7.2 The Addressing Challenge
| In PSTN, a user is typically identified by a single identifier called a telephone number (e.g., an ISDN number). The ISDN number is a string of digits that can be dialed on any phone in order to reach the user. The ISDN number is used for two purposes: as a public address to reach the user and by telecom operators as an identifier for the user (e.g., for billing purposes). The deployment of cellular networks has introduced a distinction between the public address and the private address of a user. The public address of a mobile user, also called the mobile number, is the string of digits used by any calling party to reach the mobile user. In general, in cellular networks the mobile number is not used as a private address to identify the user. As a matter of fact, in several cellular networks a mobile user can have multiple mobile numbers (e.g., for different services such as voice, data, and fax). Traditionally, the private address of a mobile user (also referred to as the mobile user identity) is globally unique, and a mobile user is assigned only one private address. The private address is used to identify the user for accounting reasons, for authenticating the user when accessing the networks, and in general for every purpose where the mobile user must be uniquely identified. Cellular networks have introduced the need to identify separately the mobile user and the terminal the mobile user is using. The reason for the separation is that mobile users need to be able to use multiple terminals and switch from one terminal to another without needing to change the public address. This applies to the case where a mobile user desires to switch to a new mobile phone while maintaining the same subscription and the same public address. Moreover, a mobile phone may get lost or stolen, and mobile users need to be able to maintain their mobile number while using a new terminal. The adoption of IP in wireless networks, and in particular in cellular networks, introduces a new set of issues regarding addressing. In this section we describe the issues related to addressing in IP wireless networks. 7.2.1 Addressing in Current Cellular NetworksIn current cellular networks (e.g., GSM, US TDMA [IS-136], US CDMA [IS-95]), a set of different identifiers is used to address a mobile node. In GSM networks, addressing is based on three identifiers:
In GSM networks, a mobile user is given a SIM (subscriber identification module) card that can be inserted into the GSM mobile terminal. It contains the mobile user IMSI and MSISDN, as well as other parameters for security and for the mobile services the user has subscribed to. The adoption of the IMSI and the SIM card allows GSM mobile users to use different mobile terminals by simply moving the SIM card from one terminal to another one, without the need to inform the mobile operator of the change of terminal. In IS-41 networks (e.g., IS-136 and IS-95 cellular networks), addressing is based on two identifiers:
IS-41 networks do not have the equivalent of the GSM IMSI. In fact, in IS-41 networks MSI and ESN are used together to identify the mobile user at registration with the network, to authenticate the user, and so on. Therefore, if a mobile user desires to change terminals, he or she needs to interact explicitly with the mobile operator (e.g., through customer services) to indicate the desire to change the terminal. 7.2.2 Layers of Addressing in IP NetworksThe basics of IP addressing were described in Chapter 2. IP hosts traditionally run different applications on top of the IP stack (e.g., e-mail, Web browsers). In addition to these well-known and widespread applications, new applications are being defined. A relevant example is SIP and applications based on SIP that allow an IP host to have multimedia communications (e.g., audio, video) with other IP hosts . Although a correspondent node can use the host IP address to address the host in applications, the IP address is typically not the most convenient and appropriate address to be used for a set of reasons:
Therefore, additional addresses have been introduced for IP hosts that allow addressing of an IP host independent of the host IP address and that traditionally have a user-friendlier format (i.e., are alphanumeric strings that are easier to remember). Examples are e-mail addresses, typically in the form username@domain, and URL addresses (e.g., Web site addresses). When considering IP networks and IP hosts, it is necessary to discriminate between two levels of addressing:
Traditionally, the IP address assigned to an IP host is considered as the host transport-level address, whereas the host e-mail address and the SIP URL are considered application-level addresses. The introduction of mobile IPv4 led to the need to identify mobile IP nodes independent from the host IP address. In fact, the foreseen shortage of IPv4 addresses due to the limited IPv4 addressing space brought out the need to have an additional identifier for the mobile node that allows the node to be uniquely identified and receive an IP address dynamically when obtaining connectivity with the network. Therefore, the NAI (network access identifier) was introduced in IP networks. The NAI is a private address for an IP user and, more specifically , for a user having a subscription with a given service provider. An IP host can have several NAIs assigned to it depending on the number of service providers the host has subscriptions with. In this complex framework, the discrimination between public and private addressing and between addressing and user identification in IP networks is a gray area that depends heavily on the specific type of IP network. In fact, several parties advocate that the sole identifier for an IP host, whether mobile or not, should be the host IP address. Other parties advocate the availability of different addresses and identifiers for IP hosts for usage in different scenarios. When making a comparison with cellular network, there isn't a clear match between public and private addresses used in cellular networks and addresses of IP networks. The GSM IMSI can be compared with the NAI, and the MSISDN can be compared with the e-mail address or SIP URL. In a similar way, the IMEI can be compared with the identifier of the network interface card the IP host uses to access the network. However, in the case of wireless networks, and in particular in the case of cellular networks, it is not clear what the network interface card identifier is. Finally, in cellular networks there is no clear equivalent for the IP address, since the delivery of services (e.g., voice calls) in cellular networks is not based on any transport-level identifier but on the MSISDN and a set of functionality in the network to keep track of the location of the cellular node corresponding to the MSISDN. Figure 7-1 compares a traditional 2G cellular terminal (GSM) with a wireless IP host. The wireless IP host has access to the cellular technology. A terminal has two types of addresses: user addresses and terminal addresses. A user address is visible to a subscriber, whereas a terminal address is internal to the terminal and required by the network for routing or identification. In the figure it can be seen that in addition to MSISDN, IMSI, and IMEI, a wireless IP host needs to have e-mail, NAI, SIP URL, and IP addresses. Figure 7-1. Addressing and identification comparison. With the application of IP to wireless networks, and in particular cellular networks, new addressing issues are introduced. These issues are mainly based on the different types of addresses and identifiers used in cellular networks and IP networks. An issue that is particularly relevant is the need to maintain a clear distinction between transport and application addresses, public and private addresses, and user/terminal addresses as described in the previous sections. |
EAN: 2147483647
Pages: 164