Television signals can be transported in three ways: broadcast via radio waves using ground antenna， by the coaxial tree network of community antenna TV network or recently via a satellite， using the so-called direct broadcast system.a
In the private domain， computer data are mainly transported by LANs （Local Area Network）. The most famous ones are Ethernet， token bus and token ring （IEEE 802 series）.e
Each of these networks was specially designed for that specific service and is often not at all applicable to transport another service. For instance， the original CATV networks did not allow rite transportation of POTS; or the PSTN does not transport TV signals; or the transfer of voice over an X.25 network is very problematic because of too large end-to-end delay and jitter on this delay.
Only in limited and special cases can service types other than the one the network was originally designed for be transported over it. This is for instance the case for the PSTN which is capable of transporting computer data at a limited speed， if modems are provided at both ends of the network.
An important consequence of this service specialization is the existence of a large number of often world-wide independent networks， each requiring its own design phase， manufacturing and maintenance. In addition， the dimensioning of each network must be done for every individual service type. Even if resources are freely available in one network， they cannot be used by another service type. For example， the peak hours in the telephone network are between 9 a.m. and 5 p.m.， whereas the peak hours in the CATV network are during evening. Since resource pooling is impossible each network must be dimensioned for its worst case traffic conditions which is the peak hour traffic.
A first step， albeit a limited one towards a single universal network， is the introduction of NISDN in which voice and data are transported over a single medium. This network cannot transport TV signals due to its limited bandwidth capabilities， so a special TV network is still required. Even in NISDN the integration of narrowband services such as data and voice can be considered as being rather limited: the user access to the network is fully integrated， either by a basic access or primary rate interface. However， inside the network there will still exist for some time a packet switched and a circuit switched network as two overlay networks incapable of transporting other traffic types and each dimensioned either for voice or X.25 data.
Another important consequence of this service specialization is the inability of the network to benefit highly from the progress made in technology and coding algorithms. For instance， current digital NISDN switches are designed for 64 K-bit/s voice channels. However， with the current progress in speech coding and chip technology， bit rates of 32 K-bit/s， 13 K-bit/s and even lower will be used in the future. The existing switches and transmission systems arc not directly suited and thus need an adaptation， or will not efficiently use their internal resources for these lower speed bit rates.
When designing the future BISDN network， one must take into account all possible existing and future services. Suppose a network is capable of transporting a specific service， e. g. a circuit switched service with a channel rate of 70 M-bit/s. Suppose also that it is specifically designed to transport this bit rate. Some years later a new teleservice of， for example， 40 M-bit/s appears on the scene. This would mean that the network designed for that service （ i. e. 70 M-bit/s） will be capable of transporting the new teleservice， but with a large inefficiency: only 40 out of the 70 M-bit/s available will be used. This example is not unrealistic. It is very likely that in the future new services will emerge which have not yet been identified， and of which the requirements are unknown today.
As can be concluded from the above examples， the networks of today are very specialized and suffer from a large number of disadvantages， the most important being:
·Service Dependence
Each network is only capable of transporting one specific service for which it was intentionally designed. Only in a limited number of cases and by using additional equipment （e.g. a modem） and with an inefficient use of its resources can it be adapted to other services.
·Inflexibility
Advances in audio， video and speech coding and compression algorithms and progress in Very Large Systems Integration （VLSI） technology influence the bit rate generated by a certain service and thus change the service requirements for the network. In the future， new services with unknown requirements will appear. For the time being it is yet unclear， e.g. what the requirements in terms of bit rate for HDTV will be. A specialized network has great difficulties in adapting to changing or new service requirements.
·Inefficiency
The internal available resources are used inefficiently. Resources which are available in one network cannot be made available to other networks.
Taking into account all these considerations on flexibility， service dependence and resource usage， it is consequently very important in the future that only a single network exists and that this network of the future （B-ISDN） is service-independent. This implies a single network capable of transporting all services， sharing all its available resources between the different services.
A single service-independent network will not suffer from the disadvantages described above， but it will have the following main advantages:
·Flexible and future-safe
Advances in the state of the art of coding algorithms and VLSI technology may reduce the bandwidth of existing teleservices. A network capable of transporting all types of services will be able to adapt itself to changing or new needs.
·Efficient in the use of its available resources