Maximise Your IoT Devices’ Battery Life

Okay, guys, let’s make a start. So yes, once again, hello, everyone. Thank you so much for being with us today for today’s webinar session. Today we’re going to be talking about a low power mindset in IoT, exploring different ways that you can maximize your IoT device’s battery life. Really great to see so many of you guys on the call today. I know it’s a really busy time, so thank you so much for giving up time in your day to join us. I suggest, you know, we have a lot to get through, so I suggest we just make a start. So if we just jump on to the next slide. Just a couple of very quick introductions. My name is Duncan. I head up the marketing team, as I mentioned, here at Soracom. And today our speaker, Markus, is going to be handling most of the presentation. So Markus is our IoT connectivity specialist here at Soracom. He’s got more than twenty years of experience working within telecom industry with a particular focus on mobile networks. In fact, he spent the last five years working in-depth with cellular modules, specifically within the LPWAN technology sector. So he’s a true power saving expert when it comes to IoT, and I, for one, am very much looking forward to his presentation today. So, hi, Markus. How’s it going? Hi, Duncan. Thanks. How are you? Doing good, thank you. Yeah, all well. A bit cold here in the UK today, but apart from that, doing all good. Thank you. I’ll tell you what, just before I hand over to you, Markus, I’d just like to give a very quick introduction to who we are at Soracom and just give a sixty second overview of who we are, what we do, and kind of the reason for presenting the webinar today. So this is who we are. We are an IoT connectivity platform for IoT applications to launch and scale globally. We can provide you with affordable, reliable IoT connectivity that can accelerate the speed to market and we make it very easy to connect you to the cloud, sending data up to AWS, Google Cloud Platform, Microsoft Azure, and we can also help reduce your data consumption, and there are many other innovative features that we provide for IoT deployments. We were founded in Japan in 2015 by former AWS and telco veterans, and today we have physical offices in Tokyo, Seattle, and London, and in fact, we’re global team distributed all over the world. We’re proud to serve more than twenty thousand startups, SMBs, and enterprises across all different industries, from agriculture to energy to healthcare and electronics, and the list goes on. In fact, you’ll find a Soracom SIM card in over four million devices around the world today. We’re financially backed by KDDI in Japan, and then we also have additional capital investment from partners such as Hitachi, Secom, and Sony. And we also have a whole wide range of tools and platform services and features that can help you reduce the amount of power that your IoT devices consume, which is the reason for today’s presentation. If you’d like to know any more, you can just come to our website, Soracom.Io, and you can learn more about the platform. You can also email [email protected]. And we also have on the live chat, if you’re watching the webinar live today rather than on demand, we’ve got Nicolas and Camilla from our team who are in chat right now. So if you want to learn more about the platform, any questions about the products that we provide, those guys are going be available to answer those in real time right now. Anyway, that’s enough from me, and that’s enough about Soracom right now. Say, have we just gone to the next slide quickly, Markus? Great, okay. I’m gonna be fairly quiet throughout the first thirty minutes or so of the presentation. What I would like to say guys is if you do have any questions at all about anything being presented today, please do use that live chat feature that you can see on your screen in front of you. We’re gonna ask you to put as many questions as you like in there. I always feel that the Q and A section of a webinar is usually the slightly more interactive and engaging part of a session, so would really encourage you to put as many questions that you’ve got in there about low power mindset within IoT, and then I promise you we’ll get the chance to answer them towards the end. So I’ll now hand over to Markus. So, Markus, over to you. Thanks, Duncan. Okay. Warm welcome from Berlin where I got a kind of stormy weather, so if it gets some noisy outside, it’s April, you know. And so I will just start right away. Today we will talk about IoT devices and their battery life. So as you might have read already in the invitation, we will try to mention something about the radio access technologies, the RATS and some power saving features within the LPWAN technologies. LPWAN is low power wide area network devices and yeah finally I will be happy to answer some Q and A’s maybe from your side. So let us start right away. Okay, a low power mindset for me it’s I think it’s my life. So I always try to save energy. So I even remove my car and use my cycle to travel. So any energy, anything you do is energy related and to reduce it and to be aware of it I think it’s the most important factor. And to show you what’s it about with