Best 802.11g and 802.11n Circuits for WLAN, LAN, DSL Connectivity

[Praisesmith]

I want to interface some of my circuits using WLAN, LAN and DSL connectivity. There is too much information out there on this subject... my answer is hidden in the forest. I want a simple solution that I can learn on; eventually I'd like to make and sell my product and so I need for the circuit to be public domain.

I want to assign a standard Device ID Name and Password to my circuit and then allow for both of them to be modified by the controlling device that connects with my circuit (e.g. a smartphone or computer).

My PRIMARY goal is to have my circuit connect with all of the most popular smartphones, and so I assume that 802.11g connectivity gives me that. So, if I can find a really small simple circuit that gives me 802.11g and some reasonable security features then I will select it.

Future Growth is important too and so adding 802.11n is desirable. If I can add this without too much complexity I will do it right away and otherwise delay the addition.

Adding 802.11b and 802.11a for backward comparability is "nice" but a low priority--I'll only add this if it is simple to do so.

My device will NOT store any personal or financial data so my security related priorities ranked in order are: Future Longevity, Commonly Used, Simplicity, Secure. Therefore I feel that I should offer something that is fairly current and common, like WPA2 and/or WPA-PSK; perhaps adding EAP if it's not too hard, and for backward compatability WEP-40 and WEP-104 if it's not too hard.

I keep reminding myself of my goals when designing this: Future Growth, Commonly Used, SIMPLE, secure.

Thanks for any advice and pointing me to the better information to focus my attention on. I'm on information overload in my searching, I need specific advice and help on how to keep things simple... or as simple as I reasonably can.

[saturation]

Making a custom Ethernet connection is pretty involved; the main advantage is reduced size and form factor for each of your devices.

The easiest way is interface the project is to your PC, and then program the PC bridge between the Internet and the device.  Otherwise, you will need to program a MCU to do many things the PC already does, but at a large form factor.

The key then is either USB or RS232 connectivity, the latter is the easiest to make; then you are set from hardware, the rest is software. 

If you don't have those interfaces, you need to buy a data acquisition module for your PC, and this makes interfacing easier; it will accept voltage inputs, which you can easily output from your devices.

[Praisesmith]

Thanks. You are correct that my desire to have my own special design is for several reasons: smaller size and COST are huge factors. I want to sell this product at some point and I want to sell it as cheaply as possible. I also want it to be small in size. And I don't want the device to have an interface attached to it, I want the control to be remotely.

This is not real-time controlling where I'm constantly making changes and monitoring--so the amount of actual data that I am pushing will often times be a few bytes, certainly less than a single K for 99% of each usage. The protocol overhead is going to be the majority of the packet data that is going to be transmitted often times.

If I have to then I'll simply pick 802.11g with WPA2 or PSK security and write the code barebones. I'm just hoping that there something public domain that offers at least 802.11g AND n, that is packed small and cheap. Or perhaps a "System on Chip" (SoC) chip that has most of the functionality built in. But I don't know who the leaders are in this field, and who would make a more basic chipset that would be easy to implement for a need like mine.

Broadcom is certainly a player in this field: http://www.broadcom.com/products/Wireless-LAN/802.11-Wireless-LAN-Solutions
and so is http://www.atheros.com/technology/technology.php?nav1=47 and http://www.marvell.com/products/wireless/8366.pdf

So I am going to look into SOC chips to see if my solution lies here--any advice is appreciated.

I think that my devices will be controlled most of the time connected near the device using the Wireless LAN (WLAN) that has been setup to connect to the device (e.g. a computer or smartphone); however, I do want the ability to occasionally control it remotely by connecting to the device if it were also connected directly to the Internet by plugging directly into the DSL, or through software that I'd build that runs on a computer and connects to the device either through a hard LAN connection or through the WLAN connection.

So I'm needing to build 2 different types of connections for the same device... one that is near remote like a WLAN, and the other Internet remote, via DSL or special computer software using LAN or WLAN.

Hopefully I can find a way to write something small and utilitarian for my needs. At worst case I can try to buy some small WLAN device for a laptop or cellphone to test with and prove my concept, but there is no longevity in those devices, they change them up in less than a year, so I'd have to buy  batches of them in bulk, and keep changing my design slightly to work with each new batch of proprietary WLAN devices. Also, I want to keep my product cost down and so this would not work very well for me unless I was buying really old discontinued stock, which is fine for hobby use, but not a very good business choice.

I don't need fast speed, I don't need fancy. I just need to lock into the standard 802.11 protocols for a very basic transmission of what will be less than 1K of data... my need is very very simple, so I hope to find a simple schematic to do the job.

[saturation]

I see.  If you want to make it that way, the Arduino folks have done a lot of controller work using fully designed devices that interface to the Internet directly & open source libraries for support.

Here's a sample:

http://www.ladyada.net/make/eshield/index.html

http://www.instructables.com/id/A-credit-card-sized-Ethernet-Arduino-compatable-co/

[Praisesmith]

Thanks. That link also led me to find these wireless links:

http://www.lantronix.com/device-networking/embedded-device-servers/wiport.html
http://www.lantronix.com/device-networking/embedded-device-servers/matchport.html

So, I'll investigate further. I assumed that somewhere someone had figured out that there's a need to connect our projects to the Internet; otherwise I was going to put my invention on hold and go invent that need first. LOL

I also found this from Microchip, and so that seems exciting: http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2884

Silicon Labs offers a nice developer kit: http://www.silabs.com/Support%20Documents/TechnicalDocs/Si4010-Keyfob-DK.pdf
http://www.silabs.com/products/wireless/pages/default.aspx

But, as you can see, soon I can get back to information overload. Any help that points me to the leaders in the market, especially for cheap solution that is easy for a beginner to get into and make it work. If it ends up being hard to do, then I'll still tackle this, I'm capable of learning anything, easy or hard, given enough time, but I don't want to choose hard if I don't have to.

UPDATED NOTE:

I'm adding this update after writing my previous notes above. I think the Microchip really looks like they are going to work very nicely for me because I can program their PIC-32 chip to work seemlessly with their special wi-fi chip that sells for about $26--perfect! Also, it looks like this same chip can work with their PIC-16 and PIC-8 chips too, which is very nice.

http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2884
http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2884

I'm also going to consider the Silicon Labs (www.slabs.com) because they seem to have a nice product too.

I want easy and cheap if I can find that.

ONE OTHER IMPORTANT FINDING:

I am adding this later from writing too. After doing more research it is obvious that the new ZigBee protocol devices are going to be a FAR FAR better solution for me. They are designed just for this type of thing. They are cheap and small and easy to use! Again, MicroChips is a leader in this field and so it's going to be very easy to work with them using MicroChip's PIC chip products with the ZigBee transmitters which can be purchased for $3 or $4; their low cost is because their architecture keeps them low cost on purpose--these are designed specifically for monitoring and controlling devices remotely, and do do so cheaply. Z Wave is their proprietary competitor--ZigBee is open source with over 150 corporate sponsors.

[saturation]

Yes, very good links.  One issue with making projects is taking on too much to make a device as clean and compact as possible; its all possible it just takes time and perseverance to design.  Many folks begin complex projects and never finish.

When I design, I need to prototype a working model in less than a day, that's all the time I have.  I do this by buying or acquiring as much premade as you can locate, then assemble it, rather than reinvent the wheel for each subcircuit. 

What is the primary activity of this design?  You wrote:

"My PRIMARY goal is to have my circuit connect with all of the most popular smartphones, and so I assume that 802.11g connectivity gives me that. So, if I can find a really small simple circuit that gives me 802.11g and some reasonable security features then I will select it."

Connect to do what?  Connect a DMM to see it remotely on a smartphone?  What you connect does determine what path you take.  All smartphones already connect via a web interface through software and can go into the Internet, using each smartphone's gateway to do the conversion.   All you need do is find a way to interface say the DMM to a server, so it can be read remotely; this is a widely already done application.  Generically, you can remotely view anything that comes out of a data acquisition board and usually it has the remote software included.  Just ideas.  Good luck!

[Praisesmith]

Yeah, I prefer less complex too; unfortunately I've got some special needs for a few smarthome products that require a very small size and they are price sensitive, so that tossed out the standard ways of WLAN; but my speed needs were low, and so this ZigBee is just the ticket; I'm still going to need a way to connect to the smartphones because I don't think the smartphones are ZigBee capable yet; I'd imagine they will be soon... in fact I'm sure of it because many interface products are going to be made with ZigBee and Z Wave.

[saturation]

Ok! If possible keep us posted with your project's progress, and as always, enjoy the trip! 

[NiHaoMike]

Buy a cheap wireless router that supports an open source firmware like OpenWRT, DD-WRT, or Tomato Linux. With some going for $10 retail, it's definitely a cheap way to add wireless to your project. I'm currently modifying a $20 Asus RT-N12 for various projects.