# Host hopping mesh network over LOR or ZigBee

I am looking to do a proof of concept for a project where each node will broadcast a set of data (GPS Related) in predefined intervals, so every other node do some processing based on that data.

I am yet to finalise my exact list of core functionalities to focus on POC. Rough overall requirements are given below. I would decide based your feedback.

• Network would be host hopping mesh network. Each node should be in coverage with at-least one other node to function. Network would be dynamic where there is a possibility of next node can go offline or off coverage.
• Maximum 10 nodes will be in the network. Each node may broadcast every 10 Seconds.
• ACK is a nice to have feature.
• Expecting coverage around 2 to 3 Km in outback/bush-lands.
• Prefer to be less than 50 USD.

I am based on Australia.

Based on my research, LORA OR ZigBee seems to be promising. Personally i am slightly inclined with LORA module because cost vs range over ZigBee.

RFM96 LORA 433Mhz Looks like good candidate. Check this product for LORA. Bundled with Atmel ATMega1284P microcontroller. This combo is resonalby priced and gives 16KB of ram to playaround.

I have found a good set of drivers and mesh network libraries for LORA RFM96.

Questions,

1. Anyone implemented mesh type network using LORA RFM96?

2. What would be practically achievable data rate using LORA RFM96? So i can decide what to leave/include in transmission.

3. What type of antenna gives more covarage and data rate for LORA RFM96?

4. I couldn't find a suitable ZigBee (Price and Range) Or Other modules in my case. Any recommendations if ZigBee or any other modules seems to be good candidate?

I believe this is not an opinion based question. I have given my research and finding and asking help to decide based on your past experience.

My background is voice and data communications. In the 1980s I worked at Ungermann Bass (Worlds Largest Networking Co. in those days) and worked on emerging network protocols like Ethernet, Token Ring, FDDI.

ZigBee has been around since the 1990s. Is supported by established reputable companies. Is an IEE802 standard. But it will not work for you with a range of kilometers.

LoRa has been around nearly two years. It's a hodgepodge mesh mess. These sorts of things take more than two years to be truly usable and interchangeable. A networking protocol would be a good thing to have.

Yes you can buy a LoRA board. It can transmit and receive data. That's it. No networking protocol. It's a toy.

Two of the ZigBee Alliance Members I checked out, TI and Silicon Labs, both have said they will not be supporting LoRa.

In my looking around I found this interesting product at Silicon Labs.
I did not check into it further but it supports IEEE 802.15.4g. ZigBee on a 433Mhz radio transceiver. Interesting. The have development kits around $300, some transceiver boards are$55.

The Toys

• Thanks for the answer. I do understand LORA is just a plain hardware. But found some libraries to manage it. airspayce.com/mikem/arduino/RadioHead/index.html – Jeyara Mar 22 '17 at 3:45
• What you really want is to plug in the board, install the drivers, and it works. Like bluetooth, WiFi. You do not want to be coding with someone else's library. Is the library supported? When (not if) you have a problem, who do you call? How good is the documentation. Why would you want to mess with a library? You want a product that works out of the box, you need not think about it, it just works. So you can go about doing the things you do, rather than wasting your time on a very likely frustrating dead end road. – Misunderstood Mar 22 '17 at 3:58
• I think you have misunderstood. I am a programmer for living. Electronics is my hobby. So understanding code is not an issue. What i mean is to leverage some ground work from libraries already tested/established. Modify if needed and then focus building solution to my problem space rather solving a problem which is solved already be someone else. I didn't expect a happy day scenario right out of the box. I believe getting help is lot improved in 25 years when i started with stereo amplifiers. Hope that enlightens you. – Jeyara Mar 22 '17 at 5:11
• Yes, enlightened, you're a masochist. I only know a couple of programmers capable of do this sort of stuff. Depending on what you are planning on doing. I'm talking writing the network layers. If you want a mesh network, you need a network protocol. I've done 3 MAC chips so I have an idea how tough it is. I'm not saying you are not capable, I do not know you. You must be if you are willing to take this on. I really like the idea of putting ZigBee on 433Mhz. That should work. But do not know how multiple networks would play together. – Misunderstood Mar 22 '17 at 7:31
• Thanks for the background information on implementing network protocol. As you suggested, i would read more on ZigBee on 433Mhz and also on Silicon Labs' EZRadioPRO. – Jeyara Mar 22 '17 at 10:45

The MRF24J40ME transceiver module has a range of up to 4000ft (1.2km) and supports zigbee.

The transceiver itself is about $16 in small quantities. https://www.digikey.com/product-detail/en/microchip-technology/MRF24J40ME-I-RM/MRF24J40ME-I-RM-ND/5252343?utm_adgroup=RF%20&%20RFID A development board with the part on it... https://www.microchip.com/DevelopmentTools/ProductDetails.aspx?PartNO=AC164143-1 ...is$34.99

https://www.microchipdirect.com/product/search/all/AC164143-1