cellular technology, this is today. Okay. Everything is related to a use case. Yeah, so in IoT we’ve got so many use cases and only a few of them are listed here. And as you can imagine, are totally different parameters which are affecting your device related to your use case. So a medium device which is in your cellar or your flat has a totally different need than pet tracking device where your PET is running around and you want to see where it is currently or where it might be in a few minutes. And so all the effects we will show you in the next half an hour are related to the use case. So you could not map one use case to the other. So you really have to build your own setup related to everything which I tell you. What is the initial use case of cellular technology? So of course, it was totally voice centric. So you did a phone call and you were working walking around. That was a pleasure to not be lined up somewhere. And the need to have was that the mobile network is continuously sending signals which are received from the cellular device and then as soon as you reach a border of an antenna, it will just hand over to the next antenna for that your call is not interrupted. From that, it went more and more to be voice and data and finally to be mainly data. As you know your smartphone is I think has more data traffic than your maybe your computer at home because it’s continuously sending data. So what was done, the voice capabilities have been extended and with multiple antennas the speed was increased for the data and with that carrier aggregation, which is a nice term for putting two pipes into into your into your setup, this speed is even increased more. What does it mean? So carrier aggregation for example or higher speeds are one possibility to save power because it’s quite easy. Imagine you order a drink, big glass, and you got a straw and you recognize, okay, maybe I’ve got not enough time to finish that drink in time. So you order a second straw and but the barkeeper only hands over these small things, you know. Your first straw is like that, your second straw is like that. It’s only half of the size. But anyway, at the beginning, you can easily use both straws and you’ve got a boost factor of two and you can empty your glass faster and that is quite useful. But as soon as the liquid is below the length of your second straw, you pull some air and which makes it maybe more difficult to use the first straw because the second does not have any effect anymore. And this exactly is that section here where you you have a kind of carrier aggregation of two straws, but only the first one is used and the second one just consumes power because you have to yeah, to use it more and finally it has no effect. But with a full glass, your second straw is always a good help. So that’s the same with carrier aggregation. Yeah, even if it takes more power, if you increase the speed, you might be in a low power mode. And this is exactly what was done with LTE carrier aggregation. So high performance LTE devices like your phone. But how does it fit to IoT and all these small devices? Everything needs to be small, cheap and sustainable and there should be millions or billions of devices. Of course, this what I’ve just told you carrier aggregation and so on does not really fit to that use case. Therefore 3GPP and the cellular technology introduce communication or machine type communication to come away from the standard voice and data centric use case to be more centric to the machines because machines are talking to each other. That’s IoT. And these devices need to be power efficient and billions of devices need to be supported and the coverage should be enhanced and definitely the complexity should be reduced to make the devices smaller and cheaper and yeah usable on your wristband or whatever. So everything needs to be small. Okay, so what is the biggest consumer, the energy consumer in cellular technology? Of course it’s the part where you’ve got the biggest resistance and this is the air. So airtime is the biggest consumer. So everything we want is reduced the airtime. And first approach which is quite easy: reduce the transceiver activity, send less and then reduce the receiver activity. Just listen less or receive less data quite easy. And then as I have shown you a mobile phone with four antennas or whatever which makes it more expensive and more power consuming because every antenna has a transceiver and the receiver part and this needs to be powered. So one antenna only would be a measure. And then these handover scenarios for a metering device which is in your cellar there’s no need to hand over between cells. So this metering device is always fixed in your cellar. It will not move. So remove these scanned off neighbor cells to be able to do this handover. And finally there’s a possibility instead of shutting down and powering on a device is putting it to sleep and waking it up. So that might be a little bit easier. Okay, what are we talking about? So just don’t be afraid of this slide just to show you what is happening. So if you switch on an LTE module, now there will be a startup phase and scan and registration phase which is similar to the sending