Business Cases for LoRaWAN Deployment

Source: IoT Space Station

Compiled by: Da Miao

Published by: IoT Think Tank

Please indicate the source and origin when reprinting

Introduction

Within the coverage area of the LoRaWAN network, the network capacity can be expanded by adding more gateways. A dense LoRaWAN deployment, coupled with optimized Adaptive Data Rate (ADR), can significantly increase network capacity while reducing power consumption and the operator’s Total Cost of Ownership (TCO) by about ten times.

Within the coverage area of the LoRaWAN network, the network capacity can be expanded by adding more gateways. A dense LoRaWAN deployment, coupled with optimized Adaptive Data Rate (ADR), can significantly increase network capacity while reducing power consumption and the operator’s Total Cost of Ownership (TCO) by about ten times.

LoRaWAN has become one of the key radio technologies for solving the challenges of Low Power Wide Area Network (LPWAN) deployment, addressing issues such as power efficiency, remote deployment, scalable deployment, and cost-effectiveness.

The LoRa Alliance has over 500 members, with heavyweight members such as Google, Alibaba, and Tencent recently joining the alliance.

The first wave of LoRaWAN network deployments primarily focused on large-scale deployments led by operators such as KPN, Orange, and Swisscom. The upcoming next wave of deployments is expected to be dominated by large private enterprises, enabling interconnection between public and private networks, especially for use cases involving LPWAN geolocation.

In IoT use cases, operators and end customers tend to prefer using LoRaWAN for connectivity. As IoT deployments grow in density, network planning inevitably becomes one of the important factors for ensuring the long-term success and profitability of operators and end customers.

A typical example is the recent contract signed by Veolia Birdz and Orange for 3 million water meters. Such large-scale projects require comprehensive and detailed network planning to achieve the required densification and service quality while optimizing costs.

Further Research on LoRaWAN Densification Technology

The LoRaWAN network utilizes existing network spectrum, and its architecture is a star topology, making network deployment and ongoing densification straightforward: when more gateways are added to the infrastructure, frequency pattern planning or reorganization is not required.

Compared to Mesh technology, single-hop to the network infrastructure minimizes power consumption, as nodes do not need to relay communications from other nodes. Another advantage is that, compared to Mesh, deploying the initial network in a sparse mode with low node density is feasible, requiring minimal node density to operate. More importantly, in the face of increasing interference, LoRaWAN is not affected by the exponential packet loss suffered by multi-hop RF network technologies.

Another unique feature of the LoRaWAN network is that messages in the uplink can be received by any gateway. The function of the network server is to eliminate duplicate elements in the uplink based on the uplink RSSI estimates and to select the best gateway for downlink transmission. This allows features such as geolocation to be easily integrated into LoRaWAN deployments. It also achieves uplink macro diversity, significantly improving network capacity and Quality of Service (QoS).

LoRaWAN also supports features such as Adaptive Data Rate (ADR), which allows the network server to dynamically change the parameters of terminal devices, such as transmission power, frequency, and signal spreading factor, through downlink MAC commands. Optimizing these settings is key to increasing terminal device capacity and reducing power consumption.

Optimization of LoRaWAN parameters and network densification can lead to a significant increase in network capacity. In fact, the capacity of the LoRaWAN network can be expanded almost indefinitely with increasing density.

Image Source: Actility

In the future, especially in urban environments, due to increased traffic, the business volume of LoRaWAN networks is expected to grow, and its future will trend towards LPWAN.

Densification: How It Reduces Total Cost of Ownership for Enterprises?

As more LoRaWAN gateways are deployed, the network achieves densification, and adaptive data rate (ADR) and power control algorithms can be intelligently applied within the network, greatly reducing the power consumption of terminal devices and lowering the Total Cost of Ownership (TCO) of terminal devices.

The following figures clearly indicate that densification can save up to ten times the power consumption and reduce the overall TCO of enterprise deployments over ten years. The labor required for battery replacement is the main cost of TCO for large-scale enterprise deployments over ten years.

Image Source: Actility

Densification leads to a significant reduction in terminal device power consumption, thereby lowering overall TCO.

Image Source: Actility

Densification leads to a significant reduction in terminal device power consumption, thereby lowering overall TCO.

LoRaWAN is typically deployed in unlicensed spectrum, allowing anyone to launch IoT networks based on LoRaWAN.There are three network deployment models:

1. Public Operator Networks

In this traditional model, operators invest in a regional or nationwide network and sell connectivity services to their customers.

2. Private/Enterprise Networks

In this model, enterprise customers typically set up LoRaWAN gateways in private locations (e.g., airports). They either own gateways managed by operators or use their own LoRaWAN network platform. The second deployment model is a game changer for dense device use cases, as network capacity and enhanced QoS can slightly increase costs. This is feasible because LoRaWAN operates in unlicensed spectrum, and gateways are very inexpensive and easy to deploy.

