A Conceptual Framework To Integrate Mobility Into Isdn

Transcript

A Conceptual Framework to integrate Mobility into ISDN J.-P. Ebert, B. Rathke and A.Wolisz# Institute of Telecommunication, Department of ElectricalTechnical University Berlin, 10587 Berlin, Germany # also GMD Fokus Hardenbergplatz 2, 10623 Berlin, Germany Abstract: We start with a general discussion clarifying the different meanings of the frequently misused term ‘mobility’. Following we consider how different types of mobility could be introduced into ISDN, focusing especially on a constraint mobility approach where we provide a wireless counterpart to the usual ISDN access interfaces. Both the technical issues of such approach and possible application scenarios are discussed. I. INTRODUCTION One of the most important trends in communication is the ongoing development to support different kinds of application - data transfer, voice, picture, video up to multiparty multimedia videoconferencing and collaborative work. Probably the most advanced approach to solve this problem is ISDN (Integrated Services Digital Networks). The importance of the ISDN concept results from several features: - Transmission according either to the packet switch principle or with circuit switching emulation, is very attractive for isochronous applications. Via combination of multiple channels for a single application the data rates can be adjusted step by step to the needs of the application, in addition the emerging broadband ISDN offers even higher data rates. - Beyond of the plain transmission services, there are several network services being stepwise introduced, like conferencing, call forwarding, automatic call back, mass calling etc. In fact ISDN seems to be one of the first realistic polygons for the practical introduction of the Intelligent network concept. - ISDN is widely proliferated. In twenty european countries, in the US and in Japan ISDN services are progressively offered to subscribers (at lower rates) almost as easily as to telephone services, at affordable rates. - A further advantage is the technical maturity and availability of low cost components for end-system integration. In consequence end-system costs are low. Both the concepts and the technical solutions for ISDN are already well understood (some of them are even classics), a recent point of view on the significance and perspectives of ISDN is presented in [1]. The other important trend in communication is mobility. Braking the old paradigm of only being able to communicate in certain, predetermined locations certainly opens a new level of user reachability, but in addition makes it possible to develop new scenarios for using the communication facilities. Surprisingly the introduction of mobility in communication systems resembles closely the history of communication systems itself. Historically, the oldest proliferated communication service was the telephone - in fact long enough data networks have been built using modems for supporting transmission of data over telephone lines. Only later the LANs appeared, and significantly later a pure wide area data communication and eventually integrated service communication became reality. Similar to that, the only service supporting mobility on a large scale today is mobile telephone service. And of course it is attempted to support low speed data communication via the mobile telephone technology. On the other hand, the wireless Local Area Networks appear and proliferate with a high impetus, while the development of the high capacity universal mobile communication is still in its infancy. The problem addressed in this paper of introducing mobility into the ISDN services is without doubt one of the major challenges in todays communication. Our discussion will be structured as follows: first we present a discussion and taxonomy for the frequently misleading concept of mo- bility (section II). Following we present general considerations of the major approaches to introduce mobility in ISDN, and justify the choice of a constraint mobility approach, supporting in-house mobility around the ISDN access points, as our main topic. Furtheron, after reminding some relevant facts on both the ISDN and wireless LANs (WLAN) (section III) we present the discussion of technical problems that have to be solved on the way to the considered solution (section V). Finally, we describe few examples of applications which could be developed on top of these new communication services. II. TAXONOMY OF MOBILITY ASPECTS Generally mobility means the possibility to use the communication services independently of the user location. In fact we believe that there are three aspects which should be considered while approaching the problem of mobility support: A. Mobility - Bindings of communication end-system to the stationary network endpoints Mobility in this case means the grade of possible displacement of communication end-systems like