data phase and afterwards we’ve got a kind of so called connected mode and finally there’s the idle mode. In the idle mode nearly nothing happens but an LTE device is listening to the neighbor cells as I said because this LTE modules they come from the mobile world and this is useful for mobility. Comparing 2G, 3G, LTE which are mostly combined in one module type, 2G is quite old, 3G is dying in the moment, we all know it’s switched off. For example in Germany, in the US it’s switched off and in many many countries. So I think we do not need to pay attention to 3G anymore. But comparing 2G and LTE, as you can see here, there’s not such a big difference after switching it on. The 2G device will register to the cell and then it will be in a kind of idle mode. And the LTE device will do it right the same. And if you pay attention to the numbers for a four minute time, They nearly have the same average of power consumption. But of course the biggest difference here is the low speed of the 2G device compared to the high speed of the LTE and the latency. So everything is quite slow in 2G and LTE is so fast. But to send only a limited amount of data 2G is still an option. You can see here, I wrote down no IP address at this stage. So after switching on the 2G device and it will attach its it will it does not have an IP address in that moment in comparison to the LTE, which has an IP address wide awake. So there’s an automatic attached procedure and registration procedure. But what you can do in this status was a 2G device is send an SMS or USSD data which is going via the signaling channel. Signaling is quite the same as the procedure while while starting up. So it is still a possibility and in many many many countries 2G is different and will be the next three four years, five years until it’s switched off. So Markus, so just a very quick question on this slide. So in this particular use case where we’re sending a low amount of data, you’d actually propose that sending data via 2G rather than LTE would be the preferable option for reducing the power consumption. That might be an option. And if you are not moving, if you’ve got better signal strength from the 2G cell, Of course, it might be better than using a 4G device, which is which needs to power much more because the LTE cell is too far away. So with LTE, you should not have a problem within in city areas, but on rural areas 2G has the better reach. Yeah. So that might be an option. And of course to find out the best solution is to measure and to try to test your device with different technologies and this is the only thing you can do. You need to try it. Okay in comparison to this low power wide area network module and there are different kinds of modules where 2G is combined with LTE M and NB IoT. Some modules have only LTE M and NB IoT. Just to show you the difference of one and the same module on the same place, so it was on my desk, Of course, somewhere else in the world you could really it look different, of course. Yeah, because on every single place you’ve got different conditions. But on my desk, I switched on the 2G device and you have seen some peaks here up to one point two amp for a short time. Then I switched to LTE M where you can see it’s much less only peaks up to five hundred five fifty milliamp for a specific time. And if you see the comparison in this idle area, it’s much lighter here because there’s already a reduction of signaling within LTE M. So there are spikes only every one point two seconds, one point two five to be correct. This is the most used setting in the moment and if you switch to NB IoT you can see that the final peaks are even lower and these small peaks in the in the idle mode are even more on the time session even more far away. So it’s two point five seconds per every spike. And if you then switch on EDRX for example, you can remove some of these spikes which will be even lower power consumption. Let’s get into detail for these RAD technology selection. This LPWAN module, how does it know which RAD technology to choose? So it’s configurable. By default, I would say all modules should choose at EM first, NB IoT and only as a fallback they will switch to two The problem is that there is no handover between these technologies. So imagine your device is connected on LTE M and you move from one place to the other. So the pet tracker, for example, so the dog is running away and there will be a cell handover, which is quite good because this is mobility, which we need for a tracking device. But as soon as the dog runs out of reach of the LTE M cell, there will not be an automatic handover, for example, to NB IoT or LTE. But the module will start to scan the networks. First, it will scan for other LTE M cells. If there is no LTE M cell anymore, it will scan for NB IoT cells. And if there is no NB IoT cell, it will scan for 2G cells. So the scanning is done quite often maybe yeah depending on your technology and is consuming power. And as you can see here it’s even worse in NB IoT. NB IoT attaches to one cell and if you move an NB IoT device, it will be connected to that cell as long as it loses the signal and then it will search for a new cell and connect to that new