that the nature of the development is bottom-up approach. Since TCP/IP protocol is used in the network, the nature of our network is fairly similar to InterNet. The discussions in PAU-Mikro-net is forwarded using E-mail (SMTP) (Postel, 1982) and news (NNTP) (Kantor, 1986) to various hosts in Jakarta, Bandung and Bogor. Indonesia has just recently becoming part of InterNet with top domain ".id". Figure 2 shows a simplified topology of our current packet radio network in Indonesia. VHF and UHF links are used to interconnect the high-speed LAN at ITB, UI, BPPT, LAPAN, STT- Telkom and P3INKOM-LIPI. The operational speed used in the VHF and UHF links is currently 1200 bps and 9600 bps. However, works are currently underway to increase the regional packet radio speed to 56 Kbps. The wireless WAN is currently coexist with some of the (expensive) UUCP links used mainly as back-up links. Furthermore, it appears most of the institutions are more favourable towards the wireless technology which would be a low- cost investment in the long run. In Figure 2, it is shown some of our international links currently in operation. These links are primarily using our commercial packet switching network (PSN), i.e.: 1. TCP/IP link to University of Aachen in Germany. 2. UUCP link to UUNET in the US. Through these two international links, we are currently able to communicate to our international counter part. Unfortunately, the Indonesian PSN is known to be one of the expensive PSN in the world. Work is currently performed to use alternative routes to get a lower cost international connection. In the next section, attempts to use VITASAT to connect our packet radio network in Indonesia to international InterNet network work will be described in more detail. LOW EARTH ORBIT VITASAT SATELLITE At the time this manuscript is written, an experiment to interchange messages between North America (InterNet) and Indonesia via VITASAT (Volunteers In Technical Assistance SATellite) has been successfully underway for several months. VITASAT is a polar Low Earth Orbit (LEO) satellite with apogee around 900 km above the Earth especially design for carrying 7 digital information using packet radio technology (Davidoff, 1990). VHF 139 MHz band and UHF 440 MHz band are used for up-link and down-link frequencies which enables a simple satellite ground station to access this satellite. A simplified topology of our experiment to exchange messages between Indonesia and InterNet via VITASAT is shown in Fig. 3. We currently have two VITASAT ground stations in operation in Indonesia, namely, ITBJKT in Jakarta and ITBBDG in Bandung. Messages from ITBJKT or ITBBDG will be carried by VITASAT to the US. A VITANET station in the US will then download the message and delivered to UUNET in the US East coast to deliver to the intended InterNet destinations. At least three additional VITASAT ground stations in Indonesia (one in Jakarta and two others in east Indonesia) will be installed by the end of 1993. Prof. Iskandar Alisyahbana group at Electrical Engineering Department ITB is currently working to built low cost VITASAT ground station for used in Indonesia. To give a ball-park figure of VITASAT ground station, excluding the PC, a VITASAT ground station can be obtained at around $1000-$2000. The equipments is very easy to find in the market and to install as it is originally based on Microsat / PACSAT design from the amateur radio (Davidoff, 1990). This approach has significant advantages in terms of overhead cost as compared to the previous approach using commercial PSN for international connection. Furthermore, it allows remote areas in many part of Indonesia (or the world if you wish) to communicate with the more "civilized" world and exchanging information to assist the development in the remote regions. This can be done without having to rely on any commercial services which is a favourable approach for the end- user to sustain and to expand network at the lowest possible costs. Cooperation and coordination between Prof. Iskandar Alisyahbana's group and our packet radio network is currently underway to integrate the existing packet radio network of "Paguyuban TCP/IP" into VITANET to communicate to InterNet. STRATEGIES TO MOVE FORWARD In the previous sections, we have reported the history and the "state-of-the art" of our wireless computer network in Indonesia. In this section, we will briefly described our strategies in sustaining the growth of the computer network in Indonesia. We are currently focusing our work on three aspects, i.e.: 1. Strategies in building our man-power. 2. Strategies in mastering the knowledge in computer network. 3. Strategies in developing the required small industries to support the required hardware. To perform these strategies, we are not relying on any authorities in the government of Indonesia to execute our projects. All of these projects are performed by committed man- 8 power in Paguyuban TCP/IP using a bottom-up approach. However, with the current political situation in Indonesia, as most of the ministers have engineering background, our tasks would be much easier. Building the required man-power is our major effort. It is currently done mainly through informal meetings, discussions, mailing lists as well as through articles in the mass media in Indonesia. In a more formal stage, the staff members at ITB (in Bandung) and UI (in Jakarta) are currently employing several students to work in building / learning the computer network as part of their final projects. Most of the members of Paguyuban TCP/IP is currently doing free consultation jobs to help others who want to initiate their institutions to integrate into our packet radio network. All of these efforts are trying to increase the awareness and self-motivation in building the computer network. Several members of the Paguyuban TCP/IP is currently learning in-depth on various practical methods in TCP/IP computer network. The references for these works are mainly from the freely available Request For Comments (RFC) documents available via FTP from various sites in InterNet. A book on TCP/IP network is currently being written in Indonesian language. In addition, at least, 3 members are currently active in programming the TCP/IP software on PC. All of these activities are performed in attempts to master the TCP/IP technology to be able to develop our own technology. Our last point, the development of the small industries to produce the required hardware to support the development of packet radio WAN. It is currently in a very early stage of development. Most of the works are currently concentrated in creating a network between Paguyuban TCP/IP and various NGO's, Development Studies group in several universities (such as ITB and University Satya Watjana) to discuss the best way in approaching the small industries in Indonesia. Several contact persons have been found within NGO's and Development Studies Group in Indonesia who are willing to help us in approaching the small industries. We are going to concentrate our effort to rely on the small industries as our primary base to support the development of computer network in Indonesia. We believe through the small industries, it would be easier for us to reach the lower income family and to increase their knowledge as well as their earning. We hope, this approach will in turn benefit not only the small industries but also the whole nation. SUMMARY In this paper, an experience in building a self-sustainable computer network in Indonesia using a low-cost packet radio technology has been presented. Although packet radio technology is not the state-of-the art technology, it has been proven to serve quite nicely in enabling us in Indonesia to build and to reverse engineering packet radio 9 technology which leads to a low cost wireless wide area network. This in turn enables us to build a self-sustainable and self- financing wireless wide area computer network to support a solid transfer of information, knowledge and science needed to have a bottom-up technological development in Indonesia without having to rely on external sources. Three different alternatives for packet radio configuration have been described. Mass production of the packet radio technology is being done in Indonesia as a result of our reverse engineering process. Thus, packet radio technology will accommodate a wide range of budget and requirements. The use of TCP/IP protocol in packet radio technology enables us to interconnect our high-speed LANs into a single WAN. Research studies are currently underway to increase the speed of our regional packet radio network to 56Kbps and also to use alternative international link via VITASAT for a low-cost connection to InterNet. The establishment of a group of knowledgable and committed people is the crucial element to be able to perform such task in building the computer network in Indonesia. Electronic computer conferencing system has proven to be an essential medium to support the transfer of knowledge needed to increase self- motivation and the knowledge of the required man-power. Exposure to the tele-conferencing in CIDA's funded CoSy project has shown its benefit few years after the exposure. It is shown that the experience learned from CIDA's funded CoSy project has lead us to use electronic mailing lists and discussion groups to perform transfer of technology to Indonesia. This in turn enables us to built and to sustain our own computer network in Indonesia with minimal support from external sources. Finally, thank you Canada to expose us to this exiting new information technology which enable us in Indonesia to help building our country. ACKNOWLEDGEMENTS We wish to thank Ms.Maria Cioni to allow the author to access CoSy in 1987 (as part of CIDA's funded project). We would like to thank Prof. Dr. T. Chapman (University of Wisconsin - Madison), Prof. Dr. Iskandar Alisyahbana (ITB), Dr. Kusmayanto Kadiman (PIKSI-ITB), Dr. S. Nasserie (ITB), Dr. Adang Suwandi (ITB), the members of ITB-ARC and the members at PAU-Mikronet for their valuable comments and encouragements during the course of the work. We wish to thank the University of Waterloo - Amateur Radio Club VE3UOW to allow the author to perform experiments in TCP/IP-based packet radio network. We wish to thank many amateur radio fellows, such as Robby Soebiakto YB1BG, Dwi YB0QC, Suryono YG1QN/N5SNN, Roger VE3RKS. The financial supports from the Indonesian Government as well as the Canadian International Development Agency (CIDA) through the Canadian Bureau of International Education (CBIE) are greatfully acknowledged. 10 REFERENCES O.W.PURBO, "An alternative approach to build low cost TCP/IP-based wide area network in Indonesia," South East Asia Regional Computer Confederation (SEARCC) '92 conference, Kuala Lumpur, 14 August 1992. J.POSTEL ed., Internet protocol, RFC 791, September 1981. J.POSTEL ed., Transmission control protocol, RFC 793, September 1981. W.STALLINGS, Handbook of computer communications standards: local network standards, vol. 2, MacMillan Book, 1987. CCITT Recommendation X.25, Interface between Data Terminal Equipment (DTE) and Data-Circuit Terminating Equipment (DCE) for Terminals Operating in the Packet Mode on Public Data Networks. PHIL KARN, KA9Q, "Amateur TCP/IP: an update," Proceedings 7th ARRL Computer Networking Conference, pp. 115-121, 1988. P.R.KARN, KA9Q, H.E.PRICE, NK6K dan R.J.DIERSING, N5AHD, "Packet radio in the amateur service," IEEE Journal on Selected Areas in Communications, vol. SAC-3, pp. 431-439, 1985. J.POSTEL, Simple mail transfer protocol, RFC 821, August 1982. B.KANTOR and P.LAPSLEY, Network news transfer protocol, RFC 977, February 1986. M.DAVIDOFF, K2UBC, The satellite experimenter's handbook, 2nd edition, American Radio Relay League, 1990. 11 FIGURE CAPTIONS Figure 1. TCP/IP based computer network architecture. Figure 2. A simplified topology of our current TCP/IP packet radio network in Indonesia as of May 1993. Figure 3. A simplified topology of the communication experiment between Indonesia and InterNet via VITASAT as of March 1993. 12