Aricent recently hosted a webinar titled “Leveraging LTE Networks for Internet of Things (IoT)”. In this webinar our experts discussed the various challenges that crop up when deploying IoT over LTE and how they can be addressed by optimizing certain network features. They also discussed how to address the various deployment and operational challenges that are faced when using LTE for IoT. They also provided insights into how LTE can be effectively leveraged for IoT through enhanced security, dynamic subscriber management and optimized subscriber and location updates.
Given the huge number of questions asked during the webinar, we were not able to answer every one of them live, so we've addressed them below.
1. Why do you feel LTE would aid IoT?
IoT applications have a completely different traffic pattern. LTE provides robust connectivity for IoT applications with different traffic patterns, even when they are constantly on the move. Also, a huge number of devices can be registered on LTE with SIM based authentication and they can be provided variable traffic patterns.
2. Will LTE replace 2G/3G IoT installations?
Aricent feels that current 2G/3G IoT installations shall continue as is since upgrading them would be cost intensive. However, all new installations will be LTE based. Gradually as operators move on from 2G, they will all adopt LTE based interfaces.
3. What are the key service related issues that need to be handled in LTE networks for IoT?
For commercial LTE networks, there are defined traffic pattern and hence all tarrifs are based on these traffic patterns. For IoT the traffic patterns are very different and hence operators need to define different billing solutions for the same. Also they need to define different policy control such that commercial networks are not affected due to IoT installations.
4. Is it possible to achieve multi-year operation with battery operated LTE modems?
For commercial phones, the traffic pattern changes and hence battery usage is not optimal, but for LTE IoT devices the traffic pattern can be defined and hence they can be signalled in most efficient manner to use minimal battery power. However, no devices that exist currently can achieve multi-year operation without batter change.
5. How will the IoT devices tariff plans be laid?
Tariff plans for IOT devices would be laid based on applications. Traffic pattern depends on the application which defines the usage of the network resources.
6. What advantages do you see in deploying parallel LTE-RAN devices for IoT?
LTE RAN running in parallel to the other commercial RAN helps in defining the QoS and priority for the IoT devices without impacting commercial usage. IoT can leverage a parallel Core which can be put in cloud and can be easily scaled-up or scaled-down, which is very essential for IOT as the devices get powered up or down on a large scale.
7. What types of applications will be used for maximizing usage of LTE for IoT?
Various applications are being perceived by service provides mainly in fleet management, home appliances, connected cars, and smart metering / smart energy.
8. Why are cloud based technologies necessary in case of LTE for IoT?
Complete core networks when hosted in cloud for LTE over IoT can help in scaling up or scaling down the systems which is extremely essential for IoT.
9. Is LTE for IoT hype or is it here to stay?
Many service providers are investing in LTE for IOT. The investments are also being done on additional infrastructure. These investments are proof that large scale LTE for IoT deployments will be leveraged extensively
10. Slide 7 seems to imply an interpretation of IoT as "sensor" network, i.e. no distributed automation is supported (for which requirements such as availability, low latency, etc. are of interest). Is this interpretation correct?
Largely yes, it refers to sensor network but LTE could also be used in conjunction with QoS supported for distributed automation networks. LTE as technology can also be used as IoT gateways for backhaul connectivity for distributed automation networks using other wireless links with the nodes like WiFi/Zigbee/wirelessHART etc as appropriate for the application
11. Seems like WiFi will beat LTE in many applications given fixed points and the HUGE data cost advantage of WiFi. What are your thoughts on this?
For fixed point devices, WiFi will be important, if they are data centric, non-real-time applications. All IoT networks are not similar. There are applications which require QoS which can be best achieved using LTE. Some IoT fixed applications use traffic pattern differently at different time of the day. For them too LTE is an advantage as one can control the strength of RAN devices based on usage. Also many IoT applications are signaling centric rather than data centric, for them too LTE is better suited than WiFi.
12. Please elaborate on MSISDN-less addressing?
IoT applications are mainly used for device to application or device to device signaling. For them MSISDN is not required. MSISDN is a network resource and hence operators can gain by not providing MSISDN to such devices, but still allow communication.
13. The cable (MSO) company is poised to lead - not the MNO. Please let us know your thoughts?
Aricent agrees with this thought that when WiFi will be deployed it will be led by cable operators rather than MNOs.
14. When is release 12 going to be commercially available?
Rel-12 could be commercially available by 2016-17.
15. Any studies on how the new low complexity/low capacity Type 0 device will effect bandwidth on sector - particularly on UL?