cell. This is really not a mobility behavior. So we can say NB IoT is more for static devices. Right, so just to confirm NB IoT will consume potentially more power on an IoT device compared to 2G and LTE M? Depends. So if you are moving around and the device is always on, it will scan so much. Might happen, yeah. Okay, what can you do to lower the scan time and to improve this scan behavior? You’ve got frequencies within your module and depending on the amount of frequencies, the time the module needs to scan is higher. So you see here on the right, I did a scan where five frequencies are enabled And I use two different mobile provider sim cards and they are nearly the same. So to scan five bands just takes much more time than scanning only two bands. It’s obvious. Yeah. But there are some bands which are not used so much. So a decision would be to remove bands or frequencies which are not so used depending on your geographic location, depending on your provider to speed up this scanning time and to reduce the energy. And then of course it would be good to have more MNOs within one country for example with a SIM card like a Soracom SIM card for example. You will have the possibility to have multiple MMOs per country, means the module will be able to find a mobility provider faster than a SIM card which is from one single provider and searches for a signal of this single provider only. As I said, this will speed up the first attach and for moving devices this continuous scan. For static devices, it’s not so important because they scan one time and afterwards they always connect to the same cell. So this is not so important for static devices. But reducing the number of bands, of course, might cause that you reduce your footprint, your worldwide footprint because some countries you might exclude from your list with disabling some frequencies. Okay, then you’ve got this efficiency. NB IoT is slow, LTE M is faster and normal LTE CAT one for example is much faster. But depending on the amount of data you send, you can decide which or you need to decide which radio access technology to use. I read a nice paper where they really made a test and they found that there is a sweet spot of choice between NB IoT and LTE M of one hundred and fifty kilobytes. If you send less NB IoT would be power efficient, more power efficient than LTE M. But as soon as the amount of data to be sent or to be retrieved is higher, you should choose a better performing technology like LTE M. But as well with that one you might come to a point where an LTE CAT one module is a better choice and right in the moment in this time there is a new generation of LTE CAT one called LTE CAT one BIS and this BIS just says that it is an LTE CAT one module with one antenna only, which makes it less complex. Again, we save some energy which makes it less expensive, which is good for IoT. So there is a decision to make how much data you have to send and which technology fits best. So but how does it work sending with less power as we have seen three slides ago, but being able to communicate even wider wide area networks. Yeah. So this is somehow strange how this is done. So the technique behind is the so called coverage enhancement. So if you are outside with an LPWAN device, you might even send with low power status. If you go light indoor, like under the roof or wherever, good reachability, good thickness range, but there’s something in between. You might need to switch to medium power and indoor maybe in your flat it’s already sending with full power. But as soon as this goes to the deep indoor, maybe into the cellar, it needs to send with full power and the coverage enhancement, which you always read in all over in all these articles is done by repetition. So it’s just like that. That’s a module is sending the same signal again and again and again. And in this example, the same signal is sent sixteen times. So if you have got three data points to send, it will be three times sixteen times, but the repetition can be up to two thousand and forty eight. So the coverage enhancement is realized by just repeating the same message in the hope that one of these sixteen or two thousand and forty eight messages will reach the final antenna. And of course this needs to be taken into account because sending the data once uses power twice uses the same power again and again and again. Okay, what can we do to reduce the power in the so called idle phase? You have seen these big idle phase in LTE and this is available in LTE M and in the IoT as well. The idle mode we can just shorten by not switching off the device, but putting it to sleep. The advantage is that at the beginning, at the startup of every module, you have this scanning phase and connection phase to get into the connected mode. This is shortened a lot if you put it to sleep and wake it up again. The disadvantage is setting up these timers, which can be quite long. So you can put a module to sleep for for hundreds of hours and you can define this active time, this idle mode active time as well to up to three hours for example. But it’s always a deal between the network and the device. So as a device, you’ll say, hey, I want to sleep one week and