telephones, PDAs, computers etc. with respect to the network access point e.g. the telephone or Ethernet wall socket. To define a given grade of mobility we are interested only in the existence of a transmission path that is sufficiently reliable between the identified components. In principle there are four grades of mobility: • • • • none, elementary, local and global. The first grade stands for the total lack of mobility like it is the case for mainframes or other systems with fixed installations. This is also the case for end-systems which are non mobile due to weight or a fixed cabling installation (cable shafts). Secondly, we can identify elementary mobility whereas the end-systems movements (while continuing the communication) are restricted to the length of the flexible cable or the range of the wireless connection with a single base station placed directly at the fixed network access point. As an example for this one can see the common home or cordless telephones, whose area of connectivity is limited to the cable length or reachability of the communication of the end-system with its base station. The crucial aspect is the private nature of this communication, the only restriction is that it should not cause any adverse influences on people and other equipment. This is precisely the reason for the proliferation of non-network-provider-accepted cordless phones. Extending further the mobility leads us to the local mode. In this case, it is possible for the end-system to move in some area belonging to a single owner/user. Such kind of mobility, restricted for example to big rooms, floors or buildings is realistic in fact only for a wireless network composed of multiple pico-cells. In this case the problems of roaming, handover etc. occur for the first time. On the other hand, the communication between the end-station and the network access point can still be considered as private in nature. Finally, a fourth variant lets us focus on the global mobility. By definition there is no restriction to the location of the communication end-system, the end-system to network access point communication is public in nature, as it covers areas belonging to different owners/users. Also the placement of end-systems (say their grouping or dispersing) normally cannot be influenced. Thus the kind of transmission has to be controlled in a much more rigid way. Global mobility, available only via wireless communication, takes place in the case of mobile telephone or satellite communication. The covering area - e.g. a district, a country or a continent is considerably larger than in the case of local mobility. The networks which support this kind of mobility could consist for example of a backbone and a number of radio cells. B. End-system addressing and routing information To supporting end-system mobility it is usually not enough just to assure its mobility, in the sense explained in the previous section, i.e. the technical possibility to transmit/receive information from the end-system. This would be the case with passive broadcast systems like radio receiver. In order to be identifiable as a required source or sink of information, the end-system has to obtain a unique address within its networking environment. The whole address scope in any network can be structured in non-hierarchical, or hierarchical fashion. A typical example for non-hierarchical address are usually hardware addresses (e.g. Ethernet addresses). This means that the end-systems can be uniquely identified in any circumstances, e.g. the Ethernet card has a worldwide unique identifier, and thus can be inserted in any computing equipment without any address adjustment. The identifier includes no information concerning the location of the end-system, which will have to be provided in addition. The other approach is followed in the internet where the whole address space is hierarchically structured into In order to communicate with a uniquely identified endsystem or group of end-systems an efficient way of routing in a mobile environment has to be found, i.e. storing and updating the information which transmission path has to be used to reach the selected, perhaps displaced end-system. This is normally equivalent to the knowledge of their recent location. The two basic approaches are routing over the home location, usually used in connection with hierarchical addressing (e.g. tunnelling in mobile IP, [2]) and updating location data banks pointing to the end-system, usually used in connection with non-hierarchical addressing. In both cases mobility aspects add significantly to the complexity of the system. C. User addressing In a classic approach users are identified only by the endsystem addresses. We can illustrate this by recalling the fixed telephone numbers for each subscriber line. End-system access security is required to avoid misuse of the endsystem (e.g. the phone owner has to pay the bills for his neighbours kids calls to New