HUMAN CONSTRAINTS FOR MOBILE COMMUNICATIONS. Prof. Dr. Pasi Tyrväinen Prof. Dr. Jari Veijalainen + - PDF

HUMAN CONSTRAINTS FOR MOBILE COMMUNICATIONS Prof. Dr. Pasi Tyrväinen Prof. Dr. Jari Veijalainen + University of Jyväskylä Waseda University, Dept. of Computer Science and Global Information and Information

Please download to get full document.

View again

of 22
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.

Data & Analytics

Publish on:

Views: 8 | Pages: 22

Extension: PDF | Download: 0

HUMAN CONSTRAINTS FOR MOBILE COMMUNICATIONS Prof. Dr. Pasi Tyrväinen Prof. Dr. Jari Veijalainen + University of Jyväskylä Waseda University, Dept. of Computer Science and Global Information and Information Systems Telecommuniation Institute P.O.Box 35 (Agora 5 th floor) 29-7 Bldng., Nishi-Waseda FIN Univ. of Jyvaskyla Shinjuku-ku, Tokyo phone: phone: fax: fax: ABSTRACT Mobile networks are proliferating currently in the world with a rather high pace. The number of mobile telecom subscribers has already exceeded one billion and the number will exceed two billions before the year The big investements needed in the telecom infrastructure raise the question, how much revenues can the operators and other players expect from the investments and what would limit them. It was shown earlier that the revenues obtained from human users of mobile networks and services are limited ultimately by the time the users are willing and able to use the services. In this paper we look closer at what are more concretely the limiting factors. The problematics is approached from five different points of views. First we analyse how much data is produced in the world in a time unit. This forms in a way the ultimate limit, because what does not exist cannot be transmitted. Second, the data volumes and types produced in organizations are analysed. This gives basis to estimate how much of the organizationally produced data could be transmitted over a mobile channel. Third, we analyze previous research that investigates how the people allocate their time among various activities, like sleep, work, watching TV, traveling to work or leisure activities, etc. Thus, only a fraction of the time of an individual can be used at the mobile channel. Fourth, cost models for communication capacity use and the key findings based on time users are willing to spend on communicating in private or work context are discussed. Last but not least, limits of human information processing from cognitive viewpoint are analyzed. Based on these + Visiting professor at Waseda Univ. until March 31, The reseach of this author was supported by Telecom Advancment Organisation of Japan (TAO) through Senior TAO Fellowship. On leave from Univ. of Jyvaskyla, Dept. of Computer Science and Information Systems until June 30, 2003 phone: , fax: different viewpoints we analyse their interdependencies and pinpoint questions that should be analysed further while determining the economic feasibility of the mobile networks. 1. INTRODUCTION Expensive mobile communication equipment were used first for professional purposes, but consumers have fast followed business users in adopting mobile telephones, hand-held computers, and PDA equipment to daily use. The number of digital telecom handsets exceeds already one billion in the world. Borderlines between a laptop and mobile telecom terminal, as well as other gadgets, such as game consols, radio receivers, and TV sets are vanishing due to the miniaturization and digital convergence of information processing and communication technologies. At the networking side, extensive use of IP protocol in all device categories and within the core network, as specified by the 3 rd Generation Partnership Project /19/ is a paramount example of convergence within networking. The digital convergence has also mixed the usage patterns of voice telephone, data communication, and computer equipment making it harder for the operators to estimate the future use of communication capacity. This estimation task has to take into account the time users are willing to spend for network-mediated communication tasks, as well as the real communication needs and capabilities of the users. We speak below about amount of information, but what we really have in mind is the volume of data the information is encoded into. Because the same information can have many different encoding, in fact an infinite number of them, measuring its volume through the encoded data is ambiguous, because a particular encoding has to be assumed, before we can speak about bytes. Below, the most commonly used encoding is assumed in the volume assessments, such as ASCII-characters for pure text, uncompressed format for telephone calls (with 64 kbps transmission capacity requirement), MPEG-2 for video, etc. Most volume estimates are given in bytes (octets). This paper elaborates the estimation task of a telecom operator, by combining and analyzing the earlier research from the following five viewpoints: First, we analyze the maximum amount of information produced and used globally by humanity, as estimated by the project How much