then I want to wake up for ten minutes. And you send this request to the network. But the network finally decides which values it will grant to you. Because as you can imagine, there are billions of devices, not within one cell, but let’s say thousands and all have the same wish for example. All say, hey, I want to sleep one week and then I want to wake up for ten minutes and then the antenna will have a problem and the network behind. So the network will just spread all these requests to a wider frame to be able to handle thousands of devices. Yeah. So in most of the cases, will not get the values which you wished or which the module requested, but slightly modified by putting a kind of slicing factor to that number. To set up these PSM values, you need a special command. And this command, we created a small calculation tool in our webpage. If you like, you can take a look at it, but only if you like AT commands. Thanks, Mark. I’ll just share the link to it in the chat as well. It was interesting. Okay. Then there’s a second possibility to reduce this idle mode and the consumption of the idle mode. The main consumer during the idle mode is the receiver. The receiver is continuously listening if there is some data to be received. So for example in SMS, if you get an SMS, the receiver continuously listens to the so called message channel and as soon as it sees, oh there’s a message for me, then the connection will be set up and the SMS will be retrieved. This is nearly the same here, but you tell the receiver to make some pauses in between. Yeah, so only listen for some pagings. That’s the word we we use for that. Listen to pagings only every twenty seconds and only within these twenty seconds only for three seconds, for example. Normally, as I said, it will be done in NB IoT every two and a half seconds and in LTE-M every one point two five seconds. And you just limit the amount of listening peaks to what is needed. And if you run an application which is not intended to receive data at all, yeah, there’s no need to receive data because you push only. So that’s depending on your use case again. If you push only, you can really minimize that idle time by setting the lowest possible values, and this will be best for your power [consumption]. Markus, how significant can those gains be when you’re using PSM and EDRX? Are we talking about small fractions or can it actually create quite a significant impact on the amount of power that a device is consuming? Yeah, of course, with PSM, if you put a device into sleep for two weeks and then wake it up for sending a data bunch instead of letting it run the whole time. And after two weeks, you send the data bunch. This is of course a big difference. Yeah. With EDRX, it’s more kind of fine tuning. Now you can minimize it, minimize it, minimize it. And this is a goal to give some more insight about some tweaks to minimize the power consumption. The connected mode here at the beginning and over there. Connected is right after this scan procedure. You remember from the one of the first pictures and after sending some data. To send some data, the module needs to connect. And unfortunately, there’s a so called RRC activity timer. So it just tells how long to be in the connected mode. As you can see here, this is a table from the GSMA. These RRC activity timer depends on the mobile provider and could be different. So some set twenty seconds. Normally in LTE, for example, it’s ten seconds. So you can say at least this connected mode is ten seconds. Ten seconds even if you just sent one bunch of data for one second, it will be connected in ten seconds doing nothing just because this timer is set to ten seconds. So these release assistance indicator is a possibility to tell the network, I send some data and I want to leave this connected mode right after sending the data because I do not expect to receive anything. Normally, they intend to be in connected mode like you send some data and you receive some data when it’s good to stay in connected mode, of course. But for some devices, depending on your use case, it might be better to leave the connected mode as fast as possible. And this of course will limit the connected mode which even comes a little bit more than the idle mode and yeah it’s good for your battery. Hardware. This is of course an amazing topic beside setting up a device. The antenna is the most important part towards the network and with a bad antenna you need more energy to transmit and receive data. So an antenna tuning is essential and we have some great tips available from antenna companies in the network. You can find it in the internet where they tell you the minimum sizes of PCBs or whatever and how far away an antenna should be from specific parts of your of your device. And this needs to be taken into account definitely. So antenna tuning, tune tuning the antenna to this to the bands which you have selected before, for example, of course, be the optimum. And there are many, many modules which have the possibility to do software defined radio, which means as soon as the module switches the frequency to another band, the antenna could be tuned to that specific band. So this is definitely one of the