Zealand while there were playing with his kids). Mobility can increase the complexity of this access security, as the mobile end-system will be possibly accessible by a larger amount of persons, under different, possibly not always easy to control circumstances. On the other hand mobility not necessarily has to be provided via end-system mobility. Alternatively it could be supported via user switching among easily available endsystems (e.g. coin-workstation on the corner). In this case, it is fairly easy for the user to be an active party in the communication. Much more complicated is the problem of assuring the end-user reachability, as contrary to the endsystem reachability addressed in the previous section. In the case of fixed end-user to end-system assignment, if the user departs from his end-system, there is no possibility to reach In the second case the estimation of the end-system positioned closest to the actual user location is required. This can be achieved in different ways. In the solution coming from the data communication, the user has be active (by a log in procedure, or simplified by inserting a keycard) to establish his use of the end-system (e.g. while going to another office). Calls will be routed directly to this end-system. In another solution coming from the telephone the user specifies in advance a list of end-system addresses under which he could be reached. The communication system tries to reach him by a sequence of call attempts. If an attempt fails the system uses the next entry in the list. Such a mechanism is used in Intelligent Networks (IN). Finally the continuous update of the user location can be supported by automatic tracing facilities. The location of users will be recognised and an incoming call will be automatically routed to the next available communication system. The recognition can be done for instance by local or global positioning systems (LPS, GPS; [3], [4]). Fig. 1 shows the taxonomy described above. mobility global local elementary none end-system addressing passive call notification active location update user addressing Fig 1. hierarchical Assuring the unique address, however, is still only one step towards supporting mobile communication. Especially in hierarchical environments the address had to be changed upon end-system displacement, because of the necessity to change its location information. The new address in general is a guest address provided in the new location. In non-hierarchical case the address remain the same and only the separate location information varies. him. There are two basic possibilities to overcome this problem: the passive call notification and the active location update. In the first case user is informed by means of using an mobile end-system for a separate simpler distribution network (mass medias - TV, radio broadcast) or individual calls (Europeeper), that a communication request with him has occurred, perhaps with an additional information which end-system (or which user) seeks the communication. After receiving such notification it is up to the user to get access to any communication end-system (e.g. the telephone cell) and become active. non-hierarchical several domains, each of them being divided into networks, subnetworks etc., and each IP address belongs to a certain subnetwork. Such address contains both, the location and the end-system identifier within the (sub)network. Mobility taxonomy III. ISDN AND MOBILITY In this section we will identify and discuss different variants of integrating mobility into ISDN. Let us first consider in detail the relevant kinds of mobility, namely the global, local and elementary mobility. Global mobility follows the idea of developing a cellular network for mobile telephone. This concepts guarantees an overall reachability and implies an active location update with automatic tracing capabilities of end-systems. Existing networks like the well known cellular telephone networks are mainly designed for audio communication. Recently a lot of efforts has been put into supporting other services like data transmission, for instance circuit switched cellular, packet-over-circuit-technology (CDPD) or private packet. Nevertheless the development of cellular networks beyond private packet networks is mainly driven by the effort to provide low cost and miniaturised solutions for the classical speech transmission, which will become very popular for millions of end-users. Therefore the support for data and multimedia services in this approach will always remain of secondary importance (except of private packet networks) an extension if possible. Neither of these networks is able to support isochronous services with a transmission rate of at least 64 kbit/s. Only future technologies like Universal Mobile Telecommunication Systems (UMTS) will be able to offer such services [5]. The other alternative follows from the observation, that there are enough customers, whose mobility while using a communication service is restricted each time to the inhouse or factory area (elementary or local mobility). Customers, who need such restricted mobility during the use of data communication services can use a growing number of WLAN products which will offer high performance communication services in the near future. A lot of WLAN offer asynchronous service capabilities, newer ones will extend this by time bounded services. approach with the WLANs. The advantages of ISDN have been discussed in the introduction. In addition recently the use of multiple 64 kbit/s ISDN-channels opened by the widely accepted de facto standard of BONDING (Bandwidth ON Demand INteroperability Group [7]) allows the provision of n*64 kbit/s to the end-user without concern for new costly dedicated infrastructure. The use of B-ISDN based on the Asynchronous Transfer Mode (ATM) will appear fairly quickly as another option. On the other hand wireless indoor approaches offer high data rates (about 10 Mbit/s), availability, low installation and maintenance costs, etc. for the limited coverage and seem to be an ideal solution to offer limited mobility around the ISDN access point. Looking from the other point of view one could argue, that in this way the limited mobility offered by individual cells of WLAN or by the whole LAN will in some time become globalised via the use of ISDN. We could describe the result as semi-global in this sense that providing a dense coverage with ISDN end-points and a restricted mobility around each of them creates a virtual global wireless network (Fig. 2), an illusion of being totally mobile. Virtual global mobility differs from global mobility in the point, that rather than offering a complete reachability coverage there are a large number of islands, and user mobility is supported only within each of these islands. Besides the relatively easy development (see the next section) such approach has numerous other advantages. Let us mention just two. The constraint mobility approach leads to much smaller electro-smog. In fact, in comparison with the global mobility the transmission in the restricted case needs much lower power per end-system, as the transmission distances are much smaller. In addition to that the “islands” aregeneral much smaller than the cells in solutions supporting the “global” mobility, thus not only the emission of local local mobility mobility WLAN We believe, that recently introduced technologies for cellular networks (e.g. cellular telephone, private packet networks) offering global mobility are not able to support the requirements for high performance services. It seems to be easier to obtain mobility with high performance services in a local mobility environment by means ofWLANs than in a global environment (e.g. Mobile Broadband Systems - MBS, [6]). Our research is focused to introduce a limited grade of end-system mobility into ISDN by creating a merger of this GSM Cell virtual global mobility virtual global mobility ISDN interconnection network a) Fig 2. b) Mobility concepts a) area of global mobility b) area of virtual global mobility individual end-system is lower but also the number of endsystems whose emissions overlap is significantly smaller. The virtual global mobility approach seems to be attractive also from the viewpoint of security, as the broadcasted, wireless communication takes place only in a restricted area. quality multimedia applications. As the matter of fact this seems at the moment the easiest way to access a bandwidth of up to 2 Mbits/s. TE1 NT2 A. ISDN In Fig. 3 the structure of ISDN is presented by the ISDNreference model [8]. Let us focus on the S reference point, with the two standardised interfaces S0 and S2M. The S0-interface specifies in fact a bus, which consists of two unidirectional broadcast paths. This bus offers two B-channels, each with a capacity of 64 kbit/s, and one D-channel for outband signalling with a data rate of 16 kbit/s. This bus, designed to support up to 8 Terminal Equipment (TE) units, could essentially be used in two ways. TEs could use the Dchannel both for sending packetised information and for signalling. Via this signalling a B-channel could have been reserved for exclusive use by a TE, in an isochronous mode. Accessing the D-channel has been controlled by a special multiple access protocol, designed according to the following principle: A TE had to observe an Echo-channel (Echannel), which is reflecting the D-channel. The idle state of the E-channel enables TEs to send D-channel data. Due to the applied signal coding, possible bit sequences from different, simultaneously sending TEs are logically ‘OR’ed. While sending each TE has to sense the E-channel and step back from sending, if the received echo is different fromthe sent data. The S2M-interface offers 30 B-channels and one D-channel. Each of them has a capacity of 64 kbit/s. The S2M-interface has been developed mainly