 |
|
|
Wireless phones are one of the few devices that are regularly sold at less than cost, and where salesmen are still handsomely compensated for selling at these ridiculously low prices. Everyone knows, of course, that payback comes to the carrier in the form of monthly bills carrying potentially lucrative airtime, long distance and miscellaneous charges. A good customer pays for their ‘free’ phone many times over. One part of this equation is unlikely to change, and that is the low selling prices. But what about the handsome commissions for signing up customers? Could that be trimmed to a more modest level?
Over-the-air Activation (OTA) is intended to do exactly that. By allowing phones to be sold at discount stores, and taken out of the store still tightly wrapped, OTA gives carriers the benefits of programming their own customers’ phones (and all the responsibilities). Discount stores will accept a much lower margin on the phones, so while the middleman is not eliminated, he is at least put on a strict diet. But, can OTA fulfill this dream? Are the technical and business challenges too great?
It is natural to want to use the radio interface to program phones, just like you would hope that chefs would want to eat their own food. The problems with a radio interface are that it is noisy, sometimes congested, and very public. OTA standards, which are available for both TDMA and CDMA digital phones, have to overcome these hurdles.
An OTA session starts out after a subscriber first takes their phone out of the box. They may have to dial a special number (a 1-800 number or a special ‘*’ feature code), or they may be automatically connected to the OTA operator when they attempt to make any call. Yes, an operator – the first step has to be to obtain the name, address and credit information for the prospective subscriber. This can only be done manually. When this phase of the call is complete, and the subscriber is approved for service, the automatic phase sets in, programming the phone with all those mysterious parameters, such as MIN and SID, that are required for operation.
One of the problems with an OTA session is that it might end abnormally at any time, due to the phone being turned off, the battery dying, loss of signal or interference. Consequently, the OTA software in the phone has to store all the information that it receives in a temporary location, and only load it into memory when the transaction is complete.
Programming a freshly picked phone is an example of the ‘bootstrap’ problem. The phone initially does not even have a valid identifier (MIN or IMSI), and one cannot be assigned right away, because the OTA session might fail, leaving an identifier stranded, or worse, in use by multiple phones. Consequently, the phone starts out with only an ESN, is assigned a temporary MIN for programming purposes, and then is assigned a permanent MIN and IMSI during the OTA session.
Another issue with OTA is security, especially if authentication keys are being programmed over the radio interface. It seems contradictory that the whole aim of authentication is to keep the secret keys from being made public by carefully ensuring that they are never transmitted across the radio interface, and yet OTA needs to transmit this very information. There are complex methods for protecting data between two partners that have no physically secure mechanism to exchange keys, but they are often vulnerable to ‘man in the middle’ attacks, where an eavesdropper sets up a false base station, receives transmissions from both parties, and then retransmits them. It is not clear how practical this type of attack is, but it does not have to be too likely to scare carriers away from using OTA to transmit authentication data.
Behind the scenes, OTA requires a variety of network communications between the MSC that is serving the mobile, the HLR that will store the new subscriber record, the AC that will store security data for the subscriber, and the OTAF (Over-the-Air Function) that acts as an interface between the wireless network and the Customer Service computers that store complete subscriber information, including billing parameters. The OTAF coordinates network operations to ensure that records are properly synchronized and that OTA session failures do not result in data left lying around. These operations are based upon the TIA/EIA-41 protocol, with heavy use of the messages that are current used to transmit short messages (SMS). TDMA OTA, in particular, attempts to program phones using short messages, as this is most likely to be transparent to intermediate network elements (such as a VLR).
The technical challenges to OTA are significant, although the standards that have been published (TIA/EIA-136 for TDMA, IS-683 for CDMA and IS-725 for network operations) go a long way to resolving them. Just as importantly, there are huge challenges on the business side of the equation, for which the answers are less forthcoming.
The problem arises because OTA-capable phones must be sold to discount stores at far below cost, perhaps even at no cost. The discount store must be able to sell phones at a lower price than traditional dealers, and they will not be sending the customer out the door with a signed contract, so a commission based on a signed contract is not possible. So, when the customer walks out the door, the carrier has lost money on them, and does not yet have a signed contract. Making money on this potential subscriber relies on a few things going right.
First of all, the customer must use the phone. If the customer decides that the purchase was a mistake, and the phone sits on the shelf, the carrier loses money. If the customer lives in an area where the carrier does not have coverage, the phone cannot be used and the carrier loses money. If the customer is able to break the subsidy lock, and activate the phone on another carrier’s network, the carrier loses money. If the customer has bad credit, and the carrier refuses to activate their phone, the carrier loses money. Worse yet is the potential for modifications to be made to these phones to allow fraud, and then the carrier will lose both on the sale and on the use of the phone.
The customer must activate their phone on the right carrier’s network in order for the carrier to make money. Carriers must carefully control the subsidy lock so that they make their money back on their phone, while not preventing the customer from changing systems at a later date. They must put as many anti-fraud checks in the system as possible, both against technical and subscription fraud, although it is subscription fraud that is the biggest concern.
If phones are sold for little money, they may often be purchased with cash, which provides no information about whether the customer has the credit rating to be a viable client for cellular service. Carriers may try to insist that phones are only purchased with credit cards, but stores may not like this special treatment for phone purchases. Because the OTA transaction is remote, it is impossible to verify photo id or signatures. Consequently, carriers will have to be much more vigilant with customers activated with OTA than with those activated through more traditional means.
The big question with OTA is whether it will really save wireless carriers money, or whether it will simply make them appreciate all those things that dealers do for them – credit checks and customer education being some of the most important. Will the burgeoning love affair with OTA last, or will carriers revert to DFA – Dealer’s Finger Activation?
© – Copyright