My design idea starts with dividing a big area into small cells (circular or hexagonal) same as used in mobile communication for frequency reuse. Here the reason behind cell division is that population density is not evenly distributed over an area. So making a single big network will waste resources while communicating to sensor nodes placed far away or behind some blank area.
Therefore I came up with small cells for data collection, may be up to 500m across, each having a master node in centers to collect data and upload it on the server. Still power may become an issue as microcontrollers and RF ICs require current in order of tens of milli-amps which will drain the battery very soon. To save power, sensor & master nodes are synchronized using RTC so that they spend more time in sleep mode consuming current in order of micro-amps and can wake up at the same time, exchange data and again get to sleep. This way they can easily withstand up to several years on a single cell.
Now, to make the solution more practical and optimized, in my design sensor network will have two parts:
1. Stationary sensor network- In this all sensor nodes will be installed at particular in such a way that crowded/residential places have more no. of sensor nodes making almost a circle or hex(called a cell), centered by a well-placed and powered master node which listens to all its sensor nodes, collect data and upload it on the server.
2. Mobile/ moving sensor nodes- Above explained stationary sensor network can be a good choice for monitoring of residential areas but there’s more in a city. To cover large roads/ highways and all, placing sensor nodes in between is really not a good solution here. But I think, there’s an efficient way of monitoring those sections of city. If we just look current affairs, how EHT(Event Horizon Telescope) managed to make a virtual telescope, size equal to the earth using just 8 telescopes to observe black holes. Similarly, if we use something that is periodically moving across the city we can easily cover a very large portion of a city with very few no. of sensor nodes. Yes, we could use metro trains, public buses, etc. One thing I want to mention here is, the design of mobile sensor nodes is different from stationary ones.
**NOTE** – We know that RF hardware requires a lot of power compared to other hardware of the system so to save more power, data collection and transmission frequency must be kept different. For example, data may be collected every hour but get transmitted/uploaded only 2-4 times a day (depending upon requirement).
What makes it innovative?
1.Using stationary & mobile sensor network in combinations we can monitor a city completely and can analyze the city’s environment in depth.
2.Area division in small cells according to population density is a great way of monitoring and saving power simultaneously.
3.RTC synchronization ensures that system stays in low power/sleep mode most of the time and all wakes up at once, do work and again get to sleep for next time. This ensures good battery life.
4.Different frequencies of data collection and data transmission will save a lot of power.
How it would be produced?
Stationary sensor nodes -
Any low power micro-controller(ATmega, STM8 or STM32, MSP430 etc.) can be utilized as they all have low power/sleep modes in which current consumption is in order of micro-amps only. RTC(Real time clock), either internal or external RTC (like DS3231) can be used. If a single cell is used (assuming Vbat = 3.7V), then we’ll need a switching buck converter IC with good efficiency (like TPS62740) which provides 90% efficiency @ 10μA, and 95% @ 10mA.
For RF communication a low cost RFIC “NRF24L01+” will be used along with a PA,LNA IC “RFX2401C” to boost its Rx/Tx capabilities. Sensor depends upon the parameter that has to be measured. For example, “for sound-quality a microphone will work with an algorithm” “for air quality SGPC3” can be used.
Optional – If battery life still becomes an issue then we can use a small solar cell with the sensor node along with an energy harvester IC “BQ25570” to recharge the cell whenever sunlight is available to it.
Mobile/moving sensor nodes-
There’s just a little difference, instead of an RTC, GPS should be connected so it will collect the location as well ,for example if it is installed on a metro train then it will submit its logged data to master node at the last station on each end. GPS may consume a lot of power but on which it is installed may provide it sufficient power or again we can use some solar cells to fulfill the power requirement.
Potential impact around the globe?
If the data collected from the system is made available for public it would create self-awareness as people can see the environment quality around them.