Episode 265's Sponsor: Silicon Labs

Silicon Labs, a leader in secure, intelligent wireless technology has launched their 2023 Tech Talk schedule. This year's Tech Talks include a dedicated technology series for Matter, Wi-Fi, Bluetooth, and LPWAN in order to help you build the development skills needed to deliver cutting edge IoT products. Join Silicon Labs experts, industry leaders for these one-hour, live virtual trainings created for developers by developers. Accelerate your device development today by registering at silabs.com.

How do you know if mesh networks are a fit for your IoT solution? Lukasz Krzak, COO at Embetech, joins Ryan Chacon on the IoT For All Podcast to provide the answer. They explore what mesh networks are, the evolution of mesh networks, mesh networks vs LPWAN networks, and implementing a mesh network.

About Lukasz

Lukasz Krzak is the Chief Operating Officer at Embetech and an assistant professor at AGH University of Science and Technology in Krakow, Poland. His interests are in embedded systems and wireless communication networks. For nearly 20 years, he has worked in cooperation with many companies, delivering hardware and software solutions, targeting mainly industrial and professional applications. He received his PhD in electronics in 2018 and has authored 30+ scientific papers. He’s currently responsible for Embetech's main product - the embeNET wireless communication network.

Interested in connecting with Lukasz? Reach out on LinkedIn!

About Embetech

Embetech designs wireless communication hardware and software for highly dependable industrial and professional applications, such as heavy duty systems in underground coal mines, railway crossings, factories, etc. We are an official partner of Cisco and STMicroelectronics.

Their main offering is targeting electronic device and system manufacturers that need reliable large-scale wireless mesh networking. Their product is called embeNET, and it is an embedded software that can run on multiple popular and cost-effective radio transceivers including chips from STMicroelectronics, Nordic Semiconductor, Texas Instruments, and others. It allows a highly scalable and reliable wireless mesh network, reaching thousands of nodes, with a superb range of communication. The network uses industry standard IPv6 and UDP protocols and is a viable alternative to popular consumer standards such as ZigBee, Bluetooth Mesh, or LoRa. Their software is backed up by a powerful network simulator that is used to validate user requirements concerning performance, timing, and energy consumption prior to any pilot installation. They also deliver additional R&D services such as radio hardware design, application software development, integration, testing, and helping with radio certification.

Key Questions and Topics from this Episode:

(01:06) Introduction to Lukasz and Embetech

(02:39) What are mesh networks?

(03:57) Evolution of mesh networks

(09:16) Mesh networks vs LPWAN networks

(14:02) Determining if a mesh network fits your solution

(15:56) Implementing a mesh network

(20:12) Learn more and follow up

Transcript:

- [Ryan] Hello everyone and welcome to another episode of the IoT For All Podcast. I'm Ryan Chacon and on today's episode, we're going to talk about mesh networks, how they have evolved over the years, how they work in IoT, and how to assess what mesh network is the right fit for your solution. Joining me today is Lukasz Krzak, the chief operating officer of Embetech. They are a company that designs wireless communication hardware and software. A lot of value here. I think you'll enjoy it a lot. Subscribe to our channel if you haven't done so already. Give this video a thumbs up and hit that bell icon so you get the latest episodes as soon as they are out. All right, before we get into this, we have a quick word from our sponsor. Silicon Labs, a leader in secure, intelligent wireless technology has launched their 2023 tech talk schedule. This year's tech talks include dedicated technology series for Matter, Wi-Fi, Bluetooth, and LPWAN in order to help you build the development skills needed to deliver cutting edge IoT products. Join Silicon Labs experts, industry leaders for these one-hour live virtual trainings created for developers by developers. Accelerate your device development today by registering at silabs.com. That's the letter S, the letter I, L-A-B-S.com. Welcome Lukasz to the IoT For All Podcast. Thanks for being here this week.

- [Lukasz] Yes, hello. Thanks for having me.

- [Ryan] Absolutely. So, let's kick this off by having you give a quick introduction about yourself to our audience, if you wouldn't mind.