with the idea of connecting (in a point-to-point mode) a PABX to a switch. This PABX should internally make the S-interfaces available to the TEs. The S2M-interface, however, can also be used in a different way, in order to access a group of 2 up to 30 B-channels. In fact a fairly popular solution BONDING [7] supports multiplexing of B-channels, thus offering a bandwidth up to 2 Mbits/s, which is quite enough for medium ET V NT12 TE2 TA ET R In this paragraph we will briefly review selected aspects of ISDN and WLANs which we consider of relevance for the integration of these technologies. LT T In the following sections we will present some of the technical problems occurring within such an approach. We decouple the aspect of user addressing from the end-system addressing, because this can be considered independently. IV. RELEVANT ASPECTS OF ISDN AND WLANS NT1 S RSTUV - Reference Points TE 1 (2) - Terminal Equipment Typ1 (2) NT 1 (2) - Network Termination1 (2) Fig 3. U NT 12 - Network Termination 12 - Exchange Termination ET - Line Termination LT ISDN-reference model B. WLAN The WLAN structure is similar to cellular telephone networks. The basic component is a cell (called pico-cell) build around a dedicated station called Access Point (AP). The AP synchronises the operation of the end-stations within the cell, it can usually act as well as a gateway to a backbone network and in some cases as a switching point. Recently techniques have been developed for arranging cells of endsystems without the central station - implementing fully distributed control. A WLAN cell can cover an area from a few meters up to several hundred meters with transmission rates of up to 5 Mbit/s, a small number offer around 10 Mbit/s [9]. The data may be transmitted in one or more channels available in multiple disjoint frequency bands. We will concentrate on the Medium Access Control (MAC) protocol as the main issue in coupling ISDN and WLANs. The radio cell can be considered as a bus. The majority of suggested MAC protocols for WLANs follow some modifications of the basic non-persistent CSMA principle. These kind of protocols have a high throughput efficiency and low access delays. The additional modifications assure robustness against high load situations. Newer proposals provide also support of time bounded services. Some proposals follow the TDMA access methods or polling methods which have better possibilities to offer time bounded services as CSMA protocols [10]. If compared to wired-environment oriented counterparts wireless MAC protocols usually have additional mechanisms in order to assure effective handling of high error rates, frequent change of network configuration, overlapping of subnets and security aspects. V. TECHNICAL CONCEPTS OF VIRTUAL GLOBAL MOBILITY ISDN ACCESS Our principle idea to introduce the virtual global mobility is to use WLANs pico-cells in order to replace the classical ISDN distribution system i.e. the S0-bus. A. WLAN pico-cell as ISDN S-bus The AP of the WLAN acts as the network termination (NT) of the ISDN (see Fig. 5). The mobile stations of a WLAN pico-cell are now ISDN-TEs.There are three aspects which have to be considered: • coupling of WLANs and ISDN access protocol, • provision of time bounded services, • required bandwidth of ISDN. 1) MAC coupling To couple the ISDN and the WLAN access methods there are the following possibilities. The first one is to extend ISDN to a new wireless ISDN access method to take over the radio part (e.g. described in [11]). Another alternative is to use an existing wireless MAC protocol to transmit ISDN data including signalling. In this case ISDN is embedded into a pico-cell of a WLAN. Two steps are necessary if we are considering data transfer from NT to TE: An encapsulation on the NT site, for instance using a Network Termination-Wireless (NT-W1), and an unwrapping on the TE site using a (NT-W2) and vice versa for the other direction. This could be done by converters located inside or outside of the ISDN-TE and ISDN-NT. Fig. 4 shows the general structure of MAC coupling using an encapsulation method. We believe that the second solution should be preferred, since this solution does not need a modification of the NT-W2 NT-W2 ISDN NT S0 NT-W1 NT-W2 S0 S0 S0 TE TE TE TE radio waves WLAN cell Fig 4. Coupling ISDN and WLAN NT-W wireless network terminatio ISDN. Several of the existing WLANs offer transmission quality sufficient to support ISDN. Furthermore it is also possible to connect non-ISDN-TEs directly to the WLAN besides of the ISDN-TEs. They build a logically separate wireless network besides the wireless ISDN. Integrating ISDN functionality (ISDN layer 2 and 3) into a non-ISDNTE which is also possible leads to a new type of ISDN-TE (wireless ISDN-TE). ISDN offers a synchronous service to transmit data from one endpoint to another. This means that this service has to be maintained over the transit media WLAN. The wireless MAC protocol needs time bounded service (TBS) capabilities like for instance suggested in the IEEE 802.11 draft standard DFWMAC [12] to support all ISDN capabilities (data, audio, video). If the wireless MAC protocol does not support TBS, transmission of synchronous ISDN data is limited [13]. The Quality of Service (QoS) in this case results from packet losses, access delays and channel utilisation [14]. The main problem is the channel utilisation, resulting in increasing access delays. Especially WLANS are characterised by high error rates. As an effect, packets are lost or have to be retransmitted, which leads to increasing access delays, channel utilisation as well as to decreasing QoS. There are various approaches to solve this problem, i.e. development of robust coding algorithms so that quality does not severly degrade under adverse environment condition or expansion of the WLAN concept which gracefully encompasses several data services disturbing the ISDN connections. 2) Bandwidth requirements Basically the bandwidth demands for each direction of ISDN are 192 kbit/s, which contain the two B-channels (2*64kbit/s) and the D-channel (16 kbit/s) plus control information (48 kbit/s). The netto data rate is reduced to 144 kbit/s. Applications requiring more than the two B-channels can access the primary rate access interface S2M with 2048 kbit/s for each direction including control information. Let us discuss the worst case in coupling the ISDN and the WLAN. Two aspects have to be considered: The bandwidth in the WLAN has to be at least 288 kbit/s or 3968 kbit/s (two direction transmission speed) for the S0 or S2Minterface with respect to the two transmission links of the ISDN S-interface. The needed transmission rates are considered without control and synchronisation information of the ISDN. This function can be provided by converters (NTW) at the edges of the WLAN. Actually the needed WLAN bandwidth is a multiple of the ISDN two direction transmis- sion speed, because as we can see in Fig. 4 in the case of TE to NT communication multiple S0-busses (one from each TE) feed the NT-W1 S0 channel. That means the bus configuration is converted into several physical point-to-point links. If, for example, eight TEs are attached to the WLAN the needed bandwidth will be at least 1296 kbit/s. We believe, that, due to an intelligent implementation of coupling, the bandwidth requirements can be significantly reduced. This is presently under further studies. According to the ISDN standard the WLAN bandwidth has to be sufficient for maximal eight ISDN-TEs. Futhermore using the NT-W1 as a PABX it is possible to attach more than eight ISDN-TEs. The maximum number of connectable ISDN-TEs is limited by the WLAN bandwidth. B. Structures for the wireless ISDN extension In Fig. 5a the basic ISDN mobility extension is given. One S0-bus is replaced by one pico-cell of a WLAN which leads to an elementary mobility for the end-user. It may also be possible to build up the logical S0-bus on the base of several pico-cells to achieve local mobility like it is given in Fig. 5b. The impact of such an configuration is the possibility to offer a wireless extension with one S0-bus in more than one separate location (e.g. several rooms in a building). In this case handover functions within the WLAN are needed. In addition a kind of active location update to locate the ISDN subscriber should be provided. We will draw a few examples to give an touch of possible application fields. ISDN telephone ISDN fax ISDN X.21 VI. APPLICATIONS The integration of virtual mobility into ISDN offers numerous advantages. First of all the standard ISDN services (telephone, facsimile, data transmission), can be made available to end-systems which are able to move with respect to constraints imposed by the Wireless LAN technology. In addition, totally new applications are possible, e.g. in the field of personal communication and computing, computer aided teaching and computer aided conferencing. ISDN AP(NT) ISDN communication socket a) ISDN telephone ISDN fax ISDN X.21 ISDN X.25 adapter communication socket b) AP(NT) ISDN Interconnection Network AP(NT) ISDN c) logical S0-bus WLAN cell Fig 5. Assume there are several such networks, as shown in Fig. 5c, building local or virtual global mobility and let us consider a customer who changes between radio cells which are connected to different S0-busses. There are two possibilities to implement the handover functions. First, the handover must take place by means of ISDN functions and inside the WLAN. In this case an extension of the ISDN standards is necessary. Second, the handover function can only be performed by the WLANs due to signalling between them to maintain the ISDN connection. AP (NT) ISDN X.25 adapter AP radio waves Access Point Structure of wireless ISDN extension The medical example: The doctor can take the documents, X-ray images, etc. and use ISDN (or 2 Mbit/s using the