These kind of devices will be heavy on signaling and can lead to lot of signaling with home networks for authentication. It is best suited if operators have a parallel overlay network for IoT.
16. USA LTE vendors’ service plans are cell phone oriented: Minutes & Gigabits per phone or per “family” (1-6 units) at a cost of $240/year. When do you see these reaching $30/year in the USA for low data IoT devices?
Aricent feels that IoT devices shall be governed by a totattly different tariff plan and will not be based on data. It can be based on number of devices, type of applications, flat rate for usage of month, etc.
17. What's the protocol used for MTC to communicate with its server, proprietary one or XMPP or MQTT?
The protocols for MTC to server communication are determined by the kind of applications and the deployments that drive it.
18. Low data usage with higher signaling needs, isn't that a wash for the operator. Why is a parallel network needed for IoT if signaling goes up and data usage goes down?
That is the reality with sensor network! The traffic patterns are different than human voice/data usage. Many IoT applications have the need to send small data packets every few minutes or hours for e.g. utility meters - few times a day; vehicular telemetry - every few minutes when moving. Given infrequent nature of data transfer and battery saving requirements, the devices go into idle mode leading to signaling needs. On the other hand, the scale of such devices in the network is expected to be large. Operators are already pricing IoT applications differently from say smartphones, so we believe this would evolve further.
19. On what basis does signaling increase in IOT devices?
IoT applications have various usages. Consider fleet management. They would have lots of handovers, which many even result in PLMN changes. At every PLMN change there is interface with home network to get new authentication credentials. At every handover, there are signaling exchanges. They may not be heavy on data and hence idle mode may be common and also may result into frequent location updates.
20. Which is your vision for the automotive in all the IOT scenario, in terms of evolution and technical challenges?
In case of automotive industry, IOT applications like synch up of data with home servers, controlling home or office servers, auto update of location, getting real time updates from neighboring areas, knowing promotional schemes, local searches, live video streaming of auto traffic etc. could be basic applications for connected cars. Other automotive applications could be where automotive acts as a device and updates the fleet updates, knows where to park, auto control of the trucks in remote areas, updates of health of cars or trucks.
21. What releases does Aricent EPC-Lite support?
Aricent EPCLite supports 3GPP release 10 for all interfaces with backward compatibility to 3GPP release 8 and 9. It is also compatible to 3GPP release 11 for WiFi interfaces.
22. Msisdn-less is still valid for automotive?
It will be application specific. If it is purely for infotainment/navigation, msisdn-less devices would work. On the other hand, for supporting emergency calls, msisdn would be required.
23. Can you elaborate a little more into the "parallel deployment of the LTE network for IOT" , do we assign a full carrier for IoT?
By parallel deployment, Aricent means that there would be a parallel RAN and core network available for LTE devices, such that the commercial networks are not hindered due to the IoT devices. These could also be centralized in cloud within data centres such that scaling is easily achieved.
24. Do you see in future Unlicensed LTE getting used for In-building IoT?
Yes, that is very much possible. We see lot of trends where operators are exploring unlicensed LTE for in-building deployments. These shall however be only used only for non-real time applications where number of devices in area can change a lot over time.
25. low latency and low priority communication together does not make sense, can you explain
What we mean is that the IoT communication would be real time, but at the same time this shall be lower in priority compared to other commercial networks. In an area if the commercial bandwidth increases, then IoT communication shall be given lower priority and hence latency may increase at that time, but when there is enough bandwidth, the applications of IoT should be such placed so as to provide low latency in signaling to the devices.
26. How will Carrier Aggregation affect IoT device?
Carrier aggregation may not be a direct affect for IOT, but will be used more in commercial markets.
27. When you mention that signaling will increase, do you foresee regular negotiations between carriers considering signaling data usage, as part of the regular data usage of roaming clients? Clients may not be aware of this & feel they are being overcharged.
No, normal commercial users would not be affected and they may be provided signaling in normal commercial networks. Only for the IoT devices, there shall be increase in signaling and hence that would be placed in parallel overlay networks.
28. Is there any specific Network Vendor Inter-Operability testing that would be required (by chipset vendors like Qualcomm and MediaTek) to support Aricent's solution?
Aricent's core network has been interoperated with various devices which also include Qualcomm chipset. Thus there are no special tests required.
29. Will any new spectrum be included in future for IOT devices since LTE already started refarming its 2G resources?
New Bands are being added, for e.g. Band 26-32 and 44 are add in Rel-11/12 compared to Rel-10. However, not sure if exclusive IoT specific bands will be specified.