- [Lukasz] Sure, so my name is Lukasz Krzak. I'm a co-founder and chief operating officer at Embetech. Both me and the company, we have some academic background. We come from one of the leading universities in Poland. We're a spin-off company with more than seven years of experience. So, Embetech has always been into developing wireless communication tools for professional and industrial applications. So first, they were in the form of hardware solutions. Now, it's all gravitating towards more firmware solutions and one of the products that we currently provide is a wireless mesh communication stack that you can use on various hardware platforms. So, our solutions have always been targeting scalability, reliability, and performance. And this is one of the solutions that we try to push to the market. Apart from that, we still do a lot of R and D services. Basically, seeking to be a partner capable of outsourcing the whole process of introducing wireless communication to our products, and from concept to design, up to certification.

- [Ryan] Fantastic. Yeah, I appreciate that overview. Yeah, it sounds like a lot of exciting things going on over at the company so that's great. So, what I wanted to do is-

- [Lukasz] Yeah.

- [Ryan] I know we had some interesting topics we wanted to discuss today and a lot of it revolves around mesh networking. From I guess for a start, would you mind kind of talking our audience through first, just what mesh networking is and then how mesh networking kind of has evolved over the years to where it is now?

- [Lukasz] Sure. So basically, when you talk about wireless communication, you either usually have something like a point to point communication where two devices talk, where the devices talk to a single point somewhere, usually called some gateway access point. And this is how many systems are organized. And in comparison to that, mesh networking gives you the opportunity for each device to be a router, which means that it basically becomes an access point to another generation of devices. So, you can actually have a network topology that spans over more terrain and each of the device extends the network. So, the capability here that you seek for is for the device being able to be a router.

- [Ryan] Okay.

- [Lukasz] To form the packets down the network and up.

- [Ryan] Gotcha.

- [Lukasz] Depending on the direction. Yeah, so actually, the mesh communication was a very hot topic several years ago. It all started with the concept of wireless sensor networks, which were like hot topic like maybe 20 years ago when the development of integrated chips actually allowed to build very cheap but capable digital circuits that had the wireless communication capabilities. So, starting from them, people started to build systems, devices that were able to operate on batteries for longer periods of time and sends the environment, send the data to some server. So, right after that, there was this need to increase the coverage and one of the methods was to introduce mesh networking. So, standards like ZigBee, for example, they became popular like 15 years ago because they provided this capability to extend the network to have more devices within the system. And later on, other standards begin to show up. So I would mention Thread for example, which is a big thing right now due to the Matter-

- [Ryan] Right, right.

- [Lukasz] protocol being introduced. But also, Bluetooth Mesh. So, a lot of these networking standards, they actually were built on let's say, the idea of accessing the wireless medium in an opportunistic way.

- [Ryan] Okay.

- [Lukasz] And this had some issues. So, back then when these systems were kind of new, kinda let's say, not mature yet, some issues were spotted. And me, myself and my team, we were trying to use these kinds of solutions back then, but we've seen some limitations. And also, people, when I talk to them right now, they kinda remember this. So, when they think about wireless mesh networking, they think about things being complex, being difficult to roll into an actual deployment. And this is still in people's heads and not without reason because these issues actually have been there from the beginning.

- [Ryan] Sure.

- [Lukasz] And later on when, for example, LoRa was introduced, so a type of networking that we now call low power wide area networking, LPWANs. This was like a relief for many people because it was easier. It was easier to set up such network, it was easier to understand how the networking actually operates, how it works.

- [Ryan] Right.

- [Lukasz] And for these reasons, I think it became very popular. So, if you think about IoT right now, a lot of applications right now are using this kind of networking because it's easy, it's well understood.

- [Ryan] Right.

- [Lukasz] But in the meantime, in mesh networking, there's had been a lot of development. So, new methods were introduced and in order to fight with these problems that there were in the beginning of mesh networking. And a lot of development was put into making things more scalable.

- [Ryan] Okay.

- [Lukasz] And one of the things that was introduced and is let's say, now pushing the limits of mesh networking is the idea of time slotted channel hopping.

- [Ryan] Okay.

- [Lukasz] So basically, the idea is that if you want to send out the message through a common medium, you either try to shout and hope that somebody will listen and tell you, "Okay, I got the message."

- [Ryan] Right.

- [Lukasz] And this type of networking, we call it opportunistic medium access because you try your best, right? But this is not scalable. If you have thousands of nodes, it's like having thousands of people trying to talk in a single room. So, in TiSCH method of communication, you divide the time and frequency into slots so you have multiple channels, multiple slots and you try to organize the communication appear on before it happens.

- [Ryan] Right.

