Nomadism means a user being able to get through to his communication and information retrieval services from different physical accesses, whether they be different terminals with different identifications and capabilities or different network access interfaces and local loop operators. Yet nomadism, unlike mobility, is not really concerned with the continuity of an access or communication session. By combining mobility and nomadism it would become possible to suspend and resume communication sessions ( i.e. with the same context of communication and the same initial identification) on different access sessions.
To this extent, mobile networks already offer the user a combination of mobility and nomadism through the possibility of establishing an access on the networks of different operators ( usually those having roaming agreements) and of accessing his own services. In the standardization of UMTS, there is a definition of the concept of ‘VHE’, or Virtual Home Environment, that has the objective of giving the user a consistent impression of the execution, the presentation and the management of his services, no matter where he may be located, no matter what type of access and terminal he is using and no matter which network is being visited. Implementing mobility and nomadism requires a combination of control of services, networks and resources according to the type of session that needs to be established and maintained, and will usually require the co-operation of several different actors.
Firstly, the relationships between different addresses will have to be established and developed, implying the designation of the user, connection to a terminal and connection to a network access. Any matches between addresses that may be made inside each of these domains will have to be ignored, and the fact that these addresses that may may be made inside each of these domains will have to be ignored, and the fact that these addresses may be under the control of different operators must be taken into account Next, the information must be found and exchanged between these domains and between operators, making sure that it is not attached to a physical location, as was the case in the PSTN.
- Metering and monitoring
Making resources available is a part of the contract between the actors, and it is therefore necessary to check that these contracts are being respected. Providing means to control and count the resources used in an access or service session is the business of service and network control, and the existence of this metering and monitoring has strong consequences for the service and network architecture.
The distribution of control and the possibility of establishing different qualities of service at the same interfaces seriously complicate the implementation of this monitoring and metering as well as the possibility of forwarding the related data without any distortion up to contractual points of reference. In the conventional circuit or packet networks, local contractual metering was considered sufficient, as long as what passed for a contract could be deduced from what actually happened on that access.
In the simplest cases, the local exchange of a PSTN measures the duration of the communications, and billing then depends on this duration and on the number dialed. The circuit-mode structure of the network automatically ensures that the service contract is respected and no other verification is necessary.
The situation became more complicated once rerouting mechanisms were introduced and the metering was effectively distributed. Signalling was used to trace the metering information back to the origin of the call, but disruptions can upset this process, and can lead to the need to abandon the service.
In routed packet (IP) networks, it is also at access that either the time or the quantity of packets transmitted can be measured , but it is very difficult to check that an end-to-end quality of service contract has been strictly respected, and it comes back to ensuring beforehand, through the resource control, that the architecture will guarantee that the contract is respected In fact, it is easier today to establish contracts making available large capacities of resources between network or service operators than to verify the dynamic and distant consumption of such resources by individual users. This situation may be improved by distributed metering and post hoc verification or payment, but the major difficulty, between integrating control and metering, remains the generalization of pre-payment possibilities to all types of services.
- terminals in the control
A further feature of NGN will be the very great variety of terminals with widely differing characteristics ( type of codec, window size, storage and display capacities, etc.) that must be known to applications servers, but also that must be accessible to the network control so that the service required can be adapted to the terminal used. This may require the use of intermediate transcoding devices , for example, between communicating terminals. These terminals will be in part, manageable by or via the network , Software could be downloaded onto them and updated, data could be synchronized etc. This will not be developed in the short-term, as it has as yet rarely been put into practice, but it seems sure that terminals equipped with everincreasing capacities ( for processing and storage) will be integrated totally into the handling of network and service control functions.
The traditional fixed telecommunications networks operate with a very high level of quality, and this value is noticed by users. Their handling of emergency calls is viewed with a high degree of confidence and they offer relatively good protection against intrusions. The is not the case with the Internet, where users do not always find their services are available and are sometimes victims of unwelcome, or even malicious intrusions. It is, therefore, of particular importance to bring the same performance levels of these conventional networks, in terms of availability and security, to packet-mode networks like the NGN. Network control through its many links with a very diversified environment ( of users, other networks, serevice providers, etc.) will obviously be an active participant in this process.
User Profile and Customer profile
The implementation of all the functions mentioned above requires the management of data related to terminals, users and contracts, if the required communication sessions are to be set up. Instrumental to this is the concept of user profile, linked to the activation of services, and that of customer profile linked to contracts and payment. The customer profile can be defined in abstract terms as the set of customer-related data that correspond to the ‘User Agent’ as defined in TINA and the user profile as the set of user-related data, whether it be for access to communication services or for the use of these services, Several user profiles can fit a single customer. For example, in a company or in a family there are usually several service users for a single customer. Moreover, several user profiles can fit a single person. These are often related to different customer profiles, often concerning the contract or to different types of access concerning usage ( e.g. mobile and ADSL access for the same person).