In general, certified modules/modems are always the best approach for many reasons like time to market, stability, support. It also depends upon the geography. For example, in US, AT&T does not allow uncertified modules/modems on its network while in europe, carriers do not have such stringent requirements.
My understanding of your definitions of modules vs modem is:
Modules: Build your own modem with third party chipsets, protocols stacks
Modems: Standard and certified product from vendors like Cinterion, Sierra, Telit etc
Let me know if you have any more questions.
As a newbie m2m developer, I see it like:
If your building a new machine with an Internet application: Embed a module.
If your adding an Internet application to a pre-Internet machine: Bolt on a "modem." AKA a gateway/terminal unit/single-board computer/cellular router with real-time OS and industrial internetworking e.g. serial ports, external power supply and antennas, etc. Examples being Digi TransPort/ConnectPort and Sierra AirLink. (Are there any cheaper alternatives to these that still have rock-solid OSes and APIs??)
If you're planning to use that modem in the US or Canada, please be aware that it will need PTCRB Certification to operate on North American Carriers. I'm not familiar with ICP DAS products, but you can ask them.
If the device doesn't have PTCRB Certification, it will cost you quite a bit of money to approve the device, especially if the cellular module inside isn't approved either.
The best solution is to integrate a pre-approved module. There are many companies providing cellular modules. On the GSM side, you should make sure the module you choose has PTCRB Certification. Otherwise, you're looking at a lot of money ($200K +). If the module is already PTCRB Certified, then the end device only goes through a small subset of tests, which are usually around $20,000.
Typically not for CDMA module, although you still need to go through Verizon or Sprint acceptance programs. Most Carriers, both CDMA & GSM, are requiring OTA Antenna Performance Measurements. This is the measurement of TRP and TIS.
For approved modules, most of the major manufacturers in the US and Europe have pre-approved modules: Telit, Motorola, Cinterion, Sierra Wireless, etc...
Hi Ganesh, if you are bringing an application to market for the first time I would respectfully suggest that you can't afford not to work with an experienced wireless design shop like Connected Development http://www.connecteddev.com/ that has an antenna engineer on staff as well as plenty of experience selecting, integrating, testing, certifying modules. The potiential cost implications of DIY mistakes here (e.g. $200,000 to get a new module registered by a tier 1 carrier) may be too high to risk.
Despite the fact that I am new to this forum, I am not new to the world of M2M. I worked in a variety of Technical/Sales roles at a Canadian carrier for 8 years, headed up Sales for a major Modem manufacturer for 4 and now am an Equity Partner / VP at a distributor of Modules and Modems in both Canada and the US. In total, a little over 13 years doing exclusively Wireless Data.
Since we sell a lot of modules and modems, I don't really have a motivation to push either, just to hopefully provide some unique insight:
Modules are usually the most cost effective when you pass over a certain number of units. The break-even point depends on how much money you will spend on development/certification, as well as a few other factors. For the most part, it is usually in the low thousands. They provide flexibility and a lot of customization capability. However, modules can have a few drawbacks. First, depending on how long the "shelf" life of your product is, you may have to switch modules mid-way through your production, so you have to factor in some development time / possible recertification time. Next, if you believe that you will need some of the enhancements that will be coming from upcoming technologies like LTE (not just speed, but reduced latency, etc), then you will have to be a little more proactive in ensuring that your modules have an LTE path, to reduce redesign costs.
Modems can be a simpler option, depending on your setup. The certifications are usually done by the manufacturer, and since they often have great relationships with carriers, their products can get done much faster. You also gain from their expertise in "keeping the product on the network", as they tend to provide more reliable connection times than many module-driven solutions. The negative part can be cost, as the cost of a modem can be several times the cost of the module and its related components. As mentioned, you may also lose the capability to make customization changes on modems, when compared to a module. If you are looking at lower volumes, it may not make sense to develop a module based solution. To answer another question, there are a number of solid lower cost options. When you lower down in price on a modem, you tend to give away three things....over the Air controllability, durability and connection stability. Another factor in pricing is speed, namely the connection speeds that are capable from the module that is inside of the modem. Buying the amount of "horsepower" that you need can make a big difference, but you also want to be sure that you are not limiting yourself for future applications (i.e. for a security monitoring application, are you eliminating the chances of providing video to your customers by using GPRS vs. HSPA+)?
Over the Air Controllability is often overlooked.....how often do you need to go out to a site when your device "falls off" the network? If the cost to send someone out is $100, and it happens 3x a year, then you have to wonder if using a lower cost modem makes sense....
