Telephony numbers are like endangered species. When they are first discovered, it appears that their numbers are infinite. Then comes the phase of wastefully using the resource and finally panic when it is discovered that the resource is going to be extinct in a few years. This has occurred not only with telephone numbers but also with MIN, SID, IMSI and ESN codes.

Telecommunications systems have an identifier for most things. In wireless, important identifiers are:

• MDN (mobile directory number), a dialable number assigned to a wireless phone; also known as MSISDN

• ESN (electronic serial number), which identifies an AMPS, TDMA or CDMA phone

• IMEI (international mobile equipment identifier) to identify a GSM phone

• MIN (mobile identification number) to identify an AMPS, TDMA or CDMA subscription

• IMSI (international mobile subscriber identity) to identify a GSM subscription and SIM card and, in the future, TDMA or CDMA subscription as well; also broadcast by GSM base stations to identify the license holder

• SID (system identifier) broadcast by AMPS, TDMA or CDMA base stations to identify the license holder.

If identifiers were assigned individually, the number assigned would match exactly the number required. Unfortunately, this is not practical because it would require a centralized assignment authority for individual numbers. Because the numbers would not directly identify the “owner” of each block, a query in a massive centralized database would be required before routing to the owner.

For example, the first six digits of an MDN usually identify the switch to which it is assigned; the first eight or 14 bits of an ESN code identify the manufacturer of a phone; and the first six digits of a MIN or IMSI identify the home system. This information allows routing to the owning system without performing a database query first. It is true that LNP does require a database query as numbers can be removed from their assigned carrier, but this is only a local query and has increased the complexity and cost of telephone networks.

Because identifiers are assigned in blocks, a minimum number can be assigned at one time, creating waste if the full amount is not required. For example, MDNs usually are assigned in blocks of 10,000 and ESN codes in blocks of 17 million. Assigning phone numbers in groups of 10,000 significantly increases the management overhead but, because of the small areas to which blocks are assigned in the United States, assigning them in larger groups would decrease the already low efficiency of assignment significantly.

Where We Went Wrong

Almost every numbering or identifying resource used in wireless has assignment problems. But if wireless hadn't been such a successful technology, the original schemes would have been just fine.

Phone numbers (MDNs) are assigned to rate centers, small geographical areas that are the fundamental billing unit in the United States. Because rate centers often are too small, and political inertia prevents them from being increased in size, MDNs have to be assigned in small groups of 10,000. Even with these small groups, they are assigned inefficiently when there are many competitors.

ESN codes have long been assigned using an 8-bit manufacturer-code prefix. This assigns almost 17 million unique identifiers to each manufacturer. Although this is good for the large companies, it means that the small are forced to waste millions of codes. The resource is approaching exhaustion, yet the number of phones that need ESN codes is far fewer than the 4 billion available. The transition to a 14-bit manufacturers code would multiply the remaining number of manufacturer codes by 64, allowing the resource to last considerably longer. Another solution is the re-use of manufacturer codes assigned more than a decade ago to manufacturers that are believed to have only produced a small number of phones.

SID codes suffered from over-enthusiastic assignment in the early days of cellular. After they were assigned in the United States by the FCC, someone sat down with an atlas and, based on the population and political boundaries of the time, assigned a group of SID codes to every country in the world, leaving few left over. Many of those countries have only GSM systems and do not need even one SID. Fortunately, the belief that every licensed area would need a separate SID has proved to be false, as many carriers have merged several systems into local calling areas using a single SID. Because of this, most of the assignments have proved to be adequate. However, the resource remains more than 90% assigned.

The MIN is a 10-digit number, so there should be 10 billion available. However, by tying the MIN to the phone number, U.S. and Canadian carriers ensured the unavailability of most of the 8 billion MIN codes that match North American phone numbers. This forced international carriers to invent the international roaming MIN (IRM), a MIN that begins with the digit 0 or 1 to keep from matching a North American phone number. The CTIA (www.wow-com.com) initiative to assign MIN codes separately from MDN will free up billions of MIN codes and give this resource a new lease on life.

The IMSI is 15 digits long, and theoretically can provide 1 billion numbers to 1 million separate carriers. How could that be exhausted? GSM carriers have tried their best. By tying routing based on IMSI to phone numbers, they have ensured that out of about 24,000 mobile network codes (6-digit IMSI prefixes) assigned to North America, only 1,000 can be used. Plus, by implementing only a 2-digit mobile network code instead of three, the number available is reduced further to 100. With the recent radical carrier consolidation, this may prove to be adequate.

Numbering is a strange and arcane field within telecommunications that seems as if it should be easy. The problem is that when identifier formats and allocation schemes first are decided upon, future usage patterns cannot be predicted. So far, we have been able to rely on the ability of engineers to come up with clever schemes to conserve the resource before it goes the way of the dinosaur.

Crowe (crowed@cnp-wireless.com) is a wireless-standards consultant and editor of Cellular Networking Perspectives, a wireless-standards and -technology bulletin.