parts in the future to optimize everything and software defined radio is one part here. And the next part would be maybe to put the application on the wireless module itself. So there are many, many modules where you can run the application on the module instead of using a microprocessor, which is running beside and doing the calculation work or whatever. So it’s possible to do that on the module right away. And then one other point might be to avoid USB for example, USB is power even if there is a power saving feature. A serial connection I think would be much better. Okay and then of course sending data causes one of the highest peaks and the more data you send, the more energy you need. So quite simple is reducing the data is one of the key topics of course. And as you can see here, how many bytes are sent if you send for example forty bytes of data in SMS USD, it’s a pure data, forty bytes. For UDP you have to add eight bytes more. TCP is having a lot more because TCP is doing a handshake before and every single message needs to be acknowledged from the other side. So there’s a communication flow between the parties and of course everything is going over the air and we wanted to avoid airtime. So TCP of course needs more, HTTP even a little bit more and then if you add security over the air, it will be much more. Doing private key exchange is so many messages will be sent securing or trying to secure the communication channel. So this is mainly here the interface as I told you between the device and the base station. So why not just leave the security in the network so we can edit later on? Soracom, from my point of view, built a perfect solution for this. You can send, for example, data via SMS or TCP, UDP, so as simple as possible to the base station over the interface to save power. It will be transferred to the Soracom back end, private back end, secured communication. And then you decide what to do with the data. You decide to push it to your server and add some encryption afterwards. So this is wired. We don’t care about the power consumption so much anymore because the machines are on a power cord. A very nice possibility is as well to encrypt your data in your own way. So to minimize it, let’s say like zipping, data zipping. Yeah, so it’s you’ve got a data string where you want to send a battery value and the temperature and so on. And you might send a whole JSON frame, which is looking like that in the output area here With the possibility of our binary parser, you can define a format and send only the data. As you can see here in the sample data, which is as soon as it reaches Soracom backbone, will be blown up to a full JSON, which then can be securely transferred to wherever you want. There’s a nice tool available on the internet as well to set up your binary parser. Another possibility would be for example to use message pack. Message pack is a very common possibility to shrink your data, but shrinking data might produce some processing power as well. Processing power will again consume some energy. You have to decide if this is necessary. It’s constant trade off, isn’t it? Between how much power is going to consume and what the Always, always, you have to choose. And finally reaching any of Azure or AWS clouds is possible via Soracom by pushing the data directly into the cloud. As we talked about, we minimize everything which is between the base station and the device. The airtime needs to be reduced. Then we’ve got a secure communication over the mobile carrier into the soracom network and then we just blow up everything which we compressed before via binary parser, via choosing the protocol according to our needs and yeah, this is Soracom. Okay. Duncan. Yeah, so I think this comes to the end of the slide. We put this slide in at the beginning because we were talking about the idea of, before we, when we were prepping for the session, we were talking about this idea of turning a light switch off at the end of the night when you leave a room or when you unplug, you know, when you turn your laptop off, you unplug it. Know, you gave some examples, Markus, ditched the car and you used the bicycle. But actually, I think having listened to the whole presentation, I really feel that actually low power mindset is about much more than just switching the light off when you exit a room, isn’t it? Definitely, definitely. As I told, for me, it’s in every single part, optimizing, squeezing here and there a little bit to optimize the solution. But it always depends on the use case. So you could not compare one with the other. There are so many variables and it depends on use case. But, you know, the changes we’ve talked about today, it’s not about just squeezing extra one percent or two percent. You know, some of the use cases we’ve seen is where you have had a device that would typically last two or three weeks on a single charge. But actually, once you’ve implemented some of the ideas that you’ve talked about today, you could extend that battery life to or potentially even years. Is that correct to say, Markus? Definitely. Definitely. Yeah. So a device that would have lasted a couple of months or weeks or months is now actually lasting for up to