- [Lukasz] So, this obviously introduces some complexity but the reward is that you can actually use the whole spectrum very effectively and not have many congestion problems. So, you don't have these collisions that may occur.

- [Ryan] Right.

- [Lukasz] And this leads the way to better scalability. So, I would say this is the major improvement that happened over the last, I don't know, five years in these kinds of systems.

- [Ryan] Gotcha. Okay, you mentioned LPWAN networks such as LoRa, so how does mesh networking kind of stand up against that? What are the positive negatives if we were kind of comparing the two or looking at the two separately?

- [Lukasz] Yeah, so first of all, in LPWANs, as I said, are not very, very popular. And I would say that still, mesh networking concept is thought of still as being more difficult and this is obviously for a reason. So actually, it requires some more understanding. It requires, maybe it requires better tools actually in order to be implemented effectively. And so, this is one of the differences. The other difference is basically in this scalability. Okay, so the LPWANs handle big number of nodes differently.

- [Ryan] Okay.

- [Lukasz] And I would say that if a system requires more performance from the large number of nodes, then probably using mesh networking is something which gives you more benefit.

- [Ryan] Okay.

- [Lukasz] Because you don't have like a single collision domain, single big room in which there are multiple nodes, you have the things more organized in the very low level of the radio interface. Also, when it comes to making the network more flexible, mesh networking gives you this ability because you can extend the network easily, right?

- [Ryan] Okay.

- [Lukasz] You just put another node and if the node is in the range of some other node, you got the communication.

- [Ryan] Gotcha.

- [Lukasz] In LPWANs, you have to extend the infrastructure by putting more access points. Now sometimes it's okay, right? Because if you know if the network is not very big, you can probably guess or have some intuition where to put the access points to have the communication going reliably. But sometimes it's not that easy. So, radio frequency communication is often viewed as a magic because you don't know how the signal goes, okay? It's counterintuitive sometimes how the signal propagates through space. So, sometimes it doesn't go through walls like we think it does. It bounces off the windows. It basically finds a path of let's say, least resistance, which is sometimes not easy to guess, to foresee.

- [Ryan] Yeah.

- [Lukasz] So, not having to think about it when you just roll the network out is something of a benefit when it comes to mesh network.

- [Ryan] Okay. Gotcha.

- [Lukasz] Yeah, so other things, maybe if you talk about data throughput, LPWANs tend to be on the lower side of things so usually, you can expect several messages maybe per hour, maybe per day-

- [Ryan] Okay.

- [Lukasz] in large scenarios, right? Because there is, the biggest difference shows. So, in mesh networking, when you have the ability to organize things better, you can actually push a lot of more information both down the network and up.

- [Ryan] Okay.

- [Lukasz] So, this is one of the differences. Unfortunately, this costs something. Okay so, usually, it costs some complexity but sometimes it's also energy because-

- [Ryan] Gotcha.

- [Lukasz] in LPWANs, you can just send out the signal and hope that somebody will hear it, but from the energy standpoint, you don't need to use a lot of energy to do that. In mesh networking, you have to be somehow synchronized with all the rest of the network. There needs to be some communication going in the background in order to support this connectivity.

- [Ryan] Gotcha.

- Sorry, connectivity.

- [Ryan] Gotcha.

- [Lukasz] So, I would think that this also shows the capability of having a two-way communication.

- [Ryan] Sure.

- [Lukasz] So, in LPWANs, it's somehow maybe more difficult to have an equal-

- [Ryan] Okay.

- [Lukasz] two streams of data being sent-

- [Ryan] Sure.

- [Lukasz] down and up the network. Whereas in mesh networking, this comes out pretty much naturally.

- [Ryan] Gotcha. So, let me ask you, if I'm listening to this and trying to understand if mesh networking or a mesh network is a good fit for kind of my solution and what I'm building, how do you kind of assess that or how do you kind of determine if it's a good fit or maybe it is not a good fit?

- [Lukasz] Yeah, that's a very good question. So, I would say that there are not a lot of differences when the application is small or you have a small number of nodes in your system. Because then if it's LPWAN or mesh networking, the differences in operation are not that significant. And in that case, I would probably go for the LPWANs as a simpler solution.

- [Ryan] Okay.