BONDING concept) connections to consult other, more experienced specialists. In the local mobility case, he can move with his equipment to the bed of the patient, to the operation room, to the emergency service room, to any place in the hospital, and from this place, close by the patient, contact by ISDN any external service, consultants etc. Another example can be taken from the area of computer aided teaching. A group of mechanics should be introduced to a new version of an engine. This engine is produced e.g. in Japan where also the developer are located. The mechanics belonging to a service station outside of Japan, perhaps in Australia. Normally the group of mechanics or the developer has to travel, which in both cases is expensive. This is not necessary if they are using wireless ISDN. Using a multimedia application the developing engineer can present an equivalent engine to the remote mechanics. Due to the wireless aspect he is free to move, showing details of the engine and demonstrating the realisation of repairs. The wireless extension offers the possibility to be flexible without being hindered from cables and there are no constraints to a certain camera position. Now let us consider a service man who is repairing perhaps cable shafts, water or electro installations. For example he got an order to switch some connections and he is looking into a distribution box and sees many plugs and cable systems. He does not recognise clearly, which connections have to be switched, because a colleague of him has changed the installation infrastructure during a previous reconfiguration. An interesting possibility is taking a laptop or PDA connected via a WLAN to the next ISDN wall socket and remaining on site to communicate with a remote database server in his parent firm.The server contain the current scheme of the distribution boxes. With the information obtained from the database, he is now able to do his work correctly. Focusing on our own problems: A group of scientists working on an common international project have an editing meeting for the deliverable. This can take place on one or maybe even several sites. Everybody is equipped with a laptop with PCMCIA interface. At the meeting site everybody obtains a PCMCIA WLAN interface card and a floppy with proper drivers and software, supporting joint editing etc. on an ad hoc WLAN. In addition if the laptops are ISDN-capable they may obtain the additional box for wireless ISDN - and access lines to get data, pictures etc. from their servers at home or conference particpicants in different sites. There is no interference with the networks on the site, stable guaranteed data rates are supported due to the nonoverlapping infrastructure for both applications. VIII. REFERENCES [1] N. Dagdeviren, J.A. Newell, L.A. Spindel and M.J. Stefanick, “Global Networking with ISDN”, IEEE Communications Magazine, Vol. 32, No. 6, June 1994 [2] IP Mobility Support, Internet Draft, Network Working Group, May 1994 [3] A. Harter and A. Hopper, “A Distributed Location System forActive Office”, Special Issue IEEE Network on Distributed Systems for Telecommunications, January 1994 [4] J. Deutrich, “Ortung und Navigation mit dem GlobalSystem (GPS), telekom praxis, Band 68, November 1991 [5] H. de Boer, “RACE mobile communications”, Electronics & Communication Engineering Journal, June 1993 [6] M. Chelouche and A. Plattner, “Mobile Broadband Systems (MBS): Trends and Impact on 60 GHz band MMIC Development”, Electronics & Communication Engineering Journal, June 1993 [7] Interoperability Requirements for Nx56/64 kbit/s Calls, version 1.0, BONDING Consortium, September 1992 [8] CCIITT Recommendations I.310-I.470, Genf 1989 [9] C.J. Mathias, “New LAN Gear Snaps Unseen Desktop Chains”, Data Communications, Vol. 23, No. 5, March 1994 [10] I, Chlamtac and A. Faragó, “Making Transmissionmune to Topology Changes in Multi-Hop Packet Radio Networks”, IEEE/ACM Transactions on Networking, Vol. 2, No. 1, February 1994 [11] P. Papantoni-Kazakos, “Accessing Protocols and Network Reconfigurations for Cellular - ISDN Systems”, Proceedings of ICCC´94, 1994 [12] W. Diepstraten, G. Ennis and P. Belanger, “DFWMAC - Distributed Foundation Wireless Access Control”, IEEE 802.11-93/190, November 1993 [13] P. Noll, V. Leesemann and G. Wessels, “Packetized Voice Transmission”, DFVLR-Mitteilungen 86-05, May 1986 [14] B. Gold, “Digital Speech Networks”, IEEE Proceedings, Vol. 65, 1977 VII. CONCLUSIONS We have discussed various possibilities to extend ISDN with mobility in order to offer improvements in the existing ISDN services as well as new applications of the well established ISDN. Basic concepts are introduced concerning a framework to virtually mobile ISDN. Even though we limited ourselves in the technical part to the case of classical ISDN, these concepts are equally relevant to the broadband ISDN. In fact the main bottleneck will be the effective throughput reachable with WLANs.