ten years on a single charge. So quite a big difference. And what I like most is reducing the amount of data to be sent, of course, is not good for our business case, you know, because the people pay per use. You only pay what you send. So less they send, the less we get. But anyway, I have the ambition to save the energy, which is and to to maybe save the environment a little bit better, which is quite important. And so for for our customers should be a win win situation by Right. The amount of data, pay less for this data, and then have the effect to save the battery power. Yeah, absolutely. Amazing presentation. Thank you for those slides, Markus. You can just go on to the next slide, which I think is just the holding slide. What I’d like to do, guys, is just open the floor up. We’ve still got another ten minutes or so for any kind of open questions that people may have. Before we jump into the Q and A, there’s a handout that I wanted to share at the beginning that I’ve got. So let me just upload that. We have, like, a Soracom overview PDF, so I’ve just shared that in the handout section of the webinar tool. So, by all means, if you’re interested in learning more about some of the things that Soracom can do to help you with your IoT deployment, please just download that PDF, and you can learn more on our website if you’re interested in discovering a bit further. If you have any questions for Markus about power saving in IoT, anything about reducing the power consumption of your devices, please do use the live chat now. We’ve had a couple of questions that have come in whilst you’ve been presenting, Markus, so we’ll go through those now. And, yeah, by all means, if you have anything that kind of springs to mind, anybody else here that wants to ask, just put it in the live chat right away. The first question we had here came from Brad, and it was when you were talking about the different cellular technology. Brad said, Is 2G going to be phasing out in the same way that 3G is? Not as fast. So 3G is phasing out much faster. I think the dates for 2G in Europe, for example, is 2025. There are of course countries where it will be faster. But in other countries in Asia, Africa, South America, I think 2G will last much longer. But for Europe, it’s scheduled to be 2025. Okay, so we still have a little bit more time to use those technologies, do we need to? Gavin asked a question, which you may have actually answered already, Markus, but he just asked, what is EDRX? So you talked a little about PSM earlier on. Perhaps you could just give a little bit more explanation about exactly what EDRX is. Yeah. Okay. eDRX is for discontinuous reception. So the receiver part, as I told, is responsible for this energy consumption. And you just tell the receiver to make some pauses. There’s no need to listen every one second or even less for for paging messages, and that isn’t connected d r x already. But this value is set from the mobile operator. And in the idle mode, you have these eDRX value and the e in the beginning is for extended. Extended just means you can extend the value to a very high number and just tell the receiver not to listen to the network, which says power. But if you have the need for download data, like you want to switch on a light or whatever, maybe this is a wrong technology because if you push something and you want to switch on the light and the receiver is not listening to the interface for two hours, then the light will light up two hours later. That might not fit to the use case. So in every use case where you need to talk to the device and need to have a fast reaction, EDRX might not be the best solution. Here’s a question that I think you’ll like, Markus, from Simon. What is better for reducing power consumption on a cellular device? LPW1, NB IoT, or GSM? Okay, NB IoT is a part of LPWAN, but LTE M, NB IoT, or GSM, of course, one of the most difficult questions. But to make a kind of thumb rule, as I mentioned in the last hour, static devices are quite good within the IoT devices which sent data only are good within the IoT, but only if you have got low data amounts. Maybe we do see it via binary whatever. Which is moving is good to have on LTE M. Moving is so mobility is also possible with with GSM, but we saw GSM has high peaks of energy because they send with much more power. So it depends. It depends, yeah. I feel like we could do a whole webinar on that one topic actually, if we really wanted to. Here from Kieran. So fairly technical one. Maya, he says, My question is with regard to protocol selection. Where will COAP and AMQP be located in the figure of comparison of power consumption? Yeah, yeah. COAP is running on UDP and has a kind of acknowledgement mechanism as well. Yeah. So not as as TCP where you you need to do a handshake and and and everything, every single message, UDP message needs to be confirmed. And if the confirmation does not arrive in a specific time frame, the message needs to be sent again. So CoAP is a good possibility on top of UDP to make a kind of acknowledged communication. But as you have seen with coverage enhancement, if you sent a message which needs to be acknowledged, you’ve got download data, so you