information /2/ at Berkeley. Second, we draw on measurements of organizational communication volume based on genre analysis reported in recent case studies. Based on these, we give a base-line for the average amount of data an employee communicates per day related to his/her assignments, whether using some technical channel or face-to-face. Employee s communication activity is divided into verbal communication, network-mediated interactive communication, and communication using usual paper or stored digital artifacts, such as digital documents or information systems. Third, we analyze previous research that investigates how the people allocate their time among various activities, like sleep, work, watching TV, traveling to work or leisure activities, etc. This brings in a further limitation on the usage of a mobile channel. Fourth, cost models for communication capacity use and their key findings based on time users are willing to spend on communicating in private or work context. Last but not least, limits of human information processing from cognitive viewpoint are analyzed. This gives yet another aspect that sets limits for the human actor. These five viewpoint are combined by analyzing their effects on interactive network-mediated communication of humans, limited mostly by the time spent by them to perform this task, and on the human communication using digital artifacts, constrained mainly by human data communication capacity. Further development is analyzed with respect of change of communication practices of organizations, and emergence of bandwidth intensive applications requiring high QoS and other emerging trends. This paper is organized as follows. In section 2 we review the estimates for volumes of data produced by humans and machines globally. These estimates establish in a sense an ultimate limit for what can be communicated over networks, because only data that exists prior to sending can be transmitted. In section 3 we look into the organizational reality and the estimates for the volume of data produced in them by humans or machines. This gives a subset of data that is at least partially eligible for transmission over a computer or telecom network in the organizational context. In section 4 we look at human communication capabilities that also set limits for the network-mediated communication. In section 5 we combine the results and try to estimate how much of the produced data would be eligible to be transmitted over a communication network. 2. GLOBAL AMOUNT OF INFORMATION AND COMMUNICATION It was estimated, that 1-2 exabytes ( bytes) of new data was produced in year 1997/1/, corresponding to about 250 Megabytes per each human being. /2/. The data is mostly produced directly by people (typing it in) or its production has required real-time human control (photographs, videos) although the actual data has been produced by a device. The output of autonomously operating devices, such as satellites that produce a considerable amount of data every day, are excluded from the table. Other estimates state that time needed to double the amount of information in the world has decreased from 50 years to 5 years, and the time needed to double documented (stored) information has reduced from 18 months even to 9 months /3,4/. Although these figures are typically very provocative presented by commercial consultancy organizations, at least the estimates produced by a research group at the University of Berkeley are coupled with the source data used. According to the project How Much Information? /2/ annual production of unique information ( original content ) consists of the following components (Table 1, figures extracted from Tables and text in Summary of /2/): Table 1. Annual production of unique information in Figures extracted from Tables 1, 2, 3, and 5 of /2/. Storage Medium / Content Type Annual Terabytes Lower Estimate Annual Terabytes Upper Estimate Annual Amount of Titles / Millions Office documents ,500 Books, Newspapers etc ,1 Paper Total Photographs 41, ,000 80,000 Other Film 17,216 17,216 X-ray: 2,000 Film Total 58, ,216 Music and Data CDs DVDs ,005 Optical Total Camcorder Tape 300, ,000 1,400 PC Disk Drives 7, , Servers 269, , Magnetic Total 577,210 1,693,000 TOTAL STORED 635,480 2,120,539 Computer-Mediated 11, ,000 Usenet 73 HTML 2.1 Usenet 6 TOTAL C-MEDIATED 11,358 Analog Communication Radio and TV 14,938 Telephone 576,000 Postal 150,000 TOTAL ANALOG 740,938 The underlying assumptions in this research have been that all data including print media is estimated in bytes (octets). The large figures are obtained by assuming a verbose encoding for the information, such as RGB encoding for images and video, whereas the minimum figures are assuming a reasonable compression (JPEG, MPEG, etc.). From table 1 we can see that when measuring by bytes of unique data, the total amount of stored data was round 0,7 to 2,1 Exabytes, i.e. 700,000 to 2,100,000 Terabytes. Most of new data was produced by individuals recording home videos or individual computer users (0,6 to 1.7 Exabytes on magnetic media). But also telephone calls amounted close to