IoT doesn’t bother to knock

Nearly every time I read some article about the Internet of Things I feel that common folks have to think we are completely nuts ūüôā Do we really believe that we’ll control temperature in room via sensors in our cloths? Or that my fridge will communicate with my wash-machine and my microwave oven, monitoring my GPS and getting ready when they’ll notice I’m coming home? It reminds me the programs for 8-bit computers which were making lists of what you needed to buy in supermarket …

And still the IoT is reality. The IoT, or M2M if you want, is not about bringing the Internet to the devices but on contrary making smarted devices which need¬†the Internet in order to behave even smarter (enough of ‘smart’, I promise). ¬†A few years ago if we wanted to connect some device (e.g. for¬†SCADA, telemetry, etc.) into our information system we would need to have a traditional radio network. Later it was possible to add a GSM module. Now the price of devices is¬†low and with LTE or in future¬†with 5G network in place the data is not a problem. Just for a fun I can buy a cheap Raspberry Pi and interconnect with LTE module. And indeed, the M2M traffic is now¬†the 2nd biggest source of revenue after the data for many mobile operators. The next year it can be the number¬†one.

Devices online per 100 inhabitants, OECD Digital Economy Outlook 2015 - © OECD 2015

Devices online per 100 inhabitants, OECD Digital Economy Outlook 2015 Р© OECD 2015

 

Of course, LTE is not the only one. For LTE-A an energy use and cost still remain concerns. More and more we can hear about alternative approaches such as Low Power Wide-area (LPWA) networking or Li-Fi.

IoT is not about standards (now)

The IoT is here already. It doesn’t wait for OMA (OMA_LWM2M),¬†GSMA standards¬†(GSMA IoT web) or Atis (5G Reimagined: A North American Perspective). This can be a surprise for some T1 operators. IoT is what we call ‘enterprise’. It’s like with the web applications. We¬†don’t need to integrate them. They are (firstly) created ad-hoc¬†to fix a problem. If you can’t do it fast, someone else can. Remember the drones? After the military usage one of the very first industries which started to use¬†Unmanned Aerial Vehicles (UAV) was surprisingly agriculture. Farmers didn’t care about how cool it was neither what were the¬†standards. They saw that it could help them with their existing problems. Btw. the agriculture is one of the IoT early adopters too.

Continue reading

IMS/WebRTC Tracing and Test Tools

IMS authors were maybe very wise but probably they were not operation engineers. If you have ever tried to trace all the messages which belong to a certain user/flow, you probably know, what I mean ūüôā¬†Basic VoLTE validation procedures are described at¬†Validating VoLTE¬†document.

In this post I’m getting on a shaky ground. I don’t know all the tools, some tools I know only briefly and I don’t want to promote any particular one. I also know that many operators and vendors create their own tools indoors (sometimes quite advanced). However¬†I hope that it makes a sense to briefly list a few tools one can encounter in practice. Would you have any favorite application, please let us¬†know.

Tracing

Wireshark

The most used protocol analyzer is Wireshark. In my experience 70 Р90 % of all the issues are somehow related to traces and Wireshark is the ultimate tool. Wireshark is free and open and all the authors and contributors deserve our thanks. If you use Wireshark often it pays-off to know a bit more about it, how to adjust the layout, filters, how to follow streams, read statistics etc. Рit can save a lot of time and energy.

Wireshark-pic1For developers it is good to know that we can write a dissector for our proprietary protocols. Actually not useful just for developers Рif we are not happy with protocol details, we can enhance already existing dissectors too.

The last year we had got a new major version 2, where the Wireshark was ported in Qt.

Continue reading