need to be listening. There’s a need to to listen to to the interface. So DRX is important here because you need to get the download data. And if this acknowledge is not received, you send it again. And maybe you’ve got coverage enhancement, but then you do not send it only once, you send it sixteen times. And you’re waiting for the acknowledge. And if the acknowledge is not coming, you’ll send it again sixteen times or two thousand times. Could be a cycle of death. So anyway, you have to try, you need to measure and you have to try to simulate as much as possible the different scenarios, like shielding your device and burying it, putting it into the earth. Yeah. So there are reduction mechanisms in the modules. If there’s no signal at all, how long should a module try to send? Yeah. So of course, there are some implementations in the module. It will try to search for a network for five minutes, then it will do a five minute rest, then it will do another five minute search, then it will do ten minutes rest and so on. So it will just enlarge in the resting time not to exhaust the battery all the time by just searching for for the network. Yeah. And this is already implemented in the module. So if your application then tries to interrupt this, oh, I have no network. I restart the module again and again. These mechanisms which are implemented in the modules are totally crashed. So it might be useful to get some details on how the module behaves in a specific situation and then adapt the application accordingly. Right. Great question, Kieran. Thank you for asking that one. And the last one that we have here, just because we’re just wrapping up on time, actually. So just from Yiru. Hey, Yiru. So her question is, What remedies do you propose if I’ve already decided on my device configuration? So she’s already deployed devices out to the market. She wants to now make changes to that device configuration. Does she have to completely revamp the hardware or does she have any other options? Depends. The hardware, hopefully not. There are possibilities. For example, it’s possible as a as a beta feature. It’s possible within Soracom to push eDRX values so we can override eDRX values from local providers in LTE networks, for example. With that possibility, you can kind of steer at least the settings of your device even if your device is already deployed in the field. If you are interested in that feature, please come back to us and we could give it a try. Reworking a device, yeah, as I said, if the application tries to steer the module too much, it might be hard because you need to change your application. Everything which is towards connectivity. So for example, provider selection within a specific country, are able so Soracom is able to to influence this selection criteria to change, for example, primary provider settings for specific countries via an over the air update of the SIM card. If somebody needs us, please come back to us and I bet we can arrange something. Can we can we We can a different operator in a specific country. We can just change the order of preference. That might be another possibility. Absolutely. I just want to squeeze in this very last one. It’s come from Daniel. So he says, how do I minimize the power consumption for large transfers to the device, such as with a software update? Do you have any? Okay. Okay. And IoT device should not have much transfers. That’s why most of the providers, they use incremental firmware updates, for example, for the modules itself, and incremental firmware updates or device updates should be one solution for your application as well. So think about doing incremental updates to minimize the amount of data. But really large transfers like for an Android phone or whatever, they are maybe boosting the power. So more powerful devices use less time. That’s one of the first carrier aggregation But depends on your use case. Great, great question. Thank Daniel. They’re really great. Well, we are pretty much at time, guys. So I think it’s time to wrap up. We’ve come to the end of the questions. All that really remains for me today is a massive thank you to you, Markus, for preparing today’s slides. So thank you so much for going into so much detail on everything presented today. And everyone who’s joined today live, thank you guys for giving up some of your time. Like I said, it is a really busy time of year, so we do really hope you found this discussion helpful, and we’d really love to see you again at another webinar in the future. If you do want to learn more about Soracom, how we can help your IoT connection connect to a application connect to a totally secure network, access tools for speeding time to market, reducing data consumption, transmitting data to the cloud, and a lot more, then you can just visit Soracom.io, or you can read more about our platform and schedule a call with one of our experts, and you can email us at [email protected], and a member of the team will be happy to have a chat with you about your application. That is it. So we are bang up on time. Thanks, everyone, again for coming, and we will see you again soon. Cheers, Markus. See you again, everyone. Goodbye. Thank you, and bye, everybody.