Sending Messages Peer to Peer

Here we are demonstrating the power of peer to peer in the field with low battery conditions and deep sleep type application;

The process is as follows:

The first chip will boot up and the Bluetooth LEDs will come on both chips and we will then demonstrate the opposite from the other chip, it boots up and turns off the Bluetooth led.  We then will show that we can send led-off and led-on messages by simply pressing a button on either chip to change the state.  This all achieved with no access point, no router, the chips are paired through their mac addresses and once they come alive are capable to send messages very quickly. the boot process is very fast, you can see below our findings.


Advantages of PXC connected embedded chips summary on savings in the table below

Technology Days* Packet size Consumption per Hour Transmission Frequency
Lora SF7 5123 10 Bytes 0.01 mA (125mA) 38ms 130khz
Ghost Low 25185** 10 Bytes 0.005 mA (80mA) 30ms 20mhz
Normal Wi-Fi 156 210 Bytes 0.24mA (90mA) 8000ms 20mhz
Lora SF12 137 210 Bytes 0.27mA (125 mA) 7700ms 130khz
Ghost High 835 1055 Bytes 0.04 mA (80mA) 15ms 20mhz

* Usage-based on 1000mAh battery

** This is beyond the life expectancy of any battery but also based on monitoring arrangements/reporting using the low energy ghost and solar power/ups type backup.

Results recorded in the table above are based on the following test criteria

Normal Wi-Fi:

  1. Connect to Wi-fi: 3500ms
  2. Send message:2400 ms
  3. Total 5900ms +circa 6secs
  4. One message = 380ma Peak
  5. 8000microseconds sending including the above start-up
  6. Sleep mode draws 80ma

The result is 156 days with an estimated consumption of 0.24mAh

  1. Below left to right 8-sec transmission duration of sending,
  2. Per hour 3600sec
  3. 90 mAh consumption during transmission
  4. 0.04mAh consumption during sleep mode
  5. Type of battery
  6. Discharge profile

Espressif limitations on their ESP processes

In station mode

  1. supports only 10 Encrypt clients
  2. Unencrypted is 20 clients maximum

Soft Access Point mode

  1. +Station mode supports 6 encrypted peers
  2. 20 unencrypted maximum at a push

EWI Tech Ghost / DNA process

  1. Scales to millions of devices for conserving power for e.g. deep sleep mode with fridges, only wakes up when a support request is required or exiting of data to an AI engine. If it is once a month exercise multiple the results by 30 times. 835 days becomes dependent on battery retention standby life as opposed consumption. Theoretically if the batter can last as long 835 days can achieve a maximum 25k days or 69 years.
  2. Connect to Wi-Fi 0ms always connected even out of a deep sleep, UDP to UART process with a exit path from PXC created
  3. Send message 200ms (36 times faster) than 8000ms


Putting the new values into the calculator:

Results are 835 days with an estimated consumption of 0.04mAh


  1. Below left to right 0.22-sec transmission duration of sending,
  2. Per hour depicted by 3600 sec
  3. 80 mAh consumption during transmission
  4. 0.04mAh consumption during sleep mode
  5. Type of battery
  6. Discharge profile

Comparing Lora WAN (courtesy of an EWI advisor and radio tester)

First some basics on Lora networks

  1. Lora only transmits with low data rates
  2. Time to transmit depends heavily on the message size
  3. Lora uses spreading factors and this influences the signal reach
  4. SF7 fastest but not reliable in critical situations
  5. SF12 slow but offers a better range

We compared the spread from 10 bytes to 200 bytes and the longer the message the longer the transmission time. We found that a long SF7 message is 20 times faster than a long SF12 for example. Lora has a maximum of 250 bytes on a single transmission.  EWI Tech Ghost/DNA is nearly always faster than Lora. Especially with larger messages, it’s not even close, 220ms compared to 1000ms.


Why does EWI Tech have a higher data rate than LORA, EWI Tech has a bandwidth of 20Mhz where Lora’s (Shannons Law) Lora’s signal only has a bandwidth of 135khz which is 150 times less.  Lora get its maximum long-range as a result of this bandwidth capability.


Lora Testing SF12

  1. 210 bytes sending using SF12
  2. Transmission 7700ms (7 seconds)
  3. Lora takes 125mA 50% more than the ESP SOC we use.
  4. We compare the 125ma with the ESP chip of 380mA but note the transmission speed is so low and Lora always has to have the transmitter on.

We had to remove some LEDs and parts on the Lora board to achieve the best we could with the poor result above.

Moving to SF7

  1. Sending 10 bytes maximum with transmission of 38ms (EWI PXC at 10bytes would achieve a 15ms transmission with the same packet)
  2. We can achieve the famous high number that Lora quotes 5123 days

Summary of EWI’s System on a Chip Technology :

  1. Suitable for long-life battery operation for fast sending
  2. Normal mode is ok when energy is not an issue and you need normal internet access
  3. When we add the peering exchange into the equation it protects the energy efficiency of the chip and gaining the fast response time
  4. The biggest win is the time it takes for a chip to connect via peering exchange as opposed to a Wi-Fi access point
  5. Speed of access is essential when a device sleeps and has to reconnect, with PXC it’s not applied as the connection is already in place out of sleep mode.
  6. PXC process has a battery life span of 835 days with Lora on 156 days with similar payloads
  7. The duration and size of transmission heavily influence consumption and battery life.

We have proven that EWI PXC DNA process is far superior when deep sleep applications are applicable, for example, the white goods analysis process and zero-touch analytics for fridges, TV’s in customer support, and data exit scenario.  Lora SF7 does have a range advantage over using ESP Soc chips but we can combine them to work with PXC/DNA and ESP to get the best of all scenarios.

However, we do have a way of building bespoke antennas:

Depending on the Antenna the ESP chip is capable to achieve a standard line of sight up to 150metres, we have expertise onboard to increase that with a bespoke antenna design, we can achieve up to 1km plus in range with our patented design and power processes.

We know that 2.4ghz behaves much more like lightwaves as it is unable to come outside of a house to be picked up through a wall for example.  We conduct all our tests within the unregulated space to determine the maximum legal permitted. So we use FCC approved chips and Effective (equivalent)isotropic radiated power limits which is a maximum of 20dbm in power. 2.4ghz has a maximum of 6dbm gain, we know that Antenna gain does not affect the range, but does affect power consumption of the ESP range of chips.  ESP transmits and receives in parallel so power has to be adjusted.  ESP patents tell us that they restrict the data rate to 1Mbits to achieve long range. We have therefore modified the chips to achieve non-standard results with gains in range reliability and bandwidth using ohms law.

One further example of Peer To Peer secure messaging


Previous Post
Next Post Coming Soon...▶