- [Lukasz] If on the other hand, also if the coverage area is small, that may be your solution may be shifting towards LPWANS. Also, if you need very, very low energy consumption, that would also probably shift you towards LPWANs. But in case of a large number of nodes where you expect more performance from the nodes in terms of data throughput, accessibility-

- [Ryan] Right.

- flexibility, these qualities I would think would gear you towards mesh networks.

- [Ryan] Okay, fantastic. That's a great overview. I mean, I think that's kind of from everything you've been saying this conversation, it helps kind of distinguish between the two and where you should be kind of focused on when you're deciding what to use for a solution. So, and this kind of ties in to that question you just answered but when you're talking with a company, what are the qualities that you really look for in mesh networking implementations? Like what's kind of the focus on that front?

- [Lukasz] Yeah, so if you decide on having mesh communication implemented in your product, you obviously want it to be scalable. Like this is the main point, right? And the tool that helps there is this deterministic network access.

- [Ryan] Okay.

- [Lukasz] So, this idea of having things more organized than it is in other LPWAN standards. So, I would go for that. And this should show up as this quality that we often call graceful degradation. So, you have the increasing number of nodes or increasing density of nodes.

- [Ryan] Sure.

- [Lukasz] And what you expect to have is that it doesn't collapse which often happens in opportunistic medium access.

- [Ryan] Right.

- [Lukasz] You may get a bit of degradation in performance because you know it's physics but it doesn't break down.

- [Ryan] Sure.

- [Lukasz] So, this is one quality. Also, it should be standard based.

- [Ryan] Okay.

- [Lukasz] Okay, so you want to have it based on actual standards that are used in the industry.

- [Ryan] Right. Right.

- [Lukasz] So, for example, I would expect this network to work in IPv6 protocol, right? Because this is like the easiest to connect to other networking infrastructure. So, I would expect UDP transport protocol or TCPIP maybe.

- [Ryan] Right, right, right.

- [Lukasz] And also, standards based security. That is something which we all rely on.

- [Ryan] Yep.

- [Lukasz] Another quality I think is, which is very important and that was clearly visible a year ago was portability.

- [Ryan] Okay.

- [Lukasz] So, a lot of companies don't want to be locked in in some kind of vendor specific solution. They want to be able to move their firmware, their software to multiple platforms. And that was very troublesome for many companies-

- [Ryan] Okay.

- due to shortages in chips-

- [Ryan] Yeah, yeah, yeah.

- over the last years.

- [Ryan] Fantastic.

- [Lukasz] So, I would mention maybe just one more.

- [Ryan] Yeah.

- [Lukasz] This ability to have things on, to be able to check how the implementation is going to actually perform in the large scale. So, think about it this way, you want to roll out a big system, okay? Multiple, I don't know, thousands of devices within the system. How can you tell whether your performance will be there? Okay, so what will be the latency? What will be the data throughput? It's not that easy to answer this, especially if you think about mesh networking because the way the network may actually organize itself will influence the way it will later perform. So, I would expect-

- [Ryan] Right.

- [Lukasz] to have a tool in place that could answer these questions and this is mainly based on simulation or better even emulation of the actual physics of the actual hardware that is in place.

- [Ryan] Right, right.

- [Lukasz] So that you can actually answer these questions before any development because at the end of the day, you don't want to invest in something that will not work.

- [Ryan] Right, right.

- [Lukasz] So, you want it checked out.

- [Ryan] Totally agree, totally agree. Fantastic. Well, thank you so much for kind of taking the time to explain all this to us. I know we haven't spoken much about mesh networks at all so being able to kind of dive into it pretty deep here today has been great. Hopefully, our audience is gonna get a lot of value out of learning just not just what mesh networks are or where they kind of, how they've grown, but just how they're being used or how they can be used, what you should be looking for, how to determine if it's the right fit and so forth. So, for our audience who wants to potentially learn more, connect, follow up on this conversation, what's the best way that they can either reach out or learn more about the company?

- [Lukasz] Yeah, so we have a lot of materials on the webpage so that would be a good starting point, I guess. And then based from there, we seek to be a partner and also showing that there is a choice.

- [Ryan] Right.

- [Lukasz] The choice that people often overlook when it comes to wireless communication that there's this mesh networking concept. So, on our webpage, we also try to push this information forward.

- [Ryan] Fantastic. All right, cool. We'll make sure we link all that up for our audience to find out more information. Thank you again for taking the time. It's been a pleasure.

- [Lukasz] Thank you.