3. Hybrid Model

This is the most interesting model for LoRaWAN deployment due to its open architecture. It is impossible (or rather difficult) to compete with other LPWAN technologies or cellular networks due to the lack of roaming models between specific spectrum and public networks. There are some solutions, such as CBRS and MulteFire from 3GPP, but they are still in development and far from maturity for large-scale IoT deployments, especially for use cases requiring 10-15 years of battery life.

In the hybrid model, operators provide lightweight outdoor coverage nationwide, but different stakeholders, such as private enterprises or individuals, can reinforce the network according to their location needs, further enhancing the network density. This model achieves a win-win partnership between private and public sectors, allowing for shared network costs and revenues, and making the densest network with the most applications and devices.

This model is feasible because multiple gateways can receive LoRaWAN messages, and the network server can eliminate duplicates. In cases where different operators or enterprises operate networks, the LoRa Alliance has approved the roaming architecture in the “LoRaWAN Backend Interfaces 1.0 Specification” to enable network collaboration.

This model significantly reduces operator investment and provides a disruptive business model to establish IoT where it is most needed.

Image Source: Actility

LoRaWAN allows for both public and private deployments, enabling cooperative models based on cost and revenue sharing, and densifying the networks where they are most needed based on IoT application needs.

LoRaWAN Densification: Key Drivers for Reducing Operator TCO

When designing and deploying LoRaWAN networks, system operators must balance the costs of dense networks and their service sensors with those of sparse networks and their service sensors.

Traditional vs. Opportunistic Network Design

In the traditional network deployment model, operators deploy LoRaWAN gateways on telecom towers. This requires leasing space from tower owners, purchasing a weatherproof outdoor gateway, climbing the tower to hang the gateway, and possibly paying additional electricity, zoning, licensing, and backhaul fees. Operators need to conduct detailed RF propagation studies and hang enough gateways to provide sufficient coverage for the sensor locations they wish to service.

Another option is to opportunistically deploy “mother” gateways among devices already deployed by operators. Gateways are stateless, so they do not add much complexity to the host devices. An 8-channel LoRaWAN reference design can be paired with the host device using USB or I2C. Operators can embed simple 8-channel gateways into ongoing WiFi hotspots, power sources, amplifiers, cable modems, thermostats, virtual assistants, or any devices already in mass production.

Calculating the number of opportunistic gateways for a specific area can be challenging. The height of the gateway has a significant impact on coverage. A gateway deployed on the 20th floor has better coverage than the same gateway deployed in a single-family home basement. Gateways deployed on WiFi hotspots on utility poles have different coverage areas than gateways deployed on utility poles. Therefore, the actual number of gateways deployed in each scenario varies widely.

When detailed designs for each type of network are completed, it is often found that the opportunistic deployment model allows operators to deploy about 100 times more gateways at about one-tenth the cost compared to the traditional third-party tower leasing model.

Water Meters: A Typical Use Case

For the remainder of this analysis, we will assume that the operator needs to deploy a LoRaWAN network to service 100K water meters. Water meters present a challenging RF propagation model. They are installed on the ground or underground, must have a 20-year lifespan, and suburban meters are often covered by grass and dirt. Let’s assume a North American deployment model where we can choose to use high-power (27dBm) or low-power (17dBm) meters.

One possible design is a tower-based approach. In a tower-based approach, operators typically deploy high-power water meters to reduce the number of expensive tower leases. To operate at high power, North American regulations require sensors to transmit over 50 channels, prompting operators to deploy 64-channel gateways.

Assuming an average distance of 3 kilometers between the water meters and the tower-based gateway, sensors need to transmit readings once a day. Therefore, many meters operate at 27 dBm with SF 10. Sensor designers include a high-power RF amplifier, calculate the energy requirements over the sensor’s lifespan, and appropriately size the battery.

Another possible design is to opportunistically deploy thousands of “mother” gateways in the area. The question boils down to: “How many mother gateways do I need to cover the required area?”

In many urban environments, the average distance between a given operator and users is 30 meters. If operators can opportunistically deploy in most sites, the distance between gateways will be as small as 30 meters. For analysis, we assume the average distance between the sensor and the nearest gateway decreases from 3000 meters to 100 meters. When the sensor is 100 meters from the gateway, it can typically operate at 17dBm (or lower) power at SF7. Clearly, network designers must consider the distance distribution between a given sensor and its nearest gateway, but overall, the energy savings are still significant.

Comparing the total capacity of a tower-based LoRaWAN network with that of an opportunistic LoRaWAN network is also instructive. The cost of 100 8-channel opportunistic gateways is about one-tenth that of a single 64-channel gateway, and the capacity of the opportunistic network is 13 times that of the tower-based network capacity. As sensor density increases, we can deploy additional opportunistic gateways, achieving approximately 130 times the network capacity for the same cost as the tower-based network.

When we compare the costs of sensors operating at 27 dBm using SF10 for 20 years and those operating at 17 dBm using SF7 for 20 years, we find that deploying a denser network can save at least $10 per sensor.

When considering other use cases, the dense LoRaWAN network model saves sensor costs for each additional group of sensors. Most sensors do not have a 20-year lifespan requirement, and the cost savings can vary, but batteries are one of the main drivers of the cost of any sensor.

Conclusion

A large-scale network deployment may require a certain number of traditional gateways to provide a “umbrella” coverage, followed by strengthening using opportunistic methods. By densifying the network, the total power budget for sensors is significantly reduced. We can also envision a deployment model where an opportunistic gateway is deployed alongside a set of services. Operators will add IoT-based services to existing services (such as voice/video/data, thermostat control, or personal assistants), allowing sensors to coexist with gateways.

What Is the Future of LoRaWAN?

When using ADR algorithms intelligently in the network, LoRaWAN has significant capacity gains and greatly reduced power consumption and TCO advantages.

The EU already has 16 spectrums, but recent regulatory framework changes have relaxed spectrum requirements, increasing transmission power, duty cycle, and the number of channels.

Additionally, Semtech has released the latest version of the LoRa chipset with the following key features:

50% reduction in power in receive mode

20% extended coverage area

22 dBm transmission power

45% size reduction, now 4mm×4mm

Global continuous frequency coverage: 150~960MHz

Simplified user interface with instruction display

New spreading factor SF5 supporting dense networks

Protocol compatible with existing LoRaWAN networks

The capacity of LoRaWAN depends on the area and morphological parameters of the use case. The aforementioned LoRaWAN features and upcoming changes to EU regulations will significantly expand unauthorized LoRaWAN networks in the coming years to meet the needs of IoT applications and use cases.

As mentioned earlier, if the network is carefully deployed and advanced algorithms (like ADR) are used, network capacity will increase sharply, and TCO will be significantly reduced. With the future demand and breadth of IoT applications, this will be one of the key factors determining the success of LoRaWAN deployment.

LoRaWAN offers innovative public or private deployment models, where operators can gradually increase capacity and supplement additional capacity through private or enterprise gateways. Typically, for cellular networks, 5-10% of IoT devices are inoperative at the network edge. This is especially true for deep indoor nodes (for example, the penetration loss for smart meters increases by 30 decibels).

Such nodes can only be covered by densification of cellular networks, but encrypting only 5-10% of IoT devices is far from meeting use case demands. One way to solve this problem is to deploy private LoRaWAN at the cell edge and use a multi-technology IoT platform that integrates LoRaWAN and cellular IoT.

On the other hand, LoRaWAN provides a cost-effective way to increase network capacity. LoRaWAN gateways are highly cost-effective and can be deployed using Ethernet/3G/4G backhaul, requiring minimal investment compared to 3GPP small cells. This allows for a cost-effective way to establish IoT. Operators can gradually expand these networks based on application needs.

The LoRa Alliance has standardized roaming capabilities, enabling multiple LoRaWAN networks to collaboratively serve IoT devices.

The LoRa Alliance has standardized roaming capabilities, allowing multiple LoRaWAN gateways to collaborate in serving IoT devices. The macro diversity used across deployments enables operators/enterprises to jointly enhance their networks, providing better coverage at lower costs, such as in industrial use cases. As described below, the future of LoRaWAN will be characterized by private/enterprise network deployments with roaming to public networks and disruptive business models.

Image Source: Actility

LoRaWAN provides horizontal connectivity solutions to meet the wide-ranging needs of LPWAN deployments for IoT applications. However, only proprietary intelligent network server algorithms from network solution providers can realize these advantages.

Previous Hot Articles (Click the title to read directly):

• “How Difficult Is It to Make Smart Locks for Shared Bicycles?”

• “Cognitive Computing, Blockchain IoT, IoT Security… Those Who Understand Will Control the Future”

• “[Heavy Release] 2017-2018 China IoT Industry Panorama Report – The Deep Transformation of Industry by IoT Has Begun”

• “[Heavy] IoT Industry Panorama Report, Pioneering the First Two-Dimensional Perspective Panorama of the Domestic IoT Industry”

• “A Cartoon Explains: Besides WiFi, Bluetooth, What Can the Recently Popular NB-IoT Do?”

• “A Cartoon Explains: What Is LoRa Behind NB-IoT That Everyone Is Talking About?”