Choosing IoT connectivity technology needs careful consideration of multiple technical and commercial factors linked closely to individual use cases. Different applications favour different technologies so that what the product is designed to do is a critical factor in the decision. Features and benefits often conflict and need to weighed carefully in the decision so these articles distil the key characteristics that define an IoT connectivity technology to help you identify those important to your particular use cases.
Here are the parameters that we think define the strength of an IoT connectivity technology:
- Quality of Service
- Battery life
- Proprietary vs Standard
As we reach the penultimate article in this series we’re turning out attention to one of the key characteristics that has held back the development of the IoT industry since its inception, the high cost of delivering a viable network and end point infrastructure. This is changing rapidly with several low cost technologies either already in place or shortly due to come to market. We’ll set out the factors that contribute to the cost of deployment and offer guidance on how key decisions might be made.
The benefit of licence exempt spectrum
One of the most significant costs is that of the network itself and it has a number of elements. The first of these is the cost of suitable wireless spectrum. Conventionally, IoT solutions have been developed around legacy telephony based GSM technologies – GPRS, 3G and LTE. Not only are these technologies not optimal for most IoT applications but spectrum licensing costs are significant. The high cost of access to the radio spectrum inevitably translates into a correspondingly high cost of data transmission for the end user.
Weightless technology is frequency agnostic and can operate across different spectrums – importantly including licence exempt spectrum. Weightless technologies are typically deployed in sub-GHz frequencies; Weightless-N and Weightless-P operate in Industrial, Scientific and Medical (ISM) bands.
Longer range, fewer base stations
Range is another element with a strong influence on the cost of network deployment. Range translates proportionately to cell size and is one factor in the determination of the number of terminal devices that can be associated with each base station. The fewer the base stations needed in any given network, the lower the overall network cost.
Both Weightless-N and Weightless-P technologies are capable of ranges equivalent to and better than GPRS, 3G and LTE. Excellent signal propagation characteristics mean long range and excellent penetration of signal into buildings. Typically, an urban range to terminals mounted internally is around 2 – 5 km whilst a line of sight rural implementation could achieve up to 30km. However, as we have seen in a previous article (Choosing LPWAN technology for range), range is usually not the limiting factor that determines the number of cells and base stations in a network, it’s capacity.
Excellent signal propagation in sub-GHz spectrum also enables smaller, less sophisticated and lower cost antennas to be used on endpoints while antenna location, which might give rise to incremental costs, can in many cases become less critical.
Whether you are developing a network that you will operate or you are designing applications to connect to an existing network managed by another operator, low network installation costs lead to lower tariffs to each user of the network. The capital expenditure (CAPEX) for a network is considerably less than for an equivalent traditional cellular technology. Commercial grade base station hardware typically costs less than USD$3000.
Weightless devices have been designed so that they do not send much data, and do not greatly load the network. Sensor-style devices may only wake up once every 15 minutes, or even less frequently. This will allow a tariff that charges only a small amount for when a device is communicating, and nothing at all when it is not, because the loading it places on the network is very small. This very low operational expenditure (OPEX) makes it possible to achieve subscription fees of a few dollars per year needed to meet typical IoT use case requirements.
Low cost hardware
Weightless specifications require no exotic, high cost components to implement at either end of the link. Commodity transceiver and microcontroller devices together with a regulator, crystal, a small number of passives and a low cost antenna mean that modules can be produced for less than USD$2 in volume. Weightless specifications have been created in such a way that they can be implemented on modules costing just USD$1 – 5 in volume.
Low cost deployment and maintenance
Fit and forget pricing of modules is complemented by the low power requirements for Weightless terminals making battery operation feasible. Battery powered devices avoid costly deployments requiring a connection to the grid whilst battery lives measured in years mean less truck roll on scheduled maintenance.
Cost of LPWAN networks is not obvious
The cost of an LPWAN network is not just the cost of a base station – in fact as we will see, far from it. For the network operator the costs per site include:
- Hardware such as the base station and antenna
- Installation, including any site engineering needed
- Site rental
- Power and telecoms to the site
There are also core network costs and the operator may also bear the terminal costs either directly or in some indirect manner. If we consider an example of a UK-wide network, we can see how these costs break down.
Modelling a scenario
A nationwide UK network might need around 6,000 base stations to provide adequate coverage. The hardware for a base station might be around GBP£3k per base station, GBP£2k for ancillary equipment and GBP£5k for site engineering. Rental might be GBP£2k/year with backhaul and comms a further GBP£2k/year. Over a 10-year span the total cost would be GBP£50k/base station or GBP£300m. Of this only GBP£18m is the base station cost – hence the hardware cost is almost irrelevant.
If we assume 10 terminals per person that equates to about 650m terminals. At, say, GBP£5 each including packaging, etc, that is GBP£3.25bn. If the operator has to pay this cost it is clearly the dominant factor by far.
This leads to two important implications:
- Minimise the number of base stations. It is better to have a base station that is twice as expensive but with twice the capacity if this can result in fewer base stations. The additional cost at one site will be about GBP£3k but the lifetime cost saved by avoiding an extra site is GBP£50k.
- Maximise the lifetime of the terminals. If replacement can be avoided during a 10-year or longer lifespan this will provide massive savings.
Maximising base station capacity is complex and the subject of a separate article (Choosing LPWAN technology for network capacity). Another factor in reducing the number of base stations is maximising range in areas where there are fewer terminals and so capacity constraints are unlikely. This can be achieved with flexible radio systems that can adapt their modulation and coding schemes according to the received signal strength and with acknowledgements such that devices on the edge of coverage can retransmit when failures occur.
Increasing the longevity of terminals is not just about robustness but also includes:
- Long battery life – not only does this mean fewer manual interventions but also fewer times that the device is opened and contacts stressed.
- Software download capability – this ensures that if there are bugs discovered in the device software they can be addressed over the lifetime of the device. This is particularly important for security flaws which might otherwise render the device useless.
- Open standards – having devices conform to open standards means there is more likelihood of continued supply of network equipment and spare parts and less likelihood of a change in direction from the proprietary supplier that does not provide backward compatibility.
Weightless technologies are designed to minimise overall network costs. High capacity base stations reduce the overall network base station count and offer modes that can deliver increased range where needed. Terminals have long battery life, the ability to efficiently download a software update and the solution is the only open standard currently available for LPWAN deployment. In the final article in this series we will explore how open standards, as opposed to proprietary technologies, uniquely deliver sustainable, low cost solutions.