the same volume, Exabytes. This estimate was based on ITU estimate of 2.5 * minutes per year for 207 countries extrapolated to 7.5 * minutes per year, or roughly 600,000 terabytes per month (with compression factor of 6 to 8). The ITU estimate would mean for the population of 7 billion individuals on earth that every one of them spoke ca. 360 min/year at phone. Subtracting those who are too young, too old, or too poor to be able to use a phone, ca. 2/3 of the population, the rest of the population would spend about 1100 min/year by speaking at phone in average. Assuming for the above voice stream no compression, and take into consideration that the connections are duplex, this would mean 960 kb/min data, i.e. 2.4* bytes uncompressed voice data in a year. In case we use unique titles as the measurement units, the largest category was photographs with 80 billion images and the second one was office documents with 7,5 billion titles accounting for 81% of all the printed material produced in the world. However, according to Davidson Consulting referred in /2/ about 400 billion pages per year are faxed annually (corresponding about 6,000 terabytes of fax data per year). It is worth of noticing that the above figures do not contain all data that was produced by autonomously operating machines, like satellites, terrestrial sensor networks (fire alarms, readings of electricity meters, elevators, vending machines, industrial process plants), car, aircraft or vessel control processors, etc. Some condensed excerpts of the information produced by these systems is presented at some point to humans. Most of this data could be transmitted wirelessly, and some can only be transmitted wirelessly. This has opened a new business area sometimes called Machine-to-Machine business (M2M). We do not treat in this context M2M area further, but want to just remind that is an area that produces more and more data every year, as more and more of analog control systems are replaced by digital ones and the number of various sensors and digital control units increases. 3. INFORMATION AND COMMUNICATION IN ENTERPISES Professional users in enterprises have been the primary target group for new information and communication technology due to their capability to adopt and invest on ICT equipment and services /10/. The adoption of new technology and services is easiest when the technology matches an existing need for communication /10/. From this viewpoint it is beneficial to analyze the communication in organizations and estimate the potential of new technology with respect to the analysis results. A recent case study performed in a high-tech organization /5/ estimated each employee of the organization to communicate about a hundred pages (92 pages) of information on an average working day. A Page was used as the measurement unit in order to unify the quantities of verbal communication, communication on paper, telephone calls and other ICT-mediated communication as well as communication using stored digital artifacts, such as digital documents and database records. Most of the communication was information received or forwarded while less than ten percent was newly created information sent or communicated to one or more receivers. Figure 1 represents the distribution of the communication according to four measures used in /5/ for each of the seven categories of mediated communication (see Figure 2). In Figure 1, the category Mediated/Semi transient refers to telephone calls and videoconferencing mediated through ICT, Analog refers to paper, VHS video etc., Digital image refers to faxes and bitmap images, Encoded refers to , office documents and other ASCII-based media, Semi-Structured refers to Lotus Notes applications and other combinations of metadata fields and encoded / bitmap data, Structured refers to XML messages and other data represented as tables, forms and other documents, and Formal refers to database information intended primarily for computer use. Figure describes the taxonomy of categories by graphical inclusion of the categories in their super-categories (adopted from /5/). Annual UI refers to number of Unique communication Instances, such as number of unique e- mails sent or received by the members of the organization in a year. Annual UI Volume refers to the volume of the content in previous, when measured in Pages, i.e. displayed content. Annual Copies and Annual Volume refer to the same figures, when including also the copies of, e.g. mail messages in the numbers and in the Pages. The figure 92 pages per person day refers to the last figures, in which about 40% of the communication goes through paper and other analog media. But this seems to be mostly due to large number of copies of paper documents if the copies are excluded, most of the unique information is communicated using semi-structured (31%) and encoded (28%) media, paper being the third with 22% of the total volume. Another case study in an university faculty results also in the same order of magnitude /6/. Figure 1. Distribution of communication forms in an enterprise for communication using medium (adopted from/5/). From our viewpoint, the distribution of communication presented in the first study is of most interest (see figure 1). When observing the total amount of communication (Annual Volume), the category Mediated/Semi-Transient communication form only 4 % of all the communication that uses some medium. The category Digital Images corresponds to fax and other bitmap images, that could be delivered using multimedia messaging (MMS), covers only about 2 % of the communication volume. Instead, the Encoded communication forms (corresponding to ASCII text in mail messages and office documents) accounted for 25 % of the volume. Further, Semi-Structured and Structured communication forms (matching with Formal Structured Semi-structured Encoded Digital image Digital medium Analogue medium Stored medium Mediated Using medium Group meetings One to One Face to face Material Database application e-business, XML Lotus Notes, , MS-Word Fax, image Paper letter, print, VHS tape, Telephone, VOX, videoconferencing Meetings,discussions Personal discussions Software, parts... Figure 2. Taxonomy of the categories of communication forms (adopted from/5/). XML, Lotus Notes applications and alike) form together almost 30 % of the volume. The rest of stored communication (41 %) was mediated using paper and other analog media Both this first case study and the second case study at the university faculty reported the proportion of paper communication going down rapidly. This implies that the new demand for ICT services will be driven from the technology applicable for replacing paper medium. The second study reported that communication on paper was a minor portion when measured with the number of communication actions and unique data, but was by far the largest communication form when measured by communication volume (see Figure 3). This seems contradictory at first glance, but is easily explained by the fact that the unique information on a sheet of paper is often copied to many people; in the case of course material the number of recipients might be hundreds of students. Assuming that the unit Page correlates with the amount of bytes to be communicated, this implies that the introduction of larger documents will also require more communication bandwith to be offered both by the internal network infrastructure and by the inter-organizational infrastructure than the previous digitization phases in organizations. This is because in earlier phases mostly text encoded with 7- or 8-bit ASCII was used, whereas now XML-encoded documents, as well as image, audio, and video data are in use. Transferring digital documents (that replace the paper documents) should require relatively limited band-with, if compared with video transmissions. Talking about videoconferences, they primarily replace face-to-face meetings and thus save peoples time and money that was spent while traveling and taking part in the meetings. A video conference can hardly replace distribution of digital documents, and hardly any document exchanges can replace a videoconference: During March-June 2003, when China and other Asian countries were hit by SARS, companies used a lot of video conferences as a replacement of face-to-face meetings, that would have required travel to/from SARS infected countries. The exchange of information was based on s, phone calls, faxes and other digitally encoded/formal documents, because these do not carry human viruses, unlike, perhaps, paper documents and 80,0 % 70,0 % 60,0 % 50,0 % 40,0 % 30,0 % 20,0 % UI U Pages Copies Copy Pages 10,0 % 0,0 % Non-stored Paper Digital document Information system because these are necessary as meeting material. During thethreat posed by SARS, companies having cooperation with SARS-infected countries increased their usage of networking infrastructure /11/. Figure 3. Distribution of communication in the university case study (adopted from [6]).Here the category Non-stored refers to all categories from material to mediated communication, Paper refers to analog communication media, Digital document refers to digital images, and encoded categories, and Information systems refers to the categories from semistructured to formal communication forms. Apart from special situations, like the SARS epidemic, the results from the Berkeley study /2/ referred to earlier suggest that video streaming will overwhelm digital document transmissions in volume also in enterprises. The reasons can only be guessed. Anyhow, the traffic patterns of delivering multi-megabyte documents (video clips or other massive documents) on demand from centralized document storages will probably have some impact on network optimization and charges. If the video streams are assumed to be played at the same time as delivered from the servers, end-to-end streaming guarantees should be provided by all the networks on the path from the server to the (mobile) client. Analysing the results in this section one can draw the following conclusions. First, any digitally encoded data is eligible to be communicated over a network. F
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks