ESP8266 Community Forum

- Thu Dec 18, 2014 4:42 am #5048 Can you share your code?

I am able to create a bridge between clients. There are the features:

1. Upto 255 client can be connected with ESP (Not 5 as it is by default. Higher device, lower performance... )
2. One client can communicate to any other client with.
3. The ESP can simultaneously be connected to LAN/WLAN and can access other node (esp or PC etc). Any node (connected with ESP-AP) can communicate to any node in LAN (ESP-STA)
4. Multiple ESP can connect each other [As of now manual connection for the issue of default AP-IP of 192.168.4.1....work in progress]. After connection one node can communicate (as of now- simplex........need to implement the routing protocol) to any remote node.

ESP8266.com Moderator
Lets get the party started and +1

[url=https://www.banggood.com/custlink/KvGGhGF4wG[/url]

I like ESP8266 and ESP8285 and ESP32

- Thu Dec 18, 2014 6:25 am #5051

I am able to create a bridge between clients. There are the features:

1. Upto 255 client can be connected with ESP (Not 5 as it is by default. Higher device, lower performance... )
2. One client can communicate to any other client with.
3. The ESP can simultaneously be connected to LAN/WLAN and can access other node (esp or PC etc). Any node (connected with ESP-AP) can communicate to any node in LAN (ESP-STA)
4. Multiple ESP can connect each other [As of now manual connection for the issue of default AP-IP of 192.168.4.1....work in progress]. After connection one node can communicate (as of now- simplex........need to implement the routing protocol) to any remote node.

This would be an awesome kickstart!

I think that the ESP8266 modules are simply the wrong hardware for that.
They are meant to be used in an environment, where you have an existing WiFi network and mains supply.
A mesh network makes sense for battery powered or energy harvesting small nodes, that transmit only small amounts of data,
where no infrastructure exists.(e.g. outdoors).
For this you'd need energy efficient micros and low power rf, like Zigbee, or BTLE.

Exactly. Imagine you have an wifi network but want to expand it within a large area such as an big warehouse, a plant site or a large terrain. Buying regular APs for this would be very costly if you are just 'playing'... I'm assuming you have no limitations on power supply.

- Thu Dec 18, 2014 12:21 pm #5088

I am able to create a bridge between clients. There are the features:

1. Upto 255 client can be connected with ESP (Not 5 as it is by default. Higher device, lower performance... )
2. One client can communicate to any other client with.
3. The ESP can simultaneously be connected to LAN/WLAN and can access other node (esp or PC etc). Any node (connected with ESP-AP) can communicate to any node in LAN (ESP-STA)
4. Multiple ESP can connect each other [As of now manual connection for the issue of default AP-IP of 192.168.4.1....work in progress]. After connection one node can communicate (as of now- simplex........need to implement the routing protocol) to any remote node.

For this type of mesh check out OLSR.
http://www.olsr.org/

Where I get my modules

- Fri Dec 19, 2014 12:03 am #5128 @ jonsmirl, Thanks. I shall be looking to the link.

uhrheber wrote:
I think that the ESP8266 modules are simply the wrong hardware for that.
They are meant to be used in an environment, where you have an existing WiFi network and mains supply.
A mesh network makes sense for battery powered or energy harvesting small nodes, that transmit only small amounts of data,
where no infrastructure exists.(e.g. outdoors).
For this you'd need energy efficient micros and low power rf, like Zigbee, or BTLE.

Exactly. Imagine you have an wifi network but want to expand it within a large area such as an big warehouse, a plant site or a large terrain. Buying regular APs for this would be very costly if you are just 'playing'... I'm assuming you have no limitations on power supply.

There are my observations:
1. ESP are the true modules for WSN and much more superior (I know, its just a perception) than other Zigbee (should I take names? ).
2. Its wrong statement that "They are meant to be used in an environment, where you have an existing WiFi network and mains supply". We can create the environment and establish our network. Mains power? Thats a paradigm. If we compare the actual-electric-power vs RF-output power, the efficiency will be too less (less than 20-30%.) for all RF in the said intended band (2.4GHz) as all are using class-A or B in both PA and LNA. Hence, if we compare Apple to Apple i.e. 0dBm TI chip vs 0dBm other chip, all will be "almost same" 10-20mA @ 3.3V. Now, if we look at ESP, it is radiating at 20dBm which means a lot and take a standard router and compare the power consumption in "RF-output power" point of view. When the power control API will be released (even if not, we can design a hardware control for that) for ESP, and it will be capable to transmit at 0dBm, we will see, it can be "Battery powered". However, as of now, the duty cycle control is enough to proceed with. The "amount" of data we can send in a second (1-8Mbps, as per my test, need review) is much more than so-called ZigBee.

3. The ZigBee and BLE are ok as "multiple technology co-exists", but WiFi (and ESP) will grow because of infrastructure and acceptability by the "USER". Lets see.


Update of my work:

Individual ESP will get individual IP in its Access-point. So, we need not to fix or customize or configure the firmware. Just flash the binary and the module will find (decide) its Access-point IP. No, its not random its mapped with chip-ID. Hence, if we "Assume" (As it is controlled by Espressif) the chips are having different Chip-ID , the AP-IP will also be different,

Next goal (may change) : Device will automatically connect to other "available" ESP module who has "Specific series of SSID" based on RSSI.