Durability factors in the ability of a modem to deal with certain environmental conditions, such as heat, humidity and direct water hits. For most applications, it may not be a serious consideration (namely if you are using the device in a controlled environment). However, for unknown conditions, such as Utility metering, vending machines that may be indoor/outdoor and video monitoring solutions, you may need to look for a modem that has a stronger resistance to these things, as well as shock and vibration.
Connection stability is vital to any successful application, as a device being unable to connect creates major issues. Many of the advanced modems (not necessarily the most expensive ones) have a strong ability to detect certain network conditions, such as a dropped connection, and are quicker to reestablish themselves onto the network. They also have a strong sense of the timing of the individual carrier's networks, and work better with them.
In terms of development houses, I won't go as far as to recommend a particular one, as I like to avoid conflicts. What I can recommend is the valuable services / time reduction that they can bring to both a module development, as well as a modem-based solution.
Finally, an often overlooked part of any solution is the antenna, which is a shame. They are the "tires" of your mobile solution, and the only part that touches the network. Using a well-built antenna can often make the difference between a successful solution and an unstable one. Few key points:
- Keep the antenna as short as possible.
- Try to avoid deploying a solution at a signal strength of less than -90 whenever possible
- Use a 2nd antenna (dual diversity) on all HSPA/EVDO deployments
- Only use Yagi and other directional antennas if you know what you are doing / having them professionally installed.
- Go up....elevated antennas (getting over the building or tree line) often saves a lot of time
Feel free to contact me directly if I can ever be of assistance.....firstname.lastname@example.org
We do work with both CDMA and GSM carriers, and each offer their own benefits. The biggest factors that I would consider are:
- anticipated geographical areas of deployment
- Amount of anticipated data usage per month
- Do you require / desire the carrier to help sell your product for you?
- Hardware cost flexibility
- Do you want to bill the airtime or have your customer accept the bill directly?
- Inventory SKUs
Areas of Deployment
- Each carrier will have their strengths/weaknesses when it comes to cellular coverage. This can be for a number of reasons, such as available real estate to place a cell tower, where the towers are aimed, etc. If you have one concentrated area of deployment, this could be a strong deciding factor as to which carrier/technology you deploy with (as you would likely choose the one with the best overall coverage in that area).
- If you are deploying across a wide geographical area, or you aren't sure where it is going, this is less of a factor.
- An option to consider may be one of the aggregators of airtime, which allows you to use multiple carriers on the same bill. These are available in both CDMA and GSM flavors.
Data usage per month
- As a general rule, GSM has often been less expensive for smaller packages of data. Again, this is a general rule, and may not be the case across all carriers.
- When you get to the higher volumes of data per unit, such as required for applications such as Video or Mobile Computing, all carriers tend to be quite competitive.
Accessing the Carrier Sales teams/channels
- If your product is one that the carriers may like to sell themselves, this can be a major source of sales activity for you. However, do keep in mind that carriers have dozens of new products brought to them each month, so you would have to have some very compelling reasons to have them to this.
- Both GSM/CDMA carriers will be open for this concept for the right product, so not much difference there
Hardware cost and flexibility
- Simply, GSM/GPRS/EDGE hardware tends to be cheaper than its 1x/EVDO counterparts, namely due to the absence of the Qualcomm royalty, as well as larger scales of production. If you are looking at the lower speed deployment, GSM offers a pretty compelling argument.
- As all of the carriers move towards 4G networks, they will look to start using similar hardware, so this advantage may go away.
- In addition, depending on the carrier, there is often less carrier-based certification required to deploy on a GSM-based network than on a CDMA one. Again, check with the carrier, as every one of them is different, but for many GSM ones, PTCRB and FCC certification is all that is required (AT&T is a major exception). CDMA carriers tend to have a more detailed software load on each device, and often require certification before accessing their network, even in small numbers (again, check with the carrier, as some CDMA carriers have unique developer programs.)
- Depending on what you are developing your products for (i.e. is it an application-driven solution that happens to use cellular), you may wish to include the cellular bill into your overall bill for the customer, for simplicity sake, and possibly as a revenue generator for you.
- Each carrier handles "re-billing" differently, so be sure to check with them. You may also look at one of the aggregators, if re-billing is a must for your application.
Again, this depends on how many carriers you want to deal with. If you were dealing with, for instance Verizon and Sprint, you would have two different SKUs or product lines for these carriers. If you are selling tens of thousands on each network, this may not be a big deal, but if you are selling smaller amounts, you need to determine if this is going to be an issue.
With GSM products, for the most part, you have the flexibility to interchange SIM cards to move between networks easily. This would assume that you are not using a locked down module, such as in the case of an iPhone or BlackBerry.
GSM, for the most part, would provide you with an easier experience of working with multiple